Discuss the various methods you use in music production, from compressor settings to equipment type.
User avatar
By n4l Wed Dec 19, 2012 9:35 am
Does anybody actually use it?
When I had MPC1k and JJ OS2 on it, I used it quite a lot.
Snares sounded so good, hi hats got that 12 bit crisp and when you did it to kicks it had the SP thump and higher freq crispiness too.
And I would be interested to know what JJ did to make it work, without any additional boards or circuits (just code)
User avatar
By Ill-Green Wed Dec 19, 2012 11:52 am
Just software ingenuity. Reducing bits and frequencies using the existing circuitry on your 1000. The secret to lo-fi grit is


therefore, giving that lo-fi gritty sound.
User avatar
By sciguy Wed Dec 19, 2012 12:08 pm
Pretty simple, actually. JJOS does not actually change the bit rate of the sample, it's still a 16bit 44KHz file.

But this is most likely what he does:

Bit reducing: Set the last few bits to 0 (or 1). i.e. 0b0110010011011010 --> 0b0110010011010000
If you have a 16-bit word in the sample, setting the last bits to 0 limits the number of possible states for that sample, effectively lowering the bit rate to 12, or whatever, bits. There may be some normalization or amplification before this step, I haven't experimented enough to try. Like, if you have a really quiet sample (only uses the last few bits), I dunno if reducing the bit rate will turn it to silence, or if JJOS will boost it and just make it louder and distorted.

Reducing sample rate is equally simple. Notice how the only options are around 44K, 22K, and 11K hertz. Not a coincidence that those divide evenly into the original 44kHz.
What JJOS must do is, for the 22kHz, copy every other 16-bit word to the next point in the sample. i.e. 0x152f;0x1355 --> 0x152f;0x152f
That would effectively lower the sample rate, as every discrete sample gets outputted for twice the amount of time, therefore half the frequency.
For 11kHz, it copies every 4th sample word to the next 3 places.

Reducing sample rate is actually something more suited to digital, software implementations, rather than external boards and circuits, since sampling is inherently digital. Analog bit-crushers do exist, but digital ones are more common.
User avatar
By TheUnfuckwithable Wed Dec 19, 2012 12:38 pm
n4l wrote:Does anybody actually use it?
When I had MPC1k and JJ OS2 on it, I used it quite a lot.
Snares sounded so good, hi hats got that 12 bit crisp and when you did it to kicks it had the SP thump and higher freq crispiness too.
And I would be interested to know what JJ did to make it work, without any additional boards or circuits (just code)


I use and like it a lot!

Combined with a LPF the bit reduction can work wonders.
User avatar
By n4l Thu Dec 20, 2012 5:43 pm
sciguy wrote:Pretty simple, actually. JJOS does not actually change the bit rate of the sample, it's still a 16bit 44KHz file.

But this is most likely what he does:

Bit reducing: Set the last few bits to 0 (or 1). i.e. 0b0110010011011010 --> 0b0110010011010000
If you have a 16-bit word in the sample, setting the last bits to 0 limits the number of possible states for that sample, effectively lowering the bit rate to 12, or whatever, bits. There may be some normalization or amplification before this step, I haven't experimented enough to try. Like, if you have a really quiet sample (only uses the last few bits), I dunno if reducing the bit rate will turn it to silence, or if JJOS will boost it and just make it louder and distorted.

Reducing sample rate is equally simple. Notice how the only options are around 44K, 22K, and 11K hertz. Not a coincidence that those divide evenly into the original 44kHz.
What JJOS must do is, for the 22kHz, copy every other 16-bit word to the next point in the sample. i.e. 0x152f;0x1355 --> 0x152f;0x152f
That would effectively lower the sample rate, as every discrete sample gets outputted for twice the amount of time, therefore half the frequency.
For 11kHz, it copies every 4th sample word to the next 3 places.

Reducing sample rate is actually something more suited to digital, software implementations, rather than external boards and circuits, since sampling is inherently digital. Analog bit-crushers do exist, but digital ones are more common.


Interesting. So if compared to SP1200, how legit sounding is the JJ OS bit reduction? ( I know it's a matter of an opinion though )
I noticed that when I used the bit reduction, the sounds got quieter, opposite what happens when you sample to SP1200?
User avatar
By sciguy Fri Dec 21, 2012 10:48 pm
You most likely aren't going to get anything near the SP1200 sound using the JJOS bit reducer.

The 12bit ≠ SP1200 grit. There are so many other factors that give the 1200 its flavor; preamps, ADCs, DACs, filters, whatever.
You'll get a grittier sound with the bit reducer, which is perfectly fine for a lot of purposes, but it probably won't be an SP1200 sound.



Also, one thing about the JJOS bit reducer: I don't believe it does any bandlimiting filtering before reducing the bit rate. What I mean, is any frequencies in the original sample that are higher than the new nyquist rate after downsampling will alias. Look up what audio aliasing for more explanation.
But basically, lowering the sample rate in JJOS (if I'm correct) will give aliasing effects, like bitcrusher pedals and stuff. So again, not the same as sampling into a vintage sampler like the SP1200.
By jimmie Sat Dec 22, 2012 1:21 am
In 'bit convert' I use the sample rate reduction (11025) to make samples cut through a mix and generally sound brighter and crunchier with the aliasing that it generates. Reducing just the bit rate alone to 12-bit doesn't do much.
By m56p87 Sat Dec 22, 2012 4:58 pm
12 bit in general gets way to much credit when it comes to vintage samplers. It's really not much of a difference between 16bit, Sciguy in the on the money on this one.
User avatar
By sciguy Sat Dec 22, 2012 10:56 pm
do I actually get money? :)
By Bathmutz Sun Dec 23, 2012 11:24 am
I found that a lot of the gritty 90's type sound comes from pitching down a sample. Recording at 45 rpm and pitched down in a sampler.

What i do now is, set up a sample/drumloop/whatever, set it to output1, pitch up an octave, the record in a other sampler s2000 (or your can re-sample your main outs), then pitch it down and record in the mpc. This way i can also use the eq on my little mixer before recording back in the mpc.

The more you pitch down, (and back up again) the grittier a sample gets.
By m56p87 Sun Dec 23, 2012 5:32 pm
Bathmutz wrote:The more you pitch down, (and back up again) the grittier a sample gets.


That's because your reducing the overall sampling rate of the sound. Unless your using a machine that's pre 1990ish that's where most of the coloring from pitch up/down comes from IMO.

sciguy wrote:do I actually get money? :)


Check is in the mail! :-D