[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/g/ - Technology

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.
  • You may highlight syntax and preserve whitespace by using [code] tags.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


Janitor applications are now closed. Thanks to all who applied!


[Advertise on 4chan]


File: Screenshot (116).png (1.66 MB, 1855x1080)
1.66 MB PNG
>Webp is only useful for low quality slop, it blows at high qua-

https://umigalaxy.com/explore/general/715
>>
Auto encoder script for the few of you who actually use their computers for more than browsing twitter.

@echo off & setlocal enabledelayedexpansion & set "psnr=30" & set "maxpsnr=60"

echo Webp vs JPG SSIM auto-encoder & echo Only input.png accepted, choose desired quality:
echo 1= 90%% & echo 2= 80%% & echo 3= 70%%
choice /c 123 /m "Enter number option on keyboard: "
if errorlevel 3 set jpg_quality=70
if errorlevel 2 set jpg_quality=80
if errorlevel 1 set jpg_quality=90

echo: & ffmpeg -hide_banner -loglevel error -i input.png -f image2pipe -vcodec ppm - | ^
cjpeg -quality !jpg_quality! > Q-!jpg_quality!.jpg 2>nul
for /f "delims=" %%A in ('ffmpeg -nostdin -i Q-!jpg_quality!.jpg -nostdin -i input.png ^
-lavfi ssim -f null - 2^>^&1 ^| findstr "All:"') do set "a=%%A"

set "a=!a:*All:0.=!" & set "ssim_target=!a:~0,3!" & echo Optimal SSIM found: 0.!ssim_target! & echo:
ren Q-!jpg_quality!.jpg 0.!ssim_target!_Q-!jpg_quality!.jpg

:loop
cwebp -preset photo -psnr !psnr! -pass 10 -m 6 -f 0 -sharp_yuv input.png -o psnr_!psnr!.webp 2>nul

for /f "delims=" %%B in ('ffmpeg -nostdin -i psnr_!psnr!.webp -nostdin -i input.png ^
-lavfi ssim -f null - 2^>^&1 ^| findstr "All:"') do set "b=%%B"

set "b=!b:*All:0.=!" & set "ssim=!b:~0,3!" & echo Webp SSIM: 0.!ssim! @ PSNR !psnr! & echo:
ren psnr_!psnr!.webp 0.!ssim!_psnr_!psnr!.webp

if !ssim! lss !ssim_target! (if !psnr! lss !maxpsnr! (del /s /q 0.!ssim!_psnr_!psnr!.webp psnr_!psnr!.png >nul 2>&1 & set /a psnr+=1 & goto loop) )
:end

for %%Y in ("0.!ssim_target!_Q-!jpg_quality!.jpg") do set /a kb_jpg=(%%~zY+1023)/1024 & for %%Y in ("0.!ssim!_psnr_!psnr!.webp") do set /a kb_webp=(%%~zY+1023)/1024
set /a saved_pct=(kb_jpg-kb_webp)*100/kb_jpg & echo [WEP]= !kb_webp! KB ____ [JPG]= !kb_jpg!
echo [WEP]= 0.!ssim! SSIM [JPG]= 0.!ssim_target! SSIM & echo cjpeg -q !jpg_quality! used as reference
echo **** !saved_pct!%% filesize reduction achieved with WebP ****
echo: & pause
>>
File: input.png (2.62 MB, 1920x1080)
2.62 MB PNG
OP source PNG and script output when selecting 80% quality:

Webp SSIM: 0.984 @ PSNR 44

[WEP]= 128 KB ____ [JPG]= 181
[WEP]= 0.984 SSIM [JPG]= 0.983 SSIM
cjpeg -q 80 used as reference
**** 29% filesize reduction achieved with WebP ****
>>
>>109034858
>the av1 schizo is a pedo
of course
>>
>>109034865
I hope you're comparing it to mozcjpeg not vanilla cjpeg?
>>
>>109034879
>Immediate sexual thoughts upon seeing a little girl
Anon, you have some serious introspection to do.
>>
File: Screenshot (117).png (1.74 MB, 1855x1080)
1.74 MB PNG
Webp SSIM: 0.966 @ PSNR 40

[WEP]= 198 KB ____ [JPG]= 313
[WEP]= 0.966 SSIM [JPG]= 0.964 SSIM
cjpeg -q 80 used as reference
**** 36% filesize reduction achieved with WebP ****
>>
File: input2.png (2.74 MB, 1920x1080)
2.74 MB PNG
>>109034947
source
>>
File: 1780677640622699.png (201 KB, 437x459)
201 KB PNG
Here's little elaboration on the script. Basically it's meant to be a jewish shortcut to running LPIPS/CVVDP/DISTS, which take ages to compute and require different score ranges for certain types of images. So I've been stuck on the idea that what if instead of chasing metric scores, we make an image codec act as a reference point for simpler and easier to compute metrics like SSIM? In my experience JPG fills that role perfectly because no matter the image type anything below -q 50 will be utter dogshit and anything above 80 will only give small quality improvements while increasing filesize dramatically. It's unusually consistent at doing this which means whatever SSIM score it spits out at that quality range seems to be pretty trustworthy.
>>
>>109034858
Wow, you saved 50 KB and made your image incompatible with tons of programs! Just imagine, you could fit like 1 more picture onto a floppy disk!
>>
AVIF is the superior image format, soon to be upgraded with AV2 compression beating webpiss even more
>>
CDNs love webp because they can quickly degrade quality of png and jpg passing through it and save on bandwidth.

for personal use, where encoding speed is not the most important thing in the world, there are better formats.
>>
File: 1777798734863447.png (759 KB, 615x767)
759 KB PNG
>>109035175
True but Webp is more universal and has the lowest decode complexity compared to modern image formats so it works no matter how lowspec a computer is.

Webp in a sense is the H264 of image codecs. Especially when you consider that it also replaces PNG and GIF.

>>109035165
I've actually just dispeoved picrel completely wrong by linking you the imigalaxy thread where I'm able to upload and post custom .webp images with better compression/quality than you would get with other websites using -m 2 -q 70.
>>
>>109035169
sorry forgot to link you to >>109035198
>>
>>109034858
>shilling av1: odd, but at least it's a decent code
>shilling complete dogshit like webp: ???
>comparing it against ancient shit like jpeg: ???
you're one weird motherfucker
at least use avif if you're incapable of pulling google's cock out of your mouth
>>
File: pb.jpg (65 KB, 900x736)
65 KB JPG
>>109034879
Your behavior is highly suspicious.
>>
Every since I made my own website, I pretty much use webp for everything.
>>
>>109035208
Can you please stop the habit of spewing diarrhea from your mouth in this technology related thread and at least have the common courtesy of reading the thread before you post anything?

See >>109035198
>>
>>109035198
>Especially when you consider that it also replaces PNG and GIF.
This is a downside, not an advantage. You have a .jpg file? You know it's lossy and not transparent. You have a .png? You know it's lossless and maybe transparent. You have a .gif? You know it's lossless but 256 color and likely animated because why else would it be saved as gif in 2026.
You have a .webp file? Is it lossy? Is it lossless? Is it animated? God only knows.
>>
>I care enough to save a few kilobytes per image
>but not enough to use modern image codecs like jxl or avif
not using webp even if the alternative is manually engraving the images on a clay tablet
>>
>>109035236
Lossy PNG exists, static GIFs get posted all the time. You will never really be able to remove the variable that causes all of us suffering: retards with computers.

Also saying this means that jpeg xl is a huge failure as well which is kinda stupid man.

>>109035239
You're not here to actually have a technical discussion about the thread topic here are you?
>>
>>109035249
No, I just wanted to shit on webp which caused me a lot of misery over the years by its blighted existence
fuck webp, all my niggas hate webp
>>
>>109035249
>Lossy PNG exists
The png format itself is never lossy. "Lossy png" is modifying an image so that it will be smaller when saved as png. By the same logic, lossy zip also exists if you decide part of your file isn't very important and replace it with zero bytes before compressing it.
>>
>>109035297
lossy zip is just zip, it'll take a chainsaw to anything beyond dead simple file structure and metadata used by 1990
>>
>>109035297
A PNG LITERALLY cannot store a JPG and people convert a JPG to a PNG all the time rendering the end result lossy.

The only way you can actually transfer a JPG with lossless transfer is via Jpeg XL.
>>
File: pepefroggie.jpg (35 KB, 780x438)
35 KB JPG
All 4 photos look exactly the same.
>>
>>109035330
Exactly you fucking inbred frog spammer.
>>
jpeg is "alien technology from the future" and can be losslessly rotated, cropped, joined and edited (re-encoding only changed blocks). with jpegtran.
can any of newfangled formats do this?
>>
>>109035311
>people convert a JPG to a PNG all the time rendering the end result lossy.
Resaving jpg as png is stupid, and of course it will perfectly preserve the artifacts that jpg created, and waste space doing so.
>The only way you can actually transfer a JPG with lossless transfer is via Jpeg XL.
It's only lossless in that you haven't created any new artifacts. You can also transfer a jpg without adding size or artifacts by leaving it as a jpg.
>>
>>109035311
what is this nonsense?
is flac lossy because converting an mp3 to flac means that this particular file is lossy?
saving a .jpg to .png creates a 1:1 pixel representation of the jpg, which means it's lossless. the source being lossy is irrelevant
>>
>>109035346
Okay so how do I get back my JPG chroma sub sampling from a PNG?
>>
File: file.png (158 KB, 360x360)
158 KB PNG
'ate telavif
'ate webp
'ate but tolerate (barely) heic
luv jxl
simple as
>>
>>109035366
>heic
Good goy.
>>
>>109035370
it's what my phone shoots in
i'm not converting existing heic into jxl
>>
>>109035377
>he doesn't know about raw mode
>>
>>109035462
>umm chuddy you should spend 30 seconds on every single photo you ever take to make colors look decent
nah
>>
>>109035549
You could also set it to shoot regular jpgs at the cost of a little more space.
>>
>>109035562
regular jpgs are obsolete for anything more than frog pictures uploaded to 4chan
>>
>>109035570
They don't require special codecs or programs to open on any modern OS, unlike HEIC, and that should be a bare minimum requirement for any format I'm taking pictures in.
>>
>>109035588
>flattens all hdr
>worse gradients
>unusable for even the most minor edits
>worse metadata handling
>worse compression for worse result
>no depth information
I'd much rather shoot heic and convert to jxl afterwards than use jpg on a modern phone
>>
>>109034858
webp clearly looks the worst in this comparison. What are you even on about lmao
>>
>>109034858
1080p is not high quality tho.. Webp can't even handle any of my high resolution fotos because of the 16kx16k resolution limit.
>>
Why would you shill webp when JXL exists?
>>
>>109034881
he's the same guy that uses dogshit encode settings with x264 when he compares it to av1 encodes.
of course he'll use the most dogshit jpeg settings he can to bias his image comparisons too.
>>
>>109035366
/thread
>>
>>109035366
yup... simple as.
>>
File: Untitled9.png (544 KB, 1424x1244)
544 KB PNG
>>109036097
Without zooming in, I honestly can't tell these apart desu. What display are you using? I'm using some a 32 inch 1080p monitor I bought a few years ago on black friday for $150.

>>109036143
WTF does image resolution have to do with quality?

>>109036179
It's been mentioned before but as great as Jpeg XL is, WebP is objectively more universal.
>>
File: jxl_proc.png (2.22 MB, 1920x1080)
2.22 MB PNG
I made a JXL about the same filesize (127 KB) as the webp on the umigalaxy thread. picrel is the PNG render of that.

https://files.catbox.moe/d52asw.jxl

I like JXL because of other things (ie HDR/losless jpg/progressive/etc) but if I'm being honest the JXL and webp at 127 KB look the same to me...
>>
File: 1781238330432_proc.png (1.8 MB, 1920x1080)
1.8 MB PNG
picrel is the PNG render of the webp
>>
>>109034858
Pedo dog whistle thread
>>
>>109037755
wtf is wrong with you man?
>>
File: 1762283457457622.png (1.44 MB, 1920x1080)
1.44 MB PNG
I save all my screenshots in avif but it really hammers my CPU to batch-convert them from png, webp is way less resource-intensive
>>
File: Screenshot (118).png (1.31 MB, 1855x1080)
1.31 MB PNG
Webp SSIM: 0.981 @ PSNR 44

[WEP]= 156 KB ____ [JPG]= 244
[WEP]= 0.981 SSIM [JPG]= 0.981 SSIM
cjpeg -q 80 used as reference
**** 36% filesize reduction achieved with WebP ****
>>
>>109037760
>go to any pragmata thread on /v/
>quarter of posts fantasizing about fucking this character
>a quarter of posts fantasizing about fucking some other child robot (8 or something)
>the occasional pizza link
must’ve been the wind
>>
>>109037996
>I save all my screenshots in avif
Why would you use avif when even google cucked out and is implementing jxl in chrome again? jxl is a lot easier on cpu
>>
>>109034865
I DONT GIVE A FUCK ABOUT THE FUCKING CODE! i just want to download this stupid fucking application and use it

WHY IS THERE CODE??? MAKE A FUCKING .EXE FILE AND GIVE IT TO ME. these dumbfucks think that everyone is a developer and understands code. well i am not and i don't understand it. I only know to download and install applications. SO WHY THE FUCK IS THERE CODE? make an EXE file and give it to me. STUPID FUCKING SMELLY NERDS
>>
File: speed.png (377 KB, 1444x1511)
377 KB PNG
>>109038602
>jxl is a lot easier on cpu
Source?

I mean that earnestly btw. Google seems to be working on an improved JXL decoder so picrel isn't a thing anymore, right?
>>
>>109038640
To be fair, you're inviting this sort of reaction if you put your shit on GitHub aka Facebook for developers.
>>
>>109038838
>it’s popular therefore it’s facebook
bitches dont know bout my password locked marginalia repos
>>
>>109038813
>cope benchmarks from 20th century
Have you considered running your own and just checking the numbers? It’s pretty trivial to do so nowadays, now rhat you can ask ai sloppa how to carry out such a test
with jxl as 100% I had avif at 283% on a macbook and 227% on linux desktop
>>
>>109038852
You know exactly why I said what I said.
>>
File: 1779752835697340.png (439 KB, 582x900)
439 KB PNG
>>109038887
You DO realize AI is just regurgitated jeet answers from reddit, RIGHT?

Anyway JXL is like 80% perfect, the only issue it has is decoding performance needs to be improved and it's automatically a clear winner.
>>
File: 1779921464893618m.jpg (64 KB, 776x1024)
64 KB JPG
>>109038640
I know this is a stupid meme but you honestly have to be the most ZOGed of windows users to not be able to use a fucking windows batch script.

Like serious, how much more dumbed down than a windows batch script can you get before resorting to a GUI?
>>
File: ZomboDroid_12062026140735.png (277 KB, 2158x1632)
277 KB PNG
>>109037711
>>109037719
I've heard that JXL is only useful for photographs like those out of DSLR raws but kinda sucks for everything else. You can bait JXL supporters with Anime screenshots but honestly that's pretty childish since anything like that could be 90% amaller if it were vectorized and saved as a .svg file.
>>
# Load zsh high-precision time module
zmodload zsh/datetime
NUM_FILES=17

# --- Encoding Settings ---
# JXL: Distance 1.0 (-d 1.0) targets visually lossless quality. (Default effort is 7).
JXL_ENC_ARGS=("-d" "1.0")

# AVIF: Quality 80 (-q 80) roughly matches JXL's visually lossless target.
# Speed 6 (-s 6) is used because AVIF's default speeds can be impractically slow
# for benchmarking and day-to-day use compared to JXL's default.
AVIF_ENC_ARGS=("-q" "80" "-s" "6")

echo "Target files: _png_1.png to _png_${NUM_FILES}.png"
echo "JXL args: cjxl ${JXL_ENC_ARGS[@]}"
echo "AVIF args: avifenc ${AVIF_ENC_ARGS[@]}"
echo "======================================\n"

echo -n "Encoding $NUM_FILES files to JXL... "
start_time=$EPOCHREALTIME
for ((i=1; i<=NUM_FILES; i++)); do
# Redirecting output to /dev/null keeps the terminal clean
cjxl ${JXL_ENC_ARGS[@]} "_png_$i.png" "_png_$i.jxl" >/dev/null 2>&1
done
end_time=$EPOCHREALTIME
jxl_enc_time=$((end_time - start_time))
printf "Done! (%.3fs)\n" $jxl_enc_time

echo -n "Encoding $NUM_FILES files to AVIF... "
start_time=$EPOCHREALTIME
for ((i=1; i<=NUM_FILES; i++)); do
avifenc ${AVIF_ENC_ARGS[@]} "_png_$i.png" "_png_$i.avif" >/dev/null 2>&1
done
end_time=$EPOCHREALTIME
avif_enc_time=$((end_time - start_time))
printf "Done! (%.3fs)\n" $avif_enc_time

echo -n "Decoding $NUM_FILES JXL files to PNG... "
start_time=$EPOCHREALTIME
for ((i=1; i<=NUM_FILES; i++)); do
djxl "_png_$i.jxl" "_png_${i}_from_jxl.png" >/dev/null 2>&1
done
end_time=$EPOCHREALTIME
jxl_dec_time=$((end_time - start_time))
printf "Done! (%.3fs)\n" $jxl_dec_time
>>
echo -n "Decoding $NUM_FILES AVIF files to PNG... "
start_time=$EPOCHREALTIME
for ((i=1; i<=NUM_FILES; i++)); do
avifdec "_png_$i.avif" "_png_${i}_from_avif.png" >/dev/null 2>&1
done
end_time=$EPOCHREALTIME
avif_dec_time=$((end_time - start_time))
printf "Done! (%.3fs)\n" $avif_dec_time

printf "JXL Total Encode Time: %8.3f s\n" $jxl_enc_time
printf "AVIF Total Encode Time: %8.3f s\n" $avif_enc_time
echo "--------------------------------------"
printf "JXL Total Decode Time: %8.3f s\n" $jxl_dec_time
printf "AVIF Total Decode Time: %8.3f s\n" $avif_dec_time
echo "======================================"


 ./avif-vs-jxl.sh
======================================
Image Encoding/Decoding Benchmark
======================================
Target files: _png_1.png to _png_17.png
JXL args: cjxl -d 1.0
AVIF args: avifenc -q 80 -s 6
======================================

Encoding 17 files to JXL... Done! (6.859s)
Encoding 17 files to AVIF... Done! (4.221s)
Decoding 17 JXL files to PNG... Done! (3.648s)
Decoding 17 AVIF files to PNG... Done! (12.774s)

======================================
Benchmark Results Summary
======================================
JXL Total Encode Time: 6.859 s
AVIF Total Encode Time: 4.221 s
--------------------------------------
JXL Total Decode Time: 3.648 s
AVIF Total Decode Time: 12.774 s
======================================


>almost 400% decoding time, ie what matters in practice
your move telavif
>>
>>109039238
>>109039247
I don't speak freetard, can you convert this to a windows batch script?
>>
hey chatgpt please convert this zsh script to powershell, using built in windows tools in place of linux ones
>not having wsl2
>>
>>109038938
>the only issue it has is decoding performance needs to be improved and it's automatically a clear winner.
It’s a pretty major issue when you open a folder with avif images and it chokes the shit out of my 9800x3d
>>
>>109039277
didn't work lol
>>
>>109039304
mmmm huh
>>
>>109039427
So you can write ssim autism but can’t manage to ask a chatbot to rewrite a simple shell script to work in powershell?
I guess you’ll have to deal with avif being proven to be slow as shit to decode then since I don’t have a winblows machine nearby to do your job for you
>>
>>109039452
im not op though

your jeetbot can't translate loonix-speek to windows for shit
>>
>trust not your lying eyes
>trust my gay chart instead
>>
>>109039460
You’re free to write it by hand. You managed ssim, you’ll manage measuring cpu time.
>>
>>109037260
Higher image quality requires higher resolution, because you cannot represent additional details without increasing the resolution. Of course that doesn't mean that a higher resolution image is always higher quality, but a higher quality image will always have higher resolution.
>>
>>109039471
I don't how to, I'm not OP.

Why don't you enlighten us and find us a jeetbot that can translate it then? Chatgpt is FUCKING DOGSHIT for this desu senpai.
>>
>>109039576
Bullshit

12MP iphone image will always be better than 100MP chinese android phone image
>>
>>109039227
it might surprise you, but jpeg xl is also a vector format.
>>
>>109040481
Really? I was under the impression that this is a lossy PNG kind of image format since it doesn't encode 4:2:0 images like Webp/AVIF.

ffmpeg -h encoder=libjxl

Encoder libjxl [libjxl JPEG XL]:
General capabilities: dr1 threads
Threading capabilities: other
Supported pixel formats: rgb24 rgba rgb48le rgba64le rgbf32le rgbaf32le gray ya8 gray16le ya16le grayf32le
>>
Here's the webp output for anyone curious.

Encoder libwebp [libwebp WebP image]:
General capabilities: dr1
Threading capabilities: none
Supported pixel formats: bgra yuv420p yuva420p


Weird thing is BGRA. What in the fuck is that thing?
>>
>>109034858
>irrrelevant time wasting post
>lust provoking image
>>
>>109040790
JXL can be lossless, lossy, lossless recompression of JPEG, collection of catmull-rom curves (this is the vector part)
>>
File: 1780731687255113.jpg (66 KB, 526x680)
66 KB JPG
>>109040834
So ffmpeg doesn't support the last part yet?

I'm not surprised desu because these COCKSUCKERS haven't added animated webp decoding yet despite how much animated webp curb stomps GIF. Encodes it fine just can't fucking decode it...
>>
>>109040790
JXL supports 4:2:0 internally. Old versions of the encoder could do it.
https://github.com/libjxl/libjxl/issues/2663
>>
>>109040871
nothing encodes catmull-rom splines now, it can only be hand-crafted.
I imagine they future-proofed the format in case future smart algorithm detects anime and vectorizes it.
>>
>>109040874
Hmmmmmmm, how do I make a 4:2:0 JXL then? FFmpeg clearly doesn't list yuv420p for JXL yet.

Shit man, is ffmpeg ever going to get replaced by something better? I'm getting sick of this "maybe it'll support it, maybe it won't te-hee XD" bullshit.
>>
>>109039701
As I said higher resolution doesn't mean higher quality. Higher quality needs higher resolution tho. A foto from my fuji gfx ii i.e. will ALWAYS have higher quality than any smartphone. The resolution will vary depending on my settings but it will be at least 11k pixels in width. A resolution of at least 32k pixels is not uncommon for high end cameras
>>
>>109041543
>As I said higher resolution doesn't mean higher quality.
>Higher image quality requires higher resolution

Man, having a stroke sucks. I know I'm going to pay dearly for eating steak tacos everyday some day...
>>
Why exactly did you use the pedo game character to showcase the encoder?
>>
File: 1779981143094328.jpg (125 KB, 774x896)
125 KB JPG
>>109041594
Diana is a pedo? I thought she was just a nazi.
>>
>>109034858
Already super-seeded by AVIF althoughbeit.
>>
>>109041594
Man, what does your raging hard on have to do with the thread topic? Take that shit to reddit.
>>
>>109041773
Hmmmmmmmmm, NAH.

AVIF sucks at lossless. JXL is the best at lossless but Webp is still a huge improvement over PNG and even oxiPNG.
>>
>>109034858
in current day most devices have devices 500gb+ storage and gigabit internet. lossy encoding of images should be illegal
>>
>>109041944
what is this, advocacy for random noise's rights?
>>
>>109041944
Are you going to pay hosting bills, you fucking imbecile?
>>
File: azxW2NN_460swp_proc.png (255 KB, 460x640)
255 KB PNG
>>109040818
>lust provoking post
>irrelevant time wasting image
What now, tough guy?
>>
>>109035359
Are you retarded? The sub-sampled chroma isn't upscaled in any fancy way, it gets nearest neighbor'd if you do something stupid like save to PNG
>but my JPEG wasn't 4:2:0
You're grasping at straws, who the fuck cares? Your lossy chroma subsampling is the least of your issues in this scenario.



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.