All times are UTC+01:00




Post new topic  Reply to topic  [ 19 posts ]  Go to page Previous 1 2
Author Message
PostPosted: Sun Feb 09, 2025 1:51 pm 

Joined: Thu Apr 21, 2011 1:52 pm
Posts: 120
Quote:
Ok, your values look similar to mine - are you using that hardware clock?? Because I just tested with a PTP master in a VMware image, which I wouldn't expect to be very accurate at all.

Either way, I want to add some code to smooth clock differences. Typically, you'll often see a large negative value followed by a similar positive value, indicating that that first clock came in (or was handled!) too late. That's easy to fix, I'll probably look at it on Monday.

In other good news: We now have an automatic test framework, not just for the processing (we already had that), but now also for the web interface. Which among others now automatically tests that all the different programs on the Pi image are successfully starting. We used to test that manually before, but we've sometimes missed things (it turns out that 10.50 Stereo Tool doesn't work on the latest version of our image, oops). More tests will be added, we have a list of things to test, and from now on, for everything that breaks we will create a test case.
yes I use that master clock


Top
   
PostPosted: Sun Feb 16, 2025 4:53 pm 

Joined: Thu Apr 21, 2011 1:52 pm
Posts: 120
Hi,
I noticed that if ST is started as a service, it does not pass the RT in the rds, I tried both as a network drive and as a real path on the network (eg. \\mypc\RDS_onair.txt ..... Z:\RDS_onair.txt)
if I start it from the .exe everything is fine.

Edit: It would be useful to also display RT in WebInterface.

@hvz this week I will be in the lab and I will try OUT AES67 vs an O9SG


Top
   
PostPosted: Mon Feb 17, 2025 1:06 am 

Joined: Tue Mar 10, 2015 9:03 pm
Posts: 241
Location: Utah, USA
The DSP version that's in the 003 beta is actually 002.

_________________
America's Country
Totally RAD Radio


Top
   
PostPosted: Tue Feb 18, 2025 1:04 pm 
Site Admin
User avatar

Joined: Mon Mar 17, 2008 1:40 am
Posts: 11404
BETA007 is posted now. Mainly AES67 output improvements, and performance improvements when running the new compressor in Peak mode. We have also re-enabled RIST for testing, and RIST will now hopefully recover from "Socket error!" messages (a bug in the RIST library itself, which has reports online from up to 2 years ago without any hint at a solution, so we decided to work around it).


Top
   
PostPosted: Wed Feb 19, 2025 1:18 pm 
Site Admin
User avatar

Joined: Mon Mar 17, 2008 1:40 am
Posts: 11404
BETA008 is posted, biggest update is that it can generate SDP files for AES67. So you don't need to make them yourself anymore! Does not yet work via the web interface.


Top
   
PostPosted: Wed Feb 19, 2025 7:56 pm 
Site Admin
User avatar

Joined: Mon Mar 17, 2008 1:40 am
Posts: 11404
We have a new version of our SD card image for the Raspberry Pi 4 and 5 (NOT 3!). This new image contains some important Raspberry kernel and firmware optimizations, which in total (on on my test Pi 5 here) gives a whopping 29% performance improvement. The CPU usage for the same processing went down from 37.5% to 29%! Also, setting the processing block size from 2048 (the default on the Pi) to 4096 has very little effect on the CPU usage.

Here's the image with version 10.51 BETA008 installed. We will officially upload this image when 10.51 is released:
https://download.thimeo.com/Thimeo_Pi45 ... ETA008.zip

You cannot get these improvements by just using the installer, it's really a fully new image. (It *might* be possible to make an installer that updates everything, but that's a bit risky, we would need to properly check for errors and handle those properly). If you want to keep your existing image, running the following commands manually should give you the same effect:

The upgrade will upgrade the kernel from 6.6.51 to 6.6.74. Note: If you're NOT using our image that runs 6.6.51, do NOT do this or it will break a lot of things.

First, run:
uname -a
Verify that you're at version 6.6.51. DO NOT CONTINUE if you're on any other version.

sudo apt-get update
sudo apt full-upgrade
sudo apt install rpi-eeprom
sudo rpi-eeprom-update -a
sudo reboot

After rebooting you should see the improved performance.

Please let us know how much performance improvement you see, on Pi 4's and 5's. It appears to matter how much memory the Pi has, so please mention that as well.

For the first few people who use our new image: Please check the contents of /home/thimeo/thimeo/log.updater/rpi-eeprom-update.log . This file contains the output of rpi-eeprom-update -a; after the first boot it should mention that the firmware was upgraded, on future boots it should just say that nothing changed:
Code:
BOOTLOADER: up to date CURRENT: Wed Feb 12 10:51:52 UTC 2025 (1739357512) LATEST: Wed Feb 12 10:51:52 UTC 2025 (1739357512) RELEASE: default (/usr/lib/firmware/raspberrypi/bootloader-2712/default) Use raspi-config to change the release.


Top
   
PostPosted: Thu Feb 27, 2025 3:20 am 

Joined: Tue Mar 10, 2015 9:03 pm
Posts: 241
Location: Utah, USA
Getting File Not Found trying to download the DSP or the VST2-32bit versions of 012 beta.

_________________
America's Country
Totally RAD Radio


Top
   
PostPosted: Thu Mar 06, 2025 10:02 pm 
User avatar

Joined: Thu Apr 07, 2011 2:37 am
Posts: 149
Location: Napier, New Zealand
This seems to be released as stable now, but the changelog hasn't been updated?

https://www.thimeo.com/documentation/ve ... story.html

_________________
I just use ST with liquidsoap to enhance a streaming radio station I like...


Top
   
PostPosted: Fri Mar 07, 2025 11:54 am 

Joined: Thu Apr 21, 2011 1:52 pm
Posts: 120
Quote:
This seems to be released as stable now, but the changelog hasn't been updated?

https://www.thimeo.com/documentation/ve ... story.html
same thing for me (also because I installed it on air)


@hvz , I don't know if you plan to create a new AGC, however, as I asked you some time ago, for those like me who use AGC (4 bands, I recently updated the advanced dynamic license) it would be nice to have the casper as a detector for the gate, and eliminate the wideband gate or implement casper as is done for the new adaptive compressor, in the AGC, naturally with a gate threshold manageable for each band and, importantly, also an adjustable output for each band, the same goes for the matrix which is manageable for each band, as in the adaptive compressor.


Top
   
Display posts from previous:  Sort by  
Post new topic  Reply to topic  [ 19 posts ]  Go to page Previous 1 2

All times are UTC+01:00


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB® Forum Software © phpBB Limited