Lemma Soft Forums

I didn't really see a thread for this, but ummm...
I was just wondering how everyone else went about making sprites for their visual novels. This is my first time working on a visual novel on my own, and when it comes to sprites, I think working on so many layers is making me nervous, hahaha!
So I'm just interested in hearing your techniques, and I'll share mine. And uhh... Hopefully we'll all learn something?
Okay, so this is what I'm working on right now. I'm making folders for different characters, just so I can keep them apart. I should probably be working on different characters on different canvases, but my sketch had all of them in one spot, so I just left them as is.
First thing I do, I do do lineart for the body, which doesn't have any facial characteristics other than the nose, just because... I want to be able to change those things later on, and I don't know if noses move around much. Then on several other layers, I make various different facial expressions (These ones are the ones titled things like Face 1 and Face 2). This part I'm kind of worried about. Should I make layers for individual facial characteristics, or should I make layers for individual expressions? Anyone have any input?
Also, hair goes over that because hair covers faces...
Umm... right now they're kind of lacking clothes, because I want to be able to make sprites that can be versatile in terms of appearance.
My sprites are kind of short. Umm... Would this be a problem?

Sorry if that was really rambly or if it didn't make sense. But I'm really interested in how you guys make your sprites as well! I'm not sure if the way I'm doing it is really efficient so...

Be careful...... eventually you'll hit Sai's 256 layer limit if you keep using that method. I'm ramming my head into the wall constantly because I want more layers but I can't add them... >_>

Woah. What do you do to get around this then? Do you have any techniques? What is Sai's layer limit anyway?

256. There is no way around it, aside from merging layers or just starting a new drawing.

Ahhh. I see, thanks very much for your help!
Generally I end up merging lineart and colour layers once I think I've finished with them though... hopefully that will be enough!

I can only provide input for Photoshop:

Separate layer folders for body, face, hair, and eyes (since eyes have so many shinies and gradients that require many layers on their own).

Each alternate facial expression has their own layer.

Each layer folder contains the base lineart, the flat colors, and the gradients / shading, and any highlights.

A separate body posture (aside from mere arm movements) can be considered as a separate body (and thus need their own face, hair, and eyes to match).

I did one big PSD file for all characters two years ago since liked you mentioned its easier to get a size comparison going, but since then the writer asked for alternate clothes so I had to re-save each character on her own psd and add an additional layer (tracing the basic posture and adjusting for the new clothes). For budget works that only need basic shading and one set of clothes, I will still make all characters on one psd file.

Oh yeah... and if you're not just sticking characters in dialogue boxes you probably will need to draw at least up until the thighs.

Be careful about having too many layers in one PSD file because it can slow down your computer considerably. At least in my case but I tend to color my characters at 300 DPI with pixel sizes up in the 1000-2000 range or more. To work around this, I will usually save different files say one file for the nude character with different facial expressions and another file with a flat layer of the nude character and work on clothes on that.

Samu-kun wrote:Be careful...... eventually you'll hit Sai's 256 layer limit if you keep using that method. I'm ramming my head into the wall constantly because I want more layers but I can't add them... >_>

Whoa,I haven't know that sai has layer limitation.
Anyways, 256 sure enough for me. XD

I didn't have a lot of experience,only share my method.
My sprites creations isn't much different from your.

Based on sai

1. Have a rough sketch layer for the pose and a layer for the color planning
2. Reduce the opacity,hide the color,and start to make the clear line ,and same as your method I drew eyes,eyebrows,and mouth in the separate layer ,each layer for each expression.
3. Color the outline , I colored all the parts but skip the eyes and mouth ,also must have to skip some highlight since some of them varies according to eyes and mouth (like grinning or eyes closed sprites)
4. Color the rest for each expression (some can share same element)

5. For another pose, I'll drag the picture of the former pose after finish rough lineart to see if they are in the same proportion ,and if not just transform or correct it until it's OK. then continue 2-4 again XD

This is how I do mine :
I use Adobe Photoshop CS2 (Which is kinda old, I know)

First. I sketch first a guideline layer. Maybe isn't the final thing. Then I make a new layer on top of it, then turn the guideline layer to 65% or any opacity that I wish, as long as it isn't as opaque as the lineart one. Then I trace the desired lineart, add hair and details and all, but I exclude the face first. I separate the face in a different layer so that it's easy to change later. Then I delete the guideline layer.

Second. I make a new layer on top of the white layer(underneath the lineart), then I fill it with a color I won't use (I often use pink). After that I make several layers for the color, then I start coloring.

Third. After the base colors, shadows first. Clipping masks make shadowing easier. Then highlights if necessary. If possible, you can use the multiply and screen blending modes in layers, then when it gets too dark or too bright, tone down the opacity of the layer. For the hair, use feathery strokes, but for the other parts, you can just scrible then blur. That's my style XD

Then after I finish details, I remove the white and pink layer underneath, clean out some mistakes then make the emotions by editing out the face layer above or below the lineart layer. Then save.

Well, that's how I do mine XD
You all have SAI... Me is jealous...

Lumen_Astrum wrote:You all have SAI... Me is jealous...

Yes, but you have talent, so I'm jealous. Can't buy talent for $50. XD

sake-bento wrote:

Lumen_Astrum wrote:You all have SAI... Me is jealous...

Yes, but you have talent, so I'm jealous. Can't buy talent for $50. XD

Definitely!! Especially when artists cost a lot :O! xD So a person who can both write and draw well... :O That's awesome right there. xD

This has been a very informative thread! Guess that's why most people do the main drawings on the computer.

Question, what size do you usually use to draw your sprites in? As in, canvas size. I don't have a huge monitor, so large pictures I find very difficult. :/

I can't help but feel as if some of you guys use more layers than me lol. Here's two screenies of two different sprites I've worked on! The first one being an earlier work, the second for the project I'm currently working on :B

My usual process is as listed:

- Rough outline, refined sketch, and then make a lineart layer over said sketch
- Fill in base colors, separate layers depending on which part (like one for skin, one for hair etc)--
- Draw the eyes on a separate layer and set that aside
- COMMENCE THE CGING. There is basically a lot of use of a blending brush and blur along with clipping layers.
- After properly coloring the main parts, I combine the clip layers since I am not concerned with changing the shading on the body or hair
- Back to the eyes layer to color them
- Then basically I have separate folders for mouth, eyebrows, and maybe another folder for different eyes. Within each of these folders are layers with different expressions in them so I can mix and match eyebrows with mouths

Wow this is a lot of work

sake-bento wrote:

Lumen_Astrum wrote:You all have SAI... Me is jealous...

Yes, but you have talent, so I'm jealous. Can't buy talent for $50. XD

Hahahahaha~ XD True dat.

Miss-Mae wrote:This has been a very informative thread! Guess that's why most people do the main drawings on the computer.

Question, what size do you usually use to draw your sprites in? As in, canvas size. I don't have a huge monitor, so large pictures I find very difficult. :/

I usually draw on the final resolution. When I do side images, I use around 200 - 350 pixels, but on sprites... maybe 350 - 400? Yah. ^_^ Well, if you draw on the computer (like I usually do) you can use this method of not resizing, but if you scan your linearts, you should first work on a bigger resolution before shrinking to the final size (since usually scanners make gigantic images from your drawings).

So 200 dpi would be okay? I usually use that, so I guess I might be okay after all. xD I hate having such a slow/small computer, it really does suck. xD Thank you very much for the response, btw. I should really check out your DA, too. *spotted your DA account name* >D <3