SCC sound output buggy?

Página 1/3
| 2 | 3

Por Edwin

Paragon (1182)

Imagen del Edwin

13-02-2008, 23:44

I've been fiddling with my SCC replayer for a while now because the sound quality was not good when played on a real SCC cartridge. But it appears that my code is not to blame, but the SCC chip/DAC itself. To test this I put together a little test.

You can start SCC blaffer NT (or probably any other scc music app) and select the same wave for all channels. A simple sine will do. Then put in the following simple pattern:

C 4  ...  ...  ...  ... 
...  ...  ...  ...  ... 
OFF  ...  ...  ...  ...
...  C 4  ...  ...  ... 
...  ...  ...  ...  ... 
...  OFF  ...  ...  ... 
...  ...  C 4  ...  ... 
...  ...  ...  ...  ... 
...  ...  OFF  ...  ... 
...  ...  ...  C 4  ...
...  ...  ...  ...  ... 
...  ...  ...  OFF  ...
...  ...  ...  ...  C 4
...  ...  ...  ...  ... 
...  ...  ...  ...  OFF

And play it.

In theory it should sound the same 5 times. But it doesn't. In fact, it even sound different on repeated plays.

I suspect this is a bug in the scc chip. And a rather bad one at that. But I find it very strange that nobody has questioned this before. So, I'm interested in whether you can hear this on all SCC chips or maybe just some versions. I tested it on both the scc in Nemesis 2 and in the SCC Flash ROM. Both have horrible results. Bifi reported that it sounds fine on an SCC-I cartridge. Some reports on others would be nice.

I have a theory on what is happening here, but I'd like to hear yours before I share mine. Also, if you find ways to avoid this, I'd love to hear them.

Login sesión o register para postear comentarios

Por sjoerd

Hero (611)

Imagen del sjoerd

14-02-2008, 00:12

I think it is because channel 4 and channel 5 share a wave. This is no fun Tongue

Por dvik

Prophet (2200)

Imagen del dvik

14-02-2008, 01:33

I'm very interested in this too and it could be the same problem I faced when doing the SCC sample player that is using 4 channels (see another thread).

Por nikodr

Paladin (750)

Imagen del nikodr

14-02-2008, 04:15

I think that this would work correctly only on scc+ chipset,on scc+ channels 4 and 5 have a different wave and do not share only one. http://bifi.msxnet.org/msxnet/tech/soundcartridge.html here it mentions that :

"The improvement of SCC is compared to the normal SCC that the Sound Cartridge has a extra waveform. Now channel 4 doesn't have to share its waveform with channel 5 (thus all five channels have a private waveform). "

and when you go further down the page it has the following information

"The difference is that now channel 5 has a private waveform. For more information about the operation of the SCC, see SCC Sound Chip.

The fact that SCC may be addressed in two modes, this does not mean that the SCC+ has twice the channels of normal SCC. If you change Sound Mode, the different addresses refer to the same registers. "

In theory though all 5 channels should work but channel 4 would sound the same as channel 5.From what you mention chanel 5 is not heard at all?

Por wolf_

Ambassador_ (10135)

Imagen del wolf_

14-02-2008, 10:13

nikodr: unrelated. It's common knowledge that 4 & 5 have the same waveform, nothing new here. Try 5 channels with the same instrument everywhere: there's a bug on 4 & 5.

Por hap

Paragon (2043)

Imagen del hap

14-02-2008, 14:19

Probably, and hopefully, related to the same 'bug': http://home.planet.nl/~haps/crap/msx_kv2.zip
This was recorded from an original KV2, it uses ch4&5 for those bells, it sounds different on all emulators.

Por wolf_

Ambassador_ (10135)

Imagen del wolf_

14-02-2008, 14:26

That could as well be normal aliasing. Ppl from OpenMSX have spent some time, quite some months ago, to perfect a resampling algo. The main test sound used to test/analyze the quality of those resamplers was a bunch o' scary sound effects from Solid Snake, full o' aliasing. If you have the opportunity: try the given example in a tracker like Blaffer or Scc-Musixx.. ^_^

Note that the 1cM's SCC implementation is without this bug.

Por hap

Paragon (2043)

Imagen del hap

14-02-2008, 14:53

I don't think so. Could someone try King's Valley 2 "start select" song on an SCC+?

Por wolf_

Ambassador_ (10135)

Imagen del wolf_

14-02-2008, 15:01

Best would be to isolate those channels on a real msx/scc .. I just don't know any bgm player that can isolate channels tho..

Por hap

Paragon (2043)

Imagen del hap

14-02-2008, 15:44

Hold on, I'll hack KV2 MSX1.

Por hap

Paragon (2043)

Imagen del hap

14-02-2008, 17:03

KV2 (MSX1 or 2) ROM, offset $8F23, change "11 FC E1 1A" to "00 00 3E 1F", with 1F being the value written to the SCC channel enable register ($988F), it gets written at the 2nd loop of the song. Disabling channel 4 or 5, or disabling channels 1,2,3 doesn't affect the bells pitch. So, unless disabling channel 4 or 5 doesn't fix the blafferthing, the KV2 problem is unrelated to it.

*edit* To be sure about it, the request of someone trying KV2 on SCC+ still stands.

Página 1/3
| 2 | 3