I am using multiple recorders with the SYS option for timecode and with chunking. First recorder is not syncing time, the rest are syncing with the fist one, via ntpd (ticking synchronisation). The problem is that after a couple of hours, apparently randomly, one of the recorders refuse to start a new chunk. The following message is shown in the recorder log:
channel[1] Target tc 01:40:37:18 not found, buffer 01:40:35:22 - 01:40:38:06
Could not start record - not all target timecodes found
How can I mittigate this?
Thank you in advance.
EDIT:
It seems that the problem is related to jumping with the chunk over midnight. Sorry for missleading you.
in MAIN log:
input0 index 0 Timecode discontinuity: -2160000 frames missing - current 00:00:00:00, last 23:59:59:24
00:06:19.072 input0 index 0 duration 22500 reached (total=22500 written=22500 dropped=0)
Is there a logic implemented for this exception when the TC IN is greater than TC OUT?
Can you please indicate the source file where it should be?
Thank you in advance.
Last edit: dvision 2018-08-13
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I think the first problem is that because SYS timecode is not locked to the video, you might get timecode values skipped or repeated occasionally. In your example:
channel[1] Target tc 01:40:37:18 not found, buffer 01:40:35:22 - 01:40:38:06
Could not start record - not all target timecodes found
I'm guessing that timecode 01:40:37:18 got skipped.
In the second problem, I think it might be because you have a pre-roll of zero. The recorder expects the start timecode to be in the buffer but if the start timecode is "now", it might not have got into the buffer yet. (Although I can't remember if chunking mode overrides your pre-roll setting.)
The "-2160000 frames missing" message at midnight is normal and can be ignored.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
By the way, for the SYS problem, you might get better results if the recorders sync to an accurate time source. I don't know if your SDI is locked to a reference but if you can get NTP from the same reference, that would be ideal.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
I am using multiple recorders with the SYS option for timecode and with chunking. First recorder is not syncing time, the rest are syncing with the fist one, via ntpd (ticking synchronisation). The problem is that after a couple of hours, apparently randomly, one of the recorders refuse to start a new chunk. The following message is shown in the recorder log:
How can I mittigate this?
Thank you in advance.
EDIT:
It seems that the problem is related to jumping with the chunk over midnight. Sorry for missleading you.
in MAIN log:
In INPUT log:
Is there a logic implemented for this exception when the TC IN is greater than TC OUT?
Can you please indicate the source file where it should be?
Thank you in advance.
Last edit: dvision 2018-08-13
I think the first problem is that because SYS timecode is not locked to the video, you might get timecode values skipped or repeated occasionally. In your example:
I'm guessing that timecode 01:40:37:18 got skipped.
In the second problem, I think it might be because you have a pre-roll of zero. The recorder expects the start timecode to be in the buffer but if the start timecode is "now", it might not have got into the buffer yet. (Although I can't remember if chunking mode overrides your pre-roll setting.)
The "-2160000 frames missing" message at midnight is normal and can be ignored.
By the way, for the SYS problem, you might get better results if the recorders sync to an accurate time source. I don't know if your SDI is locked to a reference but if you can get NTP from the same reference, that would be ideal.