Simple "no knobs" 96 kHz USB audio interface

Talk about your MIDI interfaces, microphones, keyboards...

Moderators: MattKingUSA, khz

Post Reply
CrocoDuck
Established Member
Posts: 1133
Joined: Sat May 05, 2012 6:12 pm
Been thanked: 17 times

Simple "no knobs" 96 kHz USB audio interface

Post by CrocoDuck »

Hi guys!

I am on a quest to find something like the ESI MAYA44 USB+, but I would like to go all the way up to 96 kHz or even 192 kHz (24 bit). I would like to not have knobs and other functionalities as I intend to use the interface for measurements (hence the need also for a high sample rate). Without knobs and pre-amplifiers it is much easies to calibrate the device, i.e. map the recorded digital signal amplitude to voltage values. Any of you having an interface to suggest?
User avatar
bluebell
Established Member
Posts: 1909
Joined: Sat Sep 15, 2012 11:44 am
Location: Saarland, Germany
Has thanked: 111 times
Been thanked: 116 times

Re: Simple "no knobs" 96 kHz USB audio interface

Post by bluebell »

I don't know if you know this article: https://www.xiph.org/~xiphmont/demo/neil-young.html

Of course higher sampling rates can reduce latency – in theory. The CPU has to be capable of handling it and that's the real world limit.
If your CPU handles "jackd -n2 -p32" easily at 48 kHz with your standard set of plugins then you can halve the latency by using 96 kHz and halve it again by using 192 kHz.

Mine doesn't.

Linux – MOTU UltraLite AVB – Qtractor – http://suedwestlicht.saar.de/

CrocoDuck
Established Member
Posts: 1133
Joined: Sat May 05, 2012 6:12 pm
Been thanked: 17 times

Re: Simple "no knobs" 96 kHz USB audio interface

Post by CrocoDuck »

I need the high sample rate for accurate measurements. I need the anti aliasing filter to be out of the way of the end of the audible spectrum, or what I measure will be distorted there. Moreover, I need to be able to measure nonlinear distortion, hence I need to able to sense harmonics well beyond the end of the audible spectrum, also to identify parameters of nonlinear systems accurately with high order. I don't need low latency as I can always remove it with post processing locating cross correlation peaks (it works pretty well also in case of important linear and nonlinear distortion).

It's not for music, but for "scientific" measurements.
gimmeapill
Established Member
Posts: 564
Joined: Thu Mar 12, 2015 8:41 am
Has thanked: 44 times
Been thanked: 8 times

Re: Simple "no knobs" 96 kHz USB audio interface

Post by gimmeapill »

ESI stuff looks as simple as it gets, what about this one?
http://www.esi-audio.com/products/ugm96/
j_e_f_f_g
Established Member
Posts: 2032
Joined: Fri Aug 10, 2012 10:48 pm
Been thanked: 357 times

Re: Simple "no knobs" 96 kHz USB audio interface

Post by j_e_f_f_g »

Ah, you're trying to build an oscilloscope on the cheap. Audio cards don't make for very good oscilloscopes. Usually, they have filters on the input/output to roll off anything above 20 KHz, since that isn't important for audio, and things like the op amps aren't designed to handle > 20 KHz.

Author of BackupBand at https://sourceforge.net/projects/backupband/files/
My fans show their support by mentioning my name in their signature.

CrocoDuck
Established Member
Posts: 1133
Joined: Sat May 05, 2012 6:12 pm
Been thanked: 17 times

Re: Simple "no knobs" 96 kHz USB audio interface

Post by CrocoDuck »

Actually I found more often than not that there isn't a lot of filtering in the way. For example, I successfully used my Scarlett 2i4 up to 96 kHz (48 kHz bandwidth). Both in input and output the only filtering I have seen was DC decoupling and anti-aliasing, which I think it is digitally implemented through upsampling as it is very sharp. In this post there is a measurement I made with the swept sine method (EDIT: Actually I used pseudo-random noise...). It is input filter convolved with output filter (so, each filter is perhaps half in order). I think it is pretty flat, enough for my purposes. I would expect most other sound-cards to be like that.

Anyway, this is not too much of a trouble as I have referenced measurements in mind, meaning that eventual filtering will be naturally deconvolved in the signal processing. This works very well unless you are attenuating close (or beyond) the end of the dynamic range, which happens only very close to DC or Nyquist. I tested the method with an internal soundcard of a desktop computer (a pretty crappy one) to find result well in agreement with the 01dB analyzers I use at work (a part around nyquist, which is why I want to boost up the sampling rate).

So why not to use my Scarlett?? Well, I was hoping to not have preamps in the way. Just to have a simple overall system. I might end up using it tho as I know it and, even better, it has 1 MOhm impedance on Hi-Z.
Last edited by CrocoDuck on Mon Nov 21, 2016 6:52 pm, edited 1 time in total.
CrocoDuck
Established Member
Posts: 1133
Joined: Sat May 05, 2012 6:12 pm
Been thanked: 17 times

Re: Simple "no knobs" 96 kHz USB audio interface

Post by CrocoDuck »

j_e_f_f_g wrote:and things like the op amps aren't designed to handle > 20 KHz.
Sorry I glimpsed over this. It depends on the opamp family and the way they are employed, but I would disagree on this statement for the most part. There are opamps designed to handle radio frequency, for example, but even most of the ones for audio purposes specify actually hundreds of kHz or even MHz for the bandwidth (depending on the input voltage), like this for example (opamps of this family are often in Behringher products and I used them myself for few projects). There are many examples of bad opamps made for audio that are extremely cheap (357 series for example, if I am not mistaken), but they don't tend to be very often employed in external sound-cards. They are more likely found in integrated sound-cards (which might even have a cutoff at 5 kHz, were the range of human voice ends).

Anyway, I see what you mean. The ideal solution would be to pick up an integrated class compliant USB Codec from TI or other vendors and built just few appropriate opamp buffers around it, with some high voltage protection circuitry, kinda like here (or many other similar projects). Unfortunately I don't have too much time to design such a device, although I always wanted to...
CrocoDuck
Established Member
Posts: 1133
Joined: Sat May 05, 2012 6:12 pm
Been thanked: 17 times

Re: Simple "no knobs" 96 kHz USB audio interface

Post by CrocoDuck »

gimmeapill wrote:ESI stuff looks as simple as it gets, what about this one?
http://www.esi-audio.com/products/ugm96/
Hey sorry I forgot the reply. That looks alright I think, I will keep it under my radar. It would be even better if it had 4 inputs with switchable high impedance and simple line outputs with low impedance.
j_e_f_f_g
Established Member
Posts: 2032
Joined: Fri Aug 10, 2012 10:48 pm
Been thanked: 357 times

Re: Simple "no knobs" 96 kHz USB audio interface

Post by j_e_f_f_g »

Author of BackupBand at https://sourceforge.net/projects/backupband/files/
My fans show their support by mentioning my name in their signature.

CrocoDuck
Established Member
Posts: 1133
Joined: Sat May 05, 2012 6:12 pm
Been thanked: 17 times

Re: Simple "no knobs" 96 kHz USB audio interface

Post by CrocoDuck »

That's HOT. Definitely need to get my hands on one of those, thanks! If the signal generator is flexible enough it could be even possible to implement my transfer function measurement algorithms with that!
folderol
Established Member
Posts: 2069
Joined: Mon Sep 28, 2015 8:06 pm
Location: Here, of course!
Has thanked: 224 times
Been thanked: 400 times
Contact:

Re: Simple "no knobs" 96 kHz USB audio interface

Post by folderol »

Interesting. Let us know how you get on.
The Yoshimi guy {apparently now an 'elderly'}
CrocoDuck
Established Member
Posts: 1133
Joined: Sat May 05, 2012 6:12 pm
Been thanked: 17 times

Re: Simple "no knobs" 96 kHz USB audio interface

Post by CrocoDuck »

folderol wrote:Interesting. Let us know how you get on.
Absolutely. I plan to document everything in a sort of tutorial, probably on my blog. If I get some good code out of it I will drop it in a GitHub repo or similar :wink: .
Post Reply