Paintovers for posterity

[Velocity Kendall]

I thought Diles idea of a paintover thread was a very good one, and thought of all the good ones Id seen but could no longer find. Venger Javier P and Sone1 alone had done tonnes i wish I could look at again. and it seems a shame those things get lost in the pile.
So maybe if you liked one, put it in here, and I wont have to search all over the place!

heres some i remembered; i cant find vengers ones but ill get em...

Javier P for Pavel and theGiffman




Benjaminba for Dwilliams



venger for rob_curbishley

Don't paint over a greyscale - change it to a base tone (depending on your picture)
You'll still struggle with the cool feeling though (or I know I do)
The thing is to try not to detail everything in greyscale as you'll probably end up painting on top of it on a separate layer.
I've never conquered colouring greyscale well...

Kiera for thooloo



venger and Alice Herring for David A Ray





elwell for aether technician

[vineris]

JavierP did this one over one of mine:

[EZpaint]

Nice post Velocity! It's great to see how everyone interprets things

[Velocity Kendall]

Bette for Eldinga

[sone_one]

nice idea man, heres some of those i did (no particular order).





























(quite easy due to the new and improved attachment manager )

[Velocity Kendall]

this is in my top 5 ever, easy

[Hexokinase]

Can't have a paintover thread without gregpro now can we?

[krysjez]

damn! those gregpro ones are going on my hard drive

[Aceoko]

Don't know who this Gregpro is but he sure is helpful, really goes that extra step.

[LordLouis]

Fantastic thread velocity, I have waited for one like that for a long time- I have a folder with the best crits and p/o's that I could find, but unfortunately I don't know most of the artists. I'll dump them here and maybe people will be able to attribute them (sorry!!) I would like to give credit.

1 Andrew Sonea(?)


2 This was fantastic feedback to a fightscene!
3
4
5
6
7
8
9

10 I don't know who did the PO


11 This was in response to someone young just starting off as an artist, afaik.

12
13

14 This was a very interesting critique on a children's book illustration

15
15x

16 Don't know


17 Don't know



I believe that this is the same artist doing the PO that also did the fight-scene at the bar above.

19

Don't know

21

Don't know
22

Don't know
23

24 MrCorlan


25 JavierP!



This one if JeffX's feedback to a traditional painting.


This is a very instructive gif - 1,7Mb in size..

[Velocity Kendall]

epic goodness.
EZPaint for Elemile

[Pixelestial]

Here's a few of mine I had saved, I hope they can continue to be useful to someone!

[Andrew Sonea]

Wow great thread!!! Best thing I've seen on CA in a while. I'm also super happy that one of my images was posted here

I've done a number of paintovers over time for various people...there's something very fun about doing them and it's a great feeling knowing you can help people. I actually had a thread of my own dedicated to paintovers, but I kinda abandoned it suddenly because I was/am dealing with depression...maybe one day I'll get back to it.

Anyways, here's a few I found on my computer that I've done in the past (some of these are really old and done when I was more of a beginner, so may not be very helpful!). I feel bad posting these after seeing those Greg Pro ones haha! The one with the dude lying down I tried to remove from my post but it's still there...weird. Anyways, it jsut was explaining how his strokes were laid down vs what direction his strokes should have gone.

[johnny24]

WOW love this thread!

[Deadlyhazard]

Fantastic thread. MOAR.

[sone_one]

lovely.

heres some more.

[Andrew Sonea]

Those are all brilliant sone_one! Just wondering--do you have some Reilly training in you? Your figures remind me of it a fair bit, and you have an Orbik quote in your signature...

[vineris]

Here's a few I did.

[Bette]

Here are a few i did for people

[keeptime]

Here's one Swamp Thing did for me.

[Andrew Sonea]

Just did this for someone... stupid file size limit on CA meant I had to chop it into three parts though instead of just posting one big image. Anyways, I'm no Greg Pro, but I think this shows the idea of rhythms and simplification a bit.

[Velocity Kendall]

LaCans epic crit for Pavel Sokov

"In order to be a painter you need to perceive and "think" like a painter. You currently do not.
What you do now, with most of the images posted here, is some sort of incidental photo manipulation.
There's an obvious drive and a fair deal of innate affinity on your side, but your process is working against your ambition.

Your existent approach

Here's a breakdown of your methodology as I deduced it from this thread:

1) Get a problematic reference photo
You have a weird tendency to choose photos that are not effective as a reference. Either because they are weak images according to the rules of good visual design, or not informative enough on key aspects of your motif.
Fully frontal lit photos taken with flash do not split the form nicely into areas of light and shadow.
Diffusely lit photos taken on an overcast day tend to flatten the form. They leave you with a lot of super-subtle changes in values that are hard to interpret.

2) Manipulate this photo to make it even weaker
In case of last dog photo, you tried to invent directional studio lighting by blindly dodge-tooling existing specular reflections on a diffusely lit image. This is not how light works.

3) Copy the flimsy reference by chancily painting it in photoshop. The result is a poorly designed picture.

4) When something rings not right, identify the arbitrary problematic "detail" (it's always a detail, never the whole, eh?), then either trace or do a minute photomanips to band-aid it.

5) Repeat from step 4 ad nauseam. At each futile iteration ask people on forums is it any better now.
No, it isn't. It's just running in circles. The underlying pictorial logic is what's broken in most cases. Detail fiddling won't help. The only thing to do is to toss away the image and start over.

What you have here is a process commonly known as "turd polishing".

If you insist on this methodology, you're in a danger of never rising above ranks of the common turd polishers.


Here's my quick paintover (done with mouse):



Obviously, I'm sticking with the original reference (you'll see why later). Note how the paintover is concerned with altogether different set of problems than your painting is. Instead of worrying about "somberness" or "eyelids" or any such verbal concept, my focus is on getting the value plan and the proportions right. In order to do this successfully, I had to simplify how I see the subject. Excessive detail has been eliminated from consideration in order to get a clear view on "global" structure of the image. The key is to always look at the whole forest, not the trees. There is no eyelids, no pupils, no whiskers and no thousands of individual strands of hair there, it's sloppily painted, yet the image just "works". Why do you think that is?

Let me address some of the problems that can be detected in most of your paintings.


"Amplified" local changes

By "local" I mean "small parts of the picture". This is a very common pitfall. It happens due to relativistic nature of human perception. We unconsciously tend to maximize any perceived variation inside our current narrow focus. This is observable in a couple of areas:



You need to put conscious effort in to counter this tendency. Always look at your subject and your interpretation in a non-fragmented way. Beware of the fragmentation, seek unity. The more narrow-focused your seeing is - the weaker the painter you'll be.

Building the value structure using only narrowed focus is a lot like separately applying photoshop auto-levels on small chunks of the image. Try it on some of your photo reference and compare it to your paintings. To see how prevalent this problem is just google "value drawing". Most of the images that come up will have this sort of problem.




Lack of conception of global value plan

Always start with planning the simplified value structure of the whole image. I cannot emphasize this enough. It's the foundation which makes or breaks the image. This is what the block in phase is for. Again, beware of the said fragmented perception. Squint to simplify what you see. Reduce the amount of information you take in. The less information you have to chew on the more likely the painting will succeed.

This is, of course, easier said than done. We can theoreticize all day long but it would do your painting no good. You need to internalize it. It needs to click, and it can click only if you practice a lot with the specific intent of seeing the big value picture.

By throwing values all over the place, you can easily break the plausible notan of the image. Notan is a concept from traditional Japanese art. Maybe you heard of it. It's basically a word for good (or interesting) balance of blacks and whites. This again ties in with mentioned global value plan. Each image can be reduced to a 50-50 threshold black and white version, nicely exposing the basic value structure. Here I did so for your dog painting and your references:





Cumbersome detailing

Painting is not accounting. No need to count hair strands, or measure pupil diameters. Get the values and proportion right. Do so in as detail-reduced version of the image as you can. The image then will be in the state of unity. The gestalt will take care for the details. Viewer's perception will fill in for most that is omitted. Even if you insist on super-detailing, proceed to do so only after your value plan has achieved a solid state of unity. However, you'll quickly learn that good value composition doesn't need the level of detail you're currently obsessed with. It is redundant. You now use detailing as a crutch. It won't help in improving the design flaws.


Reacting to client's remarks

Clients can feel when image is not working. However, they will not know how to articulate the problem lying in the fundamental imagecrafting area. They'll typically blurt out that the problem lies in semantically most important part. Because that's what they can easily put into words. For portrait that's probably "the eyes". You need to be less literal when interpreting the feedback. Eyes are the least of your concerns in your dog painting.


Proposition

Here's my proposition to you, if I may. It's twofold:

1) Do value studies. For start, draw 20 studies, either from life or photo reference. Let them be simple compositions, preferably still lifes. Use traditional media. Take the bluntest possible drawing tool. I'd suggest wine charcoal sticks, plain printer paper, paper towel and kneaded eraser. Charcoal lets you manipulate large areas of value quickly - both ways - darkening and lightening. It won't, however, let you habitually fiddle on details, thus forcing you to concentrate on the big picture. If you feel you don't have enough control, don't worry. Just persist. It's a sensitive yet powerful medium that needs some delicacy in handling. You'll get a grip after couple of studies.
Spend less than half an hour on each study. Possibly even less. After that time call it done, regardless of the results, and move on. See what happens after you do 20. You can post them here for revision.

2) Know your art history. It's important to know what came before you. Get this book to sharpen your tastes. Study the reproductions and read some info on the historical context. If you have time, enroll in basic online course on western art history. It'll put some things into perspective.


Hope I wasn't too blunt. It was meant as a well intentioned kick in the butt.

Cheers!"

[Black Spot]

I'm so glad that I approved that post to get it added here. This thread is epic.

[sone_one]

Those are all brilliant sone_one! Just wondering--do you have some Reilly training in you? Your figures remind me of it a fair bit, and you have an Orbik quote in your signature...

unfortunately i have no training in me... whatsoever, but i pay alot of attention to the teachings of reilly and what has grown out of it. i consider myself very lucky, ive been around when ron lemen had that portrait mentoring thread going e.g.
that quote i picked up on a blog of a orbik student, but i forgot which.

and yeah thanks man, youre doing really well yourself .

[Marlo]

posting here so i can always come back for reference, thank you all!

[LordLouis]

posting here so i can always come back for reference, thank you all!

You can subscribe without posting. At the top of the thread, click on "Thread tools" and select "subscribe to this thread" in the drop-down menu.

[LaCan]

I'm so glad that I approved that post to get it added here. This thread is epic.

Thanks! To be honest, I was not certain the post would get through. Good to see people find it useful.

[Mikael Eson]

LaCan's post felt like something of a revelation for me as well so I give my thanks

[dierat]

Awesome thread! This shit is super inspiring

Here's one Venger did for me that I personally found very helpful~

You need to decide on your primary focus (what do you want us to look at?)
At the moment you're somewhere in between the window/wall and the figure.
If you drop this to value - things aren't clear enough.
You need to be (IMO) a bit more extreme, light over dark, or dark over light are a pretty simple base techniques - you can vary this to fit the mood you are after.
Be careful with your background elements - the wing behind the head and the 'smoke'? are interfering with the read of the figure.

Quick scribble for you that isn't 100% right, but hopefully you can see what I'm getting at:

[Quike Garcia]

Thanks! To be honest, I was not certain the post would get through. Good to see people find it useful.

Thanks to you!! This is an amazing piece of advice