Apologies if this is a very strange question, but I haven’t seen it before. Here goes: what is the baseline speed a competent graphics artist use to make a high poly 3d humanoid model with a full set of textures?
What I’m trying to get at is what I need to aim for if I’m going to be considered for employment as a graphics artist. Not that I’m looking for work anytime soon, but I would like to achieve a professional level. So, how fast would I need to produce if you lot were out looking for a modeler. I guess that’s what I’m trying to gauge here.
Like, right now, I’m working on a 5 million polygon model for my own demo game, and I think it’s going to take about two or maybe even three weeks to finish. And that sounds like a very long time to me. I realise that I’m very ‘new’ as a graphic artist, so that has something to do with it. I go slow for now.
No, 2 or 3 hours PLUS texturing
Baking shouldn’t take too long (what, 2 minutes?), retopo… maybe.
But the Initial sculpt should be possible in a few hours. I can’t do it either, but its possible.
Carpe Denius has unrealistically high expectations. No one can competently make a humanoid from scratch in 3 hours. This sounds like advice from someone who hasn’t actually done it.
2 to 3 weeks is reasonable for a well done character. I wouldn’t worry about speed that much when you are starting out. First, learn how to make it good, then learn how to make it fast. Nobody is going to care that you can make a model in 3 hours if it sucks, while people will wait for good quality.
EDIT: Overall you should go to somewhere like polycount for this question, as overall a lot of people here aren’t game artists. You’ll get a much more professional opinion there.
And you sound like somebody who has never paid for it.
3 weeks full time for a single character to me sounds about right for a novice. I would expect that to include multiple revisions and have a very low hourly rate.
I am not an artist but I have learned the modelling and rigging process on DigitalTutors (so that I wouldn’t be so ignorant with dealing with them). The UV and texturing would double that time (at least, because that shit is fucking hard!). If you are sculpting and then retopologizing to lower poly add some extra time. I’d say a week max is a very reasonable compromise and what I would expect from a highly paid professional.
I can agree with that, but this doesn’t factor in everything. What if you want multiple revisions? What is the target for the atwork / the style of the artwork (is it for mobile? is it stylized or realistic?) What is the level of detail for this character? (Is it a main character shown up close in cutscenes or is it a minor npc or enemy?)
Keeping in mind all of these factors, a humanoid character really can take anywhere from a day to a month. 2 to 3 weeks was kind of just the median in my head, although I can perfectly understand why you would object to that in a lot of circumstances.
@Denius: That is called speed sculpting. It’s excellent practice, but for the most part you don’t have that level of fluidity in a real production environment, nor will anyone think that level of detail is acceptable as a final target (usually speedsculpts are either a part of a character or only start to get into secondary forms).
Focus on quality first and speed will come eventually. Nobody will be impressed with your 3-hour model if it looks bad.
A week for fully textured non-hero character is a good rough estimate. But depending on quality bar, number of revisions, art style and complexity it can vary. It’s not unusual for some of hero characters of AAA games to take months since they go through lots and lots of revisions and scrutiny.
Lots of artists make art dump threads on polycount and zbrushcentral when they wrap up work on their projects. I suggest going through them to learn more about their workflows and timeframes.
Yep, agreed. A “character” who is the main character in Uncharted is different than a character who is the 15th zombie variety in a shoot-em-up. I could see one month for a really important and detailed character. It can certainly vary!
Thanks guys. I started with Unity as a coder, and have just recently slipped into the art side. I should have gone into it from the start because I love it in a way that I never loved coding. It even made me buy ZBrush, something my partner expressed strong and firm views about let me tell you. But I want to do this. I’m an arty geek anyway, always have been. It’s why I study English at uni, and not compsci. ^^
You can see those iterations sometimes. Naughty Dogs are very open about their tec and for ubisofts (technological) disaster watch dogs you can see that the main character looks different in several trailers.
Speed will come with practise. As others said. Just like with programming - make sure you understand the basics and the tech behind whht you are doing correctly. Also: if you want to do retopology right - buy an extra program for it. Zbrush’s topology tools are either rather cumbersome or don’t give you abcolutely accurate control. Personally I love Topogun. It’s rather affordable, has a very focused workflow and is pretty easy to learn. modo apparently also has pretty good tools if you are into it (I stick to TG ) but it’s overkill to buy just for retopology. Same goes for 3D Coat.
Otherwise your question and the aswers given are still somewhat diffuse. “A character model” can still differ vastly in requirements and necessary detail. For a mobile game you won’t go into as much detail as you would do for a next-gen title. And there you already have a difference of up to a few days.
Don’t beat yourself up if others seem to be faster than they are. In the beginning you will be slower than you would like to be. You just need to keep on going.
Book recommendations for Character modeling in ZBrush:
ZBrush Digital Sculpting Human Anatomy by Scott Spencer. Excellent book! Buy this!
Realistic Game Characters by Ryan Kingslien - generally excellent as well but includes a few odd choices to keep the workflow as much inside ZBrush as possible. Still very much recommended.
Also if you want to go into character design: learn humanoid anatomy. Seriously. Do this. And use reference. Always use reference.
I asked a very well know 3D modeling guy (not in the gaming realm of the world but in the DAZ / Poser high poly world) for a very specific caricature of a famous television person dressed up as one of their famous characters from their old variety show, something you couldn’t fake, you had to know how to model, including one set of clothing and textures…and the model was exclusive for my use…
…and they came back with 40 hours give or take a couple of hours and this was without rigging. Price was $1250 - $1500 USD. So not as expensive as you think give they were self employed and their taxes, retirement, and unemployment insurance must come of that. Living in a rural area and possibly the tax burden of his particular country made that good money. I think he is making a lot of money off old products and now does it because he enjoys it.
Since it was a caricature they couldn’t use photographs to ‘model this person’s face’ and this person had already had many caricatures done of her and one always knew those caricatures were her.
So I consider that person top of the line and his estimates would be a good guide for you to use.
As jaybennett said, this professional said UVs and texturing especially were a big hassle and took as much time as the mesh to complete.
+++++
example 2)
I paid a guy cheap money to create a caricature of a nobody in zbrush to judge ‘artistic ability’. aVery impressive he got it right and in an hour or two and textured but just a bust that really wasn’t usable in a game situation without retopo and a serious reduction in polygons. He does it because he enjoys it and I’m sure as an outlet from what he’s tasked to do from his 9 - 5 job. He works Indie too and his games models are $650 - $850 a piece depending on how much work he does on them. As these are low poly game models me guess would be he starts with templates in various states of completion. An experienced modeler can do that.
+++++
example 3)
Complete amateur using Blender (myself) spent 3 arduous days one long weekend (~24 hours total) following a tutorial set to build a less than 3000 polygon model to look like a famous comic strip character. The UVs worked fine (thank goodness) but for texturing resorted to primary colors and no patterns and to rig I used Mixamo to rig it. Now I’d have to rig myself. So if you extrapolate the time I spent above to do a really nice model you’re talking 10 - 15 days of 4 - 6 hours a day. That’s 40 / 45 - 60 / 90 hours to do one good model. When you think of it as a rank beginner that’s not a bad thing to be able to do. You can look to Blendswap.org for other models from amateurs (well some seem very, very professional).
I’d learn to use a Wacom Bamboo and Photoshop and/or SketchBook Pro though for the sake of helping create textures (I just can’t use a mouse) as maybe (maybe because I haven’t install and did tuts yet) Algorithmic Substance Designer.
It is rather unrealistic if you aim for a higher level of detail. Not absolutely impossible but certainly not the norm.
Could you please link to some of the videos you are referring to?
I will post them if I find them. I don’t know where it was, I think cgsociety, but somewhere were several videos of sculpting of characters. I stopped searching for them when I got unity, because I made a meshblending script (which I have now replaced by uma) and lost the need to learn sculpting for the moment.
Not all of them were speedsculpts, but you could really see that you gain speed with enough practice.
There is one video of a skull-like character, in the time he had a very good detailed model I couldn’t get the blob to look like something other than a piece of cancer-ridden play-doh.
Before you think I answered just for the sake of posting: my art is limited (heck, I can’t even write something readable with a pen), but I’ve seen several houndreds of making ofs (not only characters, though), so I thought my answer wouldn’t hurt. I didn’t want to scare him away, I wanted to show the other side of the scale, something to look forward to.