Spread Project Ideas!

[chrisname]

Well, there's no point creating a new compression algorithm when there are existing ones that match your needs, unless you can improve them.

[sohguanh]

So Framework private project is to come out with a better existing algorithm implementation instead of a new algorithm for compression correct?

[chrisname]

No, I think his project is to compress memory on-the-fly so that you can fit more objects into memory.

I'm suggesting that he use an existing compression algorithm, especially a fast one like LZO.

[ultifinitus]

I love project threads! Not to derail the compression train.

This guy's projects:

>Android App (forwarding SMS/MMS to and from a computer)
>iPhone game (simple avoidance game)
>Improve my networking library
>Custom Threadpool

>Oh and get ready to leave in 2 weeks, to be stuck, without computers, for 2 years =] (if anyone has been around for more than a year, they may remember past posts on the subject)

[ascii]

I would like to learn C++.

[ultifinitus]

Also this: http://ifile.it/m680jcx/BlockWarrior.zip

[chrisname]

I'm seeing if I can write a speech synthesiser, using SFML to create sounds.

I'm trying to figure out how it will work. The Wikipedia article says something about modelling the human vocal tract but I have no idea how to do that, and it sounds complex. What I'm planning to do instead is have a database of recorded words that can simply be strung together to form sentences. If a word that doesn't exist in the database is found then hopefully I can just split it into phonemes and have the synthesiser try to piece them together.

As much as possible, the complexity of speech synthesis should be hidden from the user, resulting in something like this:

#include <STTS/Synthesiser.hpp>

int main()
{
	stts::Synthesiser synth(stts:SpeechDB("/path/to/speech_database.sdb"));
	synth.Append(stts::Speech("hello, world"));
	synth.CreateSound();
	synth.PlaySound();
	return 0;
}

As you can see, I've hidden SFML from the user. I was going to have it work the other way, where the user would be able to get at the SFML stuff inside the Synthesiser class so that they could apply effects, but then I remembered that SFML's Sound and Music classes don't have any effects, so I'll see how they do it in Audacity and then maybe make my own. Assuming I actually get the TTS to work first...