For all you anon artists out there, how do you go about being able to draw images with photo-realistic accuracy?I picked up some new colored pencils (Faber-Castel Polychromos), and what started a simple trial sketch turned into an endeavor to test my abilities to draw something that looks exactly like the photograph I drew it after.The drawing by itself is... fine, but for an attempt and making a photocopy in hand-drawn form, I think it's fairly poor. Details and contours of the shapes are inaccurate, or in the wrong place, flanderized, the perspectives are off, etc. I also tried to draw this with less trace-checking (overlaying a scan into art software, overlay the scan over the reference image, and changing the opacity to make sure all the details lined up)... only to do that a few times anyway, and the details still ended up being off.It's incredibly vexing, because I have what I'm supposed to be drawing right in front of my eyes, plain as day, and yet when I try to transcribe that information onto paper, I falter. It genuinely feels like I'm missing parts of my brain to disseminate and translate all the visual information properly, or my ego is getting in the way and throwing off what I should be drawing with what it erroneously and solipsistically presumes I should be drawing.What leaves me feeling the agony of Tantalus most of all is that I know photo-realism is objectively possible. I've seen many other colored pencil artists achieve it with such proficiency that they make it look easy, so I know my shortcomings and failures to accomplish what they have are entirely my fault. But I can barely even fathom what it is that I'm missing, or doing wrong, let alone how to fix it.And just to add a cherry of shame on top of the inferiority sundae, this took way, way, WAY too fucking long to draw. I spent something like 15 hours on this, and it's only 4" by 3.5".So what am I doing wrong?>"Nothing is worse than having an itch you can't scratch."
Drawing isolated:
Reference photograph isolated:
>>7901639the main thing I'm noticing is the overall lack of contrast (colored pencils blah blah blah i know). the culprit for this is your lighting and shadows. all of the darkest shadows (inside the calyx of the blueberries and the shadows in between each berry) are different values entirely. They're around the same in the original image. You are able to pick up on this contrast but it isn't applied consistently. You're also able to see the lighting correctly, but it gets lost in translation into a mess of very gradual gradients. I think the main thing to work on is getting the values of the image down in your head before acting on them. Some of the contrasty bits are also lost entirely, like at the top of the bottom blueberry in the original, the shading is fairly flat before it cuts off to the background behind it. You added a shadow gradient where there isn't one. Also, the leaves and branch in the background look like they're a culprit of you relying on drawing just what you see too much. They all kind of just blend together in a sea of green waves. You can render them very well, but visual coherency still looks off, mainly a culprit of values I think. I guess what I'm saying is work on values/luminance. Try using your phone camera while in grayscale mode if you have trouble with it, the mistakes are a lot more obvious when you take away color. Hope this helps.
Darker values are missing, you need blacks, but remember not to do pure black
>>7901639Photo-realism is a fool's errand. As a challenge it's fine, but don't get suck down this rabbit hole if you actually want to draw anime girls or whatever. Like you mentioned, it takes many hours, especially with colored pencils. In general, learning to draw accurately is a good skill to develop.For EZ mode, just trace it. You'll usually end up mindlessly tracing it, so you won't learn much but it gets the job done. Useful if you want to focus on another part of the drawing process.>grid methodOverlay a grid on the photo and your paper. Objects will overlap on the grid, use these landmarks to draw the objects in the right spot. To train your observational and accuracy skills, use a 4x4 grid, then a 3x3 grid etc. until you copy without a grid.>sight-size drawing methodPrint out the photo, put your drawing right next to it and copy it at the same size. You can use a string/pencil to line up landmarks like with the grid. double check, then triple check your measurements. the more you do it, the less often you need to double check as your skills improve.>comparative measurement methodOnce you can copy decently accurately with the other methods, you can use the comparative measurement method. Pick an object to use a a base unit of measurement. e.g. the head. Then you use the measurement to compare it relative to the head to measure the proportions of the rest of the image. This approach focuses on the ratios between elements, allowing you to draw at any scale which allows for more flexibility and further improves your skills. There's lots of ways to measure, so try them all out.
>>7901652Also, start with simpler photos then more complex ones. Copying one apple would be a bit easier than multiple blueberries for example. Draw smaller so you can finish it faster. Copy grayscale photos too, for the same reason. You'll make proportional mistakes earlier on in the drawing stage. Often you won't notice until later on, so it's better if you finish many smaller drawings to get more practice in the beginning stages of the process.
I think you're being too hard on yourself. Your blueberries look tasty and delicious, they just need deeper shadows. That is the immediate thing that pops out to me - darks not dark enough. Try mixed media if your pencils can't get there. Shading always takes a while, so don't feel bad about that either.I don't do photorealism, but if there's one thing I've learned from shading in pencil/ink, it's that the shadows, dark shapes, and especially tightly occluded areas are what make the image "pop." It doesn't take much, either. Drawing in an almost pure black cast shadow feels like a form-conveying cheat, instantly elevates the image and gives it a graphic quality.Once you come to grips with value, you will probably stop caring about making a perfect copy, and will start to interpret and exaggerate your subject.Art isn't about being a Xerox machine - you're not a scanner. It's about illusion making, interpretation, omission, heightening. You can never beat the photograph at accuracy, so try to exaggerate it in interesting ways, then call it "style."
>>7901639One classic trick is simply shrinking things down, if you don't mind a little digital manipulation.Look at the thumbnail of your images on this page - it's hard to distinguish which one is the photo and which is the drawing.So work at an even larger scale and shrink the drawing down to a reasonable size, and it'll probably look quite convincing, if not entirely so.Of course, if you're trying to make and sell traditional art as is, that trick isn't going to work.I think the big thing holding back your drawing, at its actual size, is the use of the medium itself, with scratch pencil marking being quite easily observable. Is this a bad thing? Personally I say no, and find it strange when people try so hard to hard the fingerprints of their respective medium in their works, like making a watercolour look like an oil painting for example. However, if you're trying to make an entirely convincing realistic piece, greater mastery of your medium (or higher quality supplies) are probably needed, though I won't pretend to be all that knowledgeable on working with coloured pencils.
>>7901639People who draw photorealism with colored pencils do it in a very tedious way. They literally stipple draw every pixel of the drawing. They go over a shape with regular hatching, and then they tediously fill out the little pixels which don't have enough color with little dots.Apart from this, it's just measuring and comparing. It's easier when you have your reference laying right next to the drawing, otherwise you always have to look up and down, which makes it way harder, because the brain forgets the image it saw within literal miliseconds.If you do this a lot, then your brain will get accustomed to it and your synapses will be able to better measure things. You just get a feel for the proportions after many hours of work, and you will also be able to draw cleaner (which in turn makes the proportions look more accurate too).I think your picture looks a bit desaturated compared to the ref. This is because you don't do the stipple drawing technique. Usually when using pencils, people will just add more layers of color or put more pressure on the pencil in order to make it look more satured. However when you do this, you basically destroy the paper texture. I think that's one of the reasons why people tediously fill in every pixel with little dots.
>>7901643I can't believe I forgot to do a value check. I was frequently checking the hues with a color picker tool, but the idea of dropping the chroma to zero to check the tones didn't even occur to me, despite the fact that I've been doing that with some of my recent drawings.I wonder if I'm polluting the black graphite too much. For each berry, I started with layers of the lightest colors and then worked down to the darker tones. I had to do something I've never done before and start with a layer of white in the highlighted areas, because Polychromos don't blend nearly as well as Prismacolors do. That did seem to average out the values when used darker blues and light layers of black for the shadows.I realized I messed up and added a gradient that wasn't there only after I made it. I tried to mitigate the damage by layering lighter colors on top, but it was too late. But that hits on the bigger issue I was talking about in the OP, about not being able to translate what I'm looking at directly onto the paper, and going off in different directions and just making shit up. Hues and values I can check with a color picker tool, but it's the issues with the line art itself I find myself most lacking.>the leaves and branch in the background look like they're a culprit of you relying on drawing just what you see too muchI'm confused by this criticism. I know the need to understand how the think I'm looking at is composed in order to recreate it, but, isn't the entire point to make it look like what I see?>They all kind of just blend together in a sea of green wavesAren't they supposed to? The leaves in the background are very blurry. Though the crunched values are to blame on my scanner; they didn't pick up nearly enough of the yellows I used for highlights. I also had literally only one green pencil to work with.Thank you for your critique, I'll remember the tone check for next time, but it's my inaccuracies of form that I'm more troubled by.
>>7901645>Darker values are missing, you need blacks, but remember not to do pure blackI'm still 13.I'm still a freshman in high school.I'm still in third period Drawing 101 class.I'm still being told by Mr. Hilmer to "go darker."
>>7901639Why do you want to do photorealism? Isnt it what amateur permabegs do? Why not to study academism instead? It also requires precision but it looks better than reality because its not just a copy
>>7902499When will you finally learn the lesson Hilmer was trying to teach you all those years ago.
>>7901652>Photo-realism is a fool's errandWhy?>As a challenge it's fine, but don't get suck down this rabbit hole if you actually want to draw anime girls or whatever.Implying that realism and manga are mutually exclusive? Somebody has never heard of Takehiko Inoue's Vagabond.>For EZ mode, just trace itI actually used to do trace and redraw exercises when I was teaching myself how to draw anime characters. The attached file was my first exercise. But I've tried to ween myself off that, since I ended up so frequently using a ruler to try and make every single vector of every single line perfect that it was pretty much just tracing with extra steps. It was just tedious cheating.I actually do use methods like some of those mentioned in you pic. I load just about every reference image I use into Gimp to check measurements for things like alignment and proportions, or overlay images with decreased opacity. That's what I did for the OP drawing... and still fucked up.>grid methodMan, it is disheartening as fuck to be told you essentially still need training wheels.>Pick an object to use a a base unit of measurement. e.g. the head.I've been doing that since I was 12, first learning how to draw the human body. I got that trick from none other than Stan Lee... and yet I still frequently screw up and make the head too big.>Print out the photoDon't have a printer. I do however load up reference images on my computer screen, sketch on thin, 20 lb printer paper, and periodically hold the paper up to the screen to do what I call "trace-checking," but that feels like cheating sometimes, too. Ideally, the goal should be to be able to perfectly draw what I see based off sight alone. That's what professional portrait artists do all the time, and fuck, police composite artists can draw a lifelike portrait based off nothing but artistically talentless eye-witnesses' shitty descriptions of them. Why can't I do the same thing? What do they have that I lack?
