Portrait seems boring

[joelphilippage]

My portrait seems a bit dull. the right side is completely white, which I think it should be from the light source, but it feels really empty. Also, the attitude and everything seems a bit uninteresting. I want to keep the pale skin tone though. Got any Ideas to liven this up?



I tried this as well.

[kconan]

  She looks like she should have a slightly furrowed brow.  And maybe arch the eyebrows a bit?

[joelphilippage]

How's this?

[kconan]

  How about adding just a dash of forehead lines?

[ThreeOhFour]

I don't think the pic gains anything from the extra red shade on the lips - I feel it might be nicer if you stuck to the skin palette here and shaped them a little more with shading:



I also tried to bring a bit more 'life' into the eyes and adjusted some other minor things. As for expression, depends on the sort of character she is and the situation she is in.

[joelphilippage]

I do like the improved eyes. It does give her some more life. I think adding forehead lines makes her look old, and she is not supposed to be more than 20. The arched eyebrow fits her character. I could use this as an optional expression. she looks a bit more sad in the older one.



tried a few of Ben's Ideas with a few differences. I don't want the lips to be very full. she should be kind of plain

[Khris]

The main dark shade is too saturated and the middle shade is not saturated enough.
I'd also use a neutral color close to 50% saturation as background while drawing to not screw up your color perception.

Is this a trace of a photograph?

[joelphilippage]

I enjoy bizarre colors. Does this look better?

Here is the original picture I based this off of.

[Anian]

Well the refference doesn't have that many lines either, but I saw Ben's pic and it look nicer for some reason, then I figured it's probably cause it's a tad blurred and therefore the lines are softer.
The way shades are on her lips in your pic is maybe a bit too much contrast (or maybe difference between colors), the way you used them, the illumination would require far stronger and more focuesd light (a flash or something) but on the original pic it's very diffused illumintaion.
If a flash was used, then there would be less detail on the side of the head that's in the shed (her right side), there would be no highlights in the hair etc.

So if you're not going for the high contrast, face coming out of the shadow, lightning striking nearby, I suggest you either remove shading, except for some lines that make details, or make area shades (colors) closer (in hue), so the transition so they feel less "sharp." The one before bizzare color version looks nice.

[cat]

Ben's picture is not blurred if you have the correct firefox settings...

For the pic: I like your last one best except for the raised eyebrow, I'm not sure about that. I seems to make an expression that the rest of the face does not share.

[joelphilippage]

There's a bit less of a contrast in this version and not so much of a direct light source. I added bangs for some reason.

[Sughly]

I think the white is too extreme. Just making it a really light shade of her skin tone would be better...

[loominous]

A common tip when dealing with values from a reference is to squint your eyes, which enables you to look past distracting details and get a good general impression of the value distribution.

Values don't only convey the form and material of the object, but also set the overall mood to a great extent, so it's a very delicate affair, that is easily messed up.

Like with everything else, it's usually good to start with the big picture, and work down to details. Blurring the image, equivalent of squinting your eyes, shows that due to a strong light source on the right, and a non-glossy skin, you find the main value difference on the left of her face, where the head is in shadow from the light source.



This is usually the case, unless you have very shiny objects, so it's good to remember, as people tend to instead focus and exaggerate minute highlights/shadow lines, which create weird and often mildly freakish results.

So by observing the object in this manner, you can begin with the vital areas, and work your way down to the details, confident that it'll probably turn out fine in the end, or at least that you've used a method that gives you a good chance at it.

[Mad]

I think Sughly is absolutely right, the very light skin tones are very distracting!

I actually liked the annoyed look of your previous edit the most. In the last one she looks quite timid. But then again I don't really know what your aiming for, since I guess you took the photo as a general reference(?)

So I made this paintover, adjusting some of your shades and using the colour scheme of your last edit, because it somehow works quite well imho.

 


I didn't do the neck and hair, because I ran out of time, but especially the hair needs some work: I don't think you'd get as many highlights on her right hand side, considering the angle of the light source!

Keep at it, this is a very good sprite!


Edit: Ahh, didn't see loominous' post there. Very cool method, I'll keep that in mind!

Edit 2: Ok, couldn't be a**ed to work, so I made this:



This should illustrate, what I meant about the hair. Her right hand side is shadowed and only some of the locks catch the light.

[joelphilippage]

AAAHH! Now I'll never be able to go back to my ignorance of poor shading! Thanks guys. I'll see what I can do without just copying your picture. I'll at least re-draw it for myself to get your Ideas in.



Here's my re-edit. I really like the bun in her hair, but I want to have a set amount of pixels for the portraits.

[Mad]

That is a really good edit!

And if you're going with that restricted pane, I'd lose the bun, or give her a hair style the viewer can read more easily:



I also changed a few pixels beside her nose. I am not sure there'd be a shadow there unless she had a deep crease or scar.

But as I said, you've made a fantastic improvement taking all the comments from everyone on board!

[Shane 'ProgZmax' Stevens]

Joel, I'm just curious as to why you keep destroying the form of her nose in every single edit you make?  You keep flattening the bottom of her nose which is completely unrealistic and bizarre looking.  Mad's initial edit rounded her nose quite well and showed her nostrils (much like the reference image) and then you went back and flattened it again, making her nose look almost Bullish.  Also, it doesn't seem like you've redrawn these examples, just tweaked them, making the eyes and nose considerably worse in the process.  Finally, the green shade isn't artsy, it's highly distracting like you're viewing an anaglyph image.  The prevailing light source here does not seem capable of creating either a green shadow or a green bounce light so my advice is to get rid of it entirely and focus first on a competent portrait design.  Once you have form and body and shadows and highlights properly understood then perhaps you can experiment.

[loominous]

Finally, the green shade isn't artsy, it's highly distracting like you're viewing an anaglyph image.  [...] Once you have form and body and shadows and highlights properly understood then perhaps you can experiment.

I don't see why you'd want to discourage anyone from experimenting to their heart's content. Experimentation is hardly a luxury exclusive to competent artists.

I'm just curious as to why you keep destroying the form of her nose in every single edit you make?

A not so wild guess would be that it's because noses are tricky shapes to portray, and without sufficient understanding renderings will usually fail.

[joelphilippage]

It really was a re-edit I used Mad's edit a lot. All I did was attempt to re-draw his for myself. It may not have been pixel perfect, but reproduction is often a good method for learning. Some things I may choose to keep. Thanks for the help. I will probably be improved a lot and hope to keep learning things.

[Shane 'ProgZmax' Stevens]

But experimentation here is like leaping before you've learned how to walk properly; it's like trying to draw in Dott style before learning proper perspective; it just isn't a sound approach.  Learning anatomy, shading, and perspective first gives you the solid basis with which to bend the rules and experiment with other techniques and variations thereof.  Otherwise it's like being out at sea without a compass, paddling a bit here and a bit there and unsure of your progress.  I think in Joel's case the experimentation can wait until he's tackled the more prominent issues with the portrait and reached a good level of confidence with anatomy and shading .

A not so wild guess would be that it's because noses are tricky shapes to portray, and without sufficient understanding renderings will usually fail.

I'd appreciate it if you'd let Joel answer for himself as I'm genuinely curious why he re-edits certain touched-up areas from edits (like the nose) but leaves others alone.  I'm trying to find out whether he thinks these shapes are an improvement, trying to replicate them in his own way, or what.  His reply answered this for me, thankfully!


Joel- how exactly do you approach re-drawing?  Looking at your work, it looks to me like you are drawing over the edited images directly?  Or are you placing old versions of your work next to the edits and trying to draw what you like from them by eye?  I'm curious because I think you'll get more out of the latter than doing tweaks here and there with other people's edits.  I do think you're making progress in some areas, but because I also see you grinding your wheels with certain bits of anatomy I'd like to see where you are on these questions.