The front of his feet seems to be very straight; all the toes are perfectly in line. Feet, as far as I know, generally aren't quite like that. Although it could be the angle making them look more square than they are.
What's really getting me is, what's up with his skin?
The front of his feet seems to be very straight; all the toes are perfectly in line. Feet, as far as I know, generally aren't quite like that. Although it could be the angle making them look more square than they are.
What's really getting me is, what's up with his skin?
You seem to have made him rather pale.
Thanks for the feedback! I could adjust the toes so they won't be as aligned, though I'd be covering them with shoes in the end. For the skin I'll have to tweak the lighting and SSS values with the HDR I'm using. I'll do a check when I have the final lighting setup.
There really was significant advantage in using current technical solutions to tackle anatomy. While I did have to resculpt the anatomy from reference to allow for better mesh alingment with the transfer mesh, the amount of time it took to create this was 1/10th what it would have been, had I considered a manual method to creating low resolution topology.
I still don't really know how many companies use manual retopology for body meshes, since the method I'm using can apply to any mesh regardless of its size or exaggeration and I'm also able to blend detail and form as I wish with full control over tweaking in the end through morphs.
I do wish more artists would be open about the workflows used in studios and apply that to their portfolio work to practice what they would do professionally on the job. As part of my game development program, every modeling student had to make an écorché as part of their anatomy studies, but because the majority of the jobs didn't provide any opportunity to touch anatomy, the knowledge gained from the task wasn't utilized.
The quality of the work was variable and many students were hired contrary to the incredible standard we place on candidates artwork that seemingly justifies rejections.
I'm not entirely certain about including écorché or anatomy studies for character artist portfolios, but I do want to try a few pose studies at game resolution once these characters are fully rigged.
I was in the animation stream, so I didn't have to do one, but I did have to do anatomical drawings (form and functional) in dental school with a focus on facial anatomy and cadaver dissection.
All the same I would like to see an artists anatomy skill being used in a studio, but I do see the general direction being to find more tech solutions to automate the process, so where the skill would apply is in managing those solutions.
I feel that existing tech solutions must absolutely be considered before considering any artists anatomy skill as relevant to the job.
I will not go over all the reasoning for the adoption of the "EA Sports-like" set of tools as some sort of exercise, or even as an admittedly useful shortcut to reduce some tedious aspects of model creation. That is because even though tech knowledge never hurts and can be indeed be very useful ... it is a moot point in regards to the end result of a portfolio art piece. Put differently : there's is nothing wrong with the use of a given toolset (as it is no different from someone using a certain software instead of a certain other for retopo or sculpting) ; but it is largely irrelevant in the context of an art porfolio piece.
Now IMHO the actual problem with this model (both the earlier version and the most recent one) is that one vital step seems to have been skipped : the clear establishment of the final visual target that it is supposed to live up to.
One common way to establish such a target is indeed to simply use the art style and output of the studio one wants to apply to . The point of it is not to "flex" or "to be rockstar" ; at the end of the day the point is simply to have a target to aim for, and being able to demonstrate ones ability to hit it. Erring/improvising along the way (outside of perhaps some initial style-bending research) is always detrimental.
In a team environment, this target is provided to you by the AD and your lead, passing over the work of a concept artist/designer. For games relying on real-world athletes this target could come as a set of pictures, body stats and team jersey graphics and some head scans to cleanup and process. This is pretty much what you did for your solid Khan model, by precisely referencing the actor and the movie poster. Yet no one would accuse you of "flexing to look like a rockstar" for doing so. You simply had a clear game plan.
For stylized games, the target is established by a package of drawings, paintings, expression sheets, even clay maquettes if needed. And this package tends to get refined as a stylized game project progresses, since the most successful models striking just the right look and feel end up becoming part of the visual bible of the project helping to refine the art direction gradually. This process can take months/years, and ideally but not always it is supposed to be wrapped up by the end of preproduction.
Now the apparent problem I see with this KillLaKill fan model is that you do not seem to have such a visual target established beforehand (and as said above, the tech pipeline is pretty much irrelevant to that). Yet a visual target would dictate pretty much 100% of the artistic work to be put in the creation of the model (the sculpting of the face, the choice of visual representation for hair as either "clay" or "strands", the level of stylization of the anatomy, and so on).
Of course improvising as one goes without a clear goal can have value during a research phase, or as a long-running study. But improvising *while* also juggling with the incredibly complex set of requirements of game art is IMHO a strategic error leading to weak results.
All that said, part of your target is known : it is that KillLaKill character. But the moment you decide to "give it your own spin", you have no other choice than taking the time to establish what this spin actually *is* visually. "Just make it look good" is not enough of a target.
The obvious irony is IMHO the fact that establishing such visual targets is *not* the job of a character modeler - it is the job of the art director and the concept artists, with extra input from the art leads. There is of course nothing wrong with integrating this step as part of a portfolio piece (like with the earlier mentioned Cammy and Darth Maul pieces, which did that perfectly well), but this can be very risky because it can immediately becomes the weakest link of a whole piece. Which is why, especially for people with less experience, it is more advisable to aim for the existing output of the studio they apply to as their specific target instead of inventing their own. It is not "trying to be a rockstar", it is simply showing humility.
After all being able to aim for this clear target and reaching it consistently is *exactly* what is expected of a skilled character modeler on the job.
I did use a concept for the overall look of the 2017 model, though I hadn't establish a final look of what my interpretation would look like.
I wasn't meaning to match the concept, atleast not exactly. My shading, presentation and lighting skills were definitely not at their best, and looking back I prefered the model in its zbrush render, than the marmoset result which I certainly could I have done a better job with.
I was quite flexible with the piece, and atleast after having the portfolio evaluated the detriments were with regards to my presentation, lighting and shading. As I was being considered for an associate position at EA the only visual target I was expected to reach to the best of my ability was within the art test. For my portfolio it was understood that I simply did not have anyone to direct my work, so it was assessed on technical merits necessary to qualify for the art test.
I do feel that if I were to be in the running for an associate positon again, it be better to build a character as part of a mentorship program so there is a visual target I can match that someone with more experience can provide me.
Since I am applying for a second engagement, my objective is not just the final result but also the process which atleast for EA does prepare me better for exactly what the job requires, which is versitality with using multiple tools and a lot of back and forth and reworking of a model in an efficient way.
This approach of upgrading older work, is helping me research and implement workflows, and atleast I think the result is significantly more presentable.
I do understand the importance of a visual target, not having one does make it challenging to arrive at a final result. I hope that not having a visual target doesn't work against my application for this piece, but I can see it as a acceptable risk since this would be my second engagement at EA and I should be able to get more feedback directly from reviewers before I put in an application. From my understanding it would be a combination of portfolio improvements and additional relevant work experience. I haven't yet been successful with finding work given the market, but I'm hoping the portfolio improvements and networking help with this in 2024.
I understand the risk having less experience, I'm hoping the inclusion of this piece isn't seen as a lacking humility, though honestly if that does happen I hope they do see the best in what is presented and reach out to me with any concerns.
Like with Cami, I do see the model for all its merits and given the write up I can see the intent of the artist and the humility in its execution regardless of how I feel about the result personally. I'm not sure if his humility comes from experience, but his experience does allow him to get away with a lot that might be seen as a detriment.
I am hoping at best, that the Gamagori model is judged for its own merits and establishes that I can take on more responsibility even if I'm not the best judge If what the final visual target ought to have been. I see all my work as iterative, so I'm sure there will be more upgrades and hopefully with better visual targets in the future.
I think for newer applicants, particularly ones who aren't self motivated and need direction, building their portfolio with mentorship can be meaningful. A coworker is currently pursuing this through the think tank training center, I feel his approach provides a good assessment of having visual targets and hands on guidance to work with. https://www.artstation.com/mikeforestall The pieces that were evaluated for his first engagment at EA were the ones made 3 years ago.
For the moment, I'm not sure if that's worth the investment for me personally. His workflow at the TTC uses a more conventional approach with fewer tech solutions so I'm not sure how it directly benefits working at EA. But its great to have the endorsement and motivation from peers and I do hope it will add more weight to any applications he makes.
Regarding the Khan model, a realistic model definitely gives more to work with when it comes to visual targets, and Khan also had growing pains starting in 2017. I remember being reasonably satisfied with this 2017 model as a final sculpt for portfolio and the hate, ridicule and devaluation I received from peers was quite legendary. Some are now senior artists at AAA studios so I wonder what they might feel looking back on what was said at the time.
If I had considered a proper mentorship it would have saved me from some of that vitriol and I likely would have received more beneficial feedback. but at the time I was determined to take on the learning curve and time investment to bring the Khan model to photorealism by myself valuing the private time to develop a suitable workflow.
The general consensus back then among graduates was that an applicant shouldn't even think of applying for a character artist position since its for God Tier Artists and you have to work several years as a model artist to even be considered worthy of the honor.
There was a lot of speculation on the role, its incentives and responsibilities and anyone I knew that got into the studios as model artists would clam up because of NDA's and generally looked down upon artists outside the industry unless you were the grovelling type. (And all this for 18$/hr)
At the time I had no idea about outsourced assets, automated workflows or that character art was just a department and many character artists didn't touch anatomy and worked on props related to characters, so my experience at EA was a pleasant surprise.
In addition to continuing the Khan model I did the disney and stylized pieces because there was more work at local 3D animation studios in Toronto.
I feel that my expenditure on my game dev program to immigrate into Canada stopped me for considering a formal mentorship because of the associated additional cost and I was working off savings from architecture visualization projects.
Not being a permanent resident at the time made it very difficult to find work in Toronto since the provincial government gives incentives for hiring locals and companies aren't motivated to go through the Labor Market Impact Assessment process to permanently hire juniors who don't have PR/citizenship.
But the general vibe of Toronto is really competitive and hostile, with astronomical rents, so I moved over to Montreal which has since been very accomodating.
"I wasn't meaning to match the concept, atleast not exactly."
I think everything hinges on that really. Of course when I use the term "lack of humility" I don't mean it as someone being toxic and pretentious, believing that they can do better than the source material ; I mean it more in the sense of, not having enough perspective and experience to realize that there is a *tremendous* amount of skill and design sense that went into these KillLaKill character designs and model sheets - much more so that one may realize. The Trigger designers are incredible, and decades of experiences are distilled into each of their design decisions.
So to me the correct (or rather, the most productive) approach when tackling fanart is to stick to the source as much as possible. Not as some kind of lazy easy solution, but rather as a humble acceptance of ones ignorance. We never know what we don't know. And that's partly what people mean when they advise applicants to aim for no less than matching the output of the studio they apply to. It's a high bar, but it's a very clearly defined bar.
Furthermore (and perhaps that is counter-intuitive), sticking to the source as closely as possible is a much more thorough exercise than "giving it ones own spin". Because it leads to a great amount of research : investigating how anime maquettes or 3d models deal with the inconsistencies of anime-style front/side views ; having a look at various CG movies to see how Disney and anime studios managed to solve the topic of shading and materials ; or perhaps on the contrary, going for the striking simplicity of cel shading - hence investigating the Arcsys bag or tricks. The list goes on. Whereas improvising and deviating from the source leads to much less problem solving - essentially leading to nothing new at all, since any artist naturally falls back to what they've done before when working without a clear goal. Hence no challenge, no progress, and potentially a weak portfolio piece.
I don't mean to be condescending, but IMHO the various technicalities of the toolset used to produce a model are far, far less important and much less valuable than the tremendous amount of things one can learn by successfully and faithfully giving life to a design.
At the end of the day : technical implementation (using pipeline specific tools), artistic character modeling (to capture a stylized face, anatomy, cloth stylization and costume design), and concept art/art direction (proactively deciding on an artstyle) are 3 very distinct roles and jobs.
- - - - -
(All that said I think the Cammy and Maul models are justified exceptions to this approach because there are just soooo many versions of these characters out there already. Hence another photorealistic Cammy or yet another highly detailed Darth Maul digital sculpt might actually have been boring (as well as requiring a ton of resources and effort). So from a strategic perspective I think these artists made the right choice by bending these famous pop culture icons into something a little bit different. But that takes a ton of skill.)
I wanted to give some feedback on the newer Gamagori model in this thread since it's a bit more relevant here than the other one. I think these points should be considered along with the fantastic advice Pior gave.
Unfortunately I think this is kind of a "side upgrade" it's better in some areas, but because of some decisions as a whole it feels about at the same level as your previous version. Which again, I think this is very concerning, considering these are 6-7ish (depending on how you count) years apart. I would expect a massive upgrade of an older piece with that amount of time between the 2.
Why I think it's a "side upgrade" that kind of comes out to about equal to what you had before:
You have better facial proportions and skin texture for approaching a more "realistic" style. Purely in isolation, an upgrade can be seen here
However once you consider the whole, the flaws and gaps in your knowledge start to become clear
Overall I think keeping the overall proportions is fine, it helps it read as the character you're trying to create and it is the one thing you can keep stylized if you're aiming for realism in every other sense.
However your anatomy needs work, with some muscles missing or outright made up, as highlighted in the other thread. Stylized is not an excuse for this, all stylization would do is play with the shapes of the anatomy.
In attempting to make your face more realistic, the clothing and armor pieces really stand out because they were not given the same treatment. This is where I would argue the older one is better because all of the pieces feel like they are trying to fit within the same cohesive whole. With the armor and clothing being largely untouched (minus wrinkles in the clothing but I'll get to that) it creates a dissonance where the body feels very stylized and the face has a more realistic touch.
If your target was to really bring out some realism but keep the same proportions, I would have heavily focused on clothing construction, and asking yourself how the real-life counterparts of the metal pieces would be created. For the metal I would think about things like welding seams, plating layers, bolts, insets, locking mechanisms, etc etc. I would have also given them more love in the material department, which parts would be more worn (knuckles and shoes definitely)
For the clothing, well... Kill la Kill is about Fashion and if you wanted to go real this could have been a great opportunity to really flex some knowledge on garment construction, playing around with some interesting materials (like maybe finding some cool material for the silver trim on the around the hips of the jacket, or making the black stars embroidered etc) And most importantly, get accurate cloth folds and a damn good taliored fit because this suit is designed for Gamagori specifically. That's unfortunately not what I see, the suit is vacuum-sealed pretty much, folds are pretty eh all around and don't have a lot of logic to them (especially in the arms). How does the jacket close? Is it buttons? Zippers? What style? Zoomed out, it looks very similar to your previous version even though I know some work was done. This is the area where I think your older version feels a bit better, (even though the folds could use some work) because it's simplified in a way that fits the piece in a whole and it looks like you're at least trying to match the anime. Where as for the latest version, it's a very stylized representation, not as realistic as the head makes me think it could be.
This is ultimately the pitfalls you can run into if you try to remake a piece and why it can be a double-edged sword... Sure you can show how you have improved and grown, but you can also show where you haven't, and that can come across way more strongly when you have and exact 1:1 piece you're trying to upgrade. Things are different... but they are also the same. If you're going to update a piece it has to be more than just the technical update, it has to be an upgrade in every single way... if you could literally swap out your old images for the new ones on the already-uploaded project and a person who isn't you can't spot a difference... enough wasn't done to improve the piece.
Maybe that is something I would give a shot, try and look at the piece as if it was someone else's work and be object. It's going to be hard to do, and it may even hurt, but I think it's important to step outside of our own brain a little bit and try to see this as someone else would.
I wanted to give some feedback on the newer Gamagori model in this thread since it's a bit more relevant here than the other one. I think these points should be considered along with the fantastic advice Pior gave.
Unfortunately I think this is kind of a "side upgrade" it's better in some areas, but because of some decisions as a whole it feels about at the same level as your previous version. Which again, I think this is very concerning, considering these are 6-7ish (depending on how you count) years apart. I would expect a massive upgrade of an older piece with that amount of time between the 2.
Why I think it's a "side upgrade" that kind of comes out to about equal to what you had before:
You have better facial proportions and skin texture for approaching a more "realistic" style. Purely in isolation, an upgrade can be seen here
However once you consider the whole, the flaws and gaps in your knowledge start to become clear
Overall I think keeping the overall proportions is fine, it helps it read as the character you're trying to create and it is the one thing you can keep stylized if you're aiming for realism in every other sense.
However your anatomy needs work, with some muscles missing or outright made up, as highlighted in the other thread. Stylized is not an excuse for this, all stylization would do is play with the shapes of the anatomy.
In attempting to make your face more realistic, the clothing and armor pieces really stand out because they were not given the same treatment. This is where I would argue the older one is better because all of the pieces feel like they are trying to fit within the same cohesive whole. With the armor and clothing being largely untouched (minus wrinkles in the clothing but I'll get to that) it creates a dissonance where the body feels very stylized and the face has a more realistic touch.
If your target was to really bring out some realism but keep the same proportions, I would have heavily focused on clothing construction, and asking yourself how the real-life counterparts of the metal pieces would be created. For the metal I would think about things like welding seams, plating layers, bolts, insets, locking mechanisms, etc etc. I would have also given them more love in the material department, which parts would be more worn (knuckles and shoes definitely)
For the clothing, well... Kill la Kill is about Fashion and if you wanted to go real this could have been a great opportunity to really flex some knowledge on garment construction, playing around with some interesting materials (like maybe finding some cool material for the silver trim on the around the hips of the jacket, or making the black stars embroidered etc) And most importantly, get accurate cloth folds and a damn good taliored fit because this suit is designed for Gamagori specifically. That's unfortunately not what I see, the suit is vacuum-sealed pretty much, folds are pretty eh all around and don't have a lot of logic to them (especially in the arms). How does the jacket close? Is it buttons? Zippers? What style? Zoomed out, it looks very similar to your previous version even though I know some work was done. This is the area where I think your older version feels a bit better, (even though the folds could use some work) because it's simplified in a way that fits the piece in a whole and it looks like you're at least trying to match the anime. Where as for the latest version, it's a very stylized representation, not as realistic as the head makes me think it could be.
This is ultimately the pitfalls you can run into if you try to remake a piece and why it can be a double-edged sword... Sure you can show how you have improved and grown, but you can also show where you haven't, and that can come across way more strongly when you have and exact 1:1 piece you're trying to upgrade. Things are different... but they are also the same. If you're going to update a piece it has to be more than just the technical update, it has to be an upgrade in every single way... if you could literally swap out your old images for the new ones on the already-uploaded project and a person who isn't you can't spot a difference... enough wasn't done to improve the piece.
Maybe that is something I would give a shot, try and look at the piece as if it was someone else's work and be object. It's going to be hard to do, and it may even hurt, but I think it's important to step outside of our own brain a little bit and try to see this as someone else would.
Thank you so much for the feedback! Especially on the clothing and armor. I was looking into what to do with them, since I also felt that they aren't very cohesive with the body when it comes to quality.
For my portfolio overall, I'm now focusing on hair creation, so I will take up changes to the clothing and armor after. I do mean to make applications about a month from now, so I'll likely settle for the khan and zhou yu pieces with one additional new piece I am making that would have a focus on clothing and hard surface design which should be sufficient to round off the portfolio.
hm i less concern about whether it's old or new, i mostly interested in if they encounter issue and learn how to solve them or changed their workflow just to get some result
hm i less concern about whether it's old or new, i mostly interested in if they encounter issue and learn how to solve them or changed their workflow just to get some result
I think this is the way most reviews are expected to be. I'm going to be as comprehensive as possible in breakdowns and processes. I'm trying not to rely on first impressions to sell the portfolio, regardless I'd prefer working with artists that do go into details.
Replies
I could adjust the toes so they won't be as aligned, though I'd be covering them with shoes in the end.
For the skin I'll have to tweak the lighting and SSS values with the HDR I'm using. I'll do a check when I have the final lighting setup.
There really was significant advantage in using current technical solutions to tackle anatomy.
While I did have to resculpt the anatomy from reference to allow for better mesh alingment with the transfer mesh, the amount of time it took to create this was 1/10th what it would have been, had I considered a manual method to creating low resolution topology.
I still don't really know how many companies use manual retopology for body meshes, since the method I'm using can apply to any mesh regardless of its size or exaggeration and I'm also able to blend detail and form as I wish with full control over tweaking in the end through morphs.
I do wish more artists would be open about the workflows used in studios and apply that to their portfolio work to practice what they would do professionally on the job.
As part of my game development program, every modeling student had to make an écorché as part of their anatomy studies, but because the majority of the jobs didn't provide any opportunity to touch anatomy, the knowledge gained from the task wasn't utilized.
The quality of the work was variable and many students were hired contrary to the incredible standard we place on candidates artwork that seemingly justifies rejections.
Here is an example of artwork from a graduating student
https://lazypintor.artstation.com/projects/L1aAr
I'm not entirely certain about including écorché or anatomy studies for character artist portfolios, but I do want to try a few pose studies at game resolution once these characters are fully rigged.
I was in the animation stream, so I didn't have to do one, but I did have to do anatomical drawings (form and functional) in dental school with a focus on facial anatomy and cadaver dissection.
All the same I would like to see an artists anatomy skill being used in a studio, but I do see the general direction being to find more tech solutions to automate the process, so where the skill would apply is in managing those solutions.
I feel that existing tech solutions must absolutely be considered before considering any artists anatomy skill as relevant to the job.
I will not go over all the reasoning for the adoption of the "EA Sports-like" set of tools as some sort of exercise, or even as an admittedly useful shortcut to reduce some tedious aspects of model creation. That is because even though tech knowledge never hurts and can be indeed be very useful ... it is a moot point in regards to the end result of a portfolio art piece. Put differently : there's is nothing wrong with the use of a given toolset (as it is no different from someone using a certain software instead of a certain other for retopo or sculpting) ; but it is largely irrelevant in the context of an art porfolio piece.
Now IMHO the actual problem with this model (both the earlier version and the most recent one) is that one vital step seems to have been skipped : the clear establishment of the final visual target that it is supposed to live up to.
One common way to establish such a target is indeed to simply use the art style and output of the studio one wants to apply to . The point of it is not to "flex" or "to be rockstar" ; at the end of the day the point is simply to have a target to aim for, and being able to demonstrate ones ability to hit it. Erring/improvising along the way (outside of perhaps some initial style-bending research) is always detrimental.
In a team environment, this target is provided to you by the AD and your lead, passing over the work of a concept artist/designer. For games relying on real-world athletes this target could come as a set of pictures, body stats and team jersey graphics and some head scans to cleanup and process. This is pretty much what you did for your solid Khan model, by precisely referencing the actor and the movie poster. Yet no one would accuse you of "flexing to look like a rockstar" for doing so. You simply had a clear game plan.
For stylized games, the target is established by a package of drawings, paintings, expression sheets, even clay maquettes if needed. And this package tends to get refined as a stylized game project progresses, since the most successful models striking just the right look and feel end up becoming part of the visual bible of the project helping to refine the art direction gradually. This process can take months/years, and ideally but not always it is supposed to be wrapped up by the end of preproduction.
Now the apparent problem I see with this KillLaKill fan model is that you do not seem to have such a visual target established beforehand (and as said above, the tech pipeline is pretty much irrelevant to that). Yet a visual target would dictate pretty much 100% of the artistic work to be put in the creation of the model (the sculpting of the face, the choice of visual representation for hair as either "clay" or "strands", the level of stylization of the anatomy, and so on).
Of course improvising as one goes without a clear goal can have value during a research phase, or as a long-running study. But improvising *while* also juggling with the incredibly complex set of requirements of game art is IMHO a strategic error leading to weak results.
All that said, part of your target is known : it is that KillLaKill character. But the moment you decide to "give it your own spin", you have no other choice than taking the time to establish what this spin actually *is* visually. "Just make it look good" is not enough of a target.
The obvious irony is IMHO the fact that establishing such visual targets is *not* the job of a character modeler - it is the job of the art director and the concept artists, with extra input from the art leads. There is of course nothing wrong with integrating this step as part of a portfolio piece (like with the earlier mentioned Cammy and Darth Maul pieces, which did that perfectly well), but this can be very risky because it can immediately becomes the weakest link of a whole piece. Which is why, especially for people with less experience, it is more advisable to aim for the existing output of the studio they apply to as their specific target instead of inventing their own. It is not "trying to be a rockstar", it is simply showing humility.
After all being able to aim for this clear target and reaching it consistently is *exactly* what is expected of a skilled character modeler on the job.
I did use a concept for the overall look of the 2017 model, though I hadn't establish a final look of what my interpretation would look like.
I wasn't meaning to match the concept, atleast not exactly. My shading, presentation and lighting skills were definitely not at their best, and looking back I prefered the model in its zbrush render, than the marmoset result which I certainly could I have done a better job with.
I was quite flexible with the piece, and atleast after having the portfolio evaluated the detriments were with regards to my presentation, lighting and shading.
As I was being considered for an associate position at EA the only visual target I was expected to reach to the best of my ability was within the art test.
For my portfolio it was understood that I simply did not have anyone to direct my work, so it was assessed on technical merits necessary to qualify for the art test.
I do feel that if I were to be in the running for an associate positon again, it be better to build a character as part of a mentorship program so there is a visual target I can match that someone with more experience can provide me.
Since I am applying for a second engagement, my objective is not just the final result but also the process which atleast for EA does prepare me better for exactly what the job requires, which is versitality with using multiple tools and a lot of back and forth and reworking of a model in an efficient way.
This approach of upgrading older work, is helping me research and implement workflows, and atleast I think the result is significantly more presentable.
I do understand the importance of a visual target, not having one does make it challenging to arrive at a final result.
I hope that not having a visual target doesn't work against my application for this piece, but I can see it as a acceptable risk since this would be my second engagement at EA and I should be able to get more feedback directly from reviewers before I put in an application.
From my understanding it would be a combination of portfolio improvements and additional relevant work experience.
I haven't yet been successful with finding work given the market, but I'm hoping the portfolio improvements and networking help with this in 2024.
I understand the risk having less experience, I'm hoping the inclusion of this piece isn't seen as a lacking humility, though honestly if that does happen I hope they do see the best in what is presented and reach out to me with any concerns.
Like with Cami, I do see the model for all its merits and given the write up I can see the intent of the artist and the humility in its execution regardless of how I feel about the result personally.
I'm not sure if his humility comes from experience, but his experience does allow him to get away with a lot that might be seen as a detriment.
I am hoping at best, that the Gamagori model is judged for its own merits and establishes that I can take on more responsibility even if I'm not the best judge If what the final visual target ought to have been.
I see all my work as iterative, so I'm sure there will be more upgrades and hopefully with better visual targets in the future.
I think for newer applicants, particularly ones who aren't self motivated and need direction, building their portfolio with mentorship can be meaningful.
A coworker is currently pursuing this through the think tank training center, I feel his approach provides a good assessment of having visual targets and hands on guidance to work with.
https://www.artstation.com/mikeforestall
The pieces that were evaluated for his first engagment at EA were the ones made 3 years ago.
For the moment, I'm not sure if that's worth the investment for me personally.
His workflow at the TTC uses a more conventional approach with fewer tech solutions so I'm not sure how it directly benefits working at EA.
But its great to have the endorsement and motivation from peers and I do hope it will add more weight to any applications he makes.
Regarding the Khan model, a realistic model definitely gives more to work with when it comes to visual targets, and Khan also had growing pains starting in 2017.
I remember being reasonably satisfied with this 2017 model as a final sculpt for portfolio and the hate, ridicule and devaluation I received from peers was quite legendary. Some are now senior artists at AAA studios so I wonder what they might feel looking back on what was said at the time.
If I had considered a proper mentorship it would have saved me from some of that vitriol and I likely would have received more beneficial feedback.
but at the time I was determined to take on the learning curve and time investment to bring the Khan model to photorealism by myself valuing the private time to develop a suitable workflow.
(And all this for 18$/hr)
In addition to continuing the Khan model I did the disney and stylized pieces because there was more work at local 3D animation studios in Toronto.
I feel that my expenditure on my game dev program to immigrate into Canada stopped me for considering a formal mentorship because of the associated additional cost and I was working off savings from architecture visualization projects.
Not being a permanent resident at the time made it very difficult to find work in Toronto since the provincial government gives incentives for hiring locals and companies aren't motivated to go through the Labor Market Impact Assessment process to permanently hire juniors who don't have PR/citizenship.
But the general vibe of Toronto is really competitive and hostile, with astronomical rents, so I moved over to Montreal which has since been very accomodating.
"I wasn't meaning to match the concept, atleast not exactly."
I think everything hinges on that really. Of course when I use the term "lack of humility" I don't mean it as someone being toxic and pretentious, believing that they can do better than the source material ; I mean it more in the sense of, not having enough perspective and experience to realize that there is a *tremendous* amount of skill and design sense that went into these KillLaKill character designs and model sheets - much more so that one may realize. The Trigger designers are incredible, and decades of experiences are distilled into each of their design decisions.
So to me the correct (or rather, the most productive) approach when tackling fanart is to stick to the source as much as possible. Not as some kind of lazy easy solution, but rather as a humble acceptance of ones ignorance. We never know what we don't know. And that's partly what people mean when they advise applicants to aim for no less than matching the output of the studio they apply to. It's a high bar, but it's a very clearly defined bar.
Furthermore (and perhaps that is counter-intuitive), sticking to the source as closely as possible is a much more thorough exercise than "giving it ones own spin". Because it leads to a great amount of research : investigating how anime maquettes or 3d models deal with the inconsistencies of anime-style front/side views ; having a look at various CG movies to see how Disney and anime studios managed to solve the topic of shading and materials ; or perhaps on the contrary, going for the striking simplicity of cel shading - hence investigating the Arcsys bag or tricks. The list goes on. Whereas improvising and deviating from the source leads to much less problem solving - essentially leading to nothing new at all, since any artist naturally falls back to what they've done before when working without a clear goal. Hence no challenge, no progress, and potentially a weak portfolio piece.
I don't mean to be condescending, but IMHO the various technicalities of the toolset used to produce a model are far, far less important and much less valuable than the tremendous amount of things one can learn by successfully and faithfully giving life to a design.
At the end of the day : technical implementation (using pipeline specific tools), artistic character modeling (to capture a stylized face, anatomy, cloth stylization and costume design), and concept art/art direction (proactively deciding on an artstyle) are 3 very distinct roles and jobs.
- - - - -
(All that said I think the Cammy and Maul models are justified exceptions to this approach because there are just soooo many versions of these characters out there already. Hence another photorealistic Cammy or yet another highly detailed Darth Maul digital sculpt might actually have been boring (as well as requiring a ton of resources and effort). So from a strategic perspective I think these artists made the right choice by bending these famous pop culture icons into something a little bit different. But that takes a ton of skill.)
I was looking into what to do with them, since I also felt that they aren't very cohesive with the body when it comes to quality.
For my portfolio overall, I'm now focusing on hair creation, so I will take up changes to the clothing and armor after.
I do mean to make applications about a month from now, so I'll likely settle for the khan and zhou yu pieces with one additional new piece I am making that would have a focus on clothing and hard surface design which should be sufficient to round off the portfolio.
I'm going to be as comprehensive as possible in breakdowns and processes. I'm trying not to rely on first impressions to sell the portfolio, regardless I'd prefer working with artists that do go into details.