Our FPGA core is way less accurate than Mesen! How can zero (0) dollars beat hundreds (100s) of dollars?!>https://www.youtube.com/watch?v=oYjYmSniQyM>Mesen: 118/125>Neshawk: 115/125>NES_MiSTer: 110/125>Ares: 100/125>puNes: 96/125>Nintendulator: 94/125>Nintendo Classic Mini: Nintendo Entertainment System: 94/125
zamn
>>12010231Spiritualized1997's NES core on the Analogue Pocket is 85/125. That's way less accurate than a NES Classic Mini. FPGA is a scam.
No /vr/bro actually fell for the FPGA meme, right?
>>12010283>paying for retro vidya Lol anon, pls. However, as a junior comp sci student who is NGMI, I am intrigued by this video and want to learn more.
I don't know why people think that any FPGA emulator will inherently be more accurate than any software emulator literally just because it's FPGAThe emulation accuracy depends on people writing an accurate emulator, not just the type of hardware
>>12010295It's basically audiophiles, but for emulation.If you're told that it is better and you believe it, it will be better in your head.
>>12010295Because it's part of the marketing of the FPGA scam industry. And people without any technical knowledge fall for what marketing tells them. If it's not emulation, it's like real hardware then! With YouTubers like My Life In Gaming regurgitating it, you've got yourself a million-dollar business for a product that's worse than what a nerd has been programming at home for two decades for free.
>>12010309>No emulationHow the fuck does Analogue get away with outright lying? It's hardware emulation. It literally is emulation
>>12010231I knew someone would start shitposting this here. But the FPGA fanbase is obnoxious enough that they deserve this.
>>12010309>bold-faced liesIs there a reason people still tout FPGA as “real hardware”? I mean, I guess technically it *is* real. And it *is* hardware. But it’s also a literal fucking emulator, which makes what they’re saying here 500% bullshit. Don’t lies like this usually get exposed pretty quickly in niche internet scenes?
>>12010317I like to compare it more with these "nes on a chip" things.It is hardware, it is designed by someone, but it's a product of reverse engineering and online documents, so it don't work internally like the real thing.
>>12010324>don’t work internally like the real thing Right, and that’s pretty much the crux of the issue. I don’t know shit about these things since I’d never buy them to begin with, but from all the talk surrounding them, I’d been under the impression that the MiSTer *is* 1:1 real hardware. It literally says right here that it’s emulation, except it’s also not. kek
>>12010316lol this. I could point out that those accuracy test results aren't meaningful in any way that 99% of people would appreciate, but far be it for me to carry water for FPGA evangelists.
I don't get it, isn't the entire advantage of FPGA that the gate array can be programmed to behave like any system?
>>12010339Even at the gate level, BreakNES failed a few tests.>71/125
>>12010336>dude is an emulator better the rmulators but is not an emulator.That's an statement of human folish there.>>12010338
>>12010231>>Mesen: 118/125Standalone Mesen passed 119/125 for me, while Mesen's RetroArch core passed 100/125. Not sure how he got 118.
People have reported getting different results even on real hardware. Just because a test exists does not mean it's a good test. Like that GBA test which also can and has failed on actual GBAs. The video OP linked even says this test can fail on real hardware.
>>12010231Mister is already 120% more accurate than original hardware. What more do you want poorfag?
>>12010605Not to mention that this guy boasts that his own emulator is the only one that passes every test, you know, the tests he wrote himself...(It was almost comical the way he tried to argue he could write an ACE that intentionally would crash the game if the emulator does not give the exact result his test expects, so even an edge case of a color being slightly wrong matters.)
>>12010605Do these PPU revisions actually affect anything for playing games?
>>12010628According to the guy who wrote the tests, every single one matters no matter how inconsequential or little difference it would make because... he could hack the game to put an intentional crash if it fails.Great logic on display herehttps://youtu.be/oYjYmSniQyM?t=819
>>12010339Yes and no.The advantage of FPGA is that they can be configured at any time to behave differently, making them a "one size fits all" (except not really) solution. Which is great for niche applications, since the fabrication of bespoke components would be nonsensically expensive for something you'd only need a few thousand of, and programming an off-the-shelf FPGA would be much more cost effective. And since they can be reprogrammed at any time, they're ideal for prototyping and anything that would otherwise require frequent hardware changes.So they make a lot of sense for emulation boxes. Not only because it's a relatively niche application, but also because they can be reprogrammed at any time to mimic different hardware configurations. Meaning they could theoretically be made functionally identical to any number of original consoles at the drop of a hat.The caveat that everyone marketing these boxes (intentionally or otherwise) leaves out is that they need to be told what to do. And just because they CAN be made functionally identical doesn't mean they WILL be. Accuracy is entirely dependent on the knowledge and skill of the person writing the new instruction set. And in that sense, they have the same overarching limitation of software emulators.
>>12010309Nintendo lawsuit incoming.
>>12010231its only been out for over 5 years, whats the rush
But my favorite youtuber told me fpgas have the same eletrical pathways as the real hardware...>>12010723The N64 is 30 years old. all of its hardware patents are long since expired. What would they be sued for?
>>12010712>And just because they CAN be made functionally identicalIt's always going to be hacky. They literally cannot perfectly replicate the originals. The process they were built on is different, different gate timings and quirky edge cases.
There are loads of test roms out there that pass on emulators but don't pass on real hardware. It's usually retarded devs who either don't test on real hardware at all or they blindly assume every console is identical and will always have the exact same results.If a game only runs on a fraction of consoles then it's pretty pointless, but there are a few cases of this happening though. IIRC Codemasters games won't run on early NES revisions but I don't remember why. Very early SFCs couldn't run SuperFX games.
Isn't the main benefit of FPGA that you don't get the input lag inherent to emulation?
>>12010759That's the idea but the hardware has to be told what it's meant to do at all times, inevitably adding input lag through a layer of software.People call FPGA consoles "clone consoles" but they aren't. They are generic hardware that is then run through software to feed and interpret all instructions.
>>12010753>There are loads of test roms out there that pass on emulators but don't pass on real hardware.Can you post them?
>>12010775>>12010759It's more inherent to double buffer and Oses for the most part.One could just adapt a linux to create a lag free emulator at least video wise.
>>12010637>every single one matters no matter how inconsequential or little difference it would makeI don’t think that’s true. He acknowledges that MiSTer for example is valid as far as speedrunning is concerned because the tests that it failed don’t really have any meaningful differences for the playability of NES games. And frankly, if the speedrunning autists have decided that whatever exploits you can accomplish with MiSTer are legit, then that’s probably a solid standard to consider.
>>12010231Why is this smartass using fucking 1.40 of Nestopia when that emulator still gets updates? That shit is from 2008.
>>12010605>>12010627>>12010637Kinda reminds me of that one autist that wrote a bunch of stuff about how Krizz's flashcarts will fry your console's motherboards. And then it turned out he trying to grift people into buying his own flashcarts (that he never got around to making)
>>12010874Nestopia hasn't got updates because reached a point were it won't need them, sadly people don't get that so go full retard.Updates don't mean something is good just means people are advancing one step while going backwards 3 steps
>>12010904You don't know desu exists, do you? Also krizz shit no one cares about overpriced shit as someone said "you can mod any cartridge to use any game from sd card, that's only if you care about playing games on real hardware but if you are a collector trying to show off money the hacks isn't for you because you have to mod a game you own" which was the point several did.>why pay hundreds for krizz troon shit?>if you can mod a cartridge you own.>or buy a cheap chinese everdrive with all roms included for 20 dollars.It makes no sense at all going so hard to defend your troon fag, specially when his shit is made in china and sold for 20 dollars on shein.
>>12010637>I have seen the error of my ways.Wow, what an egotistical prick! Are most assemblyfags really like this?
>share them in my discordOh god, I can already see where this is going.
>>12010712You fucking retarded clanker, you literally just gave me a rundown of FPGA without answering my question
>>12010954Lemme try to give a better shot.The short answer is no, because FPGA "cores" are still basically programs written in VHDL or other "design languages", and the precision depends on how much the VHDL programmer knows about the real hardware, and also if he can squeeze the whole functionality in the number of logic gates available, and also if the FPGA chip can run fast enough.Now the long answer is that CPUs are mostly made out of the "pseudo components" in pic related, logic gates.FGPA is a bunch of reprogrammable logic gates connected in a grid, and a VHDL compiler will turn the VHDL program into a list of connected logic gates you can use to upload to a FGPA.The main advantage is that you can run all the logic gates at the same time, so if you program say 5 CPUs into the same FPGA, they will all run in synchrony, which is pretty good for simulating a sega saturn for example.On PC, you have to simulate CPU by CPU, cycle by cycle of each of em to not break the code.Or kludge the fuck out of everything and use breakpoints or..
>>12010954How did that not answer your question?
>>12010976I heard the N64 had its Verilog files leaked, I'm wondering if that's how N64 got accurate emus.
FPGAs are the retro gaming equivalent of this shit. Just use literally any office PC produced in the past 20 years.
>>12010909My point is that Nestopia still gets updated through Undead Edition. It's like picking up Mupen64 0.5 and ignoring all the developments from Mupen64Plus.
>>12010992It was mostly the work of a angrylion dude that got an very partial job made with the MESS emulator and improved it until it was actually good.Before angrylion, there was no such thing as a full N64 emulator, the approach was always to just emulate the main CPU, and re-do the programs running in the co-processor in C inside the plugins.In theory there are very few programs for the RCP, so you can in theory just rewrite em.But 20 years of that and we never got to run all of em correctly, probably due gay ass drama.And from a preservation standpoint it was shit, because you're cheating by rewriting a portion of the game.
>>12011004No one cares about unofficial, it won't change fpga are emulators and quite shitty.
>>12011015I don't give a shit about FPGA and Nestopia UE is just as "unofficial" as other forks like FCEUX and Mupen64Plus, don't move the goalposts.
>>12010994There is a bit more than pure audiophile autism, but not much.All double buffered systems have an inherent frame of lag (if you're running at 60hz, that's 16ms).Audio is also double buffered, and you have to add even more lag because it's made with software interruptions and dreams so you need to have at least 40ms of lag on a regular OS.And then there's the Xray box elephant in the room of the CRT TV versus led displays, and the whole 240p shit, that is not easy to achieve with a PC (but can with a Wii or raspi).Now all that said.Everything PS1 and above is also double buffered and have the same input lag, including in fighting games and including on the sega saturn and FPGA won't solve it.And all 3D games look better when running in higher resolutions, even on a CRT TV. mine literally shows aliasing of 3D even at 640x480 on composite, and if i do some gay ass RGB mod, it will be even visible.And "FPGA core making" is still programming, in a weird ass language, and if you don't know how the real hardware works internally (and you don't, specially on shit like the snes), your core will still be imperfect.Also for shit like the MSX that has a zillion different models, software emulation is just much more viable, as you don't have to create 100+ cores just to simulate every machine and machine configuration, which is painfully important for MSXdev.
>>12011034If you view online arguments with strangers as having goalposts then you’re a fucking jackass.
>>12011038No. You’re a financially irresponsible dick.
>>12011043Sir, i do recommend you reading the whole post before answering,I know its a bit long, but there's a bunch of useful arguments against FPGA shit on it, which is also why i don't even have one.
>>12011048No, I’m not reading shit. I’m playing mario.nes with Nesticle. I’m hacking the graphics to make Mario bald.
>>12011064That is fucking great actually, more modern emulators should have built-in tile editors like that.Also it is pretty fucking terrifying how some emulators are retro now.Nesticle, zsnes, no$gmb, genecyst.Even shit like ultra HLE you need to use 86box to get the proper experience from the time from the weirdass "22bit color" of the 3Dfx cards.
>>12011083>Even shit like ultra HLE you need to use 86boxI managed to get it running on Win11 natively using DGVoodoo2 KEK.
>>12011092You can. these glide emulators are great. but you don't get the weird 3Dfx dither and blur from the time.
>>12010627>>12010637He's being facetious when he says that. He says multiple times throughout the video that most of this stuff doesn't matter for regular use and that even speedruns just need to be "good enough" (he points out multiple emulators that are banned but probably don't have to be).
>>12011103He even explains in detail the ones that matter for speedruns and why.I imagine that for gamedev more of em are useful, but you shouldn't be doing stupid clown tricks with illegal opcodes anyway.
>>12011103The only thing he’s being is feces.
I do like the part where this absolutely btfos the narrative of "magic nintendo emulators are always perfect".
>>12011136Literally nobody thinks this except retards who think the cops will shoot them for downloading roms.
>>12011139I seem plenty of it."you get to emulate the Wii virtual console to get the best emulation!".But i seem mostly on the N64 context, which is probably also very wrong.
>>12011148Oh, you’re ESL.
>>12011154Yes, but i'm still right, and hate being right on that.
>>12010605>unfortunatly the poorbabies had multiple revisions of copeIndeed
>>12010605I tried it on my Famicom using an Everdrive and passed 121/125. Three of the failures are known to be caused by the flashcart and wouldn't be there if I actually flashed the ROM onto a real NROM cartridge. The last is caused by a hardware revision difference between the NES and Famicom. He said he's looking into fixing it so it passes on all hardware revisions. The tests are good.
>>12010994>office PC produced in the past 20 yearsMaybe not 20 years. A 2005 PC won't be able to run Mednafen Saturn at full speed, and it won't have Vulkan support for paraLLEl-RDP
>>12010295>>12010298>>12010309FPGA has the *potential* to be more accurate than emulating on a computer because computers are subjected to a bunch of different interrupts and OS routines and so on, making it impossible to make a clock and signal accurate emulator. Impossible. That is, the clock will skew over time, what happens when the OS has to service a higher level interrupt the same instant the clock goes from high to low? And also you will get skewed signal delivery over time for the same reason.The best way to make a relatively clock and signal accurate emulator on a conventional CPU, is to run it on the bare metal without any OS at all or with an embedded real-time OS the job of which is to basically set up the device and then load and execute the emulator. Even in this configuration it will not be perfectly accurate - you will have to handle input and output routines which means servicing interrupts.With a properly done FPGA there are none of these problems. The difficulty with the FPGA compared to emulation though, is that with emulation all you need to do for highly accurate emulation is figure out what signals go to which ports, and what each device does. With this solved, you can write as efficient or simple a program you'd like, and improving it over time is easy as you can just drop in new code for a device without reworking the whole thing.With an FPGA you don't just need to figure out what each device does, you have to clone the circuit itself and even possibly bugs in each IC, which people might not even know about yet. And phantom bugs can occur when you take what should be a whole gadget like a console and shrink it down to a chip sized object, weird timing issues. There's a book called High Speed Digital Design: A Handbook of Black Magic which goes into this in great detail. It's considered a must-read for people who design chips and is still a state of the art text.TLDR? FPGA can be better but it takes a lot more work.
>>12011278Just use Project 64 1.6 you dickhead.
>>12011282Peak Dunning-Kruger.
>>12011296NTA but could you please elaborate on that or is it safe to assume that you're a fucking idiot?
>>12011282With any Unix machine (probably Windows too) you can use the POSIX 'time' command to see how much skew your emulator is experiencing over time because of OS servicing routines, other user programs getting a slice of time from the scheduler, etc. It will vary based on system load AND most likely, what you're doing in the emulator as well.[code]0m00.84s real 0m00.38s user 0m00.40s system[/code]Here, for example, I took Vice's x64sc, their accurate C64 emulator, launched it, and from the command line had it immediately enter the monitor and quit the emulator after getting it all booted up. You'll notice the program only used 0.78s during its execution loop but that it ran for 0.84s giving a 'skew' of 0.06ms. Exactly 1/12th slow.Now I'll run it again. This time after launching some a big program in the background and see how well it does. I'll run pkg_check, a program that consumes quite a lot of resources as it traverses the filesystem and performs checksums etc.[code]0m01.12s real 0m00.48s user 0m00.41s system[/code]Oh look it's slightly slower. After pkg_check is done I'll close a bunch of open programs I have running and try again.>>12011296Explain to me what I've done wrong. You can't, because I didn't do anything wrong.
>>12011338You’re being a dweeb.
>>12010295Nobody thinks that FPGAs are some magic chip that by it's mere existence will create perfect emulators, I've only ever heard that argument from people who complain about FPGAs. You still need someone that knows what they are doing to code an accurate core on one. The difference is that it can allow for inherently perfect emulation while adding no additional latency and being electronically compatible with original accessories, or even cartridges, while in most cases being easier to code a more accurate core than in software.Software emulation can't do that, you will always have additional latency, for systems like a retron or polymega the original accessories are basically converted to a USB adapter (which adds more latency) and the catridge port just simply dumps the game and then runs the dump instead of running off the cart (which means if the emulator does not support any special hardware on the cart it won't work), and software emulators can't run the system in parallel like it originally ran, making it a lot harder to keep timings accurate than doing it in FPGA.>>12010735>The process they were built on is different, different gate timings and quirky edge cases.Consoles themselves themselves have had revisions with different process nodes and differences in timings of such edge cases or even in common cases before. You don't need to re-create them at the atomic level to get a 1:1 recreation.>>12010753>IIRC Codemasters games won't run on early NES revisions but I don't remember why.Wasn't that a security chip issue? Also never heard of the SuperFX claim.>>12010759That's one of many benefits, yes, it can have no additional input lag. You can never have zero additional input lag in software.>>12010775The software isn't running the console like in software emulation, it's telling the FPGA, the hardware, how to configure itself, which then runs the game. You are trying to imply that software and FPGA emulation are the same, they aren't.
>>12011341I'm not defending any particular FPGA setup, I'm just pointing out that there's no need to seethe about them because in reality they do have the potential to be more accurate. And I pointed out that on my particular computer, which is kind of old (Thinkpad T520 with Ii7-2760QM), VICE is about 1/12th slow. It's an old computer so maybe somebody can do better. But they will never get any emulator to run without it running a bit slow, the amount of stuff the OS has to do in addition to running the emulator will ensure this. Emulation's great. But FPGA can be much better.
>>12011341>begging for attention because you feel threatened by a thread that you can’t understand It’s ok, anon. I do that too sometimes.
>>12010873He argues that the top speedrun record for Super Mario Bros was made on a MiSTer and some might argue if it's valid over real hardware, then tests it and sees that it fails a handful of his tests but that the ones it failed would have not had any effect or difference on the speedrun so it's valid.But then pretty much immediately after makes a strawman argument that EVERY test matters in a speedrun because if he wanted to he could hack a game to make it fail if the emulator fails that test for any of his tests.He not only immediately contradicts himself but his argument on why they matter is because he could intentionally code a crash in a game if he wanted to... which is a pretty silly argument. You could easily do the opposite and hack a game to only not crash if the result is wrong.That's like saying one brand of TV is better than another because he would smash the other brand if you bought it.>>12010904I don't recall him ever making his own flashcarts. He was warning that they were not applying voltage conversion, but later admitted that he had made his article a bit too sensationalist. And IIRC it did cause Krizz to revise his designs with better voltage regulation.>>12010909Updates don't mean something bad only happens either, which is what you seemed to be trying to imply here.>>12010948I initially wanted to give him the benefit of the doubt since he seems to have some knowledge on the subject, but the fact that he made a test that NO software or FPGA device can pass, even can fail on real hardware, then mentions how his own emulator passes every test... then argues that you don't need very test to pass to have a valid speedrun only to immediately turn around and basically go "Technically I could hack the game so it crashes if that test fails" sets off a few red flags. The fact that his own set of tests only passes in his own emulator was a big one.
>>12011370If I was begging for attention I’d be writing multi paragraph posts on shit nobody cares about. I’m just bullying this fucking nerd.
>>12011370My benchmark is a bit naive, it's measuring more the emulator's setup routines than anything else. I should really have it launch a test program that exercises the CPU in a loop and compare the timing to a real C64. But it serves well enough to illustrate the issue. As the emulator runs, on any modern computer, other things will also happen and the emulator will have to wait. Or maybe even worse, if the emulator is running on a multi-core machine without the ability to set processor affinity (so the program gets a whole core) it will possibly flit from core to core as the CPU sees fit which means you get to move a lot of registers and state from CPU to CPU, fun times! That could slow 'er down too.
>>12011380It's a fun topic, you're just not smart enough to understand any of it. Obviously. Emulation will never be as good as FPGA could be, if done well. It's just easier to do up an emulator in so many ways and it's been done a lot so there are high quality emulators which exist and are quite fine.
>>12011389Yeah that’s right. I’m a big dumb guy. Come here I want to give you a wedgie.
>>12011391Got another thread you're bumping? /HHG/? I've seen this behavior before with people who want to curate the board.Shoo shoo. Anti-FPGA faggotry has no place on a board that allows threads about newly-made chink garbage handhelds, which despite how accurate any emulator claims to be, will never be as accurate as a well done FPGA.
>>12011357>can be much better I asked GPT to break down what an FPGA even is and how it compares to software emulation and from what I’ve gathered, that’s absolutely correct (at least given the current way OS’s work). I guess you’d need to have some actual knowledge of how digital circuits work in order for it to seem like common sense. Emulating hardware through software simply isn’t as reliable as emulating hardware through hardware.
>>12010992Pretty sure it could be mega illegal to even be suspected of using those, expired patents or not, since they were basically stolen confidential company property.IIRC the Dolphin devs have said that none of them ever even looked at the gigaleak for that reason (it would also give you less of a legal standing because it would no longer be possible to argue it was a clean room implementation, even if you don't use any of it, just simply looking at it or being told what it says would be enough).>>12011038>And all 3D games look better when running in higher resolutions, even on a CRT TV.Have fun with the train graveyard in FF7 with it's resolution bumped upAlso you don't need to "do some gay ass RGB mod", about the only console from gens 2 to 5 that could not natively output RGB was the NES and N64. I just have a cable for my PSX.>and if you don't know how the real hardware works internallyYou know the people who wrote these decapped the chips to figure it out right? Even some software emulator devs like the ones on MAME do that. Stop whining just because you're a poorfag.>>12011048>which is also why i don't even have oneKeep telling yourself that's the reason if it helps you sleep at night.>>12011103Like I said, comes off as misguided at best and intentionally misleading at worst considering the tests can fail on real hardware but not on his software emulator that is the only thing outside of some hardware revisions that can pass.>>12011136Who ever said this? People tended to complain that Nintendo's emulators are usually shit. Did you miss how horrible N64 games ran on the NSO emulators when they first came out? OOT in particular was the poster child of that argument.>>12011250Again, just because nothing (but conveniently his own emulator) passes the tests does not mean they are good, the GBA tests are similar.
>>12011403Nice glasses, fucking nerd.
>>12011136Except he kinda sucks NERD's cock during the end...
>>12011338Yeah, though seeing how much latency you are adding doesn't mean much when you can't do anything about it without hackish methods like run-ahead that drastically increase system resources and still have other side-effects.And this is assuming you have managed to keep all the components in sync since you have to run them sequentially, fun for systems like the Saturn or Jaguar that had more processors than Bill Gate's basement.
>>12011040You didn't present any argument and missed the point, pajeet.
>>12011425>>12011428Based, I hope you both enjoy your vacations.
>>12011404>I asked GPT to break down what an FPGA even is and how it compares to software emulation and from what I’ve gathered, that’s absolutely correct (at least given the current way OS’s work).Even with a real-time OS or running on bare metal there's still some overhead in the handling of hardware interrupts between the system which will pile up over time and show up as skew in the same way that my C64 emulator runs slow on my computer which is of course, running other programs too.>I guess you’d need to have some actual knowledge of how digital circuits work in order for it to seem like common sense. Emulating hardware through software simply isn’t as reliable as emulating hardware through hardware.I think even within the realm of FPGA solutions, there can be a difference between a simulator and an emulator. Perfect emulation will include all the flaws of the original design, and will need perfect circuit-level replication of the entire working circuit including all ICs. It's a big job. Most FPGA based solutions are clearly simulations at this point, where each device is replaced with some do-alike that's good enough for horseshoes and hand grenades, obviously these companies patch their designs over time and source code for free FPGA projects also matures over time. Any FPGA based project will be a simulator at least for some amount of time, until it becomes a true emulator. Or a project may aim to enhance and fix a design and wind up with a simulator that's not really a clone but an enhancement of the original, in this case obviously, perfectly cloning a thing isn't the goal so it will always be a simulator.>>12011421Well sure you might do something about all these problems. Write better code and / or get a faster computer! hah. With FPGA-based solutions to this problem you can at least aim to make the thing 100% accurate and given the time and attention and resources etc. achieve that goal.
>>12011406>the NESThat console is weird as fuck. Technically, the only NES revision that ever had RGB was the VS. Arcade system. (Which had a much more garish palette compared to a standard NES' composite.)
>>12011431So there is no feasible way for software emulation of a circuit to function as 1:1 copy of it in terms of performance? I took an Operating Systems class a few years ago and was pretty overwhelmed with the information we were being taught, so I don’t have a detailed understanding of why these limitations exist— only that I can accept that they do because I appreciate the importance of precision and its relationship with runtime.
>>12011453There's at least one software emu I know of that aims for circuit accuracy, but you need something like a threadripper in order to run it at full speed. (Which is why it's also an FPGA core.)https://github.com/nukeykt/Nuked-MD
>>12010231Jesus Christ the absolute state,Most of these are irrelevant edge cases.We’ve been 99.9999% nes accurate for like 25 years. I could emulate it pretty well on a 486.
>>12011453even half assed emulators like zsnes work for most of the games most of the time. 1:1 emulation is in the realms of emulating pure hardware quirks which requires investing a lot of time and coding into emulating them, which is quite pedantic if you ask me, unless, UNLESS, you want to emulate CAVE games. these are probably the only case where pursuing perfect emulation is worth giving a shit.
>>12011453>So there is no feasible way for software emulation of a circuit to function as 1:1 copy of it in terms of performance?Well the other anon mentioned run-ahead but that itself comes with problems including the emulator using more system resources and weird state issues. I'm sure other people know more about this than me. But no, it's not really possible on a normal machine without some trickery which may lead to inaccurate state or other issues. Computer-based emulators can certainly be good enough though, without being perfectly accurate. Look at the Apple 2 Pi for instance. It runs an Apple II emulator which interfaces with the Apple II itself and replaces the host CPU and runs all the real hardware instead, disk drives, keyboard handling and cards, you name it. It's an incredibly impressive project that is based on an emulator replacing a real CPU as the brain in a retro computer.OTOH there's the Vampire 68080, a drop-in 68K replacement for your Amiga where it's an FPGA solution, might be that a 68040+ type CPU is too demanding to run in emulation in real time replacing real hardware yet.>>12011474Even running the biggest threadripper in the world you'll never run it at full, full full speed. You will always have some skew.>>12011476This is it right here. FPGA solutions at this point, to me at least, seem like a bad value because emulators are well good enough and quite free. Even if there is some difference you won't notice it or you'll forgive it or you'll file a bug report.
>>12011476>CAVE games Pls no Seriously though, what the fuck are CAVE games??
>>12011489>what the fuck are CAVE games??If you have to ask that here, then you should lurk a bit moar on this board.
>>12011431>Perfect emulation will include all the flaws of the original designThat's generally what any accurate emulator software or hardware will strive for. If there was a flaw in the design of one of the components for the original console that would say, make a game crash in a specific edge case instead of work, the proper thing to do would be to replicate that in the emulator, not "fix" it so the game does not crash.Though many emulators generally add the option to choose to fix this if you would rather have that over accuracy. Even the MiSTer has many optional enhancement options after all, just that they are second place to making the core as accurate as possible.>>12011453It's not a performance issue, it's a latency issue, which yes software will never be able to 1:1 match. The processing adds latency, video output adds latency, the input adds latency...
>>12011494Don’t bully me, I’ve been lurking this board for years. Years and years and years. This is one of the few hardware/emulation threads I’ve made an effort to interact with because 90% of the time I have no goddamn idea what anyone is talking about.
>>12011503You know Snow Bros., right? Same people went on to found CAVE.
>>12011505W-what did they find in the cave?
>>12011503CAVE is a developer best know for their arcade shmups, including Dodonpachi, Espgaluda, and Deathsmiles
>>12011505You’re gonna have to forgive my tremendous ignorance here senpai, but I have never heard of Snow Bros. lmao. What makes these CAVE games special in terms of emulation?
>>12011503Loser.
>>12011513>What makes these CAVE games special in terms of emulation?The amount of sprites on screen causing the game to slow down, which can fuck up the flow if there isn't any.
>>12011513>I have never heard of Snow Bros. lmao.Play it if you have a friend, it's a good co-op platformer.
>>12011514
>>12011406What is misleading? The ROM conducts objectively true tests about the emulator running it. Its a diagnostic tool for emulator developers, not some sort of competitive game.
>>12011502>>12011485>>12011431>>12011406>>12011379>>12011356>>12011338>>12011282Stop fucking spamming you retard.
>>12011539No, you don’t understand. Everybody NEEDS to know what I, SPECIFICALLY I, think!
>>12011475>We’ve been 99.9999% nes accurate for like 25 years. I could emulate it pretty well on a 486.Nesticle was NOT "99.9999% accurate", I would argue it wasn't even 25% accurate. Scored less than 5% on those tests.>>12011476>work for most of the games most of the timeThat kind of reasoning is why we have so many older romhacks and even fan translations that don't work on real hardware now that flashcarts are more common. Also doesn't matter if "most" games work (poorly) if the game you want to play is among those that does not work.>>12011485>FPGA solutions at this point, to me at least, seem like a bad value because emulators are well good enough and quite free.The FPGA devices aren't really for those that feel "good enough" and just want the game to run, they are for those who want accuracy, to connect it to a CRT, to use original accessories without additional latency, etc. It's like how a sound bar is "good enough" for most people but that doesn't invalidate a surround sound setup. But you're a fucking idiot if you buy $3000 cables or other such audiophile nonsense for said setup.
>>12011539Take your meds you retard
>>12011550N-no! I only made HALF of those unbearable posts!
>>12010874He's probably using the last "base" versions of those emus. Here's NintendulatorNRS' results, which is the most mapper-complete emu. (It supports all kinds of exotic bootlegs.)
>>12011513On real hardware each bit of what's happening is happening in a whole chip set as well as a CPU which drives these chips to handle things like drawing sprites or 3D objects or whatever. It can draw things too but in general it's much slower than the dedicated hardware an arcade machine has. So the developer will try to use the graphics hardware chip set to the greatest extent possible because doing graphics in the CPU is slower and takes precious time away from other tasks.On an emulator all the various graphics elements have to be handled by software, there is no near-real-time sprite chip in your computer. So drawing two sprites is slower than drawing one. Several scrolling backgrounds, sprites and image transformations, these all take a particular amount of time but on a real machine (or FPGA done well enough) they will happen just precisely as fast as they are supposed to. You can use some tricks of course to speed up an emulator. Or you can offload some of the drawing to the GPU. Etc. But it will still all have to be provided for and that will lead to the thing running slower as the emulator itself gets more loaded. A perfectly accurate emulator of course will run as slow as necessary to make all the chips run even if they have nothing to do. A smarter one will turn off the emulation of certain subsystems if they're not in use automatically, the code paths will simply not be taken.>The FPGA devices aren't really for those that feel "good enough" and just want the game to run, they are for those who want accuracy, to connect it to a CRT, to use original accessories without additional latency, etc.I understand, sure. But it appears that the bar of ultra high accuracy is currently set by emulators, FPGA is just harder to do. It's sensible that it should be behind a bit, there is a higher barrier to entry because of the cost and it's fiddly and emulators have been in development for a long ass time by now, their development is more rapid, etc.
>>12011552>N-no! Don't talk about things that are relevant to this thread that I have no interest in!
BTW I do believe some people can tell that the emulator is slow even if it's accurate and they might like an FPGA-based solution. I'm an amateur musician and a long time ago when VSTs were new and I finally had a computer powerful enough to run one, I extensively measured latency and discovered that I could hear / feel any latency above 8ms between my finger hitting a key on the keyboard and the sound being emitted from the speaker, that's 1/125th of a second. If you were used to the real hardware and then played on an emulator, tiny slowdowns might affect you. That's down to skew, an issue which will take FPGAs to completely 100% solve.In the far future things will be fast enough that the skew might be so miniscule that there's no point in using FPGAs any longer, who knows. But for now it seems valid if you're into it and don't mind spending some cash, and the people who just out of nowhere complain about FPGAs, especially here, seem extremely poorly informed.
>>12011568You can talk about what ever you want. But I’m going to make fun of you for it.
>>12011575For what purpose, though?
>>12011579It doesn’t matter. Fuck you.
>>12010231FPGA is like the trannies of consoles hardware, they all tell you they're 100% just like the real console and some might even pass but really deep within it's nothing but a cheap and inaccurate imitation of the real thing
>>12011379>but that the ones it failed would have not had any effect or difference on the speedrunImpossible to say without recording the control inputs from the original speedrun and playing them back precisely into real hardware and comparing the results (ie the output, a recording of the entire game) to ensure they are identical. FPGA is cheating until proven otherwise. Just like emulators.
>>12011560>But it appears that the bar of ultra high accuracy is currently set by emulators, FPGA is just harder to do.Actually the opposite, it can be a lot easier to code an emulator in a FPGA than in software, especially if it had many components running in parallel. The whole having to time all that said hardware so it's all processed at the right timings even though you are forced to make it run sequentially? Not nearly as big of a thing in FPGA, since you can just simply run it in parallel.>there is a higher barrier to entry because of the costTo be fair you need a PC for emulation too, and I have seen many people these days without computers anymore thanks to smartphones and tablets. And you can get a MiSTer for $125-185 nowadays, there are many retro games that themselves cost more than that, as well as many flashcarts. It's not like these things are mad expensive anymore.>and it's fiddlyHow so? I find it easier to configure than software emulators, and I have been using software emulators since the days of Nesticle and Genecyst.>and emulators have been in development for a long ass time by nowI don't see how this makes a difference, many things have existed for a long time then get replaced by something newer. Also, practically no emulator that has existed for a long time is still really viable for use today, who would use Nesticle in 2025?>their development is more rapid, etc.FPGA development can be just as rapid, that's not something only software has, plenty of software emulators have not had an update in a long time as well
>>12011575Cool, some autist making "fun of me" on a slow niche part of a has-been basket weaving forum will have about as much impact to me as termite's fart would have in a hurricane.
>>12011583At least ask me out on a date first, you brute!
>>12011592Quite easy to say if the speedrun in no way used the parts of the system that failed the test, or it was something that would not effect the code in any way like a color being slightly off.
>>12011590FPGA technology has incredible potential, we're hitting walls with computer speed and resorted to throwing more cores at everything decades ago. It just might not be possible to emulate every arbitrary machine at acceptable speeds on a general-purpose computer and it's possible we might top out in computer performance before we CAN emulate everything. Video game or retro-computer or whatever. Especially if you're looking at both perfect accuracy and clock-accurate speed as a goal.>>12011595>it can be a lot easier to code an emulator in a FPGA than in softwareI doubt it, just flashing the new code alone is a time delay and the code is harder to understand and deal with and organize than whatever programming language you happen to like for your emulator. Marshalling all your various bits and bobs is a design problem for emulator makers. I think you're trading one problem for another really. I suppose for certain things like sprite or audio chips which handle multiple of the same things at once, you might have a short cut there. You still need to worry about timing though, you're just looking at the various clocks instead of worrying about if your loop is both quick and accurate enough.>To be fair you need a PC for emulation tooOr phone or whatever. Most people have this already. And for a lot of people dropping hundreds of bucks on some nerd gadget that is more of a pain to deal with than an emulator is a barrier to entry.>How so? I find it easier to configure than software emulatorsCan't account for that but software emulation can be drag and drop or a pain. Depends on the emulator and FPGA software both I guess, which you're comparing.>I don't see how this makes a differenceThey're mature and quite good enough, they're a solved problem. FPGA has fewer users and fewer devs and it's slower going and harder work.>FPGA development can be just as rapidSure, given equal resources, it CAN be. Catching up takes people and time.
>>12011605Wrong, in practical fact it's impossible to do what you say, not 'quite easy.' Impossible.Only way to verify is to record the input and play it back into real hardware and see if the output states match for the entire play session. That's the ONLY way. The real hardware and the FPGA and even an emulator are forensically all black boxes and need to be treated as such. What if some arcade machine has been tampered with? It's happened before. Or allegations of this have surfaced. Only way to tell would be to perform a forensic playback of the control input and monitor the video and even sound output for the entire play session to ensure that they wind up identically, to the last millisecond.>quite easylol you have no idea what you're talking about
>>12011626SO YOU’RE SAYING THIS IS ALL JUST A PLOY TO GET ME TO BUY EXPENSIVE HARDWARE????!!? GODDAMN YOU NINTENDOOOOOOOO
>>12010283I bought a mister pi becuase nothing else at $180 could emulate as well as it did with decent 240p RBG output for my PVM, and be small enough to not cause clutter. None of the mini windows PCs have the output I wanted, just HDMI. My arcade cabinet uses a PC with CRTemudrive and an old ATI VGA card for proper RGB. That surpasses the FPGA experience but there is no small PC equivalent for CRT use. I dont want an old laptop cluttering my gaming area.
>>12011634You already own a PC you fucking idiot. You didn’t need to buy shit.
>>12011428Whatever helps make your life any less miserable.>>12011553He should have tested both just to see how it has progressed. One thing I'm curious though is how the libretro version of QuickNES handles these days compared to something like FCEUX/FCEUmm. Last time I saw I was surprised to see that QuickNES managed to be more accurate somehow.
>>12011631The amount of people who give a shit about all this is vanishingly small. They happen to mainly be big Internet nerds though so it's easy to find kvetches and special needs posting.Most people run emulators so they can simply play some video game. They're not speedrunning or trying to set records and dealing with the kinds of autistic crybaby weirdos who seem to form the core of that particular community. And I don't know if the emulator and FPGA makers themselves really are, or give a shit about speedrunners either. But spreedrunners sure have a lot to say about emulators and FPGAs and etc.
>>12010231Hi Kids!
>>12011640>Last time I saw I was surprised to see that QuickNES managed to be more accurate somehow.Sadly it only got an 84, tested it on the latest version via RA. (Maybe it's a lot better on Bizhawk, since I believe that's where QuickNES originated.)
>>12011615>just flashing the new code alone is a time delayYou do that once when setting up the core, arguing that it sets up a time delay to flash the code is like arguing about the boot time for the PC before you launch the emulator.We aren't comparing time from power on to get in-game, and even then it would be from when the emulator starts running, not the setup beforehand.Also none of this is latency, you realize that's different from speed right? You can have good speed but poor latency or vice-versa.>and the code is harder to understand and deal with and organize than whatever programming language you happen to likeDepends on the dev, I know devs that PREFER to code for the Saturn in assembly... keeping all seven of it's processors in sync with that code, than using high-level languages like C. You can always also just visually configure the FPGA instead of in code if you are that bad at it, don't be efficient though. But my point is that coding anything to run in parallel is much easier on a FPGA because you don't have to run it sequentially then time it like in software, you can just simply run it in parallel.>Or phone or whatever. Most people have this already.Emulation of anything that requires reflexes would be useless on a phone without a controller and possibly even another screen to connect it to.>dropping hundreds of bucks on some nerd gadget that is more of a pain to deal withWhere are you getting this? One of the main advantages that made many like the MiSTer is how simple and easy it is to use. It's far more work to setup software emulators. Have you ever even used a MiSTer? You sound like you are just making stuff up.>they're a solved problemThey're most definitely not, especially for complex systems like the N64, Saturn, or Jaguar.>FPGA has fewer users and fewer devsYou realize that with all these more users and devs... a lot of software emulation problems recently have been solved because it was discovered in FPGA right?
>>12011636The CRTs are in a different room from my PC. The PC is on a LG C4 OLED for modern gaming. The other pc is just a stripped down motherboard inside an arcade cabinet as mentioned above.
>>12011649>>12011605>and it's slower going and harder work.Again, no idea where you are getting this "it's harder on FPGA" argument other than you appear to have never used one and are just assuming.>Sure, given equal resources, it CAN be. Catching up takes people and time.In a fraction of the time the MAME devs had worked on CPS1 Jotego found out that their implementation was running it at significantly wrong speeds and fixed it.
>>12011626>Wrong, in practical fact it's impossible to do what you say, not 'quite easy.' Impossible.It's literally what the guy in the video does you moron.>you have no idea what you're talking aboutMost ironic statement ever>>12011636>I'm a poorfag and need everyone to know it
>>12011660The great thing about calling me poor is that even if I’m absolutely loaded there’s no way for me to prove it. It’s the ultimate non-argument. You’re such a good poster, wow you must have 50 FPGAs.
>>12011662Nah, your posting makes it obvious with how much you autistically screech that people would just emulate over using FPGAs. You didn't even try to deny it like most poorfags do insisting they own every console and hundreds of games for each.
>>12011649>You do that once when setting up the core, arguing that it sets up a time delay to flash the code is like arguing about the boot time for the PC before you launch the emulator.No I mean in the development cycle for the FPGA. Not any time delay while loading some core. You have to do this for every change, you can't change 'em in real time. It's one of the barriers to the more rapid development a computer-based emulator developer will have. I guess you could liken it to compiling though.The other issue is emulators are programmed in a language of the dev's choosing, while FPGA development takes place in VHDL, correct me if I'm wrong. That's a lot harder to use. You can use HLS tools to develop your FPGA in C so it's not a huge pain in the ass but you might not guarantee an identical circuit layout, and thus, the system will not be accurate.>Depends on the dev, I know devs that PREFER to code for the Saturn in assembly... keeping all seven of it's processors in sync with that code, than using high-level languages like C.OK? It's up to each person to decide how they want to do this.>coding anything to run in parallel is much easier on a FPGA because you don't have to run it sequentially then time it like in software, you can just simply run it in parallelIt's not necessarily harder in software it has to do with the desired performance and if your hardware is up to running the code fast enough that you don't have skew. You may have to trade accuracy or complexity to make it acceptably fast. And in an FPGA you still have to worry about a bunch of clocks and bus lines... doesn't automatically seem easier at all.>It's far more work to setup software emulators.No it's not. Don't be silly. A person can install any emulator in moments. Most proper OS's have a package manager, you can go from wanting to having an emulator in seconds.Some emulators will need a ROM file or two, maybe. But FPGAs will too to be legal and accurate.
>>12011712You’re right. You win.
>>12011679Literally making up insults just because I read you like a book isn't helping your case.
>>12011647Interesting result, thanks for testing Anon. I like using it at times because of how fast it is and with all the latency reducing options set up. This old post on Reddit showed some surprising results. https://www.reddit.com/r/emulation/comments/2o4177/android_nes_emulator_accuracy_testing_results/
>>12011660>It's literally what the guy in the video does you moron.Turns out that he's wrong. There is no way at all to verify that ANY FPGA or emulator is accurate unless you can feed it some input and get the same output that you'd get on real hardware. And in practice there is no way to truly verify that there isn't some bug or trick other than by verifying the output per a given input, frame by frame, and comparing every detail, and ensuring the visible state is the same the whole time.>>12011719If the goal of FPGA is accuracy you can't use high level synthesis. You have to clone the thing with VHDL but that's very much more difficult and involved. With HLS you can simply clone the behavior of each part of the system, sure. And get it to run as good as real hardware, maybe better! But it's not a signal and clock level accurate copy any more either, so you'd have to run extensive tests to verify that it's accurate at least in terms of observed behavior.Turns out FPGAs do worse at this at this point, compared to emulators. Probably because they are made with HLS and aren't circuit-accurate clones despite being hardware.
>>12011712>No I mean in the development cycle for the FPGA. Not any time delay while loading some core.I didn't even have to argue this, you said it yourself, it's like compiling code for software, this is no different.>FPGA development takes place in VHDL, correct me if I'm wrong. That's a lot harder to use.Again, depends on the dev, some find it much easier to code in that than higher level languages. Also, VDHL is not the only language you can code a FPGA in.>OK? It's up to each person to decide how they want to do this.So then why are you trying to argue like it's a fact that it will be harder if it can differ from person to person?>and if your hardware is up to running the code fast enough that you don't have skewAgain, this is not latency. It's inherent in software emulation and it doesn't matter if you have a 2048 core threadassripper9000, you will always have additional latency.>And in an FPGA you still have to worry about a bunch of clocks and bus linesYou realize you still have to simulate those in software emulation too right?>No it's not. Don't be silly.As someone who has used just about every emulator since the Nesticle days, even quirky experimental ones such as NES emulators that rendered purely in ASCII, yes they are. The MiSTer is by far the easiest one I have ever had to set up.>A person can install any emulator in moments.I literally have to just select update_all to install EVERY core on my MiSTer, or just drag and drop ONE file to install a core manually.>Most proper OS's have a package managerYou can try to boast about Linux all you want, but the vast majority of PCs use Windows and an even bigger majority of people who game on computers use Windows.>Some emulators will need a ROM file or two, maybe. But FPGAs will too to be legal and accurate.update_all takes care of that for me, it's no less legal than getting a BIOS file for your software emulator
>>12010317do you think there's a meaningful (game accuracy-wise) difference between my customized opentendo and my NES I took the chips off? what about my opentendo and a fpga NES?
>Too poorI dont undertsand all this talk of people being poor. The entry level for mister FPGA is very low. If you dont want a case and wifi adapter and just want a stan alone gaming board it costs fuck all. People seem to think its still 2021 with scalpers and price fixing.The average modern collector edition PS5 game costs this much now.
>>12011550>"I'm not spamming!">Continues to spam the thread with dozens of low-IQ posts.
>>12011749>Turns out that he's wrong because I said so.Sorry, but you're wrong, because I said so. No, you don't need to use a TASBOT to make sure something won't effect a speedrun, that's something someone who has no idea how anything works from a technical side will say because the only thing they understand is the simple concept of pre-recorded inputs.Plenty of ways to check by seeing what is happening in the code, do you even know what a breakpoint is? Simple to set one up if a recorded speedrun hits one of those sections of code.Go autistically screech somewhere else.
>>12011784This is somewhat related, but I find it incredible how Bri'ish kids had to do exactly that on their computers in order to enable cheats LMAO.
>>12011784>Sorry, but you're wrong, because I said so.Yeah you're wrong though. I said so because it's true. In order to verify that any emulator, or FPGA is accurate you have to test its output against the output of a known good real example of the actual hardware. Which can even differ between the actual "real" machine's revision etc. For something as monumentally important as a speedrun record of the UNIVERSE you'd have to take a recording of the input and run it on a verified known good inspected and sealed real hardware example too.>Plenty of ways to check by seeing what is happening in the code, do you even know what a breakpoint is? Simple to set one up if a recorded speedrun hits one of those sections of code.There are none. Zero. There is no actual way other than by feeding the emulator or FPGA the same input you feed into some real original hardware, and comparing the output. No other way at all. If you can think of one, go ahead and adequately describe it here. I have time. You do too. You can't do it, you won't because you can't.
Anyway you're missing the point. I'm not merely considering the accuracy test in OP, but the entire emulator or FPGA, from the instant it starts to run the game to the end of the speedrun. The only way to verify that the speedrun is good, is to plug the same exact input at the same time into real hardware and observe that all state is identical on the screen and with the audio etc. Forget about any specific accuracy test, I'm talking about an overall accuracy test. Treating the FPGA and / or emulator and / or different hardware revisions of the REAL console or say NTSC vs. PAL as black boxes. Then you'd have to use machine vision and an accurate clock to ensure that each frame is perfectly identical and displays for the same amount of time.It's the only way to truly be sure. It may sound autistic but isn't this the point? I feel like FPGA and emulator users are cheating if they try to do a speedrun or high score effort.
>>12011771That's the sad part, he can't even afford that. All he has is his father's hand-me-down laptop so he insists that anyone who has more than him is clearly dumb with their money so make himself feel better about having jack shit.
>>12011782>Low IQ poster complains that others of much higher IQ are low-IQ posters>>12011790Or in some cases, to even get the games to function like intended, such as Jet Set Willy which released a set of pokes to fix some bugs in the game that made it crash.>>12011807It's been explained multiple times to you at this point, just by tracing the code you can see what is and isn't being used, you don't need to replay the inputs. That's a very low level of understanding way to do it, as well as not even an accurate way to do it. Frames showing game same visual image don't even mean anything, values can be different in the background.
>>12010231Real gamers test using TextNES.
>>12011909Some people don't want another tech-gadget, emulation is good enough, etc. There's plenty of stuff anybody can buy, but do they want it?
>>12011720You’re still being racist. You’re implying that being ESL is a bad thing.
>>12011767You’re right. You win.
>>12011356>The difference is that it can allow for inherently perfect emulation while adding no additional latency and being electronically compatible with original accessories, or even cartridgesTrue.>while in most cases being easier to code a more accurate core than in software.This depends heavily on how smart people are and what the particular system is doing. Doubt.
>>12010268agg23 hits 110
>>12010231>brosGet back into the closet!
love me misterlove me analogue pocketlove me some /vr/ seething
>>12011987At this point, I'm here for curiosity sake. It's fun to see how well certain emus do in tests. (Even if it's for something extremely niche, like cartridge swapping.)
>>12011771>I am NOT putting the actual source or link because SHUT UP!Nice try jackass, also most people who don't live in the US are screwed because shipping it's a pain, by the way, how much did they rob you on shipping? Hmmmmm?
>>12011987Based FPGA chad. I use my MiSTer practically every day.
>>12011987I love my MiSTer. I don't see the reason for a Anallog pocket, but I don't really do mobile gaming.
>>12011994>(Even if it's for something extremely niche, like cartridge swapping.)
>>12011995NTA but that's clearly Taki's board, anyone who has even a passing interest in this thing would be able to tellhttps://retroremake.co/products/misterpi?variant=43061708685335
>>12010992It totally did, but nobody involved in N64 emudeving will ever admit it. The leak even included an OpenGL implementation of the N64 RDP, for fuck's sake. There's no way in hell something as complicated as the N64 was able to be emulated to the extent it was just mere years after its release otherwise.
Interesting how this test suite doesn't pass on real hardware. Tested on one of my real NES and only 114 out of 125 tests passed.It says quite a lot on how much these tests are actually relevant.
>>12012041Angrylion RDP Plus has existed since 2017 and Nintendo's own emulators are lacking in that aspect despite having access to said doc, Anon.>It totally did, but nobody involved in N64 emudeving will ever admit it. You can check the source code for any of them on github to corroborate your claim, but that'd require not talking out of your ass.
>>12011406>being told what it says would be enoughActually if somebody went over it and wrote down a step-by-step description of what it does, they could legally look at that and use it. That's basically one end of standard reverse engineering practices. You just have to maintain a clear separation between the people who study the original and the people who make the workalike, to be in the clear.
>>12012078No no no, you see, you need to burn it on a nrom chip and make a proper cartridge, and you can't have the wrong revision of the NES, THEN you will see that these tests are totally valid and will pass!Or you can use his emulator which 100% passes the tests! An emulator which... actually runs too fast, does not support PAL games, has no sound support, and HAS NO INPUT SUPPORT! It's an emulator designed purely to load pre-recorded Tool-Assisted Speedrun files and play them.Oh, but his emulator (which has not been updated since April) totally passes all of these tests... that he made. Clearly this means his emulator which has no PAL, sound, or input support and runs in fast forward is more accurate than your real hardware that fails his tests.
>>12012126Yeah, I remember that being how the first IBM Clones were made when I was watching an old documentary on the start of personal computers. The dude running the company (I think it was Compaq?) hired lawyers that told him exactly how to skirt around the laws without it technically being illegal.
>>12010231I almost fell for the FPGA meme, thank God I saw this
>>12012078You should email him the debug memory screen (hit select) after running each test that does not pass. He's trying to make every test pass on as many consoles as possible (if you notice, many tests have multiple passing states) but he needs volunteer data since he doesn't own 50 NES/Famicoms.
>>12012159This test itself is kind of a meme, especially with the whole "i'm gonna make the game crash if it's on an accurate emu lol" point at the end.
>>12012162No. I don’t do bitch work for free.
Every retro enthusiast knows it's real hardware on a CRT screen or bust.
>>12012167>I am going to complain on 4chan that an open source hardware research project doesn't encompass my hardware and not contribute to getting it workingDude, you cannot be more of a bitch than you already are.
>uggggh, this glitched pixel is supposed to flicker but it's not>my immersion is ruined
>>12012186that's what mistercucks say to people who use software emulators thoughever
>>12012186If the pixel in the overscan wasn't flickering, you didn't beat the game.
>>12012180Still not doing it.
>>12012209you don't even own the console
>>12012186>uggggh, this glitched pixel is supposed to flicker but it's not>my immersion is ruined
>>12012186Some people want to use original peripherals, that's legit.
>>12012078Works on my machine. Dendy doesn't count as a "real NES."
>>12012419I knew that guy didn't have a NES lol
>check out these TRUE BLACKS on crt displays!
>>12011592seems easy enough https://youtu.be/NisWoTNyfeU
>>12012445The background of the program is grey, not black...
>>12012448Neat program, glad it's available.https://github.com/RetroTournaments/Graphite
>>12010231If you really want your mind blown, check out this mapperless demo. It pulls off some neat tricks!https://www.pouet.net/prod.php?which=103928
>>12012582So I was about to test this demo on TricNES... until I realized that the latest version available on Github is 1.0.1, when the video used 1.2. Oops!
>>12012186Isn't that precisely the point? If now it doesn't matter that the FPGA isn't perfect, then why not use the more accurate, free, traditional software emulation? I think someone here is moving the goalposts, and it's not exactly the advocates of software emulation. Afraid that your million-dollar scam will collapse?
>>12012659Software emulation has like decades behind it. Eventually we may see FPGA catch up, for a new tech it sounds pretty promising.Plus it's increasingly infeasible and expensive to use original hardware largely due to LED and LCD screens, an FPGA system you can purchase brand new that has modern outs and hasn;t seen 500 14 year olds wipe their poop and cheeto encrusted hands on that also got chewed on by the family dog is a net positive. Rememebr how well the classic mini consoles sold? Because people want new retro consoles and DESU I am one of them.
>>12011651>The PC is on a LG C4 OLED for modern gaming.Stupid. An LG C4 is perfect for emulation. Enable its Black Frame Insertion and use the most accurate software emulation for each console. We've reached the point where I don't want CRTs in my house anymore with this tech.
>>12012668>Eventually we may see FPGA catch upWhy wait and pay to this scummy companies when software emulation is already almost perfect?>use original hardwareThe only real hardware I use today are handhelds (AGS-101, DSi XL, new 3DS XL, Switch and Switch 2).NES? It's Mesen, almost 100% accurate as proven by this thread.SNES? bSNES is perfect.N64? Honestly, I don't care about this console, so I don't know.NGC and Wii? Dolphin is the best experience. I also have a Wii U, so I have native HDMI output for NGC, Wii and Wii U with it.SEGA things? M2 works wonders. ¯\_(ツ)_/¯
>>12012673Its passable on an oled these days, but I have several very nice CRT and a PVM. Why give that all up to play with fake scanlines on an LED tv, even if its a nice screen?
>>12012691Because I want actual new consoles, of course.Something is lost when you emulate via PC. Something is lost when you don;t play on the original hardware too, but speaking in reasonable terms.
>>12012582Whoo ho man... Guess how many colors! HOooo hoo! Man, It's FOUR Twenty! Ho hoooo! It's FOUR TWENTY MAN! Hhaaaaaaaaaaaaaaaaaaaaaaaaaaaaa!
>>12012712>Something is lost>Somethingsee? you are just>>12010298>>12010994
>>12012691>Why wait and pay to this scummy companiesAnallog isn't the only option. If you want you can just buy the chips directly from the manufacturer and build your own board. And then you're not paying for anything related to any fpga gaming project.
>>12012721Do you think all audiophiles are the people who pay thousands of dollars just for cables?
>>12012721A fitting observation. I'm all for emulation and use it all the time, I actually literally just want a new console that can play old games, flashcarts or CD's or read directly from an SD card that I can plug into a TV or monitor using HDMI or display port whatever. That you can buy new. It can be an emulationbox or FPGA as long as it's reasonably priced and werks. Fortunately for me these things exist, but I want MOAR.
>>12012725ok, then>Why wait and pay --------------------- when software emulation is already almost perfect?>>12012728No, just the kind of audiophiles who would also buy FPGAs even knowing that software emulation is better today.
>>12012742I like software emulation but I enjoy tinkering around with hardware more. You apparently don't so FPGA just isn't for people like you.
>faggots forget about the nes-in-a-chip manufactured by bootlegging asians using a tiny ASIC back in the 1990s>wasn't 100% accurate but ran most slop>could fit hundreds in a modern fpga chipfpga is gigabloat.
>>12010262What does zombies ate my neighbors have to do with this?
>>12012781The FPGAs being talked about aren't just NES emulators, they emulate stuff up to the Saturn.
>>12011431Simulator and emulator are 2 different things, friend. Even a crappy emulator is still an emulator. FPGA emulates the behavior of chips and buses using synthesized logic gates. Software emulates with instructions running on a computer architecture.A simulator would be like reprogramming Super Mario Bros in Flash and making it look and feel like you're playing on an NESAlso, note that FPGA is in fact hardware emulation even though you're emulating a system that runs software.
>>12012781>caring about bloat on equipment intended for one purposeyou faggots don't even know what you want
>>12012691>Why wait and pay to this scummy companies when software emulation is already almost perfectDon't buy it if you're not a hobbyist, then
>>12012880Simulators are programs that mimic analog devices with digital hardware. Emulation mimics digital hardware with other, generally more powerful digital hardware. Wine is not an emulator, btw
>>12012901A hobbyist of what? Scams? I'm a hobbyist of retro video games and software emulation is more accurate for that matter. Get back to /g/.
>>12012909As in someone who is interested in the implementation of cores on an FPGA specifically. If you just want to play games then stop calling it a scam.
>>12012909>hurr durr if people buy something they want it's a scam
>>12012919All scams involve someone buying something because they believe they want it due to the buyer's misinformation.>>12012916>stop calling it a scamno
KekIt went from>FPGA hardware emulation is inherently more accurate than software emulationto>uhhhh... I bought an FPGA device because I think the implementation is interesting...
>>12012903The overall scheme of implementing a console by mimicking the behavior and relationships of its components makes it an emulator. It depends on the scope.
>>12012880>Simulator and emulator are 2 different things, friend. Even a crappy emulator is still an emulator.I disagree, you have to start with simulation and then move to emulation or you will never run the whole program. You will always discover bugs which need to be fixed for accurate emulation. Etc.> FPGA emulates the behavior of chips and buses using synthesized logic gates. Software emulates with instructions running on a computer architecture.I already know all this. But FPGA doesn't necessarily simulate the chip properly, it CAN if you have the gates to spare and a complete understanding of the CPU. But if you're using HLS tools it's almost certainly guaranteed that the FPGA will have a different circuit-level plan than the real deal. Only with an HDL can you guarantee the plans will look similar. Even then they might end up differently.Currently it looks like most FPGA cores are made using HLS so there are bound to be bizarre bugs which are impossible to fix with code, it will come down to the tools and bugs or features in them as much as the HLS language code itself.
>>12012940Shut up you actual retard. That was someone else. I've never wanted to convince people MiSTER is the best choice because it's obvious that a lot of it is still in development. All I'm saying is you should stop being a little bitch.
>>12012946Yes it's an emulator, I was clarifying the terms, that simulator is specifically about mimicking analog devices. Because >>12012947 is using some weird personal definition of "simulator"
>>12011660>It's literally what the guy in the video does you moron.If you watch to the end he said it can be non-deterministic and there is no way to predict if some bug somewhere will affect something else. Also I think it's heavily implied that there are other bugs as yet untested for.
>>12012725You're arguing with a troll, do you really think there is any meaningful discussion to be had with someone that makes such absurd statements as "Afraid that your million-dollar scam will collapse?" He's just here to be butthurt about FPGAs that he can't afford, not to make any logical arguments.
>>12012691>don't pay scummy companies>just buy a switch 2 brolol
>>12012728they're retards who can only think in memes, they've seen people make fun of audiophiles for "le expensive cable" and just assume they completely understand a topic now.
So... how did this thread go from the validity of that specific test ROM to yet another debate between software vs. hardware emus? Isn't it heavily implied that only a specific emu (the author's very own) can pass all the tests?
>>12012965>just buy a switch 2 broWhere did I say you should buy a Switch 2 to play retro games? If you can't buy Nintendo products because it's an evil company, which it is, then I understand that we can shut down /vr/ already.
>>12012971Even Sour himself doesn't want to emulate one of the tests yet until it's 100% understood.
>>12012947>I disagree, you have to start with simulation and then move to emulation or you will never run the whole program.You start with emulation in the design. For example, you could implement the NES PPU by loading and shifting data the same way the hardware does, only they're variables in your program. You'd have to design a different way to do it to simulate the PPU, but it'd still be part of your emulator. The CPU is almost always a simulation. It's actually a benefit of FPGA that cycle accurate implementations are modular. My point is that your design paradigm would have to change dramatically to call the whole thing a "simulator"
I got my Mister years ago because it was, and still is, the absolute best emulation box for the living room experience. Zero bullshit, good controller interface, good shaders, lower lag than any PC option. Would I use it on my desktop PC? Only for cores that are ahead of PC software emulation, like the Saturn.
>>12012983GroovyMiSTer had some potential, since you could use your PC as a framebuffer and use your MiSTer as a lag-free display for your CRT. Sadly the guy who created GroovyMiSTer passed away, so it's not getting active updates anymore.
>>12012951Accurate emulation doesn't need circuit level simulation and the terms are different. Older emulators were often called simulators because they performed a lot of hacks. I believe the terms were defined by the MAME people decades ago in this realm of things.This is not the same as circuit simulators such as SPICE, it's a different realm with a different lingo. I do not believe there is a single Mister FPGA core which aims for a gate-accurate and circuit-layout reproduction of any entire system. Some of it is "simulated." Non-EE lingo, just based on black box behavior simulation.>>12012960FPGAs have potential benefits, as of yet they are quite hard to come by if you just want to play games and don't care about things like CRTs and original accessories etc.>>12012980I'm just going to have to agree with your points, despite my difference in lingo. Sure. Actually you will implement skeleton functions or whatever and then fill in the gaps gradually, until something works if barely. Then iterate on that until it performs well enough and stop and do that to the next part of your project. Call whatever resulting stage what you will. Even with an FPGA. But I think with an FPGA there's a critical difference - unless you have a netlist of hardware description list of each circuit that is accurate to the original hardware, you're also just kinda running an emulator black box. HLS tools may produce an acceptable result but you're getting close to doing 'an emulator but in hardware' with 'em and newly discovered bugs in the original hardware will never show up in your FPGA core. Just for instance.
>>12012960Have a C4 OLED with a beefy PC connected to it to play retro vidya but somehow I *can't* affort a $125 FPGA. I made decisions based on facts, not feelings. Cope and seethe and enjoy your inferior devices.
I guess there are two real issues which will prevent perfect emulation or FPGA circuit-level identical clones. With emulators there is the issue of skew and latency between servicing interrupts for I/O, no matter how fast the computer gets they will be there and can be observed over time. Even with runahead, which grows more burdensome with the number of possible inputs and other state. Emulators will ALWAYS run a bit slow even if it's otherwise observably perfect because of skew alone. With FPGA the issue is, perfectly cloning a chip, while possible, is really hard because you have to somehow build a netlist or hardware description of it in a language the FPGA synthesis tools can understand. It involves describing the entire device. Which means you have to photograph the chip and trace all the connections and figure out what each part does. And even if you do this there can be bugs because the device will be laid out on the FPGA by the compiler / synthesis tools and is almost certainly guaranteed to be a different physical circuit in all its relations than the original, even if all the ports hook up properly and it mostly does the same thing. There will be gremlins.For most people they have no reason to care or invest in FPGA, it's just extra shit to buy and any old junk computer can emulate at least a SNES or PSX these days. That's fine.
Think about this - there are FPGA cores which do a whole game system. But there is no FPGA core where the entire thing is done as netlists describing the entire circuit even the CPU. Correct me if I'm wrong but parts seem to always be done as HLS or the entire project is.These are emulators being compiled and ran on an FPGA. They are not, and I do not believe any purport to be, circuit-accurate exact clones of the whole game system but just running on a chip. It's good enough for certain uses where an emulator might not be, perhaps.
>FPGA *AND* emulator fans both BTFO>hardware chads win the day as usual
>>12012998So what's with your obsession with FPGA? You destroy every thread that discusses it.
>>12012971The thread started with someone shitting on FPGAs over these tests, as well as the first few posts. The threat title is "Oh no no no FPGA bros!". It was always just another troll topic disguised as discussion from people who have been whining about FPGAs here since 2022.
>>12013021>It was always just another troll topic disguised as discussion from people who have been whining about FPGAs here since 2022.That's a genuine bummer, there's some nuanced takes ITT. Guess there's no point in salvaging a thread that was intended to be a flame war to begin with.
>>12012991>FPGAs have potential benefits, as of yet they are quite hard to come by if you just want to play games and don't care about things like CRTs and original accessories etc.I agree that they are not for everyone, if all someone cares about is a game running at all in a playable state then an ARM box is likely enough, I compared it to how for many people a soundbar is "good enough" but plenty of others want a better experience with a surround sound setup. FPGAs are like that, they are for people who want more than just simply good enough and/or want to use original accessories or CRTs. I have even seen some say they use it because it's a lot simpler and easier to setup and use than software emulators, as well as pretty much instant boot time.You also seem to be under the misconception that the FPGA is basically just acting like a CPU to emulate the system on no different than software emulation, just because you can configure it in code. That's a very incorrect view of how it works.
>>12011474It's so difficult, in chips electricity isn't flowing like through pipes but propagating like a wave. This can actually affect your device. SID chip fans will attest to this.
I've built my MiSTer over five years ago. I installed it into a repurposed anodized aluminum case that's the size of a small desktop PC. I made my own led pushbuttons and a power button with a momentary switch. Five years later there are still retards that scream like retarded brats about it because they think it's exclusively for their children's hobby. It's not. For over five years these same trolls destroy every thread that dares mention the project because the mods simply won't do their fucking job.
>>12011546>scored less than 5% on those tests.These tests are edge cases that are irrelevant for the games. NES emulation is a solved problem in practical cases and has been for decades.Decades.Twenty five years at least.If you were to go back to 1999 and make a good enough fake NES case for a PC, pump it out to CRT and hide the GUI no one would know it wasn’t a real NES.These tests are pointless.
>>12013002Software emulation can never have the same latency as the original hardware just because of how computers and running the emulation in software would inherently work, you can get it 100% accurate but having emulation that accurate also requires enormous resources. The person who wrote Nesticle was writing a GATE-LEVEL accurate NES emulator... it ran at like 0.1FPS. Then there is the NukedMD emulation of the Genesis... which runs at a similar speed. Don't even ask about the N64 attempt someone made.Problem is emulation of such old systems generally requires very very high clock speeds... something that CPUs haven't been doing for about 15 years now. We have been focusing on packing more cores and cache into the CPUs now, not just increasing the clock speed, in fact, some the higher end CPUs tend to have lower clock speeds to make it feasible to have more cores on the die.This is great for modern software... not so much for accurately emulating decades old systems.Thing is, you CAN pull that off at realtime speeds on a FPGA with basically the same design, parallel execution is a massive benefit that software emulation just does not have.https://www.youtube.com/watch?v=sMMiBEhnizEFPGAs have potential to be a 1:1 recreation, but there can be other limiting factors, both around the FPGA and the board it's on. For the Cyclone V FPGA chip itself it's really running out of space on the Saturn core (though that's also practically done), as for the DE-10 Nano board it comes on that the MiSTer project uses memory latency is a limiting factor that keeps the N64 core at more like 98% accurate than 100% (Namely, that the CPU and FPGA share the same DDR3 RAM and can have one have to wait while the other is accessing it). This can potentially cause an issue in a very small number of PSX games like Chrono Cross... but enabling options that make the GPU do more work and thus slowing it's access down fixes it.
>>12012998>Shits in every thread that ever talks about FPGA>Posts copy-pastas for years in every FPGA thread until mods started handing out vacations for it>STARTS threads shitting on FPGAs>Insists FPGAs are a "scam" and "million dollar industry" and cost $600 and all sorts of other bullshit that are beginner level trollingNot arguing on emotion, yeah, sure, and I'm the king of England.
>>12011474>but you need something like a threadripper in order to run it at full speed.Granted I've only run it on my 13700k, but I don't believe you need hardware that powerful to run it. It does sound really good. You only miss the punch that a real SC55MKII adds. However for all the other sound canvases nuked is just as good for sc55.
>>12013037>NES emulation is a solved problem in practical cases and has been for decades.>Decades.>Twenty five years at least.Wrong
>>12013040Anon, I think you're confusing what you're replying to with Nuked SC-55The thing Anon was talking about is a Sega Genesis emulator
>>12013042Ohh yea...You are correct. I forgot those are two separate things.
>>12013030>I agree that they are not for everyone, if all someone cares about is a game running at all in a playable state then an ARM box is likely enoughThey do a fine job.>I compared it to how for many people a soundbar is "good enough"No, I would say emulation winds up being better and sometimes more accurate going by the various tests. It will depend entirely on the target and etc. You can still output emulators on CRTs you know. Old Pis are great for this with the composite output. I don't see how it can be simpler than launching a file, anyway.>You also seem to be under the misconception that the FPGA is basically just acting like a CPU to emulate the system on no different than software emulation, just because you can configure it in code. That's a very incorrect view of how it works.Wrong. There are three kinds of FPGA cores, for the purposes of running vidya or retro computers. Pure VHDL, mixed HLS and VHDL, and pure HLS. None can guarantee full clock, port, and circuit accuracy.
>>12013043nukeykt has worked on tons of different stuff
>>12013048Nice. How many of those have you played around with?
>>12013046>No, I would say emulation winds up being better and sometimes more accurate going by the various tests. It will depend entirely on the target and etc.That's where I would heavily disagree. Also, these tests themselves are questionable at best, both FPGA and software emulation authors are saying so. This guy's own emulator that passes these tests is literally unplayable.>You can still output emulators on CRTs you know.With additional hardware that adds even MORE latency, doesn't do it right (It's just downscaling your PC's output and converting it to analog instead of natively outputting it like a FPGA does), and won't work with accessories that need a CRT to function such as light guns.>Old Pis are great for this with the composite output.No, they are not "great" for it, other than needing additional hardware they still have all the other issues I mentioned, on TOP of all the issues with software emulation. Plus the additional issue of not being able to run the more accurate emulators because their CPUs are too weak.>I don't see how it can be simpler than launching a file, anyway.AFTER you have set it up, sure. But even before setup, you still have to boot into an OS then boot the emulator THEN "launch a file". I don't need to start my MiSTer, wait for a desktop environment to boot, then click on "Run Emulators".>Wrong.Big claim from someone who has been wrong about everything relating to FPGAs so far and has clearly not even used these systems while arguing as if they are an authority.>There are three kinds of FPGA cores, for the purposes of running vidya or retro computers. Pure VHDL, mixed HLS and VHDL, and pure HLS. None can guarantee full clock, port, and circuit accuracy.I don't even know where to BEGIN with what's wrong with this one, but holy FUCK are you misinformed about FPGAs.
>>12013038>Software emulation can never have the same latency as the original hardware just because of how computers and running the emulation in software would inherently work, you can get it 100% accurate but having emulation that accurate also requires enormous resources.Even with infinity bajillion resources, it will never be fast enough. Clock skew from when the OS needs to service an interrupt at the same time the emulator's clock needs attention, etc. among a million other things. Yes. Pre-emptive multi-tasking. And I'm not just talking about latency, but clock skew from clock collisions where the OS is doing something while the emulator "should" be for full speed full accuracy.The gate level approach works these days though. With the magic of FPGA.The problem is NOBODY IS MAKING GATE LEVEL FPGA CORES! NOBODY! CHECK IT OUT YOURSELF! Sorry to shout and I agree FPGA has enormous potential but it's squandered with basically, HLS-coded emulators running on the FPGA. The speeds are better sure. I mean theoretically? I haven't checked. But the accuracy is worse at this point in time.>FPGAs have potential to be a 1:1 recreationOooooh that's a tough order. I will say, 0.9999999:0.9999999 or so. Even given a perfect gate-level circuit description the FPGA synthesis program will lay out the circuit differently, and smaller, than it was in real life! haha
>>12013058You don't need to replicate the actual layout of the gates to have a 1:1 accurate clone of a chip, even actual consoles had revisions of chips that changed their gates. The SNES had later revisions that combined multiple chips into one, they sure as hell didn't just make a single chip twice as big and have the exact same gate layouts. You can have a 100% copy of every single nuance of a chip with multiple different designs to it's gates.
>>12013041It was good back then.I could play NES on an emulator back then.It was fine.In fact it was so good .
>>12010231How does Neon64 score on thisI noticed he did the Excitebike 64 emulator and the Animal Crossing emulator but not Neon64
>>12013002>>12013038>>12013058This thread reads like a villain origin story of someone who went mad over emulation accuracy.
>>12013054>That's where I would heavily disagree.OK, go ahead.>With additional hardware that adds even MORE latencyMany computers have the ability to output composite or component or whatever. After all that's what the real hardware does.>No, they are not "great" for itAre.>other than needing additional hardwareDon't. Old Pis have composite NTSC video output.>they still have all the other issues I mentionedYou didn't.>on TOP of all the issues with software emulationFPGA and emulation both have issues compared to real hardware.>AFTER you have set it up, sure. But even before setup, you still have to boot into an OS then boot the emulator THEN "launch a file". I don't need to start my MiSTer, wait for a desktop environment to boot, then click on "Run Emulators".Irrelevant detail, your thing is always running. Just click the file.>Big claim from someone who has been wrong about everything relating to FPGAs so farI haven't. The cores are simply not mature or circuit-level accurate enough to bother upgrading from emulation unless you hit some wall. Which most people will never even care exists.>I don't even know where to BEGIN with what's wrong with this one, but holy FUCK are you misinformed about FPGAs.There are only two paradigms, writing or generating a netlist or hardware description list in an HDL, or using a higher level synthesis language to describe the thing. HLS is the dominant paradigm right now. All the mister cores use it. It's not circuit-level accurate.>>12013067>You don't need to replicate the actual layout of the gates to have a 1:1 accurate clone of a chipDo.>even actual consoles had revisions of chips that changed their gates.True and bugs emerged or were fixed or weren't discovered or relevant.>You can have a 100% copy of every single nuance of a chip with multiple different designs to it's gates.Can but DON'T. Name one.
Practical usage>all of the emulators are fine expect maybe the excite bike oneAutism score>if you use something not in any game then I deduct pointsOkay.
>>12013048Mostly SC55 and OPN2
>>12012991>Older emulators were often called simulators because they performed a lot of hacks. I believe the terms were defined by the MAME people decades ago in this realm of things.Doubt.jpgThe definition of emulation and simulation I'm using came from programming books in the 80s.>>12013026>Guess there's no point in salvaging a thread that was intended to be a flame war to begin with.Nah, you can have a good discussion even when OP's a faggot. Hell OP's normally a faggot, you just ignore his bullshit and carry on
>>12013072>Many computers have the ability to output composite or component or whatever. After all that's what the real hardware does.There's far more to it than just a modern system's implementation of "outputting composite">Are.Are not>You didn'tLiterally the paragraph above above>FPGA and emulation both have issues compared to real hardwareWhataboutism>Irrelevant detail, your thing is always running. Just click the file.Bad faith argument, and no, most people don't have every system running all the time>I haven't. The cores are simply not mature or circuit-level accurate enough to bother upgrading from emulation unless you hit some wall.>Admits has not used them>Still tries to argue about their faults as if he is an authority>Gets just about everything wrong because of this>Do.Don't.>True and bugs emerged or were fixed or weren't discovered or relevant.Majority of revisions saw no differences>Can but DON'T. Name one.Literally the revisions I named previously
>>12012078Hardware revisions are a thing, as well as hardware component decay. SNES systems are known to develop timing issues for example. >>12012419Yeah, Dendy and many other Famiclones don't initialise memory in the same way, and memory in the power off state may decay to a different pattern. So cold boot and warm boot behaviour can differ and affect software that checks specific addresses.
>>12013072>>12013093>>12013058>>12013054>spends the entire thread using chatgpt to create arguments with himself just to flood the board with canceryou are the dumbest and lowest iq baboon i have ever fucking seen and you need to die.
>>12013371You’re encouraging him. You’re giving him something to reply to. Stop it. The only thing that can stop him is everybody stops replying to this thread.
Here's my two cents:FPGA cores aren't inherently more accurate than software emulators. However, accurate FPGA cores can run faster than accurate software emulators.https://github.com/nukeykt/Nuked-MDhttps://github.com/nukeykt/Nuked-MD-FPGASo, while the marketing tactics of companies like Analogue are quite retarded, it's not a reason to write off the technology as a whole.As for 100th Coin, while it is alarming at first that his emulator seems to be the only one to pass his tests with flying colors, it makes more sense when you realize he designed the emulator and the test suite based on his specific revision of the NES, and his emulator matches his NES on a shit-ton of other tests too. He's just an incredibly devoted autist.https://www.youtube.com/watch?v=pK7hU-ovUso&t=800s
>>12013550People in this thread seem to think he is trying to make his emulator seem better than everyone else's by specifically tailoring the tests to be biased toward it, but the reality is the emulator and test ROM were developed in tandem as he made more discoveries about the hardware. Its only natural his emulator passes all the tests. He's not going to write a test for the ROM and then not go back and edit the emulator to pass it.
I ordered a MisterPi for its 15khz TV functionality (ever fuck with crtemudrivers or groovyarcade? They're terrible) and latency paired with a CRT. I am under no delusions of it being a magically better emulator. At the least it's a decent LAN video card via GroovyMister.
>>12010283i might in the future because all i want is a box that plays my games with analog output and no added input/output lag.i will not stoop to using coping mechanisms like runahead.
>>12013740All those words sound really gay.
>>12011634There are definitely AMD APUs that'll work for this in business minis you can buy for cheap on eBay.>>12013756>/vr/ still blows and is getting worse. Don't worry, I'm leaving soon.
>>12013761>AMD mini with 15khz 240P output???
>>12013771Yes, I can confirm the 705 AMD Pro series works for sure so long as you get one with a VGA or DVI-I port. On closer examination the ThinkCentre Tiny is not a good solution as many of them don't have the analog port. An EliteDesk 705 G3 like this would be best:https://is.gd/7Di5az
>>12013781How many women have touched your 705 AMD Pro?
>>12013781Of note, GroovyArcade actually works fairly well at outputting 15khz with most cards (AMD, Intel, and Nvidia) with a DP port through a passive DP->VGA converter. This can expand your options quite a bit if you don't mind fucking around with Linux.https://gitlab.com/groovyarcade/support/-/wikis/2-Pre-Requisites-and-Installation/2.1-Hardware-Suggestions-General
>>12013783I've honestly, unironically, lost count.
>>12013761At least on the lenovo the vga option is just a built in display port adapter.
>>12013781And its chipset is supported by crtemudriver? What gpu does it have?
>>12013789Yeah, rolling back on that rec. 705's the way to go for sure.
>>12013740>ever fuck with crtemudrivers or groovyarcade? They're terribleYeah, and it worked out of the box and never had a problem with it. As a matter of fact, it worked perfectly for two different GPU models I tried it on.
>>12012991>I'm just going to have to agree with your points, despite my difference in lingoThat's fine bro. You sound pretty smart so maybe I'm wrong anyway.
>>12013801I can tell you that the latency won't get as low as a Mister unless you use RA, run ahead, and tighten up frame delay significantly. Also boot up Chrono Cross. Tell me how great those 240p/480i transitions are.
don't care, still using zsnes
>>12013818No. Those things all blow.
>>12013837Okay. I'm glad what you have works for you, numbskull.
>>12013856Don’t call me that. I have a genetic condition that makes it so I can’t feel my head.
>>12013818I mean, that's a given. Your OS will +/- give you 2 frames of lag, but that can be mitigated as you said with runahead. And call me a retard, but I went back to back between my actual Genesis and emu and I could not tell a difference in latency (with a real Genny controller hooked to deamon bite on PC with a CRT).And switching on Chrono Cross between the 240p gameplay and 480i is just like on the console, it's just by default Retroarch is a piece of shit so you have to configure it properly, doesn't happen on other emus or when I tested arcade OSD.
>>12013948Of course your ass is sitting in a dark room.
>>12013756You're just as gay as this whole thread and every thread like it ruined by the same faggot trolls.
>>12013952>Of course your ass is sitting in a dark room.I mean he's playing on a CRTYou have to be in a completely dark room or else your screen will be covered with reflections and you'll have worse blacks than an IPS panel
>>12013948why does it fade to black? it doesn't do that when i open the menu on a ps1 or normal emulator
>>12014750Because retroarch is shit and it needs to reopen the window to switch the res. So it fades out because the 240p source is closed, then changes it to 480i with a new window that is being put on fullscreen. It should just blink in a split second like MAME or arcade OSD does, but as I said, retroarch is a piece of shit full of bloat unfinished features, with CRT switching being one of them (they still have 3 nonfunctional settings related to it, sitting for over 6 years with zero time put into them).
>>12013092That's a relief, I've been meaning to bring up more test ROMs anyway. There's this one test ROM in particular that even BlastEm gets a few wrong, and I've always wondered what Nuked-MD's results were. (VDP SPR CHANGE MIDSCREEN is one of the trickiest tests, since both sprites being displayed should have a square with an invisible line going through it.)https://gendev.spritesmind.net/forum/viewtopic.php?t=3329
>>12014895Huh, interesting. I just tried the latest 64bit Windows nightly build of BlastEm and it did even worse in that test than yours.
>>12015008That's incredibly odd, but certainly explains it. The version I used was blastem-win64-0.6.3-pre-7b3196be35f1.
>>12015015My bad, I was using an older version of the tests, got the same results as you with the newer version that was posted further down in that forum thread.
>>12015019I see the confusion, glad there wasn't a regression or anything. (Although, BlastEm has its own fair share of unrelated ones that are slowly being worked on.)
>>12015025Tried it on my MiSTer, passed every test across both set of tests except for one.
>>12015043Also trying it on the software version of Nuked-MD, which appears to have not been updated since 2023.It failed more tests already, but I have run into a problem... I have no idea how to press start to advance to the next set of tests. Doesn't help that it's also running at roughly 1 frame per 6 seconds on my PC, so whatever button it is, if there is even a button, I likely have to keep it held down for several seconds.
>>12015049Found it, just have to hold it down for a while. Saw some oddity as it's testing however, I wonder if the test actually does that but we just can't see it since it's just for like 3-4 frames at best.This could take a while...
>>12014772>Because retroarch is shitYou're telling me. I tried to make an issue on one of their emu cores because I was trying to play a lightgun game which worked fine in the original standalone emulator, but the lightgun code was broken in their version of it. So I figured, I should report it, right? That's what you're supposed to do, you know? Yeah. So I do, and the jackass dev in charge's response was like "What do you mean you want to play those games? Nobody plays those games! [WONTFIX] [CLOSED]"Retroarch sucks, and I hate that it's becoming the Master Control Program of emulators
>>12015043That's incredibly fascinating! I've only ever come across one emu that ever got the vcounter corruption test right, and it's ClownMDEmu of all things.
>>12015080It's STILL running the VDP corruption test on Nuked.So while waiting I was wondering if these tests would fail depending on your hardware revision like those NES tests, or if having a 32X and/or SegaCD connected would also change the results. Or if even a flashcart would cause some tests to fail like again on those NES tests.While I do have a Model 1 and 2 Genesis, as well as a 32X and Model 2 SegaCD, I don't have a flashcart to try it on real hardware.Closest thing I would have is the 32X core on my MiSTer. However, it has not been updated since April of 2024, and it was never meant to run Genesis games. In fact, it normally won't even load them unless you trick it into thinking they are 32X games.So I tried and the result was... bad, very very very bad.But again, the core was never meant to run Genesis games. I tried a few retail and demanding homebrew Genesis games in it anyway, and all of them except Pier Solar worked. A few had an odd issue where a specific blue color in the background would display as black (You can actually see that here) and the Titan Overdrive 2 tech demo had some graphical corruption here and there, but mostly, they all played fine.Now I am curious about real hardware.
>>12015074>"What do you mean you want to play those games? Nobody plays those games! [WONTFIX] [CLOSED]"KekWhat game was it?
>>12015195NES core, I forget which game I was interested in at the time, one of the weirder lightgun games. But I also wanted to play some Duck Hunt because nostalgia.
Is just the SBC craze all over again.The well of RPi and its derivatives is now dryed up after every single retro celeb shilled them to death and everyone that wanted one bought one, the "It'S nOt EmUlAtIoOoOn!!!" pitch is perfect for both reselling the same thing to people that already own an SBC while also giving sellers a free pass to tout the thing as sort of a "Pro" product and thus justify the higher price.Don't fall for it, don't be retarded.
>>12015074>I hate that it's becoming the Master Control Program of emulatorsthat's the entire point of all the shilling. they want to control the whole ecosystem and making the emudevs work for it by using their own standard.
>>12011771>preorderCall me when this is a cheap affordable product I can buy at any point from various sellers, FOMO tactics can eat my ass
>>12015138By the way, the old Genesis core is still available as an option on the MiSTer from before it was replaced by the much more accurate Nuked core. Mostly because it offered enhancement options that you can't have on the accurate core.It did incredibly poorly, in fact, it could not even finish the DMA tests. The VRAM to VRAM +99 test kept hitting 65535 and looping. Also I am not certain but I think the 32X core is based on this older Genesis core.
>my dogshit emulator passed my dogshit test perfectly!!!Lmao. This is like saying you have developed AGI who can pass a logic test with 100% accuracy when in reality you made the test yourself and just taught it which answer to circle for each question. MiSTer is for all intents and purposes, and to the tiniest minutia, absolutely identical to original hardware. The hardware emulation aspect of it means your CRT literally thinks and behaves exactly 100% identical as it would if it was a real NES and real OEM controller hooked up. There are ZERO documented accuracy issues, whether it be rendering, sound, timing, performance, or latency with any of the main cores.With MiSTer, you can rest assured you will be having the original hardware experience to the tiniest degree possible (as well as optional features and enhancements if you want), and every single person will have an identical experience with the hardware, no need to worry about your hardware configuration or OS or drivers muddling with things.
>>12015274Its a preorder for the 5th production batch, I got mine a year ago. They have been on the market for quite a while now, they just keep selling out because scalpers try to hoard them and put them on ebay for double the price. other vendors (QMTECH) have them in stock still. No single vendor controls the mister FPGA. You can buy it from whoever. Its jus tthe Mister PI has some of the best quality control, while QMTECH is ali express and is currently fucked by Trump tarrifs and double the price. If you are outside the US, the QMTech version is much cheaper.
>>12015302>MiSTer is for all intents and purposes, and to the tiniest minutia, absolutely identical to original hardware.R E T A R DETARDYou are all like cultist. I get it, you've been scammed as you'd get better results with the hardware you already had at home, you don't want to admit it, and you're justifying it however you can. Let it go.
>>12015302holy fucking cope lol
fPIGas btfo
>>12010231fpga is good over the long term, but tbqh emulation mogs it, and you can prove this statement by comparing the quality of cores available to emulators. emulators are still king for now.
>>12015467based anon moment.
oh yeah where's my sega model 1, 2, and 3 core? Oh I don't think THEY FUCKING EXIST!
>>12010283I just got mine and it's really great>small and quiet>svideo/composite output with an adapter with no dotcrawl >drag and drop games on it over the network>just works>can plug real controllers into it>just run update_all and it all just worksIs it more accurate than software emulation? I don't know and don't really care, but I spent too long struggling with the CRTemudriver nonsense and having something that can just switch resolutions on the fly without me having to fuck around with getting it started (CRTemudriver still flashes the desktop every time the game changes resolution which is fucking absurd) I don't care about owning original hardware, I don't want to collect games, I just wanted a little box that played more or less everything I was interested in and well enough that I never need to fuck around with anything other than running an update script once in a blue moon. The only original hardware I have any interest in owning are the controllers. Sure, I could have bought a Retro Pi 5 and gotten access to 6th console gen emulation, but PS2s, Wiis and Xboxs are so ubiquitous, cheap and easy to modify it isn't a problem for me to keep those consoles on hand and just have the mister for everything else.
I love how my FPGA is so based that it has been making poorfags here seethe and samefag since at least 2022 over something they can never have.
>>12015628fpga isn't that expensive anon.
Plus MAME is finally supporting a bunch of 3d arcade games that it couldn't, gonna be a looooong time until your fpgas catch up to that.
>>12015634I know, that's the sad part.
>He doesn't know that the MAME devs needed MiSTer developers to point out that their emulation has been shit for years and to fix it
>>12015646This. Trying to treat it as emus vs fpga "teams" is stupid. Everybody's working in the same direction.
>>12015649>Everybody's working in the same direction.Except for FPGAchuds
>>12015627pi5 barely runs a few 6th gen titels, not even worth mentioning. Dreamcast is its upper limit, and there are so many better options for that.
problem is a lot of retards say FPGA is utterly superior to emulation accross the board, at this very moment, and that's very very untrue.
>>12015302peak retard alert
>When you see poorfags desperately posting sour grapes about FPGAs because they can't afford them
>>12015661>FPGA is utterly superior to emulationFPGA IS emulation though, just a different kind
>>12015650No it's just you, troll.
>>12015690It actually sits in between emulation and hardware cloning.
Where can I have a discussion about MiSTer? /v/ and /vr/ are worthless for this topic.
>>12015786Right here, right now.
>>12015721No, it's pure emulation, just not on PC.FPGA is programmed hardware like any computer is.So programming an emulator is pure emulation regardless of the platform.Hardware is mad, not programmed, the same or alternative chips, schematics etc using same firmware.
>>12015293Goddamn, that took HOURS.Anyway, here are the results of the software version of the Nuked-MD core, three tests (including the one the MiSTer core failed) but otherwise is almost 100% accurate. Course, this literally took about 6 or so hours for my modern gaming PC to run when these tests took about 10 seconds for the same NukedMD core to run on a FPGA and get even better results.Every other supposedly accurate emulator I tried (Ares, Bizhawk, Mednafen, MAME) failed around 50% to 75% of these tests.
>>12015838Was this comment necessary?
>>12015838>programmed hardwareExactly, it's hardware that can be reconfigured. Not software.
>>12015891ExactlyInstead of software emulation, it's hardware emulationIt's still emulation
>>12015467Good speech. Promos well.
>>12015897>hardware emulation
>>12015962yes
>>12015302Nice wallpaper, did your grandma pick it out?
>>12015467>>12015469>>12015663Poorfags seething like usual
>>12015853I use the Genesis core because it allows you to overclock games. All thes inaccuracies are basically on-paper inaccuracies, they don't actually affect anything in-gameIt's frustrating that the NES core doesn't even allow overclock for slowdown removal, yet it's not even as accurate as MESEN
>>12012078Sounds like you have a worse revision.
>>12015302>CRT literally thinksCRTs dont think>>12015302>There are ZERO documented accuracy issuesyou know... except for the accuracy issues from these hardware tests which is literally the subject of this post
>>12015897Unnecessary comment.
FPGA MY ASS!!!>skew latency?!>i/o servicing interrupts?!>verilog development toolkits?!NOT MY PROBLEM (-_-)my home theater pc play the games at solid 60fps on the living room 4ktv and the 1millisecond barely noticeable input lag from the wireless xbox controller and the emulator software isnt affecting my enjoyment of playing old and dusty unc games
>>12017117Obsessed
>>12015302You're literally retarded
>>12016916>accuracy issuesWhich even the guy who wrote the tests admits are not a big deal for standard emulation, and are only accurate for his specific model hardware, other revisions of real hardware fail many of the tests
>>12011282well, i hope my analog pocket is still in working condition by the time we get there
>>12017298Which I find more interesting that MiSTer passes all those Genesis tests except for one because the new core is of a model that does not exist, so if the tests were based on a specific revision it would have failed miserably.The Nuked core is basically a frankensteined Genesis from all the best parts of all the different models and revisions of the Genesis combined into one. A revision with all those specific versions parts never existed, but it could have.Meanwhile these NES tests will fail if you have the wrong flashcart or version of the NES itself.
>>12010231>speedtranny edging cases in a TASfag emulator's cherry picked tests that he states himself mostly don't matter unless you're also trying to verify speedrunning shit>zero testing of anything of value like graphics, audio, gameplay (besides the fairly rare quirk of a single frame of run ahead), and compatibilityWhen is someone going to make a real test cart for NES emulators, and when will I ever care?
>>12017117Minisforum cost hundreds more than mister fpga does these days. Bottom end beelink mini pc is closer to what a mister costs, especially now since half the mini PCs on amazon ship from china and are fucked on tariffs.
>>12017117>1millisecond barely noticeable input lag from the wireless xbox controllerLmao sure thing bud, keepthinking that
>>12017492If he barely notices it then it sounds like it’s good enough for him.
>>12017513input lag affects you whether you notice it or not.
>>12017547If you don’t notice its effect then it doesn’t matter.
>>12017553You don't notice because you quite literally haven't experienced any better latencyReminds me of when I went to my normalfag friend's place and tried playing his PS4 and the latency was so bad cars started turning half a second late in GTA5He didn't notice anything because he was used to the ridiculous delay and just turned half a second in advance, thinking that's how video games just work
>>12017553No, it still does matter. It affects your reaction time and timing.
i hate fpga
>>12017569>>12017574No, if they don’t notice it then it doesn’t matter.
>>12013038>Namely, that the CPU and FPGA share the same DDR3 RAM and can have one have to wait while the other is accessing itso, just like a real n64 with its shitty rambus memory.
>>12017553You might not "notice" it specifically but games will become a bit harder.
>>12011356>Nobody thinks that FPGAs are some magic chip that by it's mere existence will create perfect emulatorsYou must be new here. There are many people in this hobby who think exactly this, even when the reality is broken down for them.
>>12017314can't you set those specific revisions for things like the soundchip in the mister menu under hardware?
>>12017829Show me 5 posts. Sounds like they just live in your head.
>>12015853Seeing as how the AccuracyCoin video was comparing official emulators, I'm curious to see how the Genesis Mini handles those tests. I'd love for some way to try out m2engage without needing a mini itself. I'd say its one redeeming factor is that it's the only emulator that supports M2's own Space Harrier ports.
>>12010231FPGA is more real-time accurate. Software emulators are held back by shitty pc hardware. So whilst they might be more accurate internally, they cannot bring that accuracy into the real world.
>>12017965audiophile with cable lifters thinking
>>12017858I'm not digging for 5 year old posts because you're too fucking lazy to do your homework. We've got retards in the world listening to ChatGPT to replace the salt in their diet and you think it's so farfetched that morons think the MisterFPGA is some magic black box that plays games perfectly? Go look at discussion on Analogue products, check the archives, I'm not doing your homework, faggot.
>>12017965>shitty pc hardwareI don't have that problem
>>12017969kekIt's only a matter of time before they start buying special rocks to put on their FPGA boxes to "reduce electromagnetic interference and make the timings more accurate"
>>12017586queer
>>12015891So FPGA can be better only on performance, but not for quality, not for accuracy, since configuration is programming - pure emulation, since no hardware is involved in compatibility, only software programming makes it all. So emulation is emulation, nothing more. FPGA is pure emulation lust like any emulator app for windows or whatever. If hardware is powerful enough to run it (and PC is), performance is not an issue, so FPGA is ultimately useless trinket only fragmenting emulation effort harming all. As with any scam schemes FPGA is just a crime, crime is always harmful.
>>12018171Cool schizo ramble, bro. Suggestion for future improvement: a couple of those sentences sounded like you vaguely had a notion of what you were talking about, remove that and it'll be perfect
>>12018171The funny part is there's nothing preventing you from emulating an FPGA emulating another console on a PC. It's just hardware logic gates, nothing prevents you from simulating logic gates. You could emulate FPGA entirely in Java. All the zealots who evangelize this stuff have no idea how computer hardware, compilers, or programming works, and think original hardware is Merlin's Magic Box and that FPGA is the only Philosopher's Stone that can reproduce the magic. The gaming equivalent of people who call their car old betsy and beg her to not break down like she's an animate being.
>>12017829No they don't, butthurt poorfags came up with that strawman so they can lie and go "Look how annoying FPGA people are". I have never seen a level of autism on /vr/ before that compares to those who have had nothing better to do than to shit on FPGAs for literal years here.>>12017972>There are other people in this world who are as retarded as me therefore everyone must be that retarded>>12018171>>12018213Holy schizo samefagging Batman
>>12018278CRT threads are almost as bad.I don’t care about either FPGAs or CRTs.I will continue to emulate old systems using my PC on a modern display and the games are indistinguishable from the hardware I once had.People can get super angry about all these things but it really does not matter.
>>12017801If you don’t notice that they’re harder then it doesn’t matter.
>>12018436How would you notice that if you had never played it without the lag before that?
wtf misterbros told me it was 100% accurate??? shouldn't have bought mine, i feel scammed
>>12018534That’s the thing. The guy with the Xbox controller hasn’t played it without the “lag” and is perfectly happy not going out of his way to try it. So guess what? It doesn’t matter.
>It doesn't matter because he is being completely ignorant and has no idea he's giving himself a crappy experience and doesn't know any better!Yeah, that's about the level of cope I expect out of emufags
>>12010231So FPGA is unironically worse than PC emulators? That's fucked.
the input lag is so miniscule and insignificant, that only actual autistic spergs who don't even play games care about
>>12018735It does matter because the game will be slightly harder for him. He may end up hating games that he would otherwise be able to beat if he had less latency.
>>12018775>actual autistic spergs who don't even play gamesaka MiSTerfags and CRTcucks
>>12018171gpu and sound hardware imposes hard limits on accuracy. It does not matter how accurate the software emulation is, that accuracy is impossible to bring into the real world. I consider this to be the main benefit of FPGA, perfectly synchronized low latency audio and video output.
>It does not matter how accurate the software emulation isomega-level cope
the very reason I bought and built a PC is to have all my favorite games/multimedia into one machine with modern hardware, imagine going through all those hoops just to play a fucking 40 year old NES game with "accuracy" instead or firing it up on your modern PC
>>12018787You just wrapped back around to old argument points that I already refuted. None of that matters because they are happy with how they are playing now. It doesn’t matter.
>>12018820>imagine going through all those hoops just to play a fucking 40 year old NES gameIt's literally plugging 4-5 pieces into each other like Lego (Technically, for NES you only need TWO pieces) then plugging in a SD card. It's a million times more effort to build a PC than to build a MiSTer, and if you are too retarded to handle even that you can get them pre-assembled.You have to go through dozens of more hoops to play NES on a PC.
>>12018824He could be happier.
>>12018824>It doesn’t matter.>It doesn’t matter.>It doesn’t matter.>It doesn’t matter.>It doesn’t matter.All I'm hearing is:>I'm Ignorant.>I'm Ignorant.>I'm Ignorant.>I'm Ignorant.>I'm Ignorant.
>>12018836>>12018837The man playing with his Xbox controller is happy and ignorant. He’s happily playing his video games in ignorance of some sperg’s latency numbers. He doesn’t know and doesn’t care, because he’s happy and ignorant. It doesn’t matter. Enjoy your $200 autism box that just plays Super Mario, Xbox controller guy is having fun too. It doesn’t matter.
>>12019046
>>12019052That’s the part you aren’t understanding. There’s nothing to “cope” over. The guy playing with his Xbox controller on his PC doesn’t know or care about the low latency Mario box. He doesn’t know that he is playing with latency. He is happy as he is and doesn’t look for more. His existence upsets you. His ignorant bliss infuriates you.
>>12018832>You have to go through dozens of more hoops to play NES on a PCyou're insane>download emulator>boot up the rom????
>>12019065That's some tangy sour grapes>>12019073And did that PC magically appear out of nowhere? You mentioned BUILDING a PC, so you are going to bring up building a MiSTer yet act like the PC is just there and already setup? Even if you want to argue a pre-built PC, I could build and setup a MiSTer in less time than it takes Windows to boot for the first time.Also if you get a prebuilt mister you don't even need to do the first step you mentioned, just "boot the rom"... the rom which you also skipped the step of downloading
>>12019073Don’t talk to him. He thinks buying a circuit board from a Discord user is easier than downloading mario.nes
>>12019083>buying a circuit board from a Discord userOh I see, you're retarded. That explains a lot.
>>12019081It’s sad that you don’t realize how incredibly stupid you sound when you accuse somebody having fun playing video games of being jealous of you and your loser hobby of buying YouTube influencer approved emulation boxes.
>>12019086Do you really think the people making FPGAs aren’t sitting at home grooming Discord users? They make emulators in their free time.
>>12019081I didn't build the PC solely for running 40 year old games you autist>implying downloading roms is a complicated or difficult process holy fuck you can't be serious
>>12019097Think about the type of person you are arguing with. You’re wasting your time, he’s an idiot.
>>12019091It’s sad that you don’t realize how incredibly butthurt you sound trying to toss all these heard it before troll accusations that people have been doing for years now over anything FPGA related>>12019092Do you really think you're not a fag?>>12019097>It's another "B-b-but PCs can do other things" episodeAnd? You brought up NES emulation, we were talking about NES emulation here, not "other things", if you want to bring up "other things" there there are shitloads of things you can use the DE-10 Nano for that a PC can't do beyond playing "40 year old" games as well.This is called moving the goalposts
Light guns work on PS1 mister. As far as I know no traditional emulator can run light guns
>>12018213>You could emulate FPGA entirely in Java.Not really, Nuked tried to emulate on an extremely low level in C, but it's too slow for any consumer CPU to run.
>>12017972>THEY'RE EVERYWHERE>Show me five>REEEEEEEEEEEESeek help retard.
I got one of them Lava FC consoles a few months back, not sure if anyone is interested in how it stacks up to other options on the market. I've compared it to a FC AV with NES RGB and a bone stock front loader that has fresh caps. The Lave FC has a cleaner output, esp with audio then any other option I've tired it against.
>>12019046He could be happier.
>>12019285>no traditional emulator can run light gunsDo Sindens count?
>>12019632Not on any level and I say that as a dude who got SCAMMED by buying 2 of em with recoil
>>12019648That sucks, would've had potential for certain light gun games on TeknoParrot.
>>12019650You can use Sinden with me if you wanted too, I just think they feel about as good as a wiimote light gun setup My point about mister and light guns was the timing is so close to real hardware with mister that REAL light guns work.
>buying dedicated hardware to play retro games>but worsefpgalets btfo
>>12019652In that case, it'd be worth testing out one of Namco's light guns on a MiSTer. I hear they made the best light gun the PSX had to offer.
>>12010283i have been shilled freaking exa arcadia in here lolthe state of this board.
>>12019657On the plus side, exapigs are paying extra for Earthion so I don't have to feel bad using it on a superior emulator.
>>12019659Actually, how DO you inject roms into the emu that Earthion PC uses? I want to try out some of those Genesis test ROMs on it for shits and giggles LMAO.
>>12019656Yeah, guncon work on mister no problem. That's my whole point. No traditional emulators work with real light gun
>>12019684I'm sure I mentioned this before, but there's a 50% chance GroovyMiSTer could work with a software emu as well. It's a good option if you simply want a PC-to-CRT framebuffer.
>>12019285Yup, not just PSX but any system that had a light gun. Though you will need to use an analog connection to a CRT, won't work over digital or LCD/LED.>>12019632Aren't those more like a Wiimote that controls your computer mouse? I wouldn't say so, they aren't really light guns, and require some annoying setup, while a light gun you just plug it in, point, and shoot.>>12019664I don't think anyone's bothered to try to find a way to inject roms back into Earthion. Not much reason so. The description script is simple though so it would not be hard to figure it out if you really wanted to.Careful though, a lot of the updates have been about guaranteeing the high scores are legit and avoiding cheating, it might flag you as cheating if you fuck with the stock roms.
>>12010231am i missing something here or is this guy an idiot savant who wrote tests for a bunch of edge cases that don't fucking matter?
>>12019804In terms of running SMB correctly, that's correct. In terms of running the author's own TAS-tailored emu, not exactly. However, the author also states>In my opinion, the most important tests in this ROM for determining if an emulator is good enough for speedrunning are the "Frame Timing" tests.
>>12019754>Aren't those more like a Wiimote that controls your computer mouse?Eugh, you mean to tell me modern arcade light gun games are glorified wagglers? That's disgusting on so many levels!
>>12010231HARDWARE.EMULATION.Whoa man that's really something. You mean this little box is emulating the hardware and not just hardware with software that emulates the hardware??
>>12019979It's more like cloning hardware. If you know how a chip works, you could print a new one, or you can reprogram the gates in an FPGA to work the same way. An emulator is software that can mimic a piece of hardware, but requires much more processing power than a hardware solution would. FPGAs are not quite as fast as ASICs, but are much faster than software, and they have the advantage that you can reconfigure them on the fly. Historically, they were primarily used to prototype chips that you could then print, but they've become cheaper recently. Cloning (or attempting to clone) hardware with an FPGA has some advantages over software emulation, the biggest one being parallel execution. If you set up some of the FPGA's footprint to be chip A and some of it to be chip B, you can run both of the chips at the same time, instead of having to switch between them and try to synchronize the results afterward, which is error-prone and requires far more processing power. An FPGA can make multi-chip monstrosities like the Jaguar and Saturn much easier to recreate accurately, but how accurate an FPGA version of something will be depends on the people who coded what you're running on it and the quality of the information they had access to about the original hardware.
>>12020038FPGA cores are not gate level accurate.
>>12020059The Mega Drive and Neo Geo cores for MiSTer are.
>>12020083Cool
>>12020083The MD core is 120% more accurate than gate level
>>12010231now run the same tests on some clone hardware, so we can all watch hyperkin get btfo
>>12021113I would be surprised if Hyperkin's crap does better than NesticleI'm more curious what Polymega's overpriced desktop in a console case would score in this
>>12010948don't forget to bully him for using the same audio masker as that blizzard beta tester
>>12021113>being this mad you can't even afford a famiclone
>>12020569>discordSad!
>>12013067Depends on the target, the MOS6581 sounds better than the 8580 to most people because there are traces which are too close which allows signals to leak and persist. This leads to the 'alive' sound. 8580 fixed that but ruined the liveliness of the SID, nobody wants an 8580.They do the same things on paper. But the 6581 was laid out on a whiteboard and photographed (yes really) while the 8580 was synthesized using automated chip design layout tools. Same theoretical circuit, different results.
>>12021135Audio? What audio? TricNES has none.
