Startup question: audio head / log / audio input and output

I just started with and sushi.

At the moment I don’t have any audio output.
I tried to start the pre-configurated arpeggiator:
sushi -c ~/config_files/mda-vst2-configs/config_play_arp_mda_link.json
and this leads to an error message:
“No audio frontend selected.”
I tried to define a hat with:
sudo elk_system_utils --set-audio-hat hifiberry-dac-plus-adc-pro
My hat in fact is a dac2 pro.

Is there a startup tutorial for question like:
→ where is the error log for sushi?
→ how to find out, which head is recognized and if its running correctly?
→ where can I see, if the audio inputs are recognized by sushi?

Thank you for any link or help to get a first audio output!

Add.: meanwhile I tried some more times to select a head, also I tried elk-pi
and resetted to hifiberry-dac-plus-adc-pro, but the message
still is “No audio frontend selected”.

Another observation:
I started another plugin (neuralpi): this one does not show this message:
“No audio frontend selected.”
But unfortunately it does not provide sound at the outputs of the hifiberry.

watch -n 0.5 cat /proc/xenomai/sched/stat
3 708 6 6 21 0 000680c0 0.0 sushi_b64
1 712 2 209646 623053 0 00048040 39.6 sushi_b64
0 713 1 413767 210006 0 00048042 1.0 sushi_b64
0 717 1 418570 837141 0 00048042 3.1 sushi_b64#

Started a new post meanwhile, which is more specific about my
beginners problems and ways to find out solutions:
Control sushi on boards? - #2 by RJazz

Hi RJazz and welcome to the forums.

Just to clear up some confusion. Audio frontend here refers to which software api sushi should use to connect to the audio interface (jack/raspa/portaudio), not the actual audio interface/hat connected.

When you start sushi you need to pass one of the flags -r,- j, -o or -d to select which audio frontend to use. If you are running it on a raspberry pi with Elk Audio OS, you should use the -r flag for best realtime performance. You can also run sushi --help for more info

You should also use elk_system_utils to set the correct hat for the audio driver (then a reboot is needed afaik), but you seem to have that covered already.

The log file for sushi is by default placed in /tmp/sushi.log

If sushi starts correctly, then the audio hat should be ok (the drive should complain otherwise).
The best way to test would be with a simple config file with 1 track with no plugins that just passes audio through from input to output.


Thank you Gustav, I will try this.
Before I had to and still have to solve some problems, like:
error msg for writing to /tmp/sushi.log when logged in as mind-user not as root.
I tried to install osc on another system (is it possible, to just put in the ip of
the raspberry and control it from a pc with osc?), But osc seems to need
super-user rights to install on the system, which I don’t have there.

Maybe I can see in osc then, if an input signal is detected.
Another path I was following, was to install python for using elk-pi to control
sushi. But I did not find a tutorial how to install python on
(is it only available in the development-version?).
Obviously apt-get is not available on elk, so how would it be possible to use
python with elk-py to control sushi? This would be very interesting to me.
With elk-py it should be very easy, to get detailed info, if there are inputs
or to write a diagnosis-script in python.

So, there are still some open topics for me.
In the other post I mentioned above, I already could see, that the
hat is detected with dmesg:
[ 14.616061] Elk hat: hifi-berry
[ 14.624188] audio_rtdm: driver initialized
[ 15.824332] bcmgenet: Skipping UMAC reset

So maybe it could guess, that as you say the audio hat seems to be o.k.
So most interesting to me would be to me, if it is possible to install
python (3.10?) on my elk system.

Just checked my sush-statements I used so far. One example was :
sushi -r --multicore-processing=2 -c ~/config_files/config_neuralpi.json &
from a tutorial about neuralpi plugin.
The sushi.log afterwards was empty.

O.k., some time later and one step further:
I got open stage control running on another system.
With ip I should get control over sushi.
Now still I could not find a tutorial on how to load a basic configuration into osc editor.

elk-examples/mda_jx10_vst2_open_stage_control_gui.json at master · elk-audio/elk-examples · GitHub)
seems to be something like that, but osc seems not to be able to load the
file (the file is not listed in the file dialog).

Is there any tutorial how to setup a basic sushi whith hifberry:
I mean: how to set up two input sliders, two output sliders in osc and get
sushi to react to them?

Meanwhile I have new information here:

Control sushi on boards? - #6 by RJazz

So please post any info/help to that question!