How do YOU give feedback?
polycounter lvl 9
I haven't posted anything in a while so here's something I've wondered about the way people give feedback.
So you see someone's work and you comment on it. What goes through your head to decide what to say?
So you see someone's work and you comment on it. What goes through your head to decide what to say?
- Do you have a list of criteria that you look out for? (e.g. accuracy, lighting, material definition etc.)
- Do you give feedback on something you don't feel like you have a strong knowledge about?
- How much experience do you have in that field before you feel you have an informed opinion?
- How do you go about subjective matters like design direction? What is your advice based upon?
Replies
I think this sells the piece (gets the viewer interested), and then you can start going in-depth, on functionality (like this will not work, doesn't move this way, etc) and on the technical aspects, which I consider mostly nitpicks, like texture seams, artifacts, layout of UV's and so on.
i.e. Stylized realistic hand painted hunter, Post Apocalyptic Scifi PBR Environment, etc.
This is the most important question IMO. Telling someone that the eyes are too big when they're sculpting a PowerPuff Girl is misguided and a waste of time.
2: What are the trying to accomplish/learn, i.e. Why are they doing this?
Is it a figure study? A lighting study? Texture study? Its not a study, just an idea that they want to explore and get see how people respond?
3: What do I think is wrong VS What do I know is wrong?
One is subjective (I do not think that shade of blue works) and one is objective (That shade of blue is wrong based on the concept provided)
For characters, I try and take into account whether or not the thing Im looking at falls into an acceptable range stylistically and anatomically.
4: What do I like and why do I like it?
The more I understand this question the more strongly I can build a case for my crits.
5: Whats a technical limitation and whats a skill limitation?
Ex- Bad shading, your normals are fucked / Face looks like a potato... Uhhh, whats your reference?
I dont write every crit with these questions like a checklist, these are just things going through my head when i choose to give feedback.
Edit:
Forgot one of the most important things, I usually double check what I've written and remove any emotional language if it's slipped in.
That said, I do try to provide feedback when I can pin point what appears "off", and I like to highlight what works about the piece and what doesn't.
Sometimes, I honestly don't have any critiques to point out, but I do try to make a point of calling out that I like it. Sometimes it feels a little hollow, but I like to think that everyone likes a little bit of encouragement over silence.
The more art I do myself, the more I learn and feel like I am able to provide feedback.
That's the best way to give feedback. You start with something good about the model, then just start listing everything that is bad and finish with another good thing. It really help to get your point across with people who are not used to get feedback.
If you do need to help guide their idea, i find it's best to ask what their intention is first. What point are they trying to make. What do they want the piece to say to the audience. So when do you make your suggestions, they are educated suggestions that are helping them work towards their goal. You should also always note that it is your personal preference, when it comes to changing their ideas, and that it's up to them to decide if they like it or not. That is, of course, for the extreme noobs. Experienced guys know this, so you dont have to tip toe as much. That's also the tricky part with noobs. You have to be firm but fair. And it's best to shit sandwich them. Try not to discourage them, but rather push them in the right direction and get them to keep trying.
Critiques, after all, are not about telling people exactly what to do, but rather, they are to help people learn how to make better decisions, and how to execute their ideas better and more clearly. They are suppose to teach people how to improve. They are NOT suppose to take the thinking out of it all, where they just mindlessly fix what you tell them.
Point out you like it/what you like and then note what could be done... but I guess the positive starter ends up feeling forced anyway.
To me it still feels like the kind of respect the person deserves for putting so much effort into it.
Sometimes brutal honesty is ok as well.
Oh and I like to avoid any fanboy comments that don't help anyone.
I mainly just try to make sure I am always explaining my reasoning, I absolutely hate when people post shit without explanation. If something is just my opinion then I try to state it as such, either by saying it's an opinion or by saying it's just what I would do were I in their shoes.
I don't like leaving generalized or non-specific feedback. If the best advice I can give without spending an hour writing a long post and doing paintovers is "Spend more time, study anatomy and pay closer attention to your refs" then I usually just won't comment at all.
I wonder if you can elaborate a bit on what would lead you to think something was off about a piece and what advice you would give them to improve it?
@mrhobo
This is the thing, what's the answer to this question? It's not always clear why you might like something so how do you go about breaking it down?
@marshkin
I think I'm in the same boat as you here. I don't feel well versed enough to crit with any real merit, so I just observe what other people write and try and learn from them.
@almighty_gir
Haha, but for real?
@popol
I hear you about the way to deliver feedback, I'm just more curious about the way you go about finding the good and bad.
@slipsius
Ok right, so you stick to what you know and being an animator means you can give objective animation feedback. Your points about subjective feedback are also interesting, thanks.
@panupat
I see, what would you give you a good/bad first impression then and what would you feed back to them?
@cay
Yeah I'm not sure about fanboying either. I guess I would also ask you about what might prompt you to comment on a piece?
@atticusmars
I resonate with your points here. This is Polycount, and should be about improvement more than just praise. Explaining your reasoning is something I wish I saw more of. It's kind of why I made this thread. Sometimes you just want to know how someone reached their decision. Time is also a massive factor in feedback, yes. I've spent ages before deliberating on what to say to someone only to end up saying nothing!
MAKE SURE YOU USE REFERENCE
If necessary I will also try to explain why my criticism will help them improve the quality of their work, its always good to know someones reasoning rather than just getting comments like "make the edges softer" and not knowing why that would help you get a better bake.
Well I value a piece over it's artistic merit, as in traditional values, I think it's the most important thing, above technical details or functionality.
For example, if it's a drawing, I might look at overall composition, in the classical sense (framing, point of interests, color contrast, etc), and then on details like brush strokes, plasticity, confidence. These things are easy to spot as long as you trained your eye a lot (by looking over tons of great pieces =D), and you can only analyse as far as your eye is trained. When I was just getting into art I would look at some drawings and be impressed, but years later I can see obvious mistakes and now I know what could be improved, and I can see the skill level of the artist.
For something 3D I'll look at the general flow of shapes, whether repeating patterns and large shapes mixed together are pleasing, contrast of hard lines and soft shapes, etc, and then the same color contrasts and interest points (which can even be little cool details).
And you don't really have to think a lot about these, they just pop, and if everything is good, then you can go in-depth and analyze what or how could be improved (for example changing colors, rounding off some parts, etc). But if nothing is off and it's already visually pleasing, I think changes are subjective at this point, and you can never definitely say that doing this or that is better then what the artist already did. Unless your feedback is on a technical, or functional level, not artistic. In those cases the critique can be objective.
it is a stylistic choice, or maybe I just don't bother
I think polycount used to give much more honest criticism, these days people seem loathe to comment at all.
I like to point out nice pieces because I am not a dick. nothing wrong with giving praise where it's due.
Ive mostly focussed around people trying to get their first job as they need portfolio critique the most and you can quickly summarize
a lot of things with bulletpoints, sometimes I did paintovers or webpage mockups, often a lot of text with quick bullet points as if I were in a rush.
no rules behind but packing in some praise is certainly a good idea, even if its all bad, you have to realize the current skill level
and then base some things around that. If per example texturing is below their current skill level, its important to point that out and vice versa. Its all about locating their level and then give critique around that. The better ones get very specific critique the other ones more broad suggestions and general things. People that have perfect modeling and texturing etc are getting critique on their presentation, lighting, photoshop work logo and theres always something to help with.
Overall, the best method I have found is to do paint-overs:
1. They help you get into the right mood for critiquing. The right mood being that you want to get this art piece as good as possible.
2. They help identifying problems. Even though we visualise how certain fixes can help, doing a paint-over sets them into practice, and helps identifying other issues. Furthermore, because you are interacting with the image, you see more.
3. They take far less time than you think.
4. They communicate very effectively what you think needs changing, how that would look, and most importantly, it allows the other artist to judge if that is what they want.
5. They prevent communication errors caused by unusual art-terminology. The most difficult critique I had wasn't critique that was harsh, rather it was critique I just didn't understand. Mostly because the other person was under the impression I was being touchy-feely, while I just didn't understand what the hell they were going on about.(We were both speaking normal plain english, it just seems we came from different traditions of talking about form)
6. They make you a better artist as well. Doing a paint-over requires you to go back and look at the basics so you can explain them properly. Then applying those basics onto a paint-over is a very good exercise, and allows you to re-evaluate your own skills.
I don't always do paint-overs, though I think I should do them more, especially with advanced 2d artists. Generally, I favour them due to the way they help communicate, especially when it comes to the fiddly communicating that is necessary for stylised work, because it allows you to get past the boring 'what do you mean stage' and into the goody 'oh, but I would like to solve it like this'.
You also don't have to shit-sandwich this way. Just start with weak points, finish with strong points.
Maybe not always applicable with 3d art, but it should help a lot.
Of course! Reference is so key!
@Ged
Yeah once again explaining critiques it helps so much. Say if you were talking about lighting, would you point out to the person "Your lighting is too harsh because of a,b,c. You can fix it by doing x,y,z"?
@huffer
Thanks, this is the sort of answer I was hoping for! As far as training your eye, how did you get to the stage you are at now? Was it a lot of classical art theory? I find your breakdown of 3D pieces interesting, those points would probably not occur to me at this stage. How could someone learn to create artwork with these in mind?
@dustinbrown
Yeah, I mean, that's what Polycount is all about right? Could you possibly expand upon what you might critique an artist about?
@Ruz
I know what you mean, you can only go so far with objective crits. I hope Polycount hasn't lost that vibe, it's kind of what it's known for.
@Shrike
Shame you feel that people don't give good comments, how could this be changed? I always see you giving crits on guns, what are you looking for on those pieces and how did you come to have such an understanding of them?
@Wolthera
These are some great points, they make complete sense. Seeing as you're a 2D artist, do you look for more traditional art principles upon which to base your feedback?
Also thanks for everyone's input so far, it's great!
I generally try to avoid critiquing designs, unless they're seriously flawed or off-putting.
That said, if the concept is what I consider weak, I probably won't comment at all.
Criticizing in a work environment is different again. In a learning situation people are doing things for the first time but in a commercial setting the only aim is to produce the best possible product in the shortest time. In an ideal team if I am your art-director you should know I love you and your work or you wouldn't be sitting there! Ideally then we should be able to get down to business minus the hearts and flowers. I think that is the most efficient work method, its a pity that is seldom the case though. Just about everyone needs a kind word to perform properly.
It's tricky because people have such diverse levels of knowledge. Something one person might think is common knowledge could actually be alien to another. Maybe it's best to just start with basic feedback and get more complex if the response is positive.
@JacqueChoi
That's interesting, your feedback is mostly objective? So what would you consider to be weak design, or even the opposite, strong design?
@kanga
4x 4 hour slots? With such long sessions, you must be heavily deconstructing their work. What sort of critique would you cover with a student in one of your classes?
@dustinbrown
Ah you're similar to Jacque in that you stay objective. As you're a character artist, I see on your website that you took lots of courses about anatomy and sculpting. How much of the feedback you give would you say was based on traditional artistic principles like those?
;P
Not coming from an 'Art Directors' PoV, but simply for artists looking to find work:
It's hard to break down, but it has to have an impressive level of detail, and execution, while avoiding the pitfalls of using 'style' as a crutch.
No matter how amazingly exceptionally well made you can make this fire hydrant, it will never ever be remotely as impressive as a Harrier Jet Engine:
Other times, simple things can have a TON of appeal, but I wouldn't know how to critique that. Sometimes we can strengthen a silhouette, or a design choice, or fix things feel non-functional.
For example, if someone does a handpainted model, we can analyse the style and assume it's inspired by looney tunes, or by disney princess films, or by anime, or by warcraft painted models. Like a good example of this was... a year back or so? Someone was obviously doing a texturing attempt at warcraft style, and someone else said something along the lines of 'hmz, I think warcraft has baked in the specular and ambient occlusion. Maybe it's an idea to look at DotA style models as well if you are intent on seperating the specular?'
That was a good comment because it first assesses the style(this is warcraft inspired), then points out a problem (warcraft style textures have the specular baked in, so any style separating those will look different), proposes a solution (check out the dota models, maybe?), but leaves it up to the artist to decide what they'll do.
About how direct I have to be... that depends. Eventually you will understand how to handle different people in your team, and which approach works best - however, keep in mind you'll have to work with them every day, so agressive or insulting or embarassing feedback is generally not a good idea. If there's anything like this ever required, then it should happen in private.
Furthermore, I try to find a relevant real work or conceptual example with my critiques. I find that it helps the person receiving the critique understand where I'm coming from more often. Lets say a material isn't reading properly. It's easy to say "add more surface variation to this rubber". It's more beneficial to say that and include a visual example of what you're talking about. That gives a clear and direct goal for the artist to hit. If the model is too visually noisy with panel cuts all over the place, I'll link a tutorial or a collection of artist works that show a good ratio of minimal and heavy detail (70/30 rule kinda).
I also like to point out areas that could be fixed on the next model. There are times where a model is good but there are a few little things in the high poly or UV that could be fixed on the next piece. Stuff that isn't worth spending another week going back and fixing but is beneficial to know the next time around.
Finally, I usually don't spend too much time harping on the same critique unless it's a person that I know well. Most of the time I will try to be as detailed as possible and leave it at that. It's up to the artist whether or not they want to incorporate them.
Cheers
@JacqueChoi
So it's about finding something that strikes the balance between visually appealing and technically complex. For someone to design something like that whilst also making sure it remains functional sounds like a real challenge.
@Wolthera
Oh right, that's awesome, having a broad knowledge of styles and techniques and then being able to help someone. That kind of comment can really boost some of the less experienced artists.
@Kwramm
It's a fine line you tread there! About pointing out the same mistakes, are they technical or design based (or other)?
@beefaroni
Great answer, thanks. Your kind of feedback is exactly what an artist would need in order to improve. If you were going to give someone subjective suggestions, what kind of things would you talk about?
@kanga
Thanks for the reply, are those students complete beginners then? I guess you aren't critiquing them too heavily in that case, is it more fundamentals of accuracy and technique?
Alright, thanks for your input!
@Kwramm
A checklist for quality? That sounds great, do you use one yourself?
Your main goal is: keep feedback loops short and reduce the number of iterations per feedback loop, for best efficiency.
That's really interesting, thanks for the insight