Roland S-50 .OUT files

[thared33]

Ok, predicament here. I'm fooling around with an old Mac using the SDISK equivalent called SMAC. I'm simply trying to extract some of the samples using that software, but little did I know that when exporting them using SMAC that it exports to a 15khz sampling rate. As far as I know the older S-50 type samplers could sample in either 15k or 30k, but every sample is at 15k when extracting from a disk.

When I play them back they're all an octave lower. Is this normal, and the S-50 samplers play it back at double the rate or something?

I don't know how the SDISK software works for Windows/DOS, but maybe it also exports to 15k. If I've got the 'right' stuff already there's no point in finding another way to convert it, I'll just have to pitch it up an octave in my sampler. I want to confirm that this is the way it works first though.

[Skupje]

I don't know anything about the S50.
But you pretty much sayd it.. The last thing changing would be the clock frequency on the sampler itself. So wether it's playing 30 at 15 or 15 at 30, your hands are pretty much tied.

But im confused tho, did you actually play them back on the S50?

[BertKoor]

Seems like the extraction software assumes all samples are at 15khz, ignorant of the 30khz bit flipped.
No big deal as you found out, quite easy to correct after the fact...

[thared33]

I don't have an actual S-50 sampler, but the old library was put into the public domain. I actually have Chicken Systems Translator, but I'm getting some errors and nasty sounds in the samples after converting so thought I'd try going directly to the oldschool software. They do sound better, and without errors, when using SMAC, but again there's a sampling rate issue.

Funnily enough, Translator says that the samples are 30khz samples. The only way to know what they *actually* are is to ask someone that has an S-50.

There is another piece of software called DSound that can read Roland disks, and I wanted to see what it said, but the problem there is that I need to have a disk formatted in the Roland format before it can be read/imported into that program. SMAC can't format to Roland format - you must have one formatted to Roland first before SMAC will read/write the disk. I don't know if SDISK can format a 720k floppy to Roland format, but I don't have a PC with a floppy drive, and I also don't have a USB Floppy drive either.

Somehow I've got to find a way to make a a Roland disk on an old Mac...

[Skupje]

Technically the samples should be pitched up by 147% at 44.1 kHz. Else that's no good. But the rest seems legit.

[fmr]

Ok, predicament here. I'm fooling around with an old Mac using the SDISK equivalent called SMAC. I'm simply trying to extract some of the samples using that software, but little did I know that when exporting them using SMAC that it exports to a 15khz sampling rate. As far as I know the older S-50 type samplers could sample in either 15k or 30k, but every sample is at 15k when extracting from a disk.

When I play them back they're all an octave lower. Is this normal, and the S-50 samplers play it back at double the rate or something?

I don't know how the SDISK software works for Windows/DOS, but maybe it also exports to 15k. If I've got the 'right' stuff already there's no point in finding another way to convert it, I'll just have to pitch it up an octave in my sampler. I want to confirm that this is the way it works first though.

Did you manage to extract the files? If yes, then it's easy. You just need a software that edits the sample header, and change the sample rate info from 15 kHz to 30 kHz. I think the free Audacity can do that.

[thared33]

Yep, the reason I know they're 15khz is because they exported. Now I'm wondering what the difference would be between leaving them alone and just playing higher keys on my keyboard, or editing the sample header. Hmm..

[fmr]

Yep, the reason I know they're 15khz is because they exported. Now I'm wondering what the difference would be between leaving them alone and just playing higher keys on my keyboard, or editing the sample header. Hmm..

If you leave them at 15 kHz, you will have aliasing on all content above 7.5 kHz (which is pretty low), unless you filter out all content above that frequency (which will make the sounds muffled and dull). With the sample rate at 30 kHz, you will have frequencies correctly reproduced up to 15 kHz, which will be much better.

[Skupje]

Yep, the reason I know they're 15khz is because they exported. Now I'm wondering what the difference would be between leaving them alone and just playing higher keys on my keyboard, or editing the sample header. Hmm..

What do you mean? No.. There's no such thing as 30kHz WAV files. It's either 22.5 or 44.1 kHz. Effectively you have double data for those files, it's the right thing to have.. The file has to be converted to 44.1 kHz so if you convert 30 to 44.1 kHz the factor of conversion is 1.47.

So the only thing that can be done to be nit picky, is see if the pitch of the sample you get is higher than the original (when it's converted to 44.1, this is done by your conversion program) if it is higher, it's a good thing and the conversion is solid. Else if the pitch of the sample is the same as the one on the device, thats a bad thing. But there is nothing you can do to fix it since it's a conversion error.

[fmr]

What do you mean? No.. There's no such thing as 30kHz WAV files. It's either 22.5 or 44.1 kHz. Effectively you have double data for those files, it's the right thing to have.. The file has to be converted to 44.1 kHz so if you convert 30 to 44.1 kHz the factor of conversion is 1.47.

WTF are you talking about? "It's either 22.5 or 44.1 kHz" Where did you learn that? Actually it's 22.05, not 22.5. And there are many more sample rates in the WAV format actually. And many more audio file formats besides WAV.

And who told you that S50 was using WAV files?

[Burillo]

WAV supports sampling rate from 1Hz to 4.3GHz.

[fmr]

So the only thing that can be done to be nit picky, is see if the pitch of the sample you get is higher than the original (when it's converted to 44.1, this is done by your conversion program) if it is higher, it's a good thing and the conversion is solid. Else if the pitch of the sample is the same as the one on the device, thats a bad thing. But there is nothing you can do to fix it since it's a conversion error.

Wrong again. I don't know what audio editor you are using, but it must be pretty basic. There are many resampling methods. And you can resample keeping the original pitch, or you can resample transposing. If the resampling method is by simply rewriting the sample header, then the pitch will be transposed (it's the usual method used by samplers to play a sample up or down).

But if you use a more advanced algorithm, you can resample keeping the original pitch, by interpolating samples (if resampling to a higher sample rate) or removing samples (if resampling to a lower sample rate).

[Skupje]

Wrong again. I don't know what audio editor you are using, but it must be pretty basic.

Well gee, I don't know "Quartus"?

the usual shift left add repeat on my abacus.. But 48kHz is alright too

[fmr]

Wrong again. I don't know what audio editor you are using, but it must be pretty basic.

Well gee, I don't know "Quartus"?

the usual shift left add repeat on my abacus.. But 48kHz is alright too

Don't know that one, and a quick search with Google didn't show anything

"But 48kHz is alright too"... Allright for what?

[Skupje]

Don't know that one, and a quick search with Google didn't show anything

"But 48kHz is alright too"... Allright for what?

For playing the samples. "Im purist"