nesdev.org

I downloaded a .Z7 extractor and it STILL won't extract.

Have you tried 7-Zip?

yeah

It's compressed with 7z because bandwidth isn't free, and zip makes the file 600K larger (2.9 versus 3.5 MB). Given that I expected it to be downloaded a lot, I wanted to slim it as much as possible. For smaller files where accessibility is important, I agree that zip is THE way to go. Nothing beats it in wide support, and lightweight implementations.

As for the corrupt file, I just downloaded it fine. Try the mirror link, as the nyud.net one kept cutting off on me before the entire thing downloaded.

Memblers wrote:so I'm thinking longer samples with the freqs scaled up might help it, I don't know by how much though.

Quite a bit in some cases (at least it did back when I was writing my VGM player):

mic_ wrote:The results can be quite noticable, as can be heard in this song from Alex Kidd in Miracle World:

.SPC (old frequency table)
.SPC (new frequency table)
.VGM for reference

I've just figured out why my test SPC code wasn't compiling in the past. I guess now I can program my own software mixed 2A03 emulator, so I can run SPC and 2A03 music at the same time.

How do you plan on BRR-encoding the mixer output though?
If you're thinking about outputting uncompressed samples by writing to the echo buffer, then you'll be on a pretty tight cycle budget (maybe something like 100 cycles/sample at decent quality). It seems unlikely that you'd have time to emulate the 2A03 with that approach.

mic_ wrote:How do you plan on BRR-encoding the mixer output though?

Each 9-byte block of BRR contains a couple control bits, a volume scaling amount, and an index selecting one of four IIR de-emphasis filters, followed by sixteen 4-bit signed PCM samples. One of these filters has zero for both feedback coefficients, which means the block is straight 4-bit signed linear PCM. I imagine that a mixer running on the SPC700 would encode the result to LPCM, possibly losing a bit of DAC resolution.

Mixing 4 (or 5) channels into a single sample with 4-bit resolution would probably not sound too good.

And if you wanted to get crisper waveforms by playing longer samples at a higher frequency you'd probably have to set aside a fair amount of SPC RAM for the output buffer,

Switching from 16-bit to 4-bit isn't loosing "a bit" of DAC resolution - it's massive lost.

For echo buffer writes, you have 32 cycles per sample to be computed (the output is read at 32kHz and the CPU is clocked at 1024 kHz). Considering that I think quite a few instuctions takes 2 cycles, this isn't a whole lot - barely enough for something really simple I think.

Bregalad wrote:Switching from 16-bit to 4-bit isn't loosing "a bit" of DAC resolution - it's massive lost.

The NES's DAC isn't 16-bit; it's the sum of four 4-bit DACs (and a 6-bit DAC if you're handling DPCM).

But emulating the NES's APU into samples at full speed isn't likely to happen without serious aliasing artifacts.

This is a drawing of my ponytail girl character.

http://img808.imageshack.us/img808/2377 ... ilgirl.jpg


I downloaded a trace song, and downsampling it to 16kHz to see how it sounds like.

http://www.supload.com/sound_confirm.ph ... 387279.wav


Here is my code for loading a spc700 routine. I don't know if it works or not. Please tell me If I'm doing this correctly.

psycopathicteen wrote:This is a drawing of my ponytail girl character.

http://img808.imageshack.us/img808/2377 ... ilgirl.jpg

I'd recommend a little more clothes than that, especially if you plan to have an ice world.

That's really not my type of girl at all.

Bregalad wrote:That's really not my type of girl at all.

I have to agree with that one