How does one make high quality sprites

I know HOW TO sprite, I have done countless of sprites and tested them in-game - with no problems.

But there is just one little thing.

Making custom sprites, not from importing them from a game for an example, seems to be incredibly difficult. I have Photoshop, and am skilled with it, but when it comes to making sprites, I am out of luck.

How do you make your sprite, how did ID Software make their sprites, even more so, how did Sgt. Mark IV make all of the thousands upon thousands of very high quality sprites for Brutal Doom.

You just gotta be a talented artist, I think. Additionally it helps, if you are good at pixelling. If you can draw interesting monsters on paper, you should be able to do spritework. If you can‘t... well then not so much. 

1 hour ago, elend said:

You just gotta be a talented artist, I think. Additionally it helps, if you are good at pixelling. If you can draw interesting monsters on paper, you should be able to do spritework. If you can‘t... well then not so much. 

I can draw on paper, quite well, it's just that using a mouse seems very coarse and bad. 

You could try to scan things or photograph stuff. That‘s actually what the id guys did. Photographing scale models and then cleaning the pixels up in Deluxe Paint. It is very tedious, yes.

1 hour ago, elend said:

You could try to scan things or photograph stuff. That‘s actually what the id guys did.

Some doom monster sprites were drawn from scratch with no models, I'm pretty sure. Look at the imp's cartoony face and compare it to some wolf3d/spear of destiny bosses. Adrian Carmack has a definite style when it comes to angry monster faces.

4 hours ago, elend said:

You just gotta be a talented artist

This. There's no tricks, no easy methods or how to's. You just have to be good. Tools like Photoshop, GIMP, Paint.NET and stuff are just that: tools. They can't make you into an artist.

If you can sculpt in clay or play-doh, you could set up a fixed camera, and put your model on a lazy susan, and take pictures of it at 45 degree rotations. Then you have to color each picture, and add some artistic detail. That's for one animation frame. Then, alter the model for each new standing, walking, attacking, pain, and death frame.

Of course, it's a ton of work, but it's much better than trying to draw each frame by hand. You get proper sizing/aspect ratio, hidden surface removal, and all eight rotations "for free", so to speak.

3 hours ago, kb1 said:

If you can sculpt in clay or play-doh, you could set up a fixed camera, and put your model on a lazy susan, and take pictures of it at 45 degree rotations. Then you have to color each picture, and add some artistic detail. That's for one animation frame. Then, alter the model for each new standing, walking, attacking, pain, and death frame.

 

Of course, it's a ton of work, but it's much better than trying to draw each frame by hand. You get proper sizing/aspect ratio, hidden surface removal, and all eight rotations "for free", so to speak.

That may sound good on paper, but that's a stupid idea, I already tried drawing them and taking photos, it turned out to be a massive mess. The actual sprites are quite small (only like 60x40 pixels), and trying to resize an IRL photo to that is ridiculously stupid.

And yet, that's what the id guys actually did. It is not ridiculously stupid, if you set up the model properly, if you light properly, if you adjust the final full size image properly and if you resize properly. It's still a ton of work to clean up the result pixel by pixel, but the look and feel of the outcome definitely has its advantages.

If you can shell out $15, I highly, highly, HIGHLY recommend the spriting program Aseprite. Imagine Photoshop but made specifically for spriting. Things that took days for me before can be done in five minutes now. (They're also set to be adding an Aspect Ratio mode within the next year or so, which will come in handy for Doom spriting)

8 hours ago, Death Egg said:

If you can shell out $15, I highly, highly, HIGHLY recommend the spriting program Aseprite. Imagine Photoshop but made specifically for spriting. Things that took days for me before can be done in five minutes now. (They're also set to be adding an Aspect Ratio mode within the next year or so, which will come in handy for Doom spriting)

how exactly is better than photoshop?

It is hard to make great sprites indeed. My go-to choice is to make sprites from 3D models. You can download Blender and sculpting tool for it and start to mess around. Here is example of my Heretic monster wip. I don't think it is "high quality", but I'm also trying among many others to get "there" and it is awfully annoying trial and error task, because experience is impossible get with out doing things. I also use a lot of Grafx2 for pixel level editing.

On 9/16/2018 at 8:49 AM, der_einzige said:

I know HOW TO sprite, I have done countless of sprites and tested them in-game - with no problems.

 

But there is just one little thing.

 

Making custom sprites, not from importing them from a game for an example, seems to be incredibly difficult. I have Photoshop, and am skilled with it, but when it comes to making sprites, I am out of luck.

I don't have the time or energy to write a full treatise on this (although I seem to have gotten halfway there in the process anyway) or to dig up some of the more helpful tutorials I've looked at in the past, but I'll try to at least give my basic pointers.

First, how good are you at drawing, in general?  It sounds like most of your experience is in "frankenspriting", i.e. combining and editing existing sprites.  You can pick up certain skills from doing sprite edits but if you want to learn to do sprites from scratch, that's the arena you have to practice in: drawing stuff from scratch.  There are a lot of sub-topics in this but the main things I would advise studying are:

* Anatomy.  In particular, focus on what parts of the body exist, and how they connect.  A lot of people will make a big deal about proportions but spritework is very often distorted in this respect to better accentuate the "important" details--it's more important that your proportions are consistent and visually pleasing than that they be "correct".  That's not to say that understanding the natural proportions of the body is useless, though.

* Color usage.  This is often times what makes or breaks something that's otherwise structurally sound, so for anything that's going to be in color this is worth studying up on.  In reality, things aren't always actually the colors you think they are.  Try opening some digital photos in your editor and using the color picker tool to investigate what the actual colors of light, dark and midtone areas of an object are.  You might be surprised.  Where the light and shadows go is also a frequent sticking point for beginners.

* Animation.  This is, in my opinion, the hardest and most tedious part, since you have to be able to make the sprite look like it's moving naturally and not just distorting.  Refrain from the temptation to copy-paste as much as possible, and actually redraw the frames (there are times when copying can save you work, like mirroring frames and such, but many things simply are not going to look right if all you do is select bits and shift them around in 2D space).

Learn other stuff if it's useful, but be wary of whether it really is useful.  Good books can be very useful but there are a lot of how-to books on drawing that teach little to nothing of actual value.  Avoid ones that specifically focus on superhero comics or anime in particular because they tend to be more about the tropes of the genre in question than about drawing in a more general sense.

Now when it comes to Doom sprites in particular, most 3D games that used sprites either based them off of models (physical or digital) or else took shortcuts like the Wolfenstein 3-D era games that only drew monsters facing from the front.  In the case of models, they were usually touched up a bit afterwards to emphasize the important details.  Most of the Doom creatures and weapons were based partially or entirely off of physical models that were photographed (some custom made, some using toys or other objects like the chainsaw being based on an actual chainsaw).  HacX likewised based at least some of the creatures off of similar techniques.  Raven seemed to prefer working off of digital models rather than physical ones.  You can theoretically hand-draw a full set of 3D sprite rotations from hand, but for most people (myself included) that's more work than it seems to be worth; I personally would advise learning 3D modeling if you really want to work with homemade graphics for 3D games but that's a whole other can of worms (although my advice on the most important points being anatomy, colors, and animation still stand for that as well).

Aside from that, practice is more important than some magical "talent".  Don't believe that you have to be born with mutant powers and that you should just give up if your initial attempts suck, you have to put a lot of time into studying, and studying the right things, like anything else; it's not about if you have the talent to get good so much as it's about if you have the time to get good.  A big part of improving is being able to pinpoint and understand what it is that's making or breaking a particular piece.  If you draw something and you don't like it, don't just say "this sucks" and move on.  Ask yourself why it sucks and how you could make it better.  Take notes.  Draw it again and see if it looks better.  Likewise, you can see something that looks cool, but the important part is figuring out why it looks cool so that you can adopt similar techniques.  Is it the way colors are used, contrasts, the way the character is postured?  Take notes.  See if you can draw something that uses a similar technique.  Those are the sort of things to think about.  For Doom, John Romero's made a bunch of development dumps over the years that you can dig into to look at some of the work-in-progress that was being made, which could give you some insights into the thought process as well.

For converting something you've drawn on paper to a sprite, my advised process would be: do the scan, digitally block in solid light/dark/mid colors on a layer under the scanned linework, as well as possibly strengthening the outlines if you want to keep them, then scale it down and touch it up on a pixel scale, rather than trying to pixel over a grey mess from scaling down the image without blocking in the colors first.  An alternate technique that some game companies used (not id, to my knowledge, unless they did this with Commander Keen and the like) was to overlay a clear grid on the hand-drawn pictures, and then pixel in the digital image separately just using the grid as a guide for where pixels should be.

If you want to look at a few examples of ways I've approached building stuff, you can look here:

http://ettingrinder.youfailit.net/drawing-graphics.html

These aren't really tutorials or even quite a step by step instruction but they might give you some idea.

As for what tools you use, be it software or Wacom tablets or whatever, good tools can help, but a crude tool used well is still going to give better results than a good tool used by someone with no clue what they're doing.  A mouse and GIMP have been "good enough" for much of what I've done and for the longest time there were people that still swore by MS Paint, up until some things were changed in that program at least.  Practice with the program you intend to use is certainly good to have, and fancy features can sometimes speed things up (although misapplying them as lazy shortcuts generally doesn't end well) but the underlying understanding of what you're doing in a program-independent sense is much more important, so don't rush out to buy a "magic bullet" to make your work improve, because it doesn't work that way.

20 hours ago, der_einzige said:

That may sound good on paper, but that's a stupid idea, I already tried drawing them and taking photos, it turned out to be a massive mess. The actual sprites are quite small (only like 60x40 pixels), and trying to resize an IRL photo to that is ridiculously stupid.

Do I need to say it?

I apologize for my previous post; it is not a great way to approach thread by saying, that editing and creating sprites is awful and annoying. Sorry. I have just started making sprites again after 5 month break and at the same time such an interesting thread appear and I can't say anything reasonable. :P ETTiNGRiNDER also posted very useful stuff! (how do I quote forum member @ -style?)

I decided to share one of my technique I use in Grafx2 to contribute something to the thread. Also I share intimate image of my Doom I intermission screen work-flow from 2016 to 2018 (I've worked inside that image quite long), which includes a lot of embarrassing stuff. Hopefully it'll give you something !

@Lore First of all, there‘s really nothing embarrassing about one‘s workflow, as long as you actually achieve what you wanted. What counts is imho the endresult. Secondly; Nice, that you use Grafx2. I also re installed it recently, but must admit, that it has some quirks which I did not get used to anymore. I am currently making some custom skies and still use my trusty old „Corel Photopaint“ (talk about embarrassing) because I never bothered to buy Photoshop and now I am too scared that a change will be too different. xD

On 9/17/2018 at 10:21 PM, der_einzige said:

That may sound good on paper, but that's a stupid idea, I already tried drawing them and taking photos, it turned out to be a massive mess. The actual sprites are quite small (only like 60x40 pixels), and trying to resize an IRL photo to that is ridiculously stupid.

I re-read this, and then realized what you must have done. More on that later.

As I mentioned before, this method has the following advantages:

• You get proper sizing/aspect ratio of all portions of the model (head, arms, legs, torso) all sized with proper perspective.
• Proper hidden surface removal and proportions for 8, 16, or whatever angles of rotation.
• Proper lighting and shading, regardless of source color.
• If source is colored properly, then you get proper colors with shading too.

The only issues with this method are:

• You need to be able to sculpt.
• Your model must remain pliable enough to be able to adjust it for walk, attack, pain, and death frames.
• You will need to touch up the pixels just a bit.

You said your attempt turned into a massive mess. You didn't read my post, and you didn't set up a model on a rotatable platform, right? My guess is that you just broke out your phone, aimed it at your GI-Joe doll, and winged it :(

Do you realize that the method I described is how id made the Baron (and others)?

Yeah, I guess I should have mentioned that the camera position needs to be securely fastened at the proper angle and height, the model needs to be centered on the platform, and the platform needs to be set to precise angles. Furthermore, you need fixed, adequate light of proper color to illuminate the model, and you also need a solidly-colored background, preferably of a primary color, or white or black, with no shadows. This color must be different than any colors on the model, so it can be cropped out. These are things that just seem obvious to me, to have a chance of producing high-quality master images.

The fact that you screwed it up doesn't make the method "stupid". You replied with a completely asshole-ish response to me answering your question, providing you a detailed, honest approach that is one of the fastest, easiest ways to produce great results. WTF, home?

You throw around the word "stupid", yet it was you that made 4+ separate threads asking how to make a sprite appear in Doom, regardless of dozens of available tutorials, and literally thousands of example WADs to examine.

"der" indeed.

On 9/18/2018 at 6:16 AM, der_einzige said:

how exactly is better than photoshop?

Well aside from the drastic difference in price, it has built-in features for animating, as well as a whole slew of tools tailored specifically for pixel art rather than regular photo editing. It also has RotSprite built-in to optionally use for scaling sprites, which I used to use all the time for my sprite editing. There's a laundry list of other features that it has but if there's a Trial Version that you can try out for free, I'd recommend it.

On 9/16/2018 at 3:47 PM, der_einzige said:

I can draw on paper, quite well, it's just that using a mouse seems very coarse and bad. 

This right here is your problem dude. Well at least one big problem. While it is entirely possible to make sprites and digital art with a mouse it's really not advisable, even talented artists can struggle with a mouse. If you don't want to model in 3D then I suggest you buy a fairly cheap graphics tablet, it's as close as you're gonna get to drawing on paper without paying through the roof and I assume you don't want to fork out hundreds for a pen tablet monitor. You'll have to get used to the tablet not having it's own screen though. Rotation can be a real bitch by hand so references should help a ton. Ideally you'd want a program with opacity and multiple layers so you can use the previous sprite as a guide for the next. GIMP is free and should have all you need. You can find many spriting tutorials online with a quick google search so that might help. If I were you I'd just learn 3D if you haven't already and make digitized sprites like in DOOM 64, possibly with a pixel shader to make them look more 2D. It's gotta be way easier once you know how. 

I just remembered Eriance (goes by a different username now I forget) had a tutorial on the ZDoom forums that used GIMP on the process he uses, if you can find that I'll bet it'd be helpful.

1 hour ago, Death Egg said:

I just remembered Eriance (goes by a different username now I forget) had a tutorial on the ZDoom forums that used GIMP on the process he uses, if you can find that I'll bet it'd be helpful.

He once had a tutorial to give a zombieguy a big belly, but it seems to have disappeared? Would love a link to that.

I use Photoshop and SFM for spriting. I pose what I want in SFM, export it, open up Photoshop, open the picture I've made in SFM, resize it to Doom sprite dimensions and delete the green pixels that are a leftover from the greenscreen in SFM. If needed, I add/recolor stuff and use SLADE for putting my sprites into a WAD. Though, I draw stuff from time to time, like textures.

On 9/20/2018 at 2:06 PM, Death Egg said:

Eriance (goes by a different username now I forget)

Amuscaria.  Still somewhat active in the ZDoom community but not Doomworld to the best of my knowledge.

I also recommend Graphics Gale, which has built in animation features (onion skin, etc.) that help quite a bit when animating. Also, it's now free since a couple of years.

https://graphicsgale.com/us/

You have to find little 3x3 pixel combinations that "crunch", or "facet" (think gemstones) well, yielding more visual info than an untouched photo would, plus know how to create good gradients that aren't boring or noise up. Kinda cubist, like some people (idk) claim Picasso's portraits tell more about the person they portrayed than a naturalistic/realistic approach. Difficult to explain, it helped being around when sprites first came around, going from monochrome 8x8 to Doom's massively colorful "64'ish by 64'ish" (a 128 texture's a huge image in Doom world).

Photoshop's not a great Sprite editor. Not for viewing animations & for having mostly options to do other things than painting pixel by pixel. I did the blue caco-looking thing on the left in PS, but starting on walking/writhing monsters with more complex anims I switched to Pro Motion. I'm no expert, but definetely get a dedicated tool if you think you have the fire & drive to squeeze out some diamonds.

For more info, ask the modern masters over at Pixel Joint. There's no shortcuts. Brutal Doom's achievment in spriting alone is epic, not to mention Cloud & Carmack's, who achieved a glorious combination between photo reference, sculpture & pixel art, one the world had never seen before & likely will never see again.

On 10/4/2018 at 6:07 PM, Soundblock said:

You have to find little 3x3 pixel combinations that "crunch", or "facet" (think gemstones) well, yielding more visual info than an untouched photo would, plus know how to create good gradients that aren't boring or noise up.

You're really on to something here. I'm no artist, but I know what you're talking about here (I know it exists, but I don't understand it.) Basically, if someone drew the best, most perfect Doom Imp, and then resized it to the same size, it would not produce good results. There's just something about how Adrian and Kevin's sprites that gives them that perfect balance between clean, sharp definition, and smoothness, in just the right places, to make those low-res sprites come alive.

Is there any way, without spending too much effort, that you could elaborate, just a bit? I want to understand the "3x3 crunch/facet" that you're describing. For a long time now, I've been wanting to write some sprite helper programs, but I keep running into this issue, where I know something is missing, but I can't pin it down. I know it's there, but I just can't put my finger on it.

Your post is the closest I ever been to finding someone that understands this phenomenon. You suggested that, for more info, to contact the people at Pixel Joint, and I will. But, it's not easy to put in words (as you can tell from my chicken-scratch post here :) So I thought I'd ask you for just a little bit more info, since you've described what I was trying to visualize pretty well. Thanks in advance.

23 hours ago, kb1 said:

perfect balance between clean, sharp definition, and smoothness

[...]

Is there any way, without spending too much effort, that you could elaborate, just a bit?

I think you elaborated the answer to your own question well, lol. It's hard to describe in general terms. Setting a 3x3 grid was more to get OP to think about individual pixels and how they constallate, especially when close together. Thinking in ridges & folds also helps, in order to define well. It's gotta look clean. Sharp without being too contrasted, and smooth without losing definition. Doom has the added level of how it scales based on distance, making oft-used spriting techniques like dithering an undesirable solution, since the individual pixels will start jumping around the figure when it scales, if dithered. Becoming intimate with the limited color range you're working with is key. id made some interesting choices here, going for many very subtly different browns/beiges, for instance. As a contrast, computer generated palettes to maximise the number of available colors in a limited palettes typically don't do well. This talk by Mark Ferrari is pure gold:

Btw, besides Adrian (which had the most nit-picky hands-on approach to editing his graphics, compared to Cloud - apparently Adrian held onto Deluxe Paint for a looong time, before another pixel master - Kenneth Scott - eventually converted him to Photoshop during Quake III production) my other great guru of original contemporary pixel art is Bitmap Brothers' Dan Malone. His work on Chaos Engine, Speedball, etc. is just stellar. Best artist of his time on the platforms I played (Ferrari was a later aquiantance for me, but he's top 10 pantheon too).

I think everyone summed it up pretty much, no tricks to it, be good and draw, or do what the guys at ID did and takes some photos of figurines or something.

Good image editing software goes a long way, personally I use Paint NET.

Or you know, steal. That's an option too, I suppose?

On 10/6/2018 at 6:16 PM, Soundblock said:

I think you elaborated the answer to your own question well, lol. It's hard to describe in general terms. Setting a 3x3 grid was more to get OP to think about individual pixels and how they constallate, especially when close together. Thinking in ridges & folds also helps, in order to define well. It's gotta look clean. Sharp without being too contrasted, and smooth without losing definition. Doom has the added level of how it scales based on distance, making oft-used spriting techniques like dithering an undesirable solution, since the individual pixels will start jumping around the figure when it scales, if dithered. Becoming intimate with the limited color range you're working with is key. id made some interesting choices here, going for many very subtly different browns/beiges, for instance. As a contrast, computer generated palettes to maximise the number of available colors in a limited palettes typically don't do well. This talk by Mark Ferrari is pure gold:

Thanks - I'll check it out. You know, there's some really good modern techniques for restoring and resizing sprites, like the ones built into GZDoom. They do just about as good as can be expected of a computer. But, there's just something artists do that the algorithms cannot touch. They do something special with the outer edges that just makes the sprites look natural, even at ridiculously small scales. I'm hoping to be able to create an algorithm that can duplicate some of those abilities, but I'm afraid that it requires a human's ability to know what looks right.

On 9/16/2018 at 2:49 PM, der_einzige said:

I know HOW TO sprite, I have done countless of sprites and tested them in-game - with no problems.

 

But there is just one little thing.

 

Making custom sprites, not from importing them from a game for an example, seems to be incredibly difficult. I have Photoshop, and am skilled with it, but when it comes to making sprites, I am out of luck.

 

How do you make your sprite, how did ID Software make their sprites, even more so, how did Sgt. Mark IV make all of the thousands upon thousands of very high quality sprites for Brutal Doom.

Excuse me, WHAT KIND OF ANSWER did you expect to get??

MAGIC - you know how imps can shoot fireballs?  sprite artists use dark magic...

They practice!

You are good with drawings but not with sprites?

How about sculpting?

How about 3d models?

How about oil and water painting???

These are different fields that require different methodology to practice, the sarcasm is because i don't understand what anawer the OP expected (Forgive me for guessing but I BELIEVE the OP was looking for a way to reassure himself for his laziness, you are "out of luck?", then quit tying AND instead, promote-self pity discussions ABOUT IT then and ask if there is a magical formula to get good).

OP , What have you tried in order to improve your work to get better and for how long, please share...

As far as brutaldoom , sgtmark isn't very good as well, If you were to look at the credits , you would see that mike12 and others made the good sprites , the monster deaths that sgtmark himself made are toon shaded with patches of pixels and all the detail work of the original work removed (see imp death) ,  and the places that do have details are beyond broken (see hell knight killing others).

Addressing the talent issue : Talented people in a specific field, tend to process information faster related to the subject (Talented artists can pick more information and patterns by observing), The only thing that affects, is time spent on learning / practice, In theory, everyone can reach the same level but some can reach it faster.

Talent can also be a curse, depends on the person, Some people pick on things very fast and get compliments and think to themselves ("Well damn, it is pretty good, i will settle in for that), Its pretty common.

People without talent , those who don't get burned out, serve as sink of information and on a constant path of slow improvement.

- References, Blender guru, Youtube, "'How I learned to draw' with RossDraws" and "The Art of Learning: An Interview with Concept Artist Efflam Mercier"