Audio Precision System One vs Good Sound Card

[ezalys]

Hey all,

I’m trying to decide on a solution for measuring THD and IMD as well as flatness and phase in the audio band for both audio and control loop characterization, and am not looking to spend absurd amounts of money. The older audio precision system ones and twos show up on eBay quite frequently, and there’s apparently someone who has made a usb interface for them so I can use them with a modern computer. My worry is that they’re just a bit old and I wonder if it’s just better to buy a modern sound interface to do the same tasks. Does anyone have any thoughts on the matter?

E

[Messtechniker]

  Slightly interested in this too. Just seen the prices.
Not longer interested in view of my modest needs.
I'd start by comparing published specs.

[boB]

I used the AP system one for many years.  It's a great system even though it doesn't have the newer digital interface stuff.

An audio interface with sample rate of just 48 kHz or 44.1 kHz is possibly not going to have the flat phase up to or beyond 20 kHz ?

The AP S1 is going to be quite a bit better than an auidio card but possibly a GOOD audio interface and the right software which I'm not really up on these days might do what you want.

Will have to check Ebay on this myself

[David Hess]

An audio interface with sample rate of just 48 kHz or 44.1 kHz is possibly not going to have the flat phase up to or beyond 20 kHz ?

All of the audio cards I would recommend for this sample at 192kS/s or higher anyway and lack anti-alias filtering at 20kHz.

I was going to link a review of PC sound cards for this sort of application but it is gone.

[billfernandez]

Look at the QuantAsylum QA401 audio analyzer.  Modern.  USB.  Programmable.  High performance.  Reasonably priced.

[DaJMasta]

Specs wise, the AP System One is reasonably high end when compared to sound cards, but certainly not the highest available nor even the highest price point, but there are reasons to consider it.  It's going to be tested to have much more even response, it has a wider frequency response than many low noise sound cards, it has vastly superior input protection, and the integrated software is both good and free - comparable software will cost you something on the PC side, at least.

That said, it's bigger, could have worse overall noise/THD specs, the software and interface require older computer hardware to connect to it, and if I remember right, because of the sample rates it uses, there isn't a really easy lossless audio recording output from the software that's immediately playable in regular players.

I think the moral of my story is: if you don't mind the clunkiness associated with the older software/interface and have the bench space, the unit should be sufficient for the vast majority of analog audio work you're likely to do and should have a long lifetime.  If space, power consumption, software portability, or the lowest THD/noise specifications are a concern, then maybe a higher end external sound card would be a better choice.  A couple hundred on a good sound card and a software suite license can be very powerful tools if you don't need the stability guaranteed by calibrated test equipment - which is extra true when buying used without the intention to pay for a cal.

[plurn]

Some good information here:

http://nwavguy.blogspot.com/2011/02/testing-methods.html

http://nwavguy.blogspot.com/2011/02/rightmark-audio-analyzer-rmaa.html

He covers many potential issues with using sound cards, and mentions some warning about older Audio Precision products.

[ezalys]

I mean okay... let's compare to the QA401 or something of that level rather than a sound card. I need something that's repeatable and gives proper absolute results that can be readily compared to others. I don't fully understand NwAvGuy's statements about the System One and System Two needing weird interfaces-- I think that may have been true at one point but there are USB interfaces for them now so I'm not worried about this.

[trobbins]

A bit of caution is needed when reading blanket statements, such as some from DaJMasta. 

The OP is referring to vintage AP gear, so an extra cost for calibration would need to be factored in (cal service providers don't show quote values, and courier costs would be an addition) to include comments about confidence from calibrated performance.

I can only make cursory comment on the simple USB soundcard setup I use that cost under AUD$60 - an EMU0404 USB with 100:1 P4100 clone probe and clone 10:1 probes.  I don't see how the AP has vastly superior input protection than what I have.  I use REW software which is free and allows 192kHz sampling, and I'd suggest that some aspects of REW software are going to be comparable, if not better - but given the wide range of performance capabilities of AP and REW software, any comparison is really up to the end user and intended applications.

[santiall]

AP is pretty much the standard out there if you are into professional audio. There are other systems like the A1, A2 and FX100 from NTi (Neutrik), the dScopeIII from Prism Sound and some others but at the end everybody seems to use AP, specially for production testing. Some of those other systems have better specs and are better priced so you will be perfectly OK for laboratory / R&D.

You may also check www.virtins.com, they have a multinstrument doing a bit of everything. I did use it in the past for some stuff and it was quite useful for some of the tasks I was working at that moment. The main problem with those PC audio analyzers are calibration and so on but again it depends on what you really need. Sometimes a bit of tolerance here and there doesn't really matter.

[boB]

An audio interface with sample rate of just 48 kHz or 44.1 kHz is possibly not going to have the flat phase up to or beyond 20 kHz ?

All of the audio cards I would recommend for this sample at 192kS/s or higher anyway and lack anti-alias filtering at 20kHz.

I was going to link a review of PC sound cards for this sort of application but it is gone.

That would probably work good ! Noise floor over the whole bandwidth would be higher than lower sample rate but that can be filtered if needed.

Would have loved to see a review !  Most of the A/D boxes these days are rated with A-Weighting so they usually lie a bit.

[David Hess]

An audio interface with sample rate of just 48 kHz or 44.1 kHz is possibly not going to have the flat phase up to or beyond 20 kHz ?

All of the audio cards I would recommend for this sample at 192kS/s or higher anyway and lack anti-alias filtering at 20kHz.

I was going to link a review of PC sound cards for this sort of application but it is gone.

That would probably work good ! Noise floor over the whole bandwidth would be higher than lower sample rate but that can be filtered if needed.

Would have loved to see a review !  Most of the A/D boxes these days are rated with A-Weighting so they usually lie a bit.

I should have thought of this.  Here it is on the Wayback Machine:

https://web.archive.org/web/20180507221601/http://www.clarisonus.com/Research%20Reports/RR001-SoundCardEval/RR001-PCsoundCards.html

My big concern when using a sound card is the ease of corrupting any measurements with a ground loop.  My preference would be something with a galvanically isolated USB interface.

These cards are also old and expensive making something like the mentioned QA401 more appealing.

[Loboscope]

I use this software for measuring audio stuff: https://hpw-works.com/
In combination with a good soundcard it will deliver any data you want with a very high precision. I use interfaces from RME, but any other up-to-date card will do the job.