You should be proud, and pleased with the amount of improvement you made with this. I know the new one probably took you half as long as the old one, but it really sells your stuff and has a much cleaner, and more professional feel. Your status as an artist, from the standpoint of any onlookers visiting your website, just doubled - easy.
Big improvement, good that you actually took the plunge and changed the design around this much.
What Rawrsie points out is correct I think, I was wondering why your thumbs don't seem to pop off the screen like they should. Might want to adjust contrast or something, and definitely apply and Unsharp Mask filter to your thumbs, brings them out way more.
Anyway, I just read in your opening post you'd like more art critique, just some thoughts I have:
Personally, i think you could push the renderstyle/presentation further. I read you're using Max, so I guess you're using Scanline or perhaps Mental Ray to render your stuff. That results in having them look a bit bland. Speculars are mostly missing or bland and uniform, contrast in lighting and light colors could be stronger. Just today I had to help someone out doing renders in Max, and after having spent half a year doing realtime rendering only, I found it so cumbersome and tedious to get remotely good looking results (without Vray that is, ahem).
What I'm trying to say is, learning to use some realtime 3D presentation tool could help you, not only is it more relevant for a game artist, I personally think it's way easier to get good looking results that don't scream "standard scanline render" right away.
There's Marmoset, my shader for Max and the recent free UDK, great tools that are all free and not hard to pick up.
Especially your lab assets I feel like you should just create a few more assets (railings, floor, desks in the back and floors and ceiling) and just put that in a little UDK scene. It could look amazing with Lightmass! I see you know Unreal already, so shouldn't be too hard to get up to speed with UDK.
Ive been following this post for a while and your website has come out nice, congrats on your good portfolio review. I'm going to have to second this post by Kevin though.
I don't know if it has come up or not but your poly counts are 2 - 3 times too heavy. I did read a little of the thread and saw the point made about your teacher telling you to do some things a certain way, so if he was the one that told you to follow these polygon restrictions he was wrong to do that.
Your Tri counts on your work seem a bit high for game assets. Just looking through your work it seems like you put a lot of your tri's into smaller details especially cylindrical shapes. You could definitely pull off some of these pieces with much lower tri counts without sacrificing too much detail. Specifically on your lab scene (just cause your still working on it) the center station with the robotic arm has a ton of tri's being spent on the upper mechanical workings of the arm which might not be seen closely and it wouldn't hurt to use your tri's sparingly in areas like this. It looks like there are a lot of details like indents and various shapes that don't affect the silhouette that could be captured in your texture maps as well. Just experiment with lower tri count shapes and see if they work for reducing the final in game model tri count.
Also rawrise mentioned edge highlighting in your textures I think it would really help in your color and specular maps. Just some things to keep in mind as you continue to work on your portfolio.
It looks like you're headed in the right direction keep it up and best of luck on your job search.
+1 on the poly counts being too high for game assets. Efficiency is important in game assets, especially environment pieces which often need to hold up with a greater scale and lower texture resolution than a character. A good basic check is to ask, "do these extra faces add to the silhouette in an important way?" if not, find a way to put that detail in the normal map.
There is also a good bit of redundancy in the textures. If two pieces of identical geometry have separate texture spaces that look close to identical in all maps its a red flag that the unwrap could have been more efficient. The more space you can reuse, the more pixels you have for fine detail in the diffuse and normal maps.
The biggest problem is that these same issues crop up in several pieces. The work looks good, but it needs some optimization.
As for the design, the new one is WAY better than the old one.
But theres a couple of things you could change to make it even better:
-Make the pictures take up more screenspace
-Darken the background for contrast
-Lose the drop shadow, replace with a 1px black border
-Move the whole thing a nudge down, 30 px or so.
bbob - thanks for taking the time to make this, but I think these changes are more of a matter of personal taste, and I can't really see how it would improve the function of the site. That is more of the critique I am looking for now.
bbob - thanks for taking the time to make this, but I think these changes are more of a matter of personal taste, and I can't really see how it would improve the function of the site. That is more of the critique I am looking for now.
Do you want the job or not?!!
:P jk. Keep it up I dont have any crits more than on your work because i suck at homepages, i just know enough to know that yours got everything working. There is a lot of stuff that could be done for you uv mapping and reusing bits of texture, for example stuff that you got two similar things but dont see at the same time, use the same uv space. the wooden cart got the different sides of rolled up mattrasess (4 of them) and the right and left side of the chest could use the same space and no one would notice, only that you would have more space to do more stuff in the uv or enlarging the whole thing. planks of wood...
Sure this is a little more of planning into it so that you can bake and stuff. But in game development its strict limitations (especially on textures).
Basically when you can see duplicates of stuff in your texture its a waste, if you got duplicates make sure you make them different so at least it has a purpose having stuff on the uv, somethings you cant mirror like text... ah ur smart and you get what im trying to say i hope.
:P jk. Keep it up I dont have any crits more than on your work because i suck at homepages, i just know enough to know that yours got everything working. There is a lot of stuff that could be done for you uv mapping and reusing bits of texture, for example stuff that you got two similar things but dont see at the same time, use the same uv space. the wooden cart got the different sides of rolled up mattrasess (4 of them) and the right and left side of the chest could use the same space and no one would notice, only that you would have more space to do more stuff in the uv or enlarging the whole thing. planks of wood...
Sure this is a little more of planning into it so that you can bake and stuff. But in game development its strict limitations (especially on textures).
Basically when you can see duplicates of stuff in your texture its a waste, if you got duplicates make sure you make them different so at least it has a purpose having stuff on the uv, somethings you cant mirror like text... ah ur smart and you get what im trying to say i hope.
Good Luck!
Thanks! and thanks for accepting that my site works and leaving it at that! I see what you mean about copying UVs this concept was fairly new to me when I was making this model, I did copy the UVs for half of the cart, but I guess I didn't take it far enough to do the blankets and the chest that way too. I have learned a lot since then and I definitely try and copy as much as I can now.
To be fair the only thing I don't like about it right now is that it looks too bare.
As you add more work to it I'm sure it'll look a lot better.
Really like the font on the front page and I love the saw.
Actually my crit were a matter of function as the higher contrast brings out your thumbnails, the bigger thumbnails are so they are easier to see fast, and the lowering of the content is to bring your work into a compositional butterzone on most monitors.
The drop shadow crit were a matter of taste though, but it takes focus away from your work as well.
Have to add another vote for reducing the polycount. The lab scene in particular is way too polyheavy for what you have there. It looks great, but a lot of the detail could be achieved with a normalmap and fewer triangles. The render of the lowpoly asset looks like a highpoly render Also, the robot arm is posed differently in render vs asset image. If you can, have it posed like it is in the asset image since it makes it look a bit more interesting, less straight and angular. A bit more... alive I guess you could say.
Did a quick scribble to show what I mean regarding polycounts (this is by no means set in stone and people will probably have better suggestions but I'm bored at work).
I think with a few more props and more interseting lighting + composition, it would look amazing. As it is now, your prop renders show off the assets better than the beauty render. Someone mentioned putting it in the UDK and I think that is a very good idea.
Texture wise they are a little flat, as is the portable com-briefcase. Looking at the sheets, it feels like flat colour + ambient occlusion and not much more than that. I don't mean to be mean, just letting you know what I see.
The old wooden cart shows off your skills the best at the moment, it looks really good. Texture looks great, the poly usage isn't too over the top (although it could probably be optimized a little bit ), it's a nice and polished piece.
As for bbobs suggestion on the site, a darker background would be good I think. Like he says, it will make your images stand out more and draw the eye. Most art-related sites have a dark background for precisely that reason. And if you take out the dropshadow, you don't have to redo the images everytime you change the background colour
On your resume, Corel Painter is spelled with one L. And you could put that you were the level designer on the Unreal project so whoever reads it will know what you did. Just to clarify your role a bit.
I also had a look outside your website on the links you have and only really found one thing that had me confused. On your LinkedIn page you mention. "IGDN meetings" Haven't heard of IGDN, do you mean IGDA?
urgaffel - wow you really found everything didn't you! I already know about most of these things, and given the time I will fix them, but I managed to get an art test anyway so I'm working on that at the moment, I will take all your advice and put it towards making that look good. Maybe I'll have it up in my port after its done for everyone to see it, once I make sure its ok.
And yeah I guess I did mean IGDA oops.
(oh and I'm not changing my website ANY more, can't you people ever be happy??)
urgaffel - wow you really found everything didn't you! I already know about most of these things, and given the time I will fix them, but I managed to get an art test anyway so I'm working on that at the moment, I will take all your advice and put it towards making that look good. Maybe I'll have it up in my port after its done for everyone to see it, once I make sure its ok.
And yeah I guess I did mean IGDA oops.
(oh and I'm not changing my website ANY more, can't you people ever be happy??)
We're artists, we're never completely happy (neither should you be, there's always room for improvement )
Good luck with the art test! I guess you don't want to say which company it is until it's finished one way or another but do let us know how it goes!
Regarding the lab, I think you could easily optimise the polycount without too much hassle and it would make a very cool piece with a few added bits and better lighting so I say go for it if you have the will and time to do it :poly142:
urgaffel - wow you really found everything didn't you! I already know about most of these things, and given the time I will fix them, but I managed to get an art test anyway so I'm working on that at the moment, I will take all your advice and put it towards making that look good. Maybe I'll have it up in my port after its done for everyone to see it, once I make sure its ok.
And yeah I guess I did mean IGDA oops.
(oh and I'm not changing my website ANY more, can't you people ever be happy??)
I think your portfolio structure is fine since you have just the few pieces right now. Later you might want to divide it into categories (and what those are will depend on the eventual content you create).
In my experience, it's not all that important to some to see environment props rendered in a game engine, although that is a nice touch. A good art director can spot a good quality prop from a bad one, no matter what the background you present it against.
What would be even more important is to reduce the polycounts on your models by a lot. If you are building models for a realtime 3D engine on a PC game, lots of tris/faces are an expensive commodity. A really desirable environment artist is one who can find that balance where the object almost has too few faces, yet still looks convincing to the player (who is only barely aware of it as a piece of the background if it's an exciting game, to tell the truth. It's not sculpture in an art museum to be studied closely). Distance from the camera and size on screen are going to dictate what you can get away with. Also, extremely good textures and adept lighting will make up for the realism and detail you lose when you strip a model way down to the barest essential faces needed to describe its form. This is one of the main reasons hiring managers are interested to see you list the face count for models in your portfolio.
You may even want to show three versions of a model at varous polycount levels, to show that you could create LODs (a.k.a. levels of detail), which is a task often assigned to entry-level artists. I haven't been on this forum before to know if folks know about LODs, so forgive me if this following explanation is superfluous. Basically, they're stripped down versions of an object that get drawn in the game engine to replace the object as it gets further from the camera, and doesn't need to be so detailed. LODs are made for trees, characters, buildings, etc. Their function is to conserve memory.
Replies
You should be proud, and pleased with the amount of improvement you made with this. I know the new one probably took you half as long as the old one, but it really sells your stuff and has a much cleaner, and more professional feel. Your status as an artist, from the standpoint of any onlookers visiting your website, just doubled - easy.
Many cheers and good job.
What Rawrsie points out is correct I think, I was wondering why your thumbs don't seem to pop off the screen like they should. Might want to adjust contrast or something, and definitely apply and Unsharp Mask filter to your thumbs, brings them out way more.
Anyway, I just read in your opening post you'd like more art critique, just some thoughts I have:
Personally, i think you could push the renderstyle/presentation further. I read you're using Max, so I guess you're using Scanline or perhaps Mental Ray to render your stuff. That results in having them look a bit bland. Speculars are mostly missing or bland and uniform, contrast in lighting and light colors could be stronger. Just today I had to help someone out doing renders in Max, and after having spent half a year doing realtime rendering only, I found it so cumbersome and tedious to get remotely good looking results (without Vray that is, ahem).
What I'm trying to say is, learning to use some realtime 3D presentation tool could help you, not only is it more relevant for a game artist, I personally think it's way easier to get good looking results that don't scream "standard scanline render" right away.
There's Marmoset, my shader for Max and the recent free UDK, great tools that are all free and not hard to pick up.
Especially your lab assets I feel like you should just create a few more assets (railings, floor, desks in the back and floors and ceiling) and just put that in a little UDK scene. It could look amazing with Lightmass! I see you know Unreal already, so shouldn't be too hard to get up to speed with UDK.
Anyway if you wanna get learning with this stuff, my shader can be found here: http://laurenscorijn.com/viewportshader. There's an article and a video on using it here: http://www.laurenscorijn.com/articles
Thanks a lot for the crit, i'll get going look at that stuff right away thanks!
(btw your site was my main inspiration when creating my new design)
Your Tri counts on your work seem a bit high for game assets. Just looking through your work it seems like you put a lot of your tri's into smaller details especially cylindrical shapes. You could definitely pull off some of these pieces with much lower tri counts without sacrificing too much detail. Specifically on your lab scene (just cause your still working on it) the center station with the robotic arm has a ton of tri's being spent on the upper mechanical workings of the arm which might not be seen closely and it wouldn't hurt to use your tri's sparingly in areas like this. It looks like there are a lot of details like indents and various shapes that don't affect the silhouette that could be captured in your texture maps as well. Just experiment with lower tri count shapes and see if they work for reducing the final in game model tri count.
Also rawrise mentioned edge highlighting in your textures I think it would really help in your color and specular maps. Just some things to keep in mind as you continue to work on your portfolio.
It looks like you're headed in the right direction keep it up and best of luck on your job search.
There is also a good bit of redundancy in the textures. If two pieces of identical geometry have separate texture spaces that look close to identical in all maps its a red flag that the unwrap could have been more efficient. The more space you can reuse, the more pixels you have for fine detail in the diffuse and normal maps.
The biggest problem is that these same issues crop up in several pieces. The work looks good, but it needs some optimization.
But theres a couple of things you could change to make it even better:
-Make the pictures take up more screenspace
-Darken the background for contrast
-Lose the drop shadow, replace with a 1px black border
-Move the whole thing a nudge down, 30 px or so.
a bit like this:
Do you want the job or not?!!
:P jk. Keep it up I dont have any crits more than on your work because i suck at homepages, i just know enough to know that yours got everything working. There is a lot of stuff that could be done for you uv mapping and reusing bits of texture, for example stuff that you got two similar things but dont see at the same time, use the same uv space. the wooden cart got the different sides of rolled up mattrasess (4 of them) and the right and left side of the chest could use the same space and no one would notice, only that you would have more space to do more stuff in the uv or enlarging the whole thing. planks of wood...
Sure this is a little more of planning into it so that you can bake and stuff. But in game development its strict limitations (especially on textures).
Basically when you can see duplicates of stuff in your texture its a waste, if you got duplicates make sure you make them different so at least it has a purpose having stuff on the uv, somethings you cant mirror like text... ah ur smart and you get what im trying to say i hope.
Good Luck!
Thanks! and thanks for accepting that my site works and leaving it at that! I see what you mean about copying UVs this concept was fairly new to me when I was making this model, I did copy the UVs for half of the cart, but I guess I didn't take it far enough to do the blankets and the chest that way too. I have learned a lot since then and I definitely try and copy as much as I can now.
As you add more work to it I'm sure it'll look a lot better.
Really like the font on the front page and I love the saw.
The drop shadow crit were a matter of taste though, but it takes focus away from your work as well.
Did a quick scribble to show what I mean regarding polycounts (this is by no means set in stone and people will probably have better suggestions but I'm bored at work).
I think with a few more props and more interseting lighting + composition, it would look amazing. As it is now, your prop renders show off the assets better than the beauty render. Someone mentioned putting it in the UDK and I think that is a very good idea.
Texture wise they are a little flat, as is the portable com-briefcase. Looking at the sheets, it feels like flat colour + ambient occlusion and not much more than that. I don't mean to be mean, just letting you know what I see.
The old wooden cart shows off your skills the best at the moment, it looks really good. Texture looks great, the poly usage isn't too over the top (although it could probably be optimized a little bit
As for bbobs suggestion on the site, a darker background would be good I think. Like he says, it will make your images stand out more and draw the eye. Most art-related sites have a dark background for precisely that reason. And if you take out the dropshadow, you don't have to redo the images everytime you change the background colour
On your resume, Corel Painter is spelled with one L. And you could put that you were the level designer on the Unreal project so whoever reads it will know what you did. Just to clarify your role a bit.
I also had a look outside your website on the links you have and only really found one thing that had me confused. On your LinkedIn page you mention. "IGDN meetings" Haven't heard of IGDN, do you mean IGDA?
And yeah I guess I did mean IGDA oops.
(oh and I'm not changing my website ANY more, can't you people ever be happy??)
And of course you can do whatever you want, but since you asked on how to improve the website, the crits were given :P
New website is much better, very clean.
We're artists, we're never completely happy (neither should you be, there's always room for improvement
Good luck with the art test! I guess you don't want to say which company it is until it's finished one way or another but do let us know how it goes!
Regarding the lab, I think you could easily optimise the polycount without too much hassle and it would make a very cool piece with a few added bits and better lighting so I say go for it if you have the will and time to do it :poly142:
Escalation Studios, its an iphone/wii developer here in Dallas, but the test is fun so I am excited!
Lee3Dee gave me the tip off, so I guess it pays to play here on polycount
I think your portfolio structure is fine since you have just the few pieces right now. Later you might want to divide it into categories (and what those are will depend on the eventual content you create).
In my experience, it's not all that important to some to see environment props rendered in a game engine, although that is a nice touch. A good art director can spot a good quality prop from a bad one, no matter what the background you present it against.
What would be even more important is to reduce the polycounts on your models by a lot. If you are building models for a realtime 3D engine on a PC game, lots of tris/faces are an expensive commodity. A really desirable environment artist is one who can find that balance where the object almost has too few faces, yet still looks convincing to the player (who is only barely aware of it as a piece of the background if it's an exciting game, to tell the truth. It's not sculpture in an art museum to be studied closely). Distance from the camera and size on screen are going to dictate what you can get away with. Also, extremely good textures and adept lighting will make up for the realism and detail you lose when you strip a model way down to the barest essential faces needed to describe its form. This is one of the main reasons hiring managers are interested to see you list the face count for models in your portfolio.
You may even want to show three versions of a model at varous polycount levels, to show that you could create LODs (a.k.a. levels of detail), which is a task often assigned to entry-level artists. I haven't been on this forum before to know if folks know about LODs, so forgive me if this following explanation is superfluous. Basically, they're stripped down versions of an object that get drawn in the game engine to replace the object as it gets further from the camera, and doesn't need to be so detailed. LODs are made for trees, characters, buildings, etc. Their function is to conserve memory.
Good luck, Tricia
Sincerely,
Beverly Garland