High quality sound recorder and analysis tool

Moved over here - looks to be better placed.

Have a look onto SigmaDSPs and Audio Codec, DACs and ADCs.

http://www.analog.com/en/processors-dsp/sigmadsp/processors/CU_over_SigmaDSP_overview/fca.html

MMA

Analog Microcontroller

Community Manger

Hi Valoo,

Valoo wrote:

The idea is to either store raw audio signal, or sound pressure level (every second), or both.

When you say you want to store this information, how do you mean? Store it in microcontroller memory? On a computer? On a display? I still can't quite picture exactly what this application looks like on a block diagram level.

Hi, sorry I am not keeping this thread very lively. As I am not very familiar with all the science involved in embedded computing and micro controllers, I spend a lot of time trying to catch up ...

Ok, so by "storing" I mean store data on a SDcard for exemple in order to be able to extract signal from the device onto a computer for the signal treatment part.

What I am not really sure about is how much work can be done in real-time inside the DSP and therefore how much work will have to be done on the computer for post-anlysis.

I only need 1 input channel, and no output channel for now.

I'll try to redefine my objectives for the big picture :

I want a device that can :

Record raw sound pressure (24bit/ , 51.2kHz)

Perform real-time sound pressure level (SPL) integration (like what sound level meters do) on 33 Third octave frequency bands (to get dB values in fact) every second.

Store either raw data (51200sample per second) or SPL (1 sample per second) or both

[if it can be done] perform signal treatment operations (means, correlations and convolutions mostly) without real-time needed.

hope that helps... I will go back to catching up now

The SigmaDSP does not have an interface that would easily allow you to store data. It does not have an external RAM interface or any kind of USB interface. It simply has an I2S interface (Inter-IC Sound), which is a serial audio data interface used to transmit synchronous audio data from one IC to another IC on the same PCB.

It can, however, measure the instantaneous and average level of signals (and of course it can do this for the full bandwidth or defined bandwidths, like the 1/3 octave scheme you mentioned), which can then be stored in RAM and either streamed out of the SigmaDSP via I2S to another chip, or read back from the control port (I2C or SPI) from another IC. You can see examples of that here: https://ez.analog.com/message/7669#7669

In other words, you will definitely need more than one chip to accomplish this task. Probably a SigmaDSP plus a USB microcontroller would be the best bet.

It's also important to note that SPL has no meaning in the digital world. The unit dB SPL is a measurement of acoustic pressure. In order to accurately measure dB SPL, you need to first convert acoustic pressure to a voltage, using a microphone, and then convert the voltage to a digital audio signal, using an ADC. You'll have to use the sensitivity spec of the microphone and the full-scale range of the ADC to calculate the transfer function between acoustic pressure and the digital PCM code stored in the DSP. SigmaDSP cannot calculate this for you because it has no concept of the full-scale voltage of the ADC or the sensitivity of the microphone attached to it.

Now that is very interesting !

With the few key words and hints you gave me I am starting to see clearer where I should start.

So I'll try to rephrase to see if I get it correctly.

So starting from the output audio signal of the microphone pre-amplifier, there is :

The ADC, where I get the 24bit/51,2 kHz digital signal

The SigmaDSP, which is in charge of all the real-time signal treatment AND the post-analysis signal treatments (like MLS-based transfer functions ?)

The USB microcontroller who's in charge storing data, sending data( (for post analysis), set constants for signal treatments, handle an optional LCD+buttons interface, etc...

Am I missing anything ?

I am aware of the issue about sound pressure levels. I already have a class 1 microphone with its pre-amplifier and I know the sensitivity of those. I even have a calibrator that delivers a 94dB@1kHz so I can calculate the actual sensitivity of the 2 components periodically or before every measurements.

The only issue here, for me, is the digital world...

Thank you very much for your help !

V

Does it mean I can use something like an arduino or netduino board combined with an ADC and a SigmaDSP for my project ?

Valoo wrote:

Does it mean I can use something like an arduino or netduino board combined with an ADC and a SigmaDSP for my project ?

Yes, that could be a good approach.

Valoo wrote:

Am I missing anything ?

No, that all sounds correct.

Valoo wrote:

The ADC, where I get the 24bit/51,2 kHz digital signal

Some SigmaDSPs (ADAU1701, ADAU1702, ADAU1401A, ADAU1761, ADAU1781) have built-in ADCs, so that would save you one chip.

Valoo wrote:

The SigmaDSP, which is in charge of all the real-time signal treatment AND the post-analysis signal treatments (like MLS-based transfer functions ?)

You'll probably want to check out the library in SigmaStudio to make sure it can accomplish all of the processing tasks you want to implement.

This question has been assumed as answered either offline via email or with a multi-part answer. This question has now been closed out. If you have an inquiry related to this topic please post a new question in the applicable product forum.

Thank you,
EZ Admin