Re: OSRL Presents: A World Exclusive - SCAMP has arrived

[supertimer@aol.com (Supertimer)]

CUTblakeney@home.com (Jeff Blakeney) wrote:

>cturley2@aol.com (Cturley2) wrote:
>
>>Remember to always download the SCAMP file as a raw binary file ONLY!  You
can
>>only play the SCAMP files on the IIgs with the GS/OS application Oversampler.
>
>You can probably play them with the program LongPlay too as neither
>have any special support for WAV files.  They are just treating WAV
>files as raw binary sound files and causing a little click at the
>beginning of playing back the sample due to the few bytes of header
>information that is stored in a WAV file.

Hi Jeff, I find Oversampler better than LongPlay in this regard,
mainly because of oversampling but also because the play
routine seems cleaner and you can set the file type/aux type
to make Oversampler default to a playback rate of your
choice.

I wrote an article that I posted to csa2 about conversions of
MP3 to the IIGS using a PC.

>This means that you could probably get an even cleaner sounding
>version of the MTR.WAV file by making a mono sample and you would also
>end up with a file that is only around 689,479 bytes long which is
>much better than the 1,378,958 bytes it is now.  You would also be
>playing back at the 214/215 rate so you won't need as fast a drive to
>do the 2 or 4 times oversampling.

I suggest a higher sampling rate mono file instead.  11,025
is really too low.

>Which brings me to my next point.  Why suggest playing back a sample
>with oversampling turned off?  One of the big advantages of this
>program is to allow 2 or 4 times oversampling so that the play back
>sounds better than the original file.  As long as the drive you are
>playing the sound sample from can handle it, use the 2 or 4 times
>oversampling option.  It will cut down on a lot of the noise in a
>sample.

Exactly.  This is what a CD player does, but CD players
sport even greater oversampling than 4x.  Oversampling
makes up somewhat for downsampling.  It is, if you think
about it, sort of a type of compression.  You are
throwing out some data by downsampling, then putting
back approximations by oversampling.

The HyperStudio manual gives a tip on the sampling
rate to choose.  It says tha 26320Hz (or a multiple or
half of this) results in the cleanest sounding playback
because that's the frequency the power supply operates
at so there is much less interference.  I do in fact,
find this to be the case.

One could, of course, take a 44kHz sample across.
I've done this and Oversampler is capable of keeping
up, but I'd rather save the space and run 26320Hz
at 2x or 4x oversampling.

Playback at 26320Hz or a multiple or half of it lacks
the "hum" evident at other playback rates.  Try it!

>However, even with all these points, there is no way an 11,025 Hz, 8
>bit mono sample is ever going to sound better than a 44 kHz, 16 bit
>stereo sample.  Using the oversampling and echo options in Oversampler
>can give some decent sounding play back on a IIgs but is still isn't
>as good as an MP3.

11,025Hz?  No way!

On the other hand, 26320Hz at 4x oversampling gets
pretty darn close.  Actually, it can sound a little
better depending on the speed of your CPU.  You
see, on slower computers, you can get playback
problems in real time decompression and play of an
MP3.

Lastly, there's a shift in frequency balance between
the IIGS audio and most PC sound cards.  That's why
I suggest processing the wav file using the Goldwave
program on the PC before taking it to the IIGS (and
doing a reverse process with Goldwave if taking
something digitized on a IIGS to a PC).  See the
article I posted for more information.