Digital Painting Portrait - 2nd Attempt (Moved from FF)

[BrandonSuperior]

Hey Guys! This is my Latest Piece, an Angelina Jolie Test Portrait.

I really struggled with getting her skin to look smoother and more "life-like" if you will, thought I'd register on here and get some well needed advise!

I have drawn traditional graphite pencil portraits for around 5 years. This is my second ever attempt at working digitally.

Please let me know what you think. ( Sorry if it's a bit big, I'm not sure how to downscale it)

Thanks!

[Chris Bennett]

All you've done is doodle over a photograph with a stylus while your left hand taps away at the SHIFT/sample picker. (giving the benifit of the doubt as to whether you've employed a filter as well)

Was your intention to make her look like she was in the first stages of turning into a crocodile with a piece of snot dripping from her ear? Because this is always going to be the result of doing this. That is to say, mapping colours over an image you have not internalised, not grasped as a 3D form or understood the graphic purpose of, because it is not yours to begin with.
This is not the same as working from reference, which is to gather visual journalistic information about a subject that you wish to transpose into your own imaginative, graphic epiphany.

This 'painting' is the equivalent of saying you have composed a song because you have been turning the volume up and down of a favourite track on your iPod.

Sorry to be so harsh, but if you can get over being a bit pissed of at me, it will help you far more than it wounds at the moment.

[BrandonSuperior]

All you've done is doodle over a photograph with a stylus while your left hand taps away at the SHIFT/sample picker. (giving the benifit of the doubt as to whether you've employed a filter as well)

Was your intention to make her look like she was in the first stages of turning into a crocodile with a piece of snot dripping from her ear? Because this is always going to be the result of doing this. That is to say, mapping colours over an image you have not internalised, not grasped as a 3D form or understood the graphic purpose of, because it is not yours to begin with.
This is not the same as working from reference, which is to gather visual journalistic information about a subject that you wish to transpose into your own imaginative, graphic epiphany.

This 'painting' is the equivalent of saying you have composed a song because you have been turning the volume up and down of a favourite track on your iPod.

Sorry to be so harsh, but if you can get over being a bit pissed of at me, it will help you far more than it wounds at the moment.

Hi Chris,

Now, just to clarify I am not "pissed off at you" in the slightest, if anything I am grateful that you have taken to time to look at my work and give your personal opinion. However that opinion is wrong in a few areas that I would just like to address. This piece totalled around 25+ hours for me to complete, I worked very hard in trying to create a convincing painting of her, ofcourse in order to do this I had a consistant referrence throughout. The fact you think this is the original image with a bit of paint on it and not my own rendition isn't shocking; there is no unique character and I have a tendency to try and copy the image exactly as it looks; I havent really developed much of a style in photoshop (due to the lack of tool understandings) This piece was started, and completed in an airbrush with 50% flow, no filters, no other photoshop advantages or shortcuts.

http://brandonsuperior.blogspot.co.uk/ If it would please you, I have attached here a link to my blog in which I gave a break down of this piece, and hopefully this will disprove the idea of any underlying image. Yes, I did use sample colour referrencing in the beginning stages, I am not ashamed of that at all, simply because I was worried about not involving the correct shades for the tone of her skin as I am so very used to drawing in grayscale, painting is not my forte (evidently you agree, with the crocodile skin pun! haha) but as time progressed I eventually knew which colours were needed just by simply looking at the original image and tones already used.

Just to point out, I found your "composition - turning up music" metaphor quite ammusing.

Although I feel as though your comment was nothing more than a condescending messege to someone who you probably assumed was just looking to be glorified for a painting that was cheated; I still appreciate that you had something to say, however mislead you might have been. My only disappointment is that you didn't continue your point about how I could've fleshed out the skin in a better way perhaps? or something just a bit more helpful.

Thank you again Bennett

[Black Spot]

Try using the heal tool, but be careful where you pick your source. The nose is a bit off - have another look at it. Also try not to be a photocopier - use your reference, don't be a slave to it.

I moved it from FF as it still needs work.

[Kolbenito]

The main issue I am seeing is that the skin is somewhat "painterly" but on a very fine scale with scribbly brush strokes. To my eye because of that her skin surface looks scarred. This is basically what skin that has been burned and healed looks like.

This video from Daarken is pretty good, I don't recall exactly where in it he talks about blending but it might help you out.
http://enliighten.com/blog/analysis-...-bust-part-01/

For the background can I suggest that you reference that background with some shots of resort locations or add more bokeh to drop it to just some tones? Depends on how much depth of field you want, but since it's a portrait I would suggest very shallow DOF.

GL and keep working at it!

[BrandonSuperior]

Thank you! I will try attempt this piece again in the future and not follow the reference so strictly.

[BrandonSuperior]

Thank you, after revisiting my portrait I can see exactly what you mean; I didn't want to add too much depth into the background as I was trying to keep Angelina the focus point of the piece. But thanks for the link too by the way!

[mimer]

I think the biggest question is why? Why would I look at this instead of the photo that its based on? You've essentially managed a bad copy of a heavily stylised image from a photoshoot. From the hours you've put into this, there are some rather extreme problems with the alterations you've made to what's visible of her arm and torso, just to name a few. when you're spending 25 hours on this, how do you end up with that arm and torso and those alterations to her cranium?

[sone_one]

I think the biggest question is why?

this.

in those 25 hours you could have easily done 6-12 studies. this would have forced you to ignore some details and make actual decissions, which are the hard part. you would have 6-12 pieces filled with successful and less successful decissions you could learn from and build upon in the future. all you got now is 1 piece, people think you ran a filter over, and no learning.

on a sidenote... id stay away from those glamourshots as reference. they are, in most cases badly and without proper anatomical knowledge, heavily photoshopped and therefore useless. those are rather ment to attract our sexual attention while standing in line at the checkout, then providing proper information. that shoulder, breast combo is completely redicilous and im pretty sure its in the ref aswell (post it?). its hard enough to aquire usefull information from a good photo... you might want to avoid those a graphic design trainee got wild with abusing the liquify tool.

[Star Eater]

Also, without wanting to seem like we are all ganging up on you. I have to make mention of the
size issue of the image. Resizing a digital image is one of the most basic actions there is.

If you truly do not know, look under 'Image ->Image size' at the top.

[Leonor]

The problem is you don't know what you are copying, you have no understanding of construction and anatomy and it shows. Even as flat copying is concerned you're not measuring relationships carefully. People who do "it looks like a photo" work do very meticulous measuring first.

I personally find more interesting when people learn to redraw Angelina Jolie because they studied her facial features through 3D construction, anatomy, etc. Quick sketches of her face in different poses (from film or paparazzi) is the kind of work that impresses me.

[justa]

Here are some thoughts on blending (not mine, Linda Bergkvist I believe)
http://wiki.polycount.com/TexturingT...-bergkvist.jpg


D'Arcy

[arenhaus]

This piece totalled around 25+ hours for me to complete, I worked very hard in trying to create a convincing painting of her, ofcourse in order to do this I had a consistant referrence throughout. The fact you think this is the original image with a bit of paint on it and not my own rendition isn't shocking; there is no unique character and I have a tendency to try and copy the image exactly as it looks

Work hours do not make a picture any better on their own. If you made something in an hour that makes people go all "wow", no one will care you didn't slave a week over it. If you made something in a week that makes people go all "meh", well, you wasted a week on something mediocre - and telling people that will only make them think "gee, he spent *how long* on that thing?"

In the end, with all the work hours, you've made something that looks 100% like a photo. Which is an utterly pointless exercise, because humans do not see like cameras do, and a lot of the meaning in a work of art derives from the artist's personal sight. And you are quitting your personal sight in favor of a trite shot made by a blind light-registering machine. Any potential artistic value ends up unfulfilled.

What you are doing is known as "being a slave to reference". Artists do not copy; artists compose and interpret. Being a human copy machine is pointless and self-degrading.

[Leonor]

In the end, with all the work hours, you've made something that looks 100% like a photo.

Er... What? No.

But I agree with the point you made.

[arenhaus]

Er... What? No.

If it has lighting like a photo, perspective like a photo, detail like a photo and color like a photo, I call it a photo.

[Ravatta]

Despite the hours of work supposedly (i apologize for the use of this word, it just really doesn't look very... Err, legit) put in this, it really looks a lot like a paintover... I'm especially puzzled by the lines in a few places, like her breasts and the hairline, which are clearly just a bunch of scribbles. I don't think anyone able to replicate a picture so closely there's no apparent defect in anathomy could make such bad looking paint strokes.

[StefRob]

I think some people here are quite untolerant and express good advices in a rude manner. Or simply forgot what it's like to be a beginner...
He likes spending time in reproducing pictures and can't see any probleme with that if he enjoys him. I get and agree with the common opinion, you, Brandonsuperior should start thinking of creating images, starting from scratch or at least using your own style if you reproduce photographs. You seem pretty good at reproducing but if you try you'll discover that creating from nothing an image leads to ask yourself a bunch of practical questions (lighting and shadows, anatomy, perspective and color theory). As a total beginner, I'm in this boat too and it's fascinating. Don't be discouraged.

[Star Eater]

I think some people here are quite untolerant and express good advices in a rude manner. Or simply forgot what it's like to be a beginner...

Nobody in this thread has been rude.

[bill618]

I really struggled with getting her skin to look smoother and more "life-like" if you will, thought I'd register on here and get some well needed advise!

Transcendent things like ‘lifelike quality’, effective and evocative color palettes and composition, are where real artistic sensibility kicks in, and the copycat ethic ends. Until you realize the difference, you’re going to have a hard time moving forward.

As mentioned already, what’s the point of doing this if it doesn’t look better than the traced color sampled reference photo? It can’t possibly be done as a learning exercise. I guess these contrivances of photo underlays are designed to impress artistically naïve friends and family, because they sure don’t go over well with a professional audience. Hyper-realism should be the goal. The digitally traced sketch seems pointless, as I’d bet the ref photo underlay never left the ‘Photo-shop’ document. I guess the sketch could be used in the steps breakdown to show some sort of ‘this-ain’t-a-paintover’, throw ‘em off the scent, dubious gesture of process sincerity?
You need to learn some anatomy and not rely on tracing photos. Your level of anatomical knowledge shows in the area where you tried to modify her left arm. You did not enhance it, you grossly distorted its form, and you are totally unaware of it. Her left arm looks like it would be a short, withered, deformity resembling a tiny T-Rex, vestigial limb, if seen in full.
I think your paint-by-number approach (color picking from the ref photo) is not contributing to the surface quality of the skin. You simply have to build the gradations up slowly for a smoothly blended underlayer which you can then add detail on top of (when trying to achieve a phtorealistic look).

Photorealistic digital work is self-defeating, as it will always be viewed suspiciously under the specter of PHOTO-SHOP. Photorealistic work can be impressive when done hyper-realistically in traditional media, on huge canvases, à la Chuck Close.

You actually copied that weird gray band (glitchy retouched waterline) on her lower left arm. Yikes.

Photo ref:
http://www.lowbird.com/data/images/2...lie-20-558.jpg