Pros and Cons to using "programmer art" initially in an AGS project

[Greg Squire]

In what little I seen of AGS so far it appears, that it's best to start with final (or "near final" art), before you put anything into the engine.  If not you might be doing a lot of things several times over, like defining walk areas, walk behind, areas, etc.  In the general world of making games, often "programmer art" is used and then replaced with final art once the game nears completion.  However it appears that this might not be as easy to do in an Adventure Game. 

To me it seems the best way to go about creating an AGS game (or probably one some other engines as well) would be:

1. Write the complete design first; including Story, Puzzles, List of all assets needed (rooms, characters, animations, text, cutscenes, music, SFX, etc.) May want to storyboard.
2. Create Art and Animations. (Pick a resolution and color depth and stick to it)
3. With final art in place, start building the game in the AGS engine. 
4. Add Music, SFX, and speech either before, after or during step 3.  These seem to be more easily swapped out and replaced (though speech might be best at the end, in case there were text additions or subractions in step 3.


So what are your opinions on using "programmer art" to start with and switching to "final art" later (especially if the art is contracted out). Is it too much of a hastle to do in the engine? or is it not as big of a deal?

[m0ds]

3. With final art in place, start building the game in the AGS engine.

This is the only statement I disagree with. You don't need final art to "start" working on a game. There's a quote by Ron Gilbert, which, if I hadn't already posted it about 10 times, proves that he and his teams often used preliminary art first off, just to get from game start to game end to get it all in place first. This is all people really need do.

So basically, yes, start with programmer art. It'd be a beautiful world if we could all start with the final art, but that's in a perfect world, and can take absolutely aaaaages to happen. No disrespect to artists, but often when a designer needs 100 rooms...well, you know what kind of workload we're talking about

It's no big deal if you're enthusiastic. I'm working on a game and, I rendered for eg. 3 backgrounds, did all the walkable areas, ran the game, blah de blah, and then realised the angles were wrong, etc. So, I had to re-render those three backgrounds and then re-work the walkable areas, the hotspots etc. It's no big deal to me, because...well, when you design games, make films, music etc...you know problems are going to arise. And part of your passion to do these things is your willingness to go back over them and modify, correct and improve.

[zabnat]

What if I can't do any programmer art? I mean I make the graphics for my game and I'm a bit of a perfectionist. So if I start doing any art, I'm gonna do it "properly". But doing art really does take a lot of time (just read which talent is most wanted in RAT-thread) and it creates a bit of dilemma.

[m0ds]

And then check the "Offer your services" thread, and you'll notice there are 23 musicians for every 1 artist

And not really on topic but on the subject of ratios, AGS has a 6:1 male/female ratio nowadays, which is actually really, really good! Even Dart was online last night! LOL

[Chicky]

So many people i forgot about, i lost completely lost touch with ags.

I use programmer art but i'm more of an artist than a programmer, it helps build structure and keep you interested (the irony).

[zabnat]

And then check the "Offer your services" thread, and you'll notice there are 23 musicians for every 1 artist

What I meant was that me doing art is as rare as everybody else getting art, so that's why my games tend to be on hold most of the time.

[Greg Squire]

[... It'd be a beautiful world if we could all start with the final art, but that's in a perfect world, and can take absolutely aaaaages to happen. ...

I realize that we can't always start with final art, as there will always be some tweaking that will need to be done later, and in the case of a team project, you don't want the programmer(s) sitting around doing nothing until they have final art.  I guess it boils mostly down to "Is the background/room layout final or not?".  I can see that you could do a rough sketch of a room and use that to start building the game, as long as the placement of floors, doors, and items isn't going to change a lot.  Meaning that you won't have as much to change.  However if that sketch is going to change massively, then most of the work to build that room in the engine might be wasted.

I know that programmer art can be useful in other game genres, as it may help identify art assets that won't need to be built at all (if a feature or whatnot is cut from the game), so that saves some effort there.  That could possibly be a benefit in an AGS game as well.

It seems to me that the opposite ends ("complete final art" or "almost no art") are not usually possible or practical.  I guess the ideal might be somewhere in the middle, where you use "early art" (basically sketches with little or no inking, coloring, or shading, but they are ones where the layout is mostly final). Then later the final art is dropped in and only minor tweaking to the room would be needed.

[Layabout]

I think the best thing to do would be to layout your backgrounds (using programmer art or whatever), you know the basic outlines with important details for things in your game. This will save you time if you have to, for some reason, get rid of a scene from the game (not a massive issue, but having an empty screen just needlessly drags out your game. Some scenes may go through a redisgn so they work better for the game, something you will not do with a completed background.

As for sprites, it's up to you really. You could use roger or you could do your main character sprite. Unless you do a complete redesign, there should be no reason to change this.

Another good idea is to set a style guide, so you can tell the tone of your story from that style if you know what I mean. Sometimes you may find some art goes well whilst others do not, this can all be set into the style guide so you don't have this problem. Esp things like colour selections for common elements like Wood, grass, etc.

I know for a fact Dave uses programmer art, but this may also be out of necessity of getting commercial artists to produce the art whilst he is scripting the game.

[Shane 'ProgZmax' Stevens]

Dave doesn't use programmer art much anymore, at least he didn't really need to for convergence.  I know ECC used some unfinished sketches for sprites and backgrounds in spots, but those were done by the artists.  

As far as 'getting it done', the only method that matters is the one that works for you.  I don't 'do' programmer art because I'm already doing all the art for my game so it would be absurd to do a bunch of poor sprites and animations and then go back and redo everything properly!  There's nothing wrong with doing the art yourself in a pinch, provided you're sure you can find someone to get it done the way you want.  

I think what designers will see most of all going into a project is its evolution; practically every aspect of the game will evolve as you go along in an amateur project.  The art will steadily improve, possibly to the point where you'd like to go back and redo everything to match the improvements the artist has made.  The music will probably change from the start of the project to the end, and you'll find better ways to script things and organize your game as you go along.  It's very much an evolutionary process, and it's fun to go back to early stuff from a project and see how much you've improved since then -- just as long as you can resist the urge to redo everything .  This happens with programmer art too, by the way.  I've seen people who claimed to have no art skill improve over the course of a single game to having a respectable style of their own.

[Dave Gilbert]

Oh, i used plenty of programmer art.  Here's a neat example.

For at least six months, the Minetta tavern pictured here:


used this lovely piece of programmer art:



I knew that the room needed a back door, a bar, an exit, and the portrait of Joe Gould on the wall.  Anything else was just superfluous   

[Ali]

That's very impressive Dave! Are we to assume from the 'yar' that at one point Joe Gould was a pirate?

[Ghost]

I always feel better when I have at least a main character's static poses in the game, but I tend to use placeholders for minor characters, objects and inventory a lot. Same with rooms, mostly it's outlines until I really know where things are supposed to be. Cutscenes without animations look boring, yeah, but once you got the text in, and know that you really need to have two characters doing some violent stuff in the boxing ring, then it's time to add the fine details. That's where I see the real use of rough art or whatever you want to call it.

[Tyr]

I've found not needing to redo walkareas et al if you just  put in a changed version of the existing image. For some reason.

[Ghost]

I've found not needing to redo walkareas et al if you just  put in a changed version of the existing image. For some reason.

That's because as long as you do not use a different sized background, everything you already imported stays if you import a new background- walkbehinds, objects, everything. Which makes using a rough sketch/programmer's art all the easier.

[edmundito]

One of the key things about designing anything good, including a game, is to do things in three simple steps, aka the iterative process:

1. Think of an idea
2. Try it out
3. Keep changing it until you feel it's good enough

Trying it out is crucial, and programmer art allows you to do this. Once you see something in practice you'll notice flaws and things that could be improved than just theoretically working in a vacuum in Photoshop or whatever. The prettier you make it, the less you are going to want to change it because you've invested so much time on it, and therefore it costs a lot more (time, money, and emotional connection) to change it.



In additon, I'll give you a lengthy story from the trenches, as I've been working with Grundislav on this new game. And definitely, "programmer art" has come in handy.

I'm doing a lot of the scripting, he does the final art. The rooms was a tricky situation because I needed something asap before Grundislav could take good time to paint it. So we came up with layouts of the room. Depending on how complex the use of the room would be, we would either come up with the room as a sketch or as a top-down blueprint-type plan of what it would look like, which I would just import into AGS to start scripting.

The best part about doing this method is that you have a few chances to iterate on the room. If I needed something to be there, I'd open it up in something like mspaint and just add whatever I needed or roughly moved stuff around until I was satisfied with the design. Bluecup (sprite 0) was also my best friend for the time being.

In AGS, I would be very rough with the walk areas, mostly using the square tool, and never mind the walk behinds! The nice thing about AGS is that when you import the final background you can just nuke the walk areas completely and do a nice final pass with scale, lighting, walk behinds, etc, which is exactly what we did as well.

One handy thing we did was that instead of using Character.ChangeRoom(room, x, y), we would call Character.Change Room(room) and then set the characters position on the room script:

Code: ags

function room_Load() {
  if (player.PreviousRoom == 5) {
    player.x = 50; player.y = 100;
  }
}

This made it much easier to reset the entrance positions than to have to wonder where the ChangeRooms were.

Of course, there's always the amount of time you have to do this though. This is the "con" you have to fight with. We only picked the rooms because we felt they were a lot of the action took place, but we also are working in our spare time and we have full time jobs, so we couldn't get too picky. And it was handy to divide up the work when two people are working on the game.

[Dualnames]

Wow dave, you shouldn;t have hired artists if you can draw with orange!!

[auriond]

I've been having problems with programmer art. I had a room done up fairly quickly so a scriptor (Dualnames) could work on it, but in my haste I screwed up the perspective. It's complicated to explained what happened next but the upshot of it is that a character's movement was done along a 2D side-to-side plane when it was supposed to move through a 3D space. This was fine because my bad perspective hid much of the floor, so you couldn't see where the character's feet were.

Then I drew the final room according to its original dimensions, and discovered that it was much longer than the programmer art, AND you could see the character's feet nice and clear. That meant an almost complete overhaul of the scripting. It was completely my fault, of course, and I learned an important lesson from it: programmer art doesn't mean an approximation of the final room, it means an ACTUAL outline of the final room to be used, otherwise be prepared to do things over.

[Huw Dawson]

The primary reason I never make progress on my games (despite having a great deal of plot/puzzle work done) is that I always end up getting half way through a background and giving up... Sprites are worse, of course. I think any budding game maker who has got as far as sprites usually ends up getting demoralised, because the sheer amount of work that is required for one walk cycle is just not fun...

Of course, then you go and pinch RON sprites.

Really, if you use programmers art and stolen sprites and actually finish the game, you can advertise that on these forums and an artist will be far more likely to step forward, I feel. Then it's simply a case of helping the artist to plan out the graphics and spend the time when they're helping with the graphics to dedicate to testing/refining.

- Huw

[Igor Hardy]

To be honest I can't imagine a reasonable gameplay design method other than constantly play-test and develop your ideas at the programmer art stage. If I would just stick to design documents I wrote, the gameplay would feel artificial, boring, and I'd probably mostly reuse ideas from other titles anyway.

[Radiant]

In what little I seen of AGS so far it appears, that it's best to start with final (or "near final" art), before you put anything into the engine.  If not you might be doing a lot of things several times over, like defining walk areas, walk behind, areas, etc.

That's not actually a problem: a room with programmer art doesn't need walkbehinds, and setting the walkable areas for a room with programmer art takes about thirty seconds.

So yes, I do use what I call "scribbles" (similar to what Dave posted) for rooms, a big for portraits, or Roger for walkcycles, and plug in the finished art when it's available. AGS makes that easy.