Need some C&C on sketch

[Eplekongen]

try superpositioning the images.

The mouth are needs work. The line under the left cheekbone should be straight, and the line under the right cheekbone should be more round. The mouth is either not wide enough or too large high-wise. The teeth needs to be darker. And you show too much of the of the "gommes" (not sure about this word) over the teeth.

my take on it anywho

[DArcy1]

This may help:

http://farm4.static.flickr.com/3176/2940576152_06f1c42caa_o.gif

Basically you weren't careful enough about your original tracing of the image;; despite all your ?locator points you didn't get the shape of the skull right. Using overly thick and choppy tracing lines never helps, especially in a face where we are hardwired to notice tiny differeces between people. Finally, as you can see you didn't get the shadows right, which are what give you the form of the face - more important than just the outlines.

Let's see an update !

D'Arcy

[stuh505]

Thanks. I'll see what I can do tomorrow ;) Looks like my major problem was dropping the top of her hairline too low, which made her skull look shorter and squatter.

[CybrGfx]

The mouth is crooked, and the teeth are incorrect.

Your greatest mistake, though, is simply one of Anatomy and Value.
Highlights go on the Higher parts of a structure, Shadows go in the lower parts...
You have not provided the proper bone and musculature values.

This includes the eyes, which you gave really goofy, "fake" specular light reflections, and solid colored eyeballs. Eyeballs a spherical, which means they shade just like a ball...

Here's a quick Paintover in Photoshop, adjusting the mouth, evening out the skintones a bit, and adjusting the hair. You do not need to obsess over hair. A bit of "soft edge" is actually desirable, so the viewer doesn't obsess over looking at it...
http://img522.imageshack.us/img522/7749/redhead2fd8.gif


HAHAHAHA D'Arcy beat me to it!
~C

[CybrGfx]

Paintovers can be like drunken blind dates...Even with the best of intents, you can sometimes wake up the next morning, look over, and go, "OMFG, what WAS I thinking?"

Today, the PO looks like Miley Cyrus...gah.
But, in my defense, I was trying NOT to drastically change the foundational image as posted. Rearranging, resizing, and reshaping the basic features too much "hijacks" a paintover, making it a "start over..."
So, the Irish miss has the "sparkle" of Life, as requested by stuh505, it's just a bit extreme, and does not compensate for the anatomical differences.

Speaking of which, stuh505, you now see some of the problems with your measuring system...
There is an old saying that applies to many disciplines, from cooking, to sewing, to carpentry:
"Measure twice, cut once."
The second measure need not be the same as the first...

With your match point measuring, you got quite close with your anatomy, although not close enough. Imagine how much MORE accurate you would have gotten, had you used a SIMPLE grid on your image? Until you have drawn and observed and studied enough human forms to train your eye to naturally see accurate proportions, you use tools to help you train your eye. Regardless what ANYONE EVER tells you, a grid is a tool like a hammer. You can use it to pound nails until your fist (or fingertip) is developed enough to do it kung-fu style...

Try using a grid along with your point grid next time. I bet you'll amaze yourself.
You show good patience to do portraits. I know that just from the time I spent on the Miley Cyrus paintover (about 85 minutes)...lol!

Look forward to seeing your next effort, or an update to this one.

~C

[stuh505]

Well...I don't want to cheat. I want to be able to say, "I drew this" not "I traced this." I mean, I could just set the true image underneath with some transparency and then draw on top and it would be perfectly accurate...but this feels wrong. I would be embarrassed to admit that's how I did it. So I want to do work that I can be proud of and learn from at the same time.

In previous female sketches (see profile) I have not used any form of measurement, and just eyeballed it. I usually end up making gross anatomical errors which aren't easy to spot until I start to finalize it with shading etc...which makes it a pain to fix, and I end up re-drawing most parts of it 5...6...10 times over.

Using the measurement dots this time helped a lot, however I still ended up re-drawing certain aspects many times. The mouth has gone through probably 15 iterations already :/

Anyway, I'm not sure about this grid structure. I've used it once before to do a charcoal drawing and the proportions came out well, however it seems pretty close to tracing. If you have a computer, and you only care about the end product not how you got there, then why wouldnt you just trace it?

But I DO care about how I get there, not just the end product...which is why I haven't overlaid the reference with my sketch. I want to be able to see the differences and correct it without cheating. Sometimes after getting frustrated for 15 hours of painting I will overlay it just to check and see how much I am off.

I don't know...I'm curious how other artists feel about this subject. Hopefully I'm not the only one who has trouble getting proportions exactly right. I would feel pretty stupid if everyone else could just freehand a sketch and have the proportions look perfect!

[CybrGfx]

Excuse me, I did NOT say the word "trace" anywhere in that post.

You DO know how to use a grid to transfer, don't you?
Maybe not...

Your comment about "cheating," is absolutely hilarious. What on earth do you call your ineffectual dot-match grid, son? That's "cheating," too, you know!
You already are showing the bad habits of "elitist" thinking and defeatest arguing.

You have already proven that your eye is NOT sufficiently developed to "eyeball it" when attempting to draw the human figure, no?

One does NOT learn ANY skill by just plowing into it, with the attitude that using TRAINING TOOLS is "cheating."
NO ONE says you have to use "cheats" (that is SO tunnel-visioned an attitude!!!) any longer than it takes you to LEARN how to accurately SEE and render believable images...Do you SERIOUSLY expect your learning works to be such "masterpieces" that they are somehow "sullied" if you make them look better by gridding, or even *shudder* tracing a photograph? Michaelangelo figured out the basic "cheating" proportions of the human body, and was a strong proponent of gridding out works. If you want to be thoroughly anal about the topic, even using photo refs is considered "cheating" by the rabid art students. NOTE: the keyword is "students." Any artist actually producing and selling their art will tell you that you use ALL the TOOLS available at your fingers to get the image you are striving for...The true artistic SKILL is in using them so well, no one would ever suspect that you montaged 5 photos and the background in Photoshop...

I'd rather have a viewer appreciate my artwork, regardless how it was achieved, rather than have them wince at the bad anatomy, just so I could brag, "I drew that myself, with no visual aids whatsoever..."

If you care so much about how you get there, you are missing 90% of the process of creativity by purposely impeding your development.

NO, everyone cannot get proportions correct WHEN THEY ARE STARTING OUT. That is why they "cheat," and use grids, and tracings, and LOTs and LOTS of practice. If you want other opinions, do a forum search, and break down and do a web search. There are numerous discussions about "cheating." And you'll find that the attitude is almost directly proportional to the skill level...These forums are for actual artworks, not ineffectual debates on the relative merits of using tools to help you learn...Hmmm....I wonder if books and tutorials could also fall under that categorization...

So you go ahead and learn the hard way by repeating mistakes and correcting them. IF you ever decide to actually learn proactively, rather than reactively, you now have a rough guideline to help you help yourself.

Best of success to you.
~C

[stuh505]

Settle down there, mate.

Now, as I stated, I have previously tried to not use any tools at all. This is not because I think I'm superior, but because I want to maximize my learning potential. I figure that given practice, we can learn to do exactly what we practiced doing. I think I'll learn faster if I practice everything 100% freehand than if I use tools. I also get more satisfaction out of it when it turns out good.

Now, it's already apparent that I relaxed and used a tool on this one as an experiment. For some reason you're saying I should have used the grid method, but that is really not any different from the control point method I was using -- both methods simply give you some more reference points reducing the amount of guesswork you have to do by eye. You could say that the grid method makes it easier, but that would imply that it required less free-handing, which is exactly what I'm trying to practice doing.

[strobel]

hm thats ok if you want to go the hard way. Of course one can learn from freehand drawing only. But any method will do the trick if you learn from it. For me, freehand only just won't work. Why? because i can draw all the wrong shapes i want, if i dont get the feeling for the right thing to do down pretty early, I'll get the hang on "how to draw wrong" which will longen the learning effect alot :argh:
but well...thats just my experience

If you dont like grid drawing:
I'd have to suggest you use the upside down method. I'm not good at drawing at all so i simply won't give my useless advice on the image itself. But i think i can provide another of the basic methods to train the eye.

You simply flip the photo upside down and then draw freehand. This forces you to focus on exact shapes as your brain wont simply correct drawing flaws at first glance.
From my expirience first time upside down pictures of human faces really look...well funny. (worse than any attempts not flipped) however it shows (me) I'm not capable of actually drawing the shapes i see but my brain usually trys to cover that up by simply suggesting a more believable (yet still wrong) shape. Which it does because its "build" to do it. our brains ar real masters in analysing faces. for recognition only. Drawing them must be trained.

Its not really hard to get a certain likeness in a drawing. thats because our brain simply covers all the flaws up (at first glance). Its more trying to go beyond a certain likeness to really drawing a portrait.

The method can be combined with grid drawing. (which actually isnt about giving you more reference points but to distract your view from "the whole" to "the real" as my teacher puts it)


if you want to learn face/skull atanomy, I believe portraits arent really the way to go for it. why bother with colors, details and likeness if you just want to train your sense for anatomy?
If you think you have the anatomy down, but want to do a portrait...why making the job harder than it is? I'd use a grid, lay out a simplyfied "dummy" to get the perspective and proportions right, then remove the grid and define the shape in greys. I'd then pic up colour and decide how I'd turn my portrait from "you could have done that with a cellphonecamera" in "wow thats art!" (IF I'd be any good at painting ...still got way to go though)


its your choice :)

[CGTalk Moderation]

This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.