Jump to content

Audio Equipment Testing


Brian

Recommended Posts

Is anyone using a PC and some sort of interface to test audio equipment? Note that I'm not asking about any testing that involves 'acoustics', just 'electronics'. So I'm looking at alternatives to things like Audio Precision or Lindos kit.

 

I've seen several bits of software online which use external USB 'sound cards' but everyone seems to use expensive 24-bit 192k units. I'm not looking for audiophile performance so 24 bits is over the top, 16 is probably a bit low. Does anyone do 20-bit units? Or are there any decent 24-bit units which are a more reasonable price?

 

I can buy secondhand 'proper' audio test sets for around £400 so I guess that would be my budget for a more modern software and hardware solution.

Link to comment
Share on other sites

Smaart and a good audio interface ought to make a reasonable job of most tests, though not sure if it can do THD?

 

Most interfaces now can do 24 bit, so I doubt you’ll find anything sufficiently flat response with 16 bit, and I’m not aware of anything 20 bit.

Link to comment
Share on other sites

I would have thought you can get a 24-bit interface for well inside £400, so the question is how much of that is needed for the software (which I have no experience). I'm currently using a Roland UA-101 which was bought S/H (one at the moment here). That was chosen because I wanted 192kHz sample rate and Linux support. Now I would probably go with a Focusrite Scarlett series interface (192kHz / 24-bit / USB Audio Class 2). For measurement you probably only need 2 channels, so that should be about £100 of interface, leaving you a budget for analysis software. I have used the free Mat's Audio Analyzer for some things, but it does tend to sit more as a collection of bits to build from than a completed product. Never tried to to THD etc. with it.

ADC design is now in a space where adding bits costs significantly less than designing the analogue circuitry to make use of them, hence you can buy 32-bit converter ICs (but with no better than 22-bit dynamic range) for very low prices. Once you accept that the bottom bits are noise, you can still get a lot of dynamic range of these parts - or none at all if you fail to get the PSU/reference voltages clean! Because bigger number sell, there is no market for kit which honestly says "only 20 bits usable" - the pressure is to sell as 24-bits, and most users won't be able to hear the difference!

Link to comment
Share on other sites

I bought a USB audio interface with 2 inputs and 2 outputs with BNC connectors. It's used for testing frequency response, etc with some software that wasn't too expensive. I can't recall the brand / model off the top of my head but can check at work later today.
Link to comment
Share on other sites

I've found Visual Analyse to be a useful piece of software over the years. I've never used it for THD + N, but I know it's capable of it. You can do loopback calibration of whatever interface you use, and there's a fair number of 24/96 interfaces out there these days for less than £100. Even dropping to 24/48 you've still got an effective bandwidth of 24kHz. Also, Just checked out the website on reading your post and there's a new beta up. Edited by vlfaudio
Link to comment
Share on other sites

I've found Visual Analyse to be a useful piece of software over the years. I've never used it for THD + N, but I know it's capable of it.

A bit idiosyncratic, but a superb piece of donate-ware for general audio measurements.Certainly a big step-up from my home-made "mini-Lindos".

Edited by sandall
Link to comment
Share on other sites

I was hoping to use Visual Analyser to check out some ultrasonic "nasties" on a piece of gear. The spectrum analyser section can sample at up to 96kHz, giving a display range to 48kHz, but even the decent 96kHz & 192kHz USB interfaces only quote a 20Hz - 20kHz bandwidth. Does this mean the audio gets heavily filtered, or will any of them actually pass a signal of up to 48kHz or 96kHz?
Link to comment
Share on other sites

I was hoping to use Visual Analyser to check out some ultrasonic "nasties" on a piece of gear. The spectrum analyser section can sample at up to 96kHz, giving a display range to 48kHz, but even the decent 96kHz & 192kHz USB interfaces only quote a 20Hz - 20kHz bandwidth. Does this mean the audio gets heavily filtered, or will any of them actually pass a signal of up to 48kHz or 96kHz?

 

That's a good question. It will depend on the interface, the limiting factor would be the bandwidth of the analogue I/o . If you already have a 96 or 192 kHz interface, VA has a fairly comprehensive signal generator section, so you could generate a sweep and do a loopback test to find out.

Edited by vlfaudio
Link to comment
Share on other sites

It will depend on the interface, the limiting factor would be the bandwidth of the analogue I/o . If you already have a 96 or 192 kHz interface, VA has a fairly comprehensive signal generator section, so you could generate a sweep and do a loopback test to find out.

Unfortunately I only have a variety of 44.1/48kHz ones, & obviously don't want to buy a high-res one unless it's got the extra analogue bandwidth. The sweep function in VA only seems to work if you have the Pro version. I've noticed that whatever input I apply, the noise-floor drops by about 20dB above about 20kHz, so I wonder whether there is some other filtering going on, in which case I may be wasting my time.

Link to comment
Share on other sites

It will depend on the interface, the limiting factor would be the bandwidth of the analogue I/o . If you already have a 96 or 192 kHz interface, VA has a fairly comprehensive signal generator section, so you could generate a sweep and do a loopback test to find out.

Unfortunately I only have a variety of 44.1/48kHz ones, & obviously don't want to buy a high-res one unless it's got the extra analogue bandwidth. The sweep function in VA only seems to work if you have the Pro version. I've noticed that whatever input I apply, the noise-floor drops by about 20dB above about 20kHz, so I wonder whether there is some other filtering going on, in which case I may be wasting my time.

It's likely your interface has a brick wall anti - aliasing filter which would be at the Nyquist frequency; 20/22kHz for 44.1/48kHz. I had a brief look at the specs of a few interfaces, and you're into Focusrite Red territory to get something with a quoted bandwidth above 22kHz. Cheaper in that case to get a PC scope/function generator combo ( which I'd expect would come with better software ) or trawl fleabay for second hand scope and sig gen for what you're wanting to do.

I hadn't used VA in a while. Just tried out the beta and I remembered why :-/ . Couldn't get it to talk to my X32 Producer. It worked fine with an ancient Soundblaster Audigy 2 zs Cardbus I have for my laptop and can generate sweeps, but I stopped carrying that in my gig kit when I got the X32 . AFAIK there is still no Pro version, though it was mooted there would be one in the future.

I'd mostly been using it as a no cost pink noise/ spectrum analyser and I've got all that in my desk these days.

 

Link to comment
Share on other sites

It will depend on the interface, the limiting factor would be the bandwidth of the analogue I/o . If you already have a 96 or 192 kHz interface, VA has a fairly comprehensive signal generator section, so you could generate a sweep and do a loopback test to find out.

 

I thought I would try this with my UA-101. The resulting sound files are here:

https://soundcloud.c...-1/source-sweep

 

(go to More > Download file on each track to download the uncompressed 192kHz AIFF files, the streaming is resampled down).

The summary is that (provided that your software stack is transparent) this interface goes most (but not quite all) the way to 80kHz when set to 192kHz sample rate. There are some artefacts visible in the highest frequency range.

I did this both under Linux (using jack) and under Windows (using WASAPI and MME audio APIs) with pretty much identical results (which was a slight surprise).So it looks like this card has plenty of bandwidth (note this card has a hardware sample rate switch, so there is no doubt what rate it runs at).

 

I also borrowed a Focusrite Scarlett 2i4 interface. This is 24bit / 96kHz interface. Under Windows with MME I had to switch Audacity project rate down to 96kHz to get playback. So using the same sweep to 80kHz, the last section will be lost because the Nyquist frequency is 48kHz. However the recording cut out at just below 23kHz (not very cleanly).

I then opened the Windows Sound control panel and changed the Properties > Advanced > Default Format from 44100Hz to 96000Hz on both recording and playback. Then re-do the test. On the second pass, the bandwidth has now increased to more than 40kHz.

So the take-away message is that (depending I'm sure on your software), if your interface sample rate is software selected, you need to change the default rate in Windows Sound control panel as well as the rate in your software in order to take advantage of a high sample rate interface to make wide bandwidth measurements. Without changing the settings (which defaulted to 44.1kHz), the signal was clearly being resampled down and back up again, either on recording, playback or both.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.