Experimenting with new scale
Sunday, August 28, 2011 - 23:27
New blog post here:
http://clintbellanger.net/rpg/blog/20110828
Please share your thoughts.
Currently I'm tempted to wrap up the current demo (plus a couple more tilesets and creatures) into a shorter game, just to have something released.
But maybe it's worth it to aim higher. Maybe I should start working with higher-detailed models immediately.
Do it now :) It's so much better visual quality. Also there probably will be more work required at later stage.
Polish what you have a bit, spend some time on docs, release a demo. Build some buzz, get some help, start again at full quality.
New scale is definitly better, I would go for it better now then later if you plan to do it at some point anyways.
One point though, you mention "high quality normal-maps" in the post... I guess one of the nice things about FLARE and rendered 2D games is that you can save yourself the trouble of UV mapping and real texture painting. This is such a speed up of workflow that I wouldn't sacrifice that for the limited amount of extra detail (but basically you should be able to just make the render mesh more detailed also right?).
What would be really cool though are normal maps allpied to 2D sprites like showcased there:
http://www.asantee.net/ethanon/
Keep up the great work!
--
http://freegamedev.net
I think the new scale looks great and it's a good opportunity to improve the quality of the overall assets.
And for the question whether you should change it now or later, I'd rather do it now - because once the demo game is finished, it will probably be much harder to change the complete artwork without messing up the demo. Just my humble opinion, though.
Why would old art become unusable? Because it lacks details? Or would it just have to be re-renderd?
I would say: stick to the current art for flare and the next game will get better gfx.
@qubodup: Some of the old art won't be the correct scale. Especially stuff that is supposed to be relative sized to the human e.g. manmade object like benches, containers, doors, etc. Some art like the cave walls/floors should look fine without change (but the minecarts will look too small).
For example: the dungeon tile set has dungeon doors/portcullis that are good for the current scale, but would be too small for the proposed new scale. New doors would have to be significantly taller. This probably means all the walls will need to be taller too.
@CruzR: the idea would be to release a small stand-alone game with the current assets. It wouldn't be a demo. It might exist in the same "world" as future games, but that's about it. Most of the engine and a bit of the art will be reused for the next game.
About the Scale, can't you just render the old meshes in a higher resolution to get the required results?
When rerendering everything you can also take that opportunity to give everything a better lighting.
The method that I would use is:
First set a percentage how much art you are willing to change.
Then look at all art and list.
Now you can make a time estimate. Part three, and that is the most important part, is making a decision that YOU feel comfortable with. Don't let anyone dictate you!
@MrBeast: some things can be scaled up. But the things that can be scaled up easy are going to be the trivial pieces, e.g. stand-alone objects. If I need taller walls, for example, stretching the tiles vertically would probably not work.
@wokste: good idea. Let me catalog this.
Dungeon Tile Set: throw this out, make something much better.
Cave Tile Set: keep the walls/floor. Redo all the "mining" decorations and containers.
Grassland Tile Set: make the docks, houses, containers, trees larger.
Base Male: already have a new version.
Base Female: upscale the old version, do a decent sculpt to create a good normal map.
Weapons: Add textures and UV maps to all of these.
Armors: Probably create better ones from the ground up.
Goblin: I'll try to just upscale this and add textures but it might be too cartoony. Might have to create a new model from scratch.
Skeleton: I'd like to make a much more stylish skeleton from scratch.
Antlion: The base model is okay, I might add some good details and a better texture and render it larger.
Minotaur: The model is good, I will touch up some textures to look better at the new resolution.
Zombie: Maybe a good texture treatment can save it.
Spells: once everything else is updated, the spells will probably be lagging in quality. I'll probably aim to get these better.
As for monster models, maybe sculpting tools (like in blender or sculptris) can be used to add more details. Sculptris is good at adding extra geometry details. Especially when polycount isn't that important when prerendering.
It seems like a lot of work to finish up the grasslands tileset then have to redo it later. This scale issue keeps coming up and it seems like it will continue being a nagging thought in the back of your head until it is fixed.
Maybe make a quick (just the basics) hi-res tileset, and put it in-engine and get people to test it? Just scale up some antlions for quick enemies, use the new base male and see how it feels compared to the old size.
Or: Finish the grasslands tileset at the current scale, then make a short game (recruit 5-6 people to each make a dungeon or something... I'll volunteer for one) and release it with just the three current tilesets. Then, redo all the art at the larger scale and work on the bigger (both in visual size and in content ;) ) game with nicer art.
Here is what I think should happen.
Stop creating Lo-Scale (old scale) models and tilesets.
With the currently finished assets we do a Dungeon Jam for 48 hours or a week.
During this Jam using new and old maps we make a story. The first part of a serial story.
Package it all up and ship it as installment 1 of the franchise. Please come back and check for new installments often.
Then the shift to going Hi-Scale should start.
Create New/Replace incompatible art.
Finish replacements + plus a couple new tilesets and enemies
Do Installment 2 Dungeon Jam which includes updated graphics for installment 1
Package and Release
Don't change scale again :) Rinse Repeat
You and I have talked about this plenty and I know I have been one of the nagging voices urging a switch to a larger sprite size for the sake of detail and player immersion. I think the comparison speaks for itself. It will take extra time to make models (which I can help with) but there are other modeleres in the community and blender users outside of the free gamer community who would likely be willing to help especially if we had awesome concepts and a good looking game to show for it. Detail isn't everything, the aesthetic of the game overall and it's gameplay is whats important, but I know the extra room for detail would allow us to do some cool stuff, and make textures more visible.
I totally understand and respect your desire to finish out the Averguard story arc, and I think you should do it if it would make the transition to more detailed visual assets more natural for you. If creating the short game first takes the worry and personal controversy out of making the new sprites, I say go for it.
You've already got enough stuff to do a three level demo, if you polish it up and make it shine you'll be able to release it, releasing is the bane of open source games. Once you have a released product for people to get excited about the helpers will start emerging from the woodwork. How long do you really want it to be before you have *something*? Redoing all the graphics at a higher res sounds more like a sequel, if you get into that cycle you'll never end, it'll be like duke nukem forever. Just focus on getting something finished, the shiny parts can come afterwards.
I'd like to see things in the current scale released as a game, to put those old assets to use rather than just throwing them out, to generate some hype for FLARE and/or OGA, and to keep morale up. Having something released will at least give you something you can point at and demonstrate as a finished product, even to yourself. It could also be a nice test bed when more people play it and give feedback that can be used for the next (higher-res) FLARE game, and could get the lore/story started if the games are connected. I could certainly see meeting Scribe Guill as an old man in the sequel.
A lot of valid points have been raised!
I like the ideas of just doing a demo/small game, and then upgrading to the new scale after it is released. But how many people will base their options of a future Flare game on the graphics of a demo/small game? (Do first impressions count for game sequels?)
Another point to consider is other people who create art for Flare. (or wish to make their own games using Flare)
I have seen a variety of art on OGA that is stated at being "flare compatible", or to the likes there of.
The longer you postpone the transition, the larger the amount of incomparable art may be generated.
I suppose Flare could be modified to use a user definable scale....
If this is a definite goal for the future, the longer it is postponed, the more time people will have to spend improving/creating new assets for the game.
Some of the most noticeable features of any game are the visual aspects, better visual quality is likely to attract a larger audience.
Weighing up the pros and cons seems a bit difficult as the time factor needs to be considered.
If converting to a better scale does not require too long a time period, then the sooner the better!
If there is a target release date for Flare / Flare demo, maybe postpone it for the next incarnation of Flare.
I for one am eagerly anticipating the future version/s with this improvement!
Honestly, though bigger means more detail on individual actors, smaller means more stuff on the screen. No matter how good graphics are, they get old after looking at them for a while. In an isometric game, smaller is better.
@anonymous: I haven't looked at it from that perspective, but I must admit, you're right.
I think the Duke Nukem Forever reference is most appropriate. If you ask me, work with what you've got and make an awesome game out of it. After that you can make an even better sequel with the new art scale and many exciting new features.
I think there are already a number of reasonable RPG engines out there in the open-source world but the main problem is lack of a polished and complete game to actually make use of the technology. You're a great content creator, Clint, and that's what the open-source community needs right now.
After throwing those thoughts out there: Do what you feel comfortable with. You don't owe anyone anything so continue working on Flare (or abandon it) as you see fit. Thanks for sharing what you've achieved so far.
Thanks to everyone for the valuable input!
I think the correct move here is to polish the hell out of a small game with the current assets.
Part of my long-term plan for this engine is to build several similar games. Each time I change or add a feature it's an opportunity to refactor, generalize, improve. Do this over enough iterations and the final result could be a very usable action RPG engine. If I try to build a "perfect" first game I don't get this useful iteration. The second game could be really great if we learn lessons from the first one.
With a small game out and some good-quality preview art for a second game, we should be able to attract attention from artists and devs.
Sign me up for creating some of the content for the first game. If you've got any specific maps in mind I can do, just let me know.
Also, if it helps, I'll turn my map making into a tutorial so that others can create more content too.
It will be a lot easier to get a steady influx of potential contributors if there is already a game to play. It doesn't have to be vast, or perfect. Already the releases have been pretty fun to hack and slash through.
Also it could be an opportunity to 'hand off' some responsibility. There may well be people happy to take over finishing off or adding further content to Flare RPGv1.
http://forum.freegamedev.net/