Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Sample Rate Signal Chain (Read 2293 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Sample Rate Signal Chain

I have a small home studio and due to it's integration with my DAW, a signal can be sampled many, many times. Kind of like a virtual patchbay - where every step is sampled again and again (I've got 72 I/O, I may was well use them).  Generally this means an input can be sampled as many as 8 times.

Now, does a higher sample rate offer a benefit in this scenario? I record at 48kHz to give me some headroom for processing - but would going higher offer me any advantage in this setup? I don't use VSTs or or in the box processing at all, I'm just wondering if the process of sampling the same signal over and over is improved (even theoretically) by a higher sample rate.

I don't think there is, but I'd just like other's thoughts.

Thanks!

Re: Sample Rate Signal Chain

Reply #1 –
Quote
I have a small home studio and due to it's integration with my DAW, a signal can be sampled many, many times. Kind of like a virtual patchbay - where every step is sampled again and again (I've got 72 I/O, I may was well use them).  Generally this means an input can be sampled as many as 8 times.
Sorry, I don't understand that...  

Are you saying you run the signal into 8 analog-to-digital converters?   Do you really have 72 ADCs?   That doesn't sound like a "small home studio".    Or, you just have 8 copies of the same digital data?   You'll get very-slightly different data with separate ADCs, but not significantly different as long as they all use the same clock.   Either way, I don't see the point.

You wouldn't want a "chain" of unnecessary analog-to-digital-to-analog-to-digital...   Conversions.

Either way, a higher sample rate mainly means more data to deal with.   There is generally no quality-advantage of anything greater than 44.1kHz (or 48kHz for video).   But, the "pro studio standard" is 24-bit/96kHz.

Re: Sample Rate Signal Chain

Reply #2 –
Quote from: DVDdoug on
Quote
I have a small home studio and due to it's integration with my DAW, a signal can be sampled many, many times. Kind of like a virtual patchbay - where every step is sampled again and again (I've got 72 I/O, I may was well use them).  Generally this means an input can be sampled as many as 8 times.
Sorry, I don't understand that...  

Are you saying you run the signal into 8 analog-to-digital converters?   Do you really have 72 ADCs?   That doesn't sound like a "small home studio".    Or, you just have 8 copies of the same digital data?   You'll get very-slightly different data with separate ADCs, but not significantly different as long as they all use the same clock.   Either way, I don't see the point.

Yes, I suppose 72 ADCs and 72 DACs, although it's really three separate units with 24 of both. It is certainly a small home studio (excuse the mess, we're reconfiguring). What I'm saying is that I don't run a patchbay in order to change my signal chain, I reconfigure the DAW to sample one device, then output it to another, then another.

Example channel would be: Drum machine to DAW, Daw to distortion unit, distortion unit to DAW, DAW to reverb, reverb to DAW.  Although this seems unnecessary, this allows me huge flexibility in my work from which can chain a lot of devices together.

Quote from: DVDdoug on
You wouldn't want a "chain" of unnecessary analog-to-digital-to-analog-to-digital...   Conversions.
Well, I agree - but I do not want patchbays in my studio because 1 they take up too much rack space, 2 they are unreliable 3 I find my workflow is better like this.

Quote from: DVDdoug on
Either way, a higher sample rate mainly means more data to deal with.   There is generally no quality-advantage of anything greater than 44.1kHz (or 48kHz for video).   But, the "pro studio standard" is 24-bit/96kHz.

They do not use the same clock, they may well be on three separate (internal) clocks, one inside each bank of 24IO. Adding an external clock means introducing more jitter into the chain unfortunately.

Re: Sample Rate Signal Chain

Reply #3 –
Quote from: Funkstar De Luxe on
They do not use the same clock, they may well be on three separate (internal) clocks, one inside each bank of 24IO. Adding an external clock means introducing more jitter into the chain unfortunately.
You do realise that ADCs have jitter? And that every time you pass a signal through an ADC it bakes in additional jitter. So if you're worried about adding jitter, it would be a VERY good idea not to keep passing your signals through extra DAC-ADC steps.

Just a thought.

Re: Sample Rate Signal Chain

Reply #4 –
Quote from: cliveb on
Quote from: Funkstar De Luxe on
They do not use the same clock, they may well be on three separate (internal) clocks, one inside each bank of 24IO. Adding an external clock means introducing more jitter into the chain unfortunately.
You do realise that ADCs have jitter? And that every time you pass a signal through an ADC it bakes in additional jitter. So if you're worried about adding jitter, it would be a VERY good idea not to keep passing your signals through extra DAC-ADC steps.

Just a thought.

Yes I do. I also know that my setup measures less jitter when using the internal clock rather than an external clock. I find jitter very difficult to hear anyway (in fact, I'm not sure I've ever been aware of it)

None of this is really answering the question. I'm trying to understand if increasing the sample rate would have tangible benefits to my particular situation.

Also, I'm sure that a little jitter is much preferable to multiple noisey patchbays.

Re: Sample Rate Signal Chain

Reply #5 –
Quote from: Funkstar De Luxe on
None of this is really answering the question. I'm trying to understand if increasing the sample rate would have tangible benefits to my particular situation.

Probably not.  Most likely doubling the sampling rate will just halve the decimation rate in the A/D-D/A chain, which in effect will not change anything other than how the data is represented.  Could test it and see though in case there is some non-ideal behavior going on.

BTW, I don't think you should worry about jitter, but using a common external clock might still be a good idea so that your channels run at exactly the same frequency.  While the internal clocks are probably very close to each other, you might have some minor drift between them. 

Re: Sample Rate Signal Chain

Reply #6 –
Quote
Now, does a higher sample rate offer a benefit in this scenario? I record at 48kHz to give me some headroom for processing - but would going higher offer me any advantage in this setup?
No.   (And you don't get more "headroom" with a higher sample rate...  You get more dynamic range with higher bit depth, which you can use to get more headroom by recording/processing at a lower level.)    

Quote
Example channel would be: Drum machine to DAW, Daw to distortion unit, distortion unit to DAW, DAW to reverb, reverb to DAW.  Although this seems unnecessary, this allows me huge flexibility in my work from which can chain a lot of devices together.
If you want to do analog/external effects/processing, it would be best to make a single analog chain and avoid the extra D/A/D conversions.    But practically speaking it shouldn't be a problem and the electronics in the analog processors are doing more "damage" than the extra conversions.   (i.e. Any analog processor is going to add more noise than a similar digital plug-in, unless the plug-in is trying to simulate an analog processor.)    I'm not saying the analog noise will be audible or that digital processing is "better" if you like the sound of the particular external-analog processors....   

Quote
Yes I do. I also know that my setup measures less jitter when using the internal clock rather than an external clock. I find jitter very difficult to hear anyway (in fact, I'm not sure I've ever been aware of it).
Right, you are not going to hear jitter (or the effects of jitter).     In general, I'd say it's more important to have a clock that's accurate over the long-term than it is to have a low-jitter clock.    i.e. If you are synchronizing over the length of a movie or a concert you don't want to be off 1/10th of a second by the end of the program.   Expensive professional master clocks have good accuracy and low jitter

If you are recording from more than one interface in parallel it's better to have all the samples perfectly aligned from the same master clock from the beginning to the end of the recording.    

In series it's not as critical to have matching clocks but it's more important to have accurate clocks, since the timing will drift a little with the extra D/A/D conversions and presumably different tracks will be processed differently.   And, the timing errors will accumulate with each A/D/A conversion.




Re: Sample Rate Signal Chain

Reply #7 –
Quote from: saratoga on
Quote from: Funkstar De Luxe on
None of this is really answering the question. I'm trying to understand if increasing the sample rate would have tangible benefits to my particular situation.

Probably not.  Most likely doubling the sampling rate will just halve the decimation rate in the A/D-D/A chain, which in effect will not change anything other than how the data is represented.  Could test it and see though in case there is some non-ideal behavior going on.

BTW, I don't think you should worry about jitter, but using a common external clock might still be a good idea so that your channels run at exactly the same frequency.  While the internal clocks are probably very close to each other, you might have some minor drift between them.

Excellent, thanks for the post. I have tried an external clock before, but got measurably (not audibly) worse results.

DVDdoug, thanks to you too. I did use the wrong term 'headroom'. Really I meant more data with an increased sample rate which may result in different sound when processing (i.e., slowing down audio). As for the clock, I didn't get the results I hoped for. My recordings are rarely more than 10 minutes, so it doesn't need to be super stable.

Cheers gents!