Redrum pitch latency question

This forum is for discussing Reason. Questions, answers, ideas, and opinions... all apply.
User avatar
Purpleb
Posts: 114
Joined: 17 Jan 2015

Post 28 Jan 2015

There was an old post in the old forum a while back that kind of talked about this.

This might be a bit of a ramble, b/c I can't really explain it greatly, but I am sure one of you will know exactly what I am talking about and help me clarify things some.

Example:
when you adjust redrum's pitch knob it does not create latency (if it is does it is an extremely small bit)
Now when you use something like Polar or Neptune it does create latency (latency that is usually pretty audible)

question is: Why?

This is where I can't remember the exact reason or explanation, I think it was something about polar and neptune calculating the pitch shift in real time and redrum's pitch adjustment is sample based or something like that... is that right?


User avatar
eusti
Moderator
Posts: 2768
Joined: 15 Jan 2015

Post 28 Jan 2015

Might it be because with the REs the signal has to be processed on the fly, while the samples in any kind of sample player or in the audio track are set to a certain pitch and then just played at that pitch... Not sure though!

D.

User avatar
MirEko
Posts: 274
Joined: 16 Jan 2015

Post 29 Jan 2015

With the redrum you have adjusted the pitch before the sample is triggered so the audio is already adjusted before it leaves the output of the device. Also redrum is a sound generator and doesn't process sound as a audio signal but instead creates the audio signal.

When the audio enters polar or Neptune the incoming audio has to be copied to a "buffer" so that the RE can manipulate it as a sample. A good example of this is the Beatchop RE where you can see the process in action, with Beatchop you have to record to the buffer before any manipulation of the incoming audio can happen. The latency in Neptune and polar is this process happening.
:reason: :record: :re: :ignition: :refill: :PUF_take: :PUF_figure:

User avatar
Olivier
Moderator
Posts: 1248
Joined: 15 Jan 2015
Location: Amsterdam

Post 29 Jan 2015

These are all basically great answers :)
To pitch down in redrum, all that is needed is to slow down the sample playback speed. If that means the sample sounds twice as long. Redrum doesn't care. Doing that takes very little extra CPU time.
If you want to pitch up by an octave, you have to play the sample faster. This means you will run out of sample information faster. If previously the sample lasted 1 second, if you pitch it up one octave, it will last half a second.
To hear this effect really well just reset an NNXT and put in one sample and play around on your keyboard. So thats the easy method. It takes little to no additional DSP over normal Redrum operation.

A pitch shifter however is an entirely different beast. To see what needs to be done the entire process is very well explained on this site: http://www.guitarpitchshifter.com
Jump straight into the actuall process here : http://www.guitarpitchshifter.com/algorithm.html

This is also the process that has to be covered in for example Polar or Neptune (Certain details can ofcourse be handled differently.. its Propellerhead after all ;) ). If you read that site you can imagine its going to take lots more time (and thus add latency) then simply changing playback speed.
:reason: V9 | i7 5930 | Motu 828 MK3 | Win 10

User avatar
Purpleb
Posts: 114
Joined: 17 Jan 2015

Post 29 Jan 2015

eauhm wrote:These are all basically great answers :)
To pitch down in redrum, all that is needed is to slow down the sample playback speed. If that means the sample sounds twice as long. Redrum doesn't care. Doing that takes very little extra CPU time.
If you want to pitch up by an octave, you have to play the sample faster. This means you will run out of sample information faster. If previously the sample lasted 1 second, if you pitch it up one octave, it will last half a second.
To hear this effect really well just reset an NNXT and put in one sample and play around on your keyboard. So thats the easy method. It takes little to no additional DSP over normal Redrum operation.

Does this have a name? Slowing down and speeding up the sample to change it's pitch.

Excellent answers everyone. Exactly what I was looking for. Thanks.

User avatar
selig
RE Developer
Posts: 10145
Joined: 15 Jan 2015
Location: The NorthWoods, CT, USA

Post 29 Jan 2015

Purpleb wrote: Does this have a name? Slowing down and speeding up the sample to change it's pitch. Excellent answers everyone. Exactly what I was looking for. Thanks.
If anything, you can say it's the same effect as changing tape speed/pitch (vari-speed). With Polar (a "pitch shifter"), and unlike Reason's samplers, there's no change in time, only pitch. But with a tape machine they are both tied to each other. This is the more typical relationship between time and pitch - higher pitch, shorter time. Longer time, lower pitch. Polar is "newer" technology that attempts to break that bond, but at a cost (processing time). 

Polar changes pitch without changing time, something folks request for Reason's samplers btw. And Polar has NO idea as to what audio (if any) it will be presented with (why you want multiple algorithms to choose from for different audio signals). Polar's ability to adjust pitch separately from time and "on the fly" is a complex process that can take many milliseconds to compute. Sampler pitch, OTOH, is more like oscillator pitch - there's a value that represents the playback rate and all the sampler has to do is spit out samples at that rate when triggered. There's no further math that needs to be done, unlike with Polar etc. (Pitch Shifters). Well, there's sample rate interpolation, but that doesn't affect latency. 

Clear as mud?!?
;)
Selig Audio, LLC

User avatar
Purpleb
Posts: 114
Joined: 17 Jan 2015

Post 29 Jan 2015

Purpleb wrote: Does this have a name? Slowing down and speeding up the sample to change it's pitch. Excellent answers everyone. Exactly what I was looking for. Thanks.
selig wrote:
If anything, you can say it's the same effect as changing tape speed/pitch (vari-speed). With Polar (a "pitch shifter"), and unlike Reason's samplers, there's no change in time, only pitch. But with a tape machine they are both tied to each other. This is the more typical relationship between time and pitch - higher pitch, shorter time. Longer time, lower pitch. Polar is "newer" technology that attempts to break that bond, but at a cost (processing time). 

Polar changes pitch without changing time, something folks request for Reason's samplers btw. And Polar has NO idea as to what audio (if any) it will be presented with (why you want multiple algorithms to choose from for different audio signals). Polar's ability to adjust pitch separately from time and "on the fly" is a complex process that can take many milliseconds to compute. Sampler pitch, OTOH, is more like oscillator pitch - there's a value that represents the playback rate and all the sampler has to do is spit out samples at that rate when triggered. There's no further math that needs to be done, unlike with Polar etc. (Pitch Shifters). Well, there's sample rate interpolation, but that doesn't affect latency. 

Clear as mud?!?
;)
Great explanation! I get it now.

  • Information
  • Who is online

    Users browsing this forum: enalena and 7 guests