Radioberry Rpi4 Performance

[Peter]

Hello everyone 

I have been testing Radioberry now on RPi4 with 2Gb RAM  and it have woorked good but if I try to use more than 96000 sample ie 192 or 384 i get hacked sound or a interrupted sound so I supose Rpi has hard to catch up.

Do anyone have experice with this if there is better to have Rpi4 with  8 Gb or is the 

speed limit the same, i have not seen any CPU clock speed change on Rpi4 

73's 

Peter / SM6NTR

[pa3gsb]

Peter

More memory will not help.

https://github.com/pa3gsb/Radioberry-2.x/wiki/Radioberry-performance

If you really want more speed; http://www.pa3gsb.nl/2021/05/31/radioberry-juice/

(i have still 2 juice boards available)

73 Johan

PA3GSB

Op vrijdag 25 februari 2022 om 08:11:54 UTC+1 schreef sm6...@gmail.com:

[Paulh002]

I use radioberry on RPI4 without any problem on 384 khz.

But I use my own developed software, You can find it at https://github.com/paulh002/sdrberry

73,

Paul

Op vr 25 feb. 2022 om 08:12 schreef Peter <sm6...@gmail.com>:

--
You received this message because you are subscribed to the Google Groups "Radioberry" group.
To unsubscribe from this group and stop receiving emails from it, send an email to radioberry+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/8792cc0b-95b9-4ab1-b663-4c7edd9ac22en%40googlegroups.com.

[Peter]

I have NO problem at all to use 384 if i connect over IP to windows for example SDRConsole  

The problem is only local with the built in audio interface  so maybe it is little strange because it should go faster locally than over network  ?

//Peter 

[Paulh002]

I only use a local audio device, that is I have a USB audio adapter, because the PI does not have microphone support.

The problem I have in my software is that the decimation factor is not constant, The radioberry is not running exactly 384 Khz and 

the Raspi audio card is not exactly 48 Khz. This will result in buffer over- or underruns. What I saw in the code of pihpsdr that it drops data when the decimation is not aligned.

PC software is maybe more advanced. I now measure the buffersize and adjust the decimation accordingly. For me this is good enough, but a more advanced feedback loop would be better.

You also do not want the buffer to grow too large that is causing unwanted delays.

73

Paul   

Op vr 25 feb. 2022 om 12:03 schreef Peter <sm6...@gmail.com>:

To view this discussion on the web visit https://groups.google.com/d/msgid/radioberry/8457efda-972f-4e82-be40-93f8f2335d53n%40googlegroups.com.

[Brad Jeffries]

Hi Peter,

First of all thanks for such a wonderful project!   I am interested in the juice board you say you have 2 available? 

73 

Brad 

[Brad Jeffries]

I meant to say Thanks Johan.  But thanks Peter too :)