Playing and recording sound

[JackB]

I’m thinking of getting a Q403 to test some prototype scientific equipment. I will want to do many of the same tests that are needed for audio equipment, but also some extras. In particular I will need to:

• Send two signals to the outputs with precisely controlled amplitude and relative phase
• Record the response from the thing I’m testing
• Analyse that response using external software, probably a python script
• Use the results to calculate a new signal to send to the outputs
• Go back to step 1 with the new signal

I will need to do that in a fully automatic way, no user interaction, no clicks, preferably ten times a second. One way I could do that is if I could set up the QA403 so that it appears as a soundcard in windows. Another way would be if there is a software API that I can use. A third way would be if I can write a wav file, then call the QA403 software with a command line argument to play that WAV file and record the response as another WAV file, and read that wav file back into my script.

I can’t work out whether this kind of thing would be supported, and the QA403 page mentions a proprietary communications protocol, which makes me think that it may not be. Does anyone know if this will work, or am I barking up the wrong tree?

[JackB]

I’ve now found some forum posts I missed before. Looks like the bare metal or REST interfaces are what I’m looking for. Are these 1) available on the QA403? 2) documented anywhere?

[matt]

Hi @JackB,

The first thread below might be a good starting point. I have a lot of changes to the PyQa40x repo coming that will hopefully get pushed up today or this weekend. The aim for me is an environment for quick prototyping of algorithms using the QA403 and python. I’m not a python guy, but with llm can make it do what I need.

The second link is the PyQa40x repo.

If you can deal with discrete communication events (the bullets you outline suggest you can), then what is in place currently should work. Just define and output (DAC) and input (ADC) buffer, and call send_receive() and that will handle everything for you. When that’s done, you can process the input ADC data. See the readme at the PyQa40x repo.

The send_receive() will give you a consistent timing relationship between the ADC and DAC, which is often very helpful when first trying algorithms.

Python and QA40x

Here’s a standalone program for looking at energy decay in a room. There are just two calls to the QA40x Python analyzer code: The init() and the sendreceive() for getting the calibrated time-domain data in and out of the QA403. The rest is handled in the code below. Much of the code below doesn’t care about the absolute calibration, except for the instantaneous RMS estimate. I used an SM58 in the home office, and should have used the Earthworks M23R. But it was at work. I used GPT a lot for th…

GitHub - QuantAsylum/PyQa40x: Python library for bare-metal interface to QA40x...

QA40x_BareMetal

Very nice, thank you for the example code. I’d like to see an audio driver (like getting the ASIO QA401 driver running for current hardware). This could be a start. The commands are hard-coded, no definition file? Some must be missing. How is the sample rate set, how to use the front panel I2S, how to read power stats, etc? There are 6 endpoints? So far, got the relays clicking in Python, with this piece of code: #!/usr/bin/env python3 import time # just for test import struct import libusb_…

[JackB]

That looks like just the thing, thanks. We’ll go ahead and order