General Project Thread & Feedback

Wouldn't minor updates be what Twitter is for (not that I have a Twitter account, but from what I've heard it sounds like the right place for this sort of thing)?

 

...So what was wrong with the sound engine anyway?

 

I'll be happy to answer this question since it's something I haven't heard before and is actually related to the game. The sound engine had some things it didn't do correctly, mainly with OGG files. The sound engine resamples the audio to fit the output sampling rate, but it didn't do this correctly for OGG files. But now it does =)

 

https://www.facebook.com/sonic2hd

 

Excuse me if this has already been answered, but what's the current issue?

Spoiler

I mean, I can just go download all of the frameworks and play a different demo by now.

 

Hamneggs, I'll assume you're talking about the Sonic 3 project, which is no way related to our work at all. We wish them the best in their endeavors. As to our main issue at the present, LOst and Saxman are currently battling integration issues between the game engine and the sound engine. Separately, they both work fine, but, needless to say, that doesn't make for a fully playable game. We took about a week's break working on the project in lieu of spending time with family for the holidays. And on that note:

 

Didn't want to post this in the Tech thread since it doesn't specifically relate to Allegro.

I know a lot of effort (and heartache, etc) has been put into the sound engine, but considering you're writing an engine from scratch as opposed to scripting something like GM/MMF2, is there a particular reason you aren't simply using audio engine middleware like fmod or Wwise? Both are legally free for non-commercial use (granted, a remake of an existing IP is somewhat dubious but ultimately I don't think they'd care) and honestly speaking why reinvent the wheel?

I know there's nothing like being able to say you did it all yourself, but if it really is the main thing holding progress back, why not? Not to mention it gives the guys in charge of audio/music a shit ton of control over balance/presentation/etc without having to constantly pingpong back and forth since the toolset more or less is already there. There's also the fact that knowing how to work with audio middleware is quite a big thing to put on a resume; perhaps as appealing as having written a simple audio engine from scratch.

 

Oh god... wouldn't that just be a dream .

 

There's no problem with our audio engine. We've recently determined the problem lies in how we're working with DirectSound. It has proven to be a learning curve for us, but we're getting there.

As for using another engine: our audio engine is specially taylored to S2HD's needs, unlike the generic brands that give us a whole lot of stuff we have no use for and not enough of the stuff we actually do need. We have YM2612 and PSG sound emulation, OGG playback and looping support, and we even have some digital signal processing stuff built in such as compression and interpolation. We've been able to do some really interesting things with it. One example that comes to mind is the spindash: we're able to manipulate the pitch of a single spindash sound to produce the 11 other reving versions identical to that of the Genesis. That's one file for all 12 versions of the spindash, not twelve files. We couldn't do that with another audio engine. Nothing does that, except the original games.

 

Are you even aware of Wwise's feature set? You're making it sound like pitch modulation is some amazing concept no one has implemented in games for the past twenty years.

edit: not to mention looping, ogg support and DSP.

Glad to hear you've worked out what's wrong though.

 

I'm aware of looping and DSP being available in many different programming libraries. But we only need a subset of features, not an entire massive library filled with useless stuff we do not need for our software. I like not to bloat a program needlessly. Or at least that's my personal point-of-view on it.

As far as the spindash stuff, Sonic games use additive pitch modifications, not multiplied as would normally be used in DSP. Because we're also working with Genesis sound emulation, we're able to directly manipulate the values that come together and make the more complex FM sounds we hear. So yes, I'm claiming our audio engine does things that nothing else will.

 

I think this is incredible.

I think it is incredible that you wrote a sound driver that would do exactly what you need, no more, no less. It's great to see the effort that is being put in and the amount of care that is being taken.

I understand what Falk is saying regarding the other pre-existing sound engines but I think programming one yourself with everything you need with it is just incredible. No bloat, no waste.

Keep up the good work guys and I look forward to seeing more updates.

 

You know if you even just used samples and an existing audio library for everything at the expense of a few mb and maybe a negligible amount of ram usage, people would be just as happy. If that means getting this done a lot sooner, it would probably be worth it, at least for now. This applies not just to the audio engine, but whatever else you're doing for the game's programming.

Yet on the other hand, most of my respect for this project is the very fact that you're all willing to put a godly amount of effort into making this the best it can possibly be, sparing no expense and wasting nothing. That said, lets just consider the me from one minute ago that wrote the first half of this post a total moron.

 

But the sound engine we have is completed. It has no problems. The only issues we are having is with DirectX itself, and really I think we have pretty much got it figured out. We also have every single sound ripped from Sonic 2, partly with the help of a little tool I wrote to strip dumps of scrap data and only include data needed for the sound. It was easier doing that than it was to rip PCM data. The files we're using are of GYM and VGM variety. Don't worry, we've got it covered =)

 

...So what's next? And also, why was it that you opted to use DirectX?

 

I'm also curious. OpenGL sounds way more beneficial to me, but I haven't worked with either.

 

For audio, the equivalent would have been OpenAL. DirectX is a collection of libraries. It has DirectSound, DirectMusic, DirectInput, and Direct3D. There are some others, but those are the core libraries.

For the cross-platform path, DirectX would be replaced with SDL, OpenAL, and OpenGL usually. Alternatively, DirectX can be replaced entirely with Allegro 5.x.