PGA gain in supercollider

giuliomoro

I was and you are missing -I after CL= (see my
amended example above)

scientificsound

that works, thanks!

are these functions documented somewhere? I wasn't able to find any info on learn.bela.io, and it would be nice to access other missing functions like turning the led off.

giuliomoro

Well ... they are documented here, but the list is out of date because it doesn't get updated regularly. The most up-to-date list is by running any Bela C++ program with --help. Just pick any C++ example from the IDE and add to the command line options box in the Project setttings tab of the IDE --help . The program will run but exit immediately and print all available options.

RafaeleAndrade

Hi everyone,

I'm happy to read these previous post about the pgagain stage, but I having some troubles setting up mine

I want to optimize the gain stage for individual channels on the CTAG + Bela for the input channels [0,1,2,3] . At this moment I dont see any difference on the first channels 0,1 if I change on the scd code patch:

s.options.adcLevel = 1; // workaround so .pgaGainLeft/Right works (must be out of range > 0) --> (https://forum.bela.io/d/2810-pga-gain-in-supercollider/3) s.options.pgaGainLeft = 30.0; // THIS WORKS! (default is 10 dB) s.options.pgaGainRight = 0.0; // (but only for positive numbers...!)

Does the CTAG has a different approach?

Also, if I try to create a file at '/root/.bela/belaconfig ' it says this place doesn't exist. Since I have 4 input channels, I was trying:

nano /root/.bela/belaconfig

and then:
CL=-I2,30,3,30

Is this the correct way to write it? Where would be the right adress to create the file?

giuliomoro

there's no pga gain on the ctag unfortunately.

Also, if I try to create a file at '/root/.bela/belaconfig ' it says this place doesn't exist.

You need to create /root/.bela first with mkdir -p /root/.bela

RafaeleAndrade

Hey people,

I would like to ask the same question actually 6 months later but now using Bela Multi Gem.

Is there a way I can set the digital pregain 10 dB louder for 8 input channels?

giuliomoro

it should still work the same - if it worked then. Did it work then?

RafaeleAndrade

By following this conversation, I did follow these steps on my Bela multi Gem:

Access bela:ssh debian@bela.local
Create /root/.bela first with: sudo mkdir -p /root/.bela
Create a file with: sudo nano /root/.bela/belaconfig
Write on the file: CL=-I2,10,3,10,4,10,5,10,6,10,7,10
Use the adcLevel hack( s.options.adcLevel = 1) on sc patch. The gain of the first two channels can be set via pgaGainLeft/Right from the scd file.
sudo reboot

However, when I reboot only the sound of 0 and 1 works. The remaining inputs are almost muted (I only hear the noise gate from them)

giuliomoro

RafaeleAndrade Access bela:ssh debian@bela.local

If you are stil using debian@bela.local for real, then you are not up to date. Get one of the release images https://github.com/BelaPlatform/gem-image-builder/releases/ to make sure you are up to date. Not to say that then it will work, but to say that I can't support you on that outdated image.

try CL=--verbose and see if you get a verbose log like this one:

Warning in SC_BelaDriver::DriverSetup(): hardware buffer size (16) different from SC audio buffer size (64). Changed the hardware buffer size to be equal to the SC audio buffer size .
Bela_initAudio()
Starting with period size 64 ;analog disabled
DAC level 999999.000000 dB; ADC level 999999.000000 dB; headphone level 999999.000000 dB
Detected hardware: BelaRevC
Hardware specified at the command line: NoHardware
Hardware specified in the user's belaconfig: NoHardware
Hardware to be used: BelaRevC
Codec mode: 0 ()
gFifoFactor: 1
core audioFrames: 64
Stopping PRU1_1
PRU memory mapped to ARM:
digital: 0xffffb5439000 0xffffb5439400
audio: 0xffffb543a000 0xffffb543a200 0xffffb543a400 0xffffb543a500
analog: (nil) (nil) (nil) (nil)
analog offset: 0xffff00004abc8000 0xffff00004abc8000 0xffff00004abc8000 0xffff00004abc8000
Mcasp parameters:
inSerializers: { {axr: 1, mask: 0x3} }
outSerializers: { {axr: 0, mask: 0x3}, {axr: 6, mask: 0xff} }
numSlots: 8
slotSize: 32
dataSize: 16
bitDelay: 0
auxClkIn: 9.6e+07
ahclkFreq: 2.4e+07
ahclkIsInternal: 1
aclkIsInternal: 0
wclkIsInternal: 0
wclkIsWord: 0
wclkFalling: 0
externalSamplesRisingEdge: 0
r.outChannels: 16
r.inChannels: 2
r.outSerializersDisabledSubSlots: 0x5550
r.inSerializersDisabledSubSlots: 0x0
Mcasp registers:
      pdir: 0x08000041
     rmask: 0x0000ffff
      rfmt: 0x000080f4
   afsrctl: 0x00000400
  aclkrctl: 0x00000000
 ahclkrctl: 0x00008003
      rtdm: 0x00000003
     xmask: 0x0000ffff
      xfmt: 0x000080f4
   afsxctl: 0x00000400
  aclkxctl: 0x00000000
 ahclkxctl: 0x00008003
      xtdm: 0x000000ff
    srctln: 1 2 0 0 0 0 1 0 0 0 0 0 0 0 0 0
  wfifoctl: 0x00010202
  rfifoctl: 0x00010101
SchedulableTask WSClient_signalMonitor40100_bin starting
Bela_cleanupAudio()
SchedulableTask WSClient_signalMonitor40100_bin exiting
SC_BelaDriver: >>DriverSetup - Running on PRU (1)
Configured with
 (8) analog input and (8) analog output channels, (16) digital channels, and (0) multiplexer channels.
 HeadphoneLevel (-6.000000 dB), pga_gain_left (10.000000 dB) and pga_gain_right (10.000000 dB)
 DAC Level (0.000000 dB), ADC Level (0.000000 dB) oscilloscope channels (8)
Speakers are not muted.
SC_AudioDriver: sample rate = 44100.000000, driver's block size = 64
SuperCollider 3 server ready.
Requested notification messages from server 'localhost'
localhost: server process's maxLogins (4) matches with my options.
localhost: keeping clientID (0) as confirmed by server process.
Shared memory server interface initialized

If you don't see the verbose log, there may be something deeper not working . If it works, then do CL=-I10which should set all channels to 10dB gain and see if that works. But again, first of all get a release image.

RafaeleAndrade

Good morning!

I’ve been delayed in installing the new image mostly because I had a lot of concerts, and I didn’t want to risk a full re-patching process, especially with the TrillCraft UGens, WiFi setup, and all the wiring on my instrument.

Now, though, I’ve finally tried the new image on a fresh micro USB card, but I’m running into issues using SuperCollider with the audio input channels. I always rely on the SoundIn UGen to capture the signals from my pickups—currently, I’m working with 9 inputs: 8 electromagnetic pickups (very low gain) plus one Arduino mic input (which is louder). When I run my simple code to listen to all these channels, I only get sound from channel 0 and 1. Even after modifying the server options (setting the input bus channels to 10 and output bus channels to 10), it still doesn’t work. I checked other patches, and they run fine, so it seems isolated to this one SuperCollider patch.
Do you have any idea what might have changed in the new image or if there’s something I missed in the configuration? Any advice would be super appreciated. Thanks so much!
Best,
Rafaele

giuliomoro

RafaeleAndrade I checked other patches, and they run fine, so it seems isolated to this one SuperCollider patch.

Do you mean that other supercollider patches can read channels beyond 2 through 9? Or that examples in other languages can ?

RafaeleAndrade

giuliomoro I tried the example patch from c++ of multichannel delay reading channel 3 instead of 2 for example. That didn't work as well.

What it works in other patches and projects is the output channel configuration. If I send a sinewave to 8 outputs channels for example, it works pretty fine.

giuliomoro

what's the revision of your Multi cape? A2 or A3 ?

RafaeleAndrade

giuliomoro I have revision A3

RafaeleAndrade

I wanted to give you an update on my testing. I wrote a simple C++ patch to check the MultiGem channels, but it didn't work in the new image builder from February or January. To double-check, I rolled back to my old November image,It works as expected: I can hear sounds from SoundIn.ar([2,3,4,5,6,7,8]). The channels from 2 to 8 respond fine, but when I try to adjust the gain in the old image, the behavior breaks down similarly to the new image as well.

Could we take a look together or see if something changed in the signal processing chain?
Let me know! Thanks so much.

giuliomoro

with that version of the board you need either a hw rework of the board or the following software patch:

echo "CPPFLAGS+=-DGEMMULTI_REVA3_NOREWORK" >> /root/Bela/CustomMakefileTop.in
touch /root/Bela/core/Gem_Multi_Codec.cpp /root/Bela/include/BelaContextManager.h
make -C /root/Bela lib

After this, try again and you should have all 10 input channels working - at the very least in C++/PureData and in the signal monitor.

RafaeleAndrade

Sorry Giulio , when I run this code on the ssh terminal, I get a different response than in the bela platform terminal, here is the print screen for you

giuliomoro

RafaeleAndrade

Running these 2 lines at the terminal doesn't solve the problem on my bela

giuliomoro

Run this program, look at the scope and look at the signal monitor:

#include <Bela.h>
#include <libraries/Scope/Scope.h>

Scope scope;

bool setup(BelaContext *context, void *userData)
{
	scope.setup(context->audioInChannels, context->audioSampleRate);
	return true;
}

void render(BelaContext *context, void *userData)
{
	for(unsigned int n = 0; n < context->audioFrames; ++n)
	{
		float vals[context->audioInChannels];
		for(unsigned int c = 0; c < context->audioInChannels; ++c)
			vals[c] = audioRead(context, n, c);
		scope.log(vals);
	}
}

void cleanup(BelaContext *context, void *userData)
{
}

RafaeleAndrade

giuliomoro I
get:

giuliomoro

It feels like you put some garbage in CustomMakefileTop.in. Do this:

echo "CPPFLAGS+=-DGEMMULTI_REVA3_NOREWORK" > /root/Bela/CustomMakefileTop.in
touch /root/Bela/core/Gem_Multi_Codec.cpp /root/Bela/include/BelaContextManager.h

RafaeleAndrade

giuliomoro running this code and rebooting doesnt give me any progress as well :/

« Previous Page Next Page »