New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
RDCatch Recording: nothing in cut after apparent successful recording #785
Comments
Confirmed here. |
I've noticed that tonight's package builds would include a pair of commits related to RDCatch problems, one of them including this...so I've checked this by re-scheduling a recording. I'm not sure whether should I close this issue, since may be other commits related to all this queued to be pushed and tested... (I will open another, separated issue, related to yet another one-shot issue regarding recordings in rdcatch, as I thing those are totally unrelated) |
This specific issue should be fixed by c22a040. That said, in the course of fixing this, I uncovered a bunch of problems lurking in the RDCatch subsystem. Stay tuned for more fixes! |
OK... so, I think I better close this particular one, so the issues list gets cleaner. Thanks! |
Hi again... I'm reopening this. However, today, I've realized in RDLibrary that the recording cart was 'red' ... so I looked at the cut, and no audio present was reported... and it should. From GUI perspective, apparently, it works: RDCatch vumeter at the record deck goes live, and everything seems to be working fine.
Looks good! ... it even seems it is handling the temp file as it ends generating the cut on the cart...however, nothing ends on the cut. How could I troubleshoot this? EDIT 1: Listing /var/snd for record files shows some old recordings there... maybe old troublesome pieces of recordings... I think they're safe to delete:
EDIT 2: Checked wether the temp file does actually grows in size as a new attemtp goes on
So, apparently, audio (or something) went to the file... but then seems as if the actual import into the cart/cut never happened or failed silently |
Unable to reproduce the problem here. Any chance that you could post a screenshot of the 'Edit Recording' dialog in |
Hi Fred... sure. What's most surprising is that actual data ends into the temp file... it very much seems as it fails at the very last step. EDIT: I removed also the 'normalize' option just to ease the ingest of the recorded audio (which is actually recording) And it worked!!!! I have to make further tests to try to have a consistent behaviour... at this point, seems that if normalize AND autotrim are unchecked, the recording works... but I would like to test several times every combination |
Hi again... I've done some further testing Here's an screenshot of RDCatch event scrolled right for details of event being shown (Nice feature I didn't notice!!! ... I however notice the trim field is showing -4000dB instead of -40.00dB, and also I see there's no column for the normalization field) regarding codec, channels, bitrate, etc Early results seem to point at a problem with NORMALIZATION. EDIT: |
Still unable to goose it here. Could you configure your Syslog setup to record DEBUG level messages and then send me the output during the entire course of the event? That should flag exactly where this is going off the rails. |
HI Fred... I'm trying to get DEBUG messages from rivendell, but I'm failing to do so. EDIT: OK ... I fugured out that /etc/rd.conf defaults to normal user log, and has to be set accordingly to match LOCAL7 facility as provided in rivendell's rsyslog.d conf file! The, I took a look at /etc/rsyslog.d/10-rivendell.conf , uncommented the line as suggested: So... Time to repeat the bugged recording -> export test Related lines on /var/log/rivendell/debug:
Related lines on /var/log/rivendell/operations
Apparently, it all goes well... however, I got a 'red cart' (no audio in cut) whnever normalize is checked. |
Should be fixed in current 'qt5' branch. Please test! |
Hi fred. During this August holidays I've been pushing my mockup VM quite a bit... I've had a ton of fun setting it up kinda close to a real radio (I've even named it :-P), and I've been theese weeks testing a whole setup of automated recordings of certain parts of the grid, together with their automatic upload and publishing on a podcast feed... so far, it works perfect! recordings are done and upload and it is working for weeks, super stable (as far as I've managed to test)... but still, the normalize checkbox must be kept untiked, otherwise, the temp wav audio record never gets to the target cut. |
'Kay, thanks for testing! I'll pound on it a bit more here. |
'Kay, I've been testing this heavily on Linux Mint 21 (basically, Ubuntu Jammy with XFCE), using a JACK-based setup. It's been rock solid. I'm not sure where to go from here. Any chance that I could get access to your instance and see this in realtime? |
Hi! ... The VM is quite isolated, very hard to get SSH Anyhow, chances are that what I'm experiencing being a problem very specific to this install for whatever reason. For the sake of following through involved setup, RDCatch record event looks like this: And the host RDLibrary settings being: I don't recall of anything else that may affect the final audio ingest into the Library... If so I'm missing it now. I've actually checked, during recording, at filesystem level, that actual .wav data is actually recorded in the temp file.. I mean, the actual audio is there. Moreover, without noormalization, it works fine. Cheers! EDIT:
|
The other way around, actually; the settings in 'RDLibrary' are the defaults, which can then be modified on a per-event basis. The screenshots are very helpful. Any chance you could post the one for RDAdmin->ManageHosts->[hostname]->RDCatch as well? |
No worries; I see you have already! (Time for coffee...) |
I'm closing this one... I feel we may get into this in the future as I strongly believe (after the Magick feed image issues) that all this could be simply a matter of media libraries on Debian Bullseye being more recent that the ones in CentOS (specially for those using deb-multimedia packages). It is just a matter of being on alert and recall on this, provide eventual future problems or issue reports on blank recordings. Thank you for your presence. |
Hi.
I've been playing with rdcatch, trying to do some recordings.
My setup is a VM with just Jackd, so only card 0, wich is jack.
I have setup card 0 as a switch device (just to allow for routing as I learnt from RD 3.X)
I have setup a recording event in RDCatch.
I've selected card 0 / port 0
Format is mpeg layer 2 / bitrate 128kb
Switcher Host to NONE
This doesn't work, and following line appears in syslog:
Bitrate=0 caught my attention... is that right?
I've managed to make it to do something by seting the switcher host to the host itself, then choosing the jackd card switcher...but it seems to trigger a sound loop/feed-back at the input with lots of noise.
No recording to file anywaus whatsoever
The text was updated successfully, but these errors were encountered: