We are looking for an alternative to our current DSP solution. The ADAU1452 looks good but the development of the Sigma series seems to go faster than the documentations.
Our questions are:
1. Can we build some complex function blocks using the provided basic blocks, save them in a custom library and use them as such afterward? (hierachical design)
2. It does not seem possible to write assembly for signal processing but is it possible to write lines of code to build custom interfaces?
3. Do we know where the SigmaStudio software blocks place the parameters in the DSP (possible to check the log of the compiler?) and can we access and modify them from oustide, via SPI, using a MCU? We would like to change filters parameters, gains, delays, etc. possibly at run time.
4. Is it possible to load a binary without using ADI tools, and store it to EEPROM? This is for software upgrade by the client, for example using USB->SPI interface.
5. Is it possible to make the audio input port as master? (DSP master clock being derived from input port rate, or else DSP running asychronously, knowing the incoming sampling rate and delivering the samples to the output following the rate of the input IRQ --- Not speaking of ASRC here)
Or is it only possible to have the DSP master over input and output ports?
6. Are there blocks to detect the incoming sampling rate ? We would need to switch from one processing to another depending on the input rate.
7. Can we know how the function blocks are built? (for example, for an 2nd order IIR filter, whether it is a direct form 1 or 2. And for double precision blocks, are the parameters in DP as well as the MACs and is the output of the block reduced to SP (dithering yes/no, which type?) or remains DP?
8. Do we have control over the words scaling (the 24bit audio data cast in the 32/64 bit word)? or shall we waste cycles using gain blocks?