I am using the DVI / I2S interface to play (and sometimes receive) audio over the voice channel. It works very well in general. The only problem is that I would like to increase the gain for the uplink (outgoing) voice path.
So far the only way I have is by doing a PSEL=1 and then HFCFG=1,900,1. But then if I switch back to PSEL=0 I was expecting it to drop back to the ‘unadjusted /default’ gain but it remains high.
The documentation on audio profiles is very poor. I saw a reference to using audio profiles 2 or 3 for a car for example somewhere in this forum. A car is not my application’s environment but my point is the AT commands reference says nothing about PSEL=2 or 3 other than suggesting that these are ‘alternative profiles’ that can be configured and saved.
In short, what is the recommended way to increase / decrease the audio gain over the DVI (digital audio front end) on a per call basis?
The only way to increase the gain in DVI mode is trough the PSETcommand and <scal:in> parameters so for example AT#PSET=6 allow to increase the gain by 5dB
HFCFG is not a command designed for the gain set up.
Thanks, that’s perfect
I have noticed that occassionally when I play audio into the DVI interface that the audio fades for a number of seconds. It starts off well for about a second, and then fades (as listened to at the far end). It seems to get attenuated. Then at a particular point it recovers and is fine.
On the other hand, my "old" implementation which played audio over the main uart interface is always solid.
Is there any difference between the audio path via the UART and the audio path via the DVI in terms of gain or automatic gain control or filtering?
The first chunk of voice that I play is recorded on a very rudimentary circuit and may contain components at the original sampling freq 16kHz. I would have expected this to be filtered out successfully.
it’s very strange indeed. Since I believe you are using #PSEL with parameter other than zero to set up the volume via #PSET do you enable the AGC maybe? I mean AT#SHSAGC
Or do you sittlk use HFCF?
The second idea I can have is your component at 16KHz can be view as a tone which is faded by GSM usually. this should happens also via UART. The only difference between the two audio path is that DVI pass trough a BIQUAD filter while UART is routed directly to the internal buffer of ADC, even if I can’t see a good reason for this to let the 16KHz to pass trough or being downsampled …
Thanks Luca for the reply. I found a problem on my own audio driver (in a decompression algorithm) and now this problem is resolved. Nevertheless it was useful to learn more about the audio paths. Thanks.
Hit enter to search or ESC to close
Knowledge Base & Download Zone