Opposing Camps? (Low vs. Hi-Res Discussion)

[SpacePirateCaine]

Hi folks, I don't usually start a lot of threads on the AGS forums, but earlier in the year, there was a lot of debate, I believe partly due to a contest in the Competitions forum that I hosted, about the differences between high and low resolution artwork and their usage/viability in adventure games.

Note:
The below thread topic is very wordy and very tangental in places. Much of it was written as sort of a stream of thought in the wee hours of the morning, so I'd like to cut to the chase right here for people that don't want to slog through the enormous post below:

"What are the real differences between High and Low Resolution graphics, and what can the members of each camp do to help raise 'awareness', and perhaps assist the other group to really understand and be able to work together?"

Now, it's a no-brainer to ask whether hi-res adventure games exist or not - 2.5D games like Grim Fandango, Vampyre Story and the lot-so-lauded Escape from Monkey Island, and full 3D games like the new Sam 'n Max games or Dreamfall aside, we do have games like CMI and I know that we have lots of people in both camps, and double-agent people like myself who dabble on both sides (though in the last few years, I'll admit I've been leaning a lot further to the pixel-pushing side), but I'd just like to start up a topic for some open discussion about the merits of both sides, how they can be used within adventure gaming - particularly in our own little neck of the woods that is AGS.

I've done plenty of work in high resolution, but in the AGS scene, I consider myself almost on the hardcore end of pixel guys. I post art on Pixel Joint and have an unused account at Pixelation, and most of my work at Deviant art is also low-res sprites. I think I've fallen as much as many others in to the mindset that pixel=low-res. I suppose a lot of this stems from my extreme exposure to low res work here on AGS - almost all of the games released on this platform are low-res, but there are some very outstanding exceptions developed in our very own backyard that I think, if not revolutionizing our engine and rebranding it as a more versatile tool for both high and low res games (Please stop me if I'm wrong about Nelly, Ali), have proven that it certainly does work.

That said, are we more suited to low-resolution games here at AGS? CJ has done an awesome job of implementing usage of alpha channels, extreme color-depths and higher available resolutions, but how many of us are using them? I'd also like to talk about whether this medium of ours really is anachronistic or obsolete or not? I'm of course in the camp that would say pixel art is alive and well, and provides a great challenge (as I believe, at least, that it's a lot more difficult to really convey a lot of emotion in a character with an extremely limited palette, etc, and I love the challenge that it provides). But what about our high-resolution brethren?

High Resolution
We have some very talented hi-res guys here. loominous, zyndicate and Misj' stand out recently as very active contributors, and were key faces in the above-linked debate about the sprite jam representing the high resolution side of the coin. I really respect their work, but at times, I question the viability of their sprites based on my own preconceptions of how a sprite should fit on the screen, so to speak.

Let me mention again, in game development, I'm mostly low-res. I open up the images in my sprite editor of choice, GraphicsGale (Which you'll note is designed for pixel art) and end up with some very interesting issues.

1) 2) 3)
(Apologies to loominous for using these images without asking first)

1: Pristine .PNG image, clipped from loominous's most recent Sprite Jam entry
2: What happens when I use a 0-tolerance fill tool on the background, attempting to make a transparent solid-color background)
3: Who knows what happened there. Looks like something went majorly screwy with the alpha during a copy-paste.

Basically, this is one of those technological issues that can trip up someone, especially who is used to working with low resolution. My natural want to work with .gif files as an animator, and the higher potential for problems in high resolution sprites can really trip up someone who isn't prepared.

Is this actually a problem? No, not at all. With prodigious use of an alpha channel, the first image looks clean and reads beautifully. When I attempt to convert it to a .gif file, problems abound. To avoid this issue with a 'halo' (an issue I think a lot of us have seen with non-alpha sprites) or worse, the team at Lucasarts for Monkey Island used a technique to solidify the edge of the sprites by giving them a one-pixel-wide black outline:



I believe Igor uses the same technique in Ignac (another hi-res AGS venture that I'd love to see in action again). It works like a charm, and looks great in-game. There are various methods to avoid having hi-res sprites reading badly in a lower-resolution environment, so from the pure standpoint of a sprite as being something that can be used in a game, I suppose that there really is no issue with any of the high-res sprites that we see. Now, where was I going with this?

Basically, the question is, how well-adjusted is AGS to the hi-res scene? And are we really equipped to make adventure games in hi-res? I certainly believe that it's possible - a sprite is a sprite is a sprite, but how can we apply ourselves as a community in the creation of 'current generation' graphics... Or at least graphics that fit a little better with the times, as it were.

I'd like especially to invite the members of the community who have successfully completed and released high resolution projects (Ali, Le Woltaire, etc) to educate us on their techniques in smoothly integrating high resolution graphics in their games. I believe that any insight that the rest of the community can gain from your triumphs would be a great asset and help us all learn a lot.

Low Resolution
Now, I've spent quite a bit of time on high resolution. Let's talk a little about low-res. This is a lot more my speed, I think. As primarily a pixel artist, of course, I'm going to be biased about the usefulness and ease of working with low-resolution sprites. I've been doing low-res since I started doing pixel art for a failed team project competition with Geoffkhan, Cameron, Rui and Custard a number of years back. One of the main draws of low-resolution sprites in my mind is the ease of animation. With a tool geared towards that very task, like GraphicsGale (Linked above), animation becomes a piece of cake. In my current project, I'm regularly making in-game character animations of up to 100 frames per loop (Thank you, CJ, for taking the limit off of frames per loop). I can cut a lot of corners by copy/pasting certain parts of a frame, and only making minor pixel adjustments which would (and I may be misinformed), in my mind, require complete redraws of each individual hi-res sprite per frame to achieve the same amount of fluidity.

Is low-res more convenient for our purposes? I'd like to make note that there really don't seem to be all that many games that really fit the bill - most of the high resolution adventure games out there jumped straight to 3D. With the exception of a few games like CMI, there are actually very few games of any genre that I can find that make use of hand-drawn 2D graphics. This leads me to question, why not? It's not like there weren't a few years where this was considered cutting-edge; so was it a matter of 3D just being there at the same time and easier to create lots of various animations without having to redraw frames?

I know of course that it's impossible to discount all of the animation that has been done since the dawn of time which was all hand-drawn and looks just marvelous, but those (with a few exceptions) were done by a whole studio of artists, and took a damned long time to do, in general terms, and also weren't used in adventure gaems. How does the hi-res crowd compensate for this? Of course, nobody ever said that an AGS game has to be animation-heavy, but I'm a real sucker for a lot of fluid movement (Limey Lizard is a great example of what I think visually you can really accomplish with a motivated animator in an indie project, even in a very short period of time).

For the members of the community interested in working in high-resolution, does hi-res make animation more difficult? Or is it just a difference in approach? As a low-res artist, I do all of my work on a digital medium. I actually don't take a traditional approach to sprite development almost at all - I start and finish all on the digital format, at the sprite level (no concept drawings - or if there are, they generally aren't used as part of the final product).

As an independent game development community, how well-equipped are we to develop a full hi-res game? I've seen plenty of concept art coming out of our hi-res members, but most of the actual product leans very heavily to the low-res spectrum. Someone mentioned during the debate I linked at the top of this thread that the tendency towards low resolution in AGS is because we're generally focused on the adventure games of the '90s. I wouldn't say that's incorrect, but at the same time, I am willing to bet that a number of us would be more than happy to develop or play a game done in high resolution like CMI or Toonstruck, though it seems a bit inaccessible.

What are everyone else's thoughts about the differences between high and low resolution?

[Misj']

I'll probably add some more comments in the future

But some (high-res) traditionally animated adventure games:
Curse of Monkey Island (LucasArts)
Torin's Passage (Sierra)
King's Quest VII (Sierra)
Broken Sword (1&2) (Revolution)

I know it aren't many (in comparison to the low-res period) but that was caused by a number of things: for one the decline of adventure games started when the screen-resolution was high enough to support high-res (and the data-carriers were big enough to support the amount of resources).

As for the animation aspect...in traditional art you can either have full animation or limited animation. Full animation looks better because fewer loops are used and the character is animated (more or less) as a whole...it's used in most 2D animation movies, and many of the classic shorts. Limited animation on the other hand is much cheaper and faster. It's used in a lot of TV-animations (most of Disney's TV series use this approach). You basically only animate those things that are important, and the rest is maintained unchanged over a longer series of frames. Disney's Kim Possible is a great example of both techniques: in the slower scenes you can see that less important characters have a very limited movement, and even the body is moved less predominantly than the arms. On the other hand, the action-scenes have a lot of full animation in them, where every frame has to be redrawn fully. Compared to the classic shorts (Bugs, Road Runner, Tom and Jerry) very few modern cartoons are any fun to watch frame-by-frame by the way...there's a real decline in animation over recent years. Part of it is - I believe - caused by tools like flash which 'help' the animator too much, but results in a more boring (less snappy) animation.

Ps. I wish I had time to create any game...hopefully one day.

[Shane 'ProgZmax' Stevens]

Misj-

By 'limited animation' do you mean the technique Filmation used to use in their cartoons to re-use animations for existing characters?  One example would be their Star Trek animated episodes, where they would reuse the same head tilts and walk animations for each character over and over to save on money and time.  I actually have these on dvd and one of the episode writers remarked how cheap Filmation was in those days and how they didn't like to make any animations they couldn't reuse in later episodes.  It's a shame, too, because they clearly had some talented animators at their studio.

[Misj']

My apologies for the length of this post...and it doesn't even really answer much of the original question. I haven't had time yet (wrote most of this post on my Palm-PDA in the train).

An important concept in animation - and particularly TV-animation - is, that every original frame costs additional money; the more low-budget the product, the more important this rule. As a consequence, animators have found several approaches to reduce the costs by reducing the amount of frames required in creating an animated movie. Below I'll describe a few of the approaches taken, and also look at the possible impact this might have on an adventure game that is based on traditional animation.

The first thing that is often applied is lowering the amount of frames per animation. So if an animation is 24-frames per second (for AGS this value would be 30) you can reduce the costs by creating a 12-frame per second animation and holding each frame for two frames. This is referred to as on twos, and works generally well for most but fast animations. Using on threes or fours is also possible but starts to get obvious to the viewer. Reducing the number of frames per animation is common practice in adventure games as well, where a walk-cycle can be as few as four frames.

Another common practice is re-use of animation footage; which is what ProgZ referred to. As mentioned animations cost money, and if footage is used twice, the technical costs of that animation is halved. There are two ways to re-use an animation: either directly or traced. To exemplify the two, I've added a few images that could be found here.

Direct re-use

Tracing

It should be mentioned that there is nothing wrong with this approach (as long as you own the original copyrights...even if you trace), and both approaches reduce the production costs; although tracing is - of course - slightly more expensive than a direct copy. Both methods are valid to adventure games as well: the less frequently an animation (or its backbone) is used, the more expensive it is. So in order to lower the costs, and approach often taken is to have for example a single 'take object' animation that is independent of the object taken. Even though specialized animations are more rewarding to the player, they are not a practical approach...just imagine the difference in workload between a character who pulls up his shoulders if he doesn't know how to interact with an object, or if you had to create a special animation of every object the player can interact with in the game.

This brings us to another difference between a game and a cartoon: interactivity. In a cartoon you know and control exactly what the character will interact with. Consequently you can use that knowledge to keep specialized and stock animations mixed quite well to provide enough 'newness' to the viewer (balanced throughout the entire experience). In an adventure game the player can do anything at any time. This means that as a designer you can much less control the frequency of a new animation. Also, since a lot more interaction is possible, stock footage is visible on-screen more frequently. During design, it can be important to think about key animations (that should be special) and common animations to decide on which ones - and at what positions of a game - should be spend more (or less) time.

Loops (or cycles) are a special type of re-using animation footage. Loops are repetitions of an - often short - animation sequence. A common example in adventure games are the walk-cycles. Loops are a great tool to extend the length of an animation without the need to create additional frames. Loops are extremely useful to put some movement and livelihood into a background/scene (think about falling drops or waves hitting the shore). The problem is, however, that once a loop gets noticeable as such, they actually take away from the illusion of life. While more extensive/more complex loops take longer before they start to feel repetitive, they also require more frames and thus more money. Therefore simple loops are often preferred.

I've grabbed a small part from the intro of Filmation's New Adventures of Gilligan to show the value of a loop (in four frames, on fours):

-- the animation speed is reduced for clarity --

At this point I should mention another difference between cartoons and adventures games. Where cartoons are foreground oriented, adventure games are background oriented. Many TV cartoons can do with very simple background (sometimes just two lines or even merely a gradient) because the background shouldn't compromise the readability of the characters. Also, cartoons can use character close-ups; which in adventure games are scarce because close-ups generally prevent the player to interact with the game-world. Lastly, in adventure games backgrounds are often presented in an overview shot (even scrolling backgrounds) with a more-or-less static camera as opposed to cartoons where camera movements are easily implemented and a cheap way to add movement to a scene. All these differences - apart from other functions - have the effect that loops take more time to be noticed in cartoons than in adventure games, where the player dwells for a long time at a single background, exploring every aspect of it.

This brings me to the difference between full and limited animation. Before I go into it, I first have to say that traditional animation thinks in layers (literally in the old days where different cells were overlayed (layered) on op of each other). This concept is particularly obvious for limited animation. Below I've added two short talking animations from Disney's Kim Possible. On the left you see full animation, and on the right a (quite extreme) example of limited animation:

-- the animation speeds are reduced for clarity --

There is a drastic difference between both animation styles as you can see...full animation is much more dynamic than limited animation. I have to repeat though, that this is an extreme example of limited animation (played at a very low pace); but I have chosen it because it presents the basic idea very clear.In limited animation characters (and everything that is animated is in a way a character) is split into multiple layers where for example one (the mouth) is animated, while another is held over multiple frames (the face). One advantage of splitting the character into layers is, that it is easier to have multiple animators work on a single character. One might for example animate the face, while another will add dynamic hair. However, when I talk about splitting the character into different layers, these layers are not static. Just because layer two contained eyes for frames one to twenty doesn't mean that in the next frames it can't contain the character's hand.

Limited animation is also quite visible in the clip below:

-- the animation speed is reduced for clarity --

-- I didn't loop this animation, so you might have to reload it in the browser --

First of all, the characters that are not within the focus of the animation are not animated at all. Secondly, the woman's dress changes to a hold at 1/3rd of the animation...even though this at some points causes her upper and lower body to appear slightly disconnected. The other characters that are not the point of focus of this animation are not animated at all for most of the scene. In adventure games this latter use of limited animation is scarcely used...most characters - even if they are not the focus - have some idle animation. But the use of 'layered' characters that are kept mostly still is common practice (it's the copy-paste thing pixel-artists use).

Since limited animation helps reducing the drawn frames, it could theoretically also be used to reduce the amount of resources. To do this, however, the engine should be able to have multiple layers for a single character. This is not a requirement to use limited animation though...only to make use of it's resource reductive effects.

Ok...so much about reducing animation for traditional art...this was an extremely long post (even for my standards)...

[Dataflashsabot]

Well for what it's worth, the original broken sword is beyond any doubt my faviourite adventure game of all time. and that would not have been as good in lo-res. on the other hand, monkey island works really well in lo-res, and just wouldent have the same at higher resoloutions. so i say, to each their own, whichever works. there are however, as has been said, various practical issues which must be considered.

[Babar]

It is kinda sad that when PCs got the capability to do hi-res, 2D hi-res stayed for such a short time before everything went to 3D.

The ONLY reason for almost ALL mainstream commercial games made today being 3D seems to be that it is cheaper than 2D hi-res. Otherwise, it really isn't any sort of natural progression. Although you've got to wonder...if taking into consideration the techniques (which is probably cutting corners, ) Misj mentioned, would it still be as expensive?


As far as the AGS community, there is no denying that most of us find it easier to work in lo-res, mostly because it is easier and faster for us.

[Shane 'ProgZmax' Stevens]

That was pretty informative, Misj!  I was particularly surprised to see the tracing between Christopher Robin and Mowgli since I didn't think Disney needed to use such methods with such profitable stories and talented artists.

[ManicMatt]

Neat thread.

I prefer hi-res 2D graphics over lo-res 2D or 3D graphics when it comes to adventure games. A lot of 3D adventure games look ugly, or seem to be badly designed, with a few exceptions. Broken Sword didn't have crate pushing when it was 2D. However I had more fun playing one of ProgZmax's lo-res C64 style games than I did pushing wooden crates around as George Stobbart.

[miguel]

I feel the same ManicMatt, George Stobbart was much fun to play when he didn't have to push crates
I think hi-res 2D can be the way to rescue adventure games.
And thanks for the post Misj', great work!

[buloght]

I generally work in pixel art because of the pixel charm (well that is an excuse haha, if I could make something similar to CMI I would, but that is a big commitment and talent requirement). What I like about pixel animation is that you can draw only one sprite, and then create animation in parts, (even up until when everything moves). You do not have to redraw every frame, everything can be played with and bugged with. (two extreme examples to demonstrate, THIS and THIS)

Gives a massive advantage over traditional animation I imagine.

(I'm referring to some animations (idle, background, etc), of course most other animations in pixels still require key frames, etc, but even then sometimes still easy to simply modify the sprite)

[MashPotato]

I generally work in pixel art because of the pixel charm (well that is an excuse haha, if I could make something similar to CMI I would, but that is a big commitment and talent requirement).

My feelings exactly :)  I make pixel art, and I enjoy working in pixel art; I also paint in hi-res, and enjoy that too.  However, working in hi-res 2D for an entire game isn't even an option due to time and ability constraints.  Resolution is more than just a number! ;)

btw, as an aside, another example of a hi-res 2D game that came to mind is Guilty Gear http://www.arcade-history.com/images/game/3974_2.png

[Shane 'ProgZmax' Stevens]

Wow, Guilty Gear's characters use very minimal shading compared to fighting games like, say, King of Fighters (which comes off overshaded and pillowy at times, imo).  I notice you point out some of Fool's animations, Buloght, which ties in somewhat with this thread.  As brilliant as his art is, I think Fool has a tendency to waste a lot of frames (he doesn't do as much per frame as he could, bloating the animation size and slowing it down) which is definitely a concern when you are making a game for a memory-limited platform (like a phone or older pc games).  Making the most of every frame is important, imo, even if you've got unlimited storage space.

[Alarconte]

One important reason for the shot duration of hi-res in adventure games in market, was the interest of the sector; The adventures start to making in 3D so soon than they can, 'cause a very important sector of the market only buys games in 3d, without knowing the genre. Is the shootem'up fever.

The hi-res epoque of the adventure games (has a lot of titles, including second-line adventures as Shelork Holmes, Dr. Moureau, o in the spanish situation, the games Skulls of the Toltecs and a lot more.

The thing was that in the hi-res epoque, the adventure genere was down. 'Cause this the need to abandon the 2d to run into the 3d and the multi-genre games (even Dreamfall has combats!)

I love low and hi-res, (and 3d adventures like Dreamfall) but I know I'll never be capable of do a game in hi-res... xD

[buloght]

Wow, Guilty Gear's characters use very minimal shading compared to fighting games like, say, King of Fighters (which comes off overshaded and pillowy at times, imo).  I notice you point out some of Fool's animations, Buloght, which ties in somewhat with this thread.  As brilliant as his art is, I think Fool has a tendency to waste a lot of frames (he doesn't do as much per frame as he could, bloating the animation size and slowing it down) which is definitely a concern when you are making a game for a memory-limited platform (like a phone or older pc games).  Making the most of every frame is important, imo, even if you've got unlimited storage space.

I agree with you, I know that you make your animations very efficient (and with few frames). That is why I said "extreme" examples, I just thought I'd mention how one can just simply animate in parts, and make everything separately and then make it whole, I don't like animations where some parts are copy-pasted and only one part move. I tried to follow this principle with my avatar walk, it's only 8 frames, but I animated everything separately trying to give it life like the head, hair and shirt (I am not good animator though)

In pixel animation one can make everything feel fluent without really redrawing anything, I really like that in pixel animation, it gives everything life.

Mash: I completely love Guilty Gear's graphics.

[edit] Ben: Yes! Metal slug is excellent example of awesome pixel animations.

[ThreeOhFour]

Guilty Gear has some absolutely fantastic animation! I took notes on doing animations while playing that game!

But then again, the Metal Slug games have absolutely awesome animation as well

[SpacePirateCaine]

Guilty gear certainly is a great example of good pixel work 2D animation used in-game. It is very well put together, but despite looking on par with what you may find in CMI or the like, it is actually pixel art - the screenshot MashPotato linked seems to back up my point. It also doesn't seem to follow the Street Fighter series' extreme use of Selective Outlining, and I don't even notice a lot of Anti-aliasing.

I would almost go so far as to say that well-done art in this style could make for a very easy-on-the-eyes adventure game. And at only roughly about 200 pixels tall, wouldn't be so difficult to animate either. A great recent example of hand-drawn art used in a 2D environment would be the newest installment of the Street Fighter II series, as redone by UDON.

[LimpingFish]

The Metal Slug series is probably my favourite example of low-res pixel animation on a grand scale, with Metal Slug 3 the pinnacle. The Last Blade 2 and Garou: Mark of the Wolves are other good examples.

Games that also impressed me are Undercover Cops and GunForce 2, from Irem Corp. Both, coincidentally, developed by talent that would later go on to form Nazca Corporation.

Nazca Corporation = Metal Slug.

[Joshua]

I'm not sure why everyone considers high-res to necesarily involve longer production times or larger talent. A very simple way to quickly create animated characters is to use vector graphics. Here is an example...

The image is 'high-res' (definitely not pixel-graphics), but it was very quick to create. I think this kind of thing would be a good entry point for many animators and also a good method for those without a lot of time - creating good pixel art requires a lot of skill. The penguin is also a very simple example - a combination of hand drawn and vector can be effective too (I don't have any examples though).
Also, in terms of fast but effective high-res animation production Nelly Cootalot is a great example - look at Nelly, only her very basic arms move when she walks and yet it is an effective walking animation. Talking character's heads often move from side to side (very easily done with vector graphics) and only the mouth needs to be redrawn.
So, I believe low-res is not the only option for those without much time or just starting out.

P.S - I am very new to animation and am certainly not in any position of authority on the subject. I just find it interesting that high-res actually seems easier to me. This is my first character animation- I actually created it in Powerpoint then copied it over to Photoshop... yes I realise it's not the best way but it was an easy way to start out.

[Andail]

Joshua, no offense, and you may have a point, but most people demand more from their sprites and animations than flat geometric shapes bobbing up and down.

[cat]

For vector animation you should take a look at Synfig http://synfig.org/Main_Page