I've been struggling with HEARTBEAT_TIMEOUT for a long time.

[MomoBrut]

During longer recordings, let's say longer than 3-4 minutes, my IMU trigger, which is supposed to start and then end the recording and the GoPro is supposed to turn off after 5 seconds (!SI6D25H5!O), crashes and ends with a camera reset. It doesn't happen all the time, just occasionally, but it's very annoying and causes problems in production. After the reset, the IMU trigger is triggered again and the process repeats, if the recording is short, it's usually accompanied by a correct end and the camera turning off. I'm sending parts of the log, the first one is with a reset, the second one is with a correct trigger termination following each other, as explained above.

[2022-03-05 03:02:33 --% ] STARTCAPTURE:0
[2022-03-05 03:02:34 --% ] start capture:IMU 0.021g+47dps threshold = 0.032g+15dps
[2022-03-05 03:07:11-delayed ] wake reason:HEARTBEAT_TIMEOUT
...
[2022-03-05 03:02:37 --% ] STARTCAPTURE:0
[2022-03-05 03:02:37 --% ] start capture:IMU 0.003g+1dps threshold = 0.032g+15dps
[2022-03-05 03:02:45 --% ] STOPCAPTURE:0
[2022-03-05 03:02:45 --% ] stop capture:IMU 0.008g+2dps end threshold = 0.032g+15dps

So the problem is in HEARTBEAT_TIMEOUT. Can someone please explain to me what the error is and how to avoid it? It's quite confusing, as if it's some kind of loss of time synchronization, I don't understand it. Mainly, it's a sporadic, not always recurring error, it's confusing me. I have 4 pieces of Hero11 and 2 of them are naked, but with good cooling. And all pieces have this problem regardless of whether they are naked or not.

Thanks in advance for your help.
Momo

[dnewman-gpsw]

HEARTBEAT_TIMEOUT just reporting that the capture locked-up in the previous session, and the heartbeat timer rebooted the frozen camera. The problem is not the heart beat (it is good thing,) but why are your cameras locking up. You will need to consider the other Labs extensions like BITR (H11 reliable ~180 for 24/25/30fps, 160Mbs for 50/60), or anything else that could be adding stress to the camera. When configured with reasonable settings, you should not crash the camera.

[MomoBrut]

Thanks for the quick response and explanation.

I'm testing on 5.3k/50fps/10b settings, I don't use BITR, so just the standard High setting. My permanent LABS settings, in addition to the already mentioned IMU trigger triggered during BOOT, are the following: WAKE=2, LOGB=400, NR01=20, EXPS=1 and FRNT=1. I must admit that during long-term testing I found out that if I permanently use TUSB=1 and BYPS=1, the aforementioned rebooting becomes almost regular. As for SD cards, I tried different types, but they didn't have any clear effect on the behavior. And one more observation, the reboot tends to happen during the first start of the camera, after the reboot the camera more often works as it should, only exceptionally does another reboot occur. It's strange, but that's how it could be described statistically.

Are any of those settings problematic, where to look for the problem? At the same time, I don't need audio recording, the image is enough, so DAUD=1 could help or is it pointless?

Momo

[MomoBrut]

I tried to solve it myself. I did a lot of tests with the Hero 11 over the last few days, it took several hours, fortunately there are other things you can do with it.

I kept recording 5K3/50fps/10b with High settings lasting at least 10 minutes. And I did it with both the standard version and the naked version with active cooling. I started with a clean install of the latest GoPro Labs firmware and I didn't have any macros or settings activated. I gradually added the settings one by one, tested separately and then together in this order LOGB=400, NR01=20, FRNT=1, EXPS=1, WAKE=2 and finally I added the IMU trigger (!SI6D25H5!O). I always let the camera cool down after each test and deleted the SD card. And the result is quite clear, none of the settings cause a reboot during the first 10 minutes of recording, at least in the number of tests I did. The instability will only start after the IMU trigger is introduced. The problem is that disabling the IMU trigger will cause me a pretty major problem in the need to manually start and stop recording, which is simply not possible sometimes. So disabling it will solve the reboot problem, but in some situations the camera is not usable at all.

Probably because the IMU trigger starts constant logging of the gyro and accelerometer values, the camera is then at the limit of what it can simply handle. That is my interpretation of the problem. Is there any way to reduce the frequency of logging values ​​so that this process is not so demanding? I think that it is not entirely necessary that these values ​​are known with a time interval smaller than 1s, for my purposes an interval of 2s will be completely acceptable (even 5s is still acceptable). Could this be implemented somehow? To have the ability to set this logging interval?

Thanks.
Momo

[dnewman-gpsw]

Yes, you are using the original IMU trigger, when there is a newer scriptable solution, but it is when more complex (for more power.)

You could do some think like this (untested):
!9N!9N!7N<r0>a0.1!S+!R9>r0<a0.01=A+1+!1N+!R9>A5!E+!1O

!9N!9N!7N - Delay 25s
<r0>a0.1!S+!R9 - if not recording and accelerometer > 0.1g, Start recording repeat from character pos 9
>r0<a0.01=A+1+!1N+!R9 - if recording and accel < 0.01g, A=A+1 once a second repeat
>A5!E+!1O - if A is greater than 5 (5 seconds), end capture, shutdown

[MomoBrut]

Thanks a lot for the explanation and sample scripting.

In the meantime, I tested another naked Hero 11 and with the verified settings that did not reboot, the third one reboots again. But I figured out what caused it, and it's strange. On the first two cameras, I had Rear display set to turn off after 1 min and Front display to Never. On the third camera, I had Rear set to turn off after 1 min and Front display to Same as Rear. And this causes the reboot. The reboot occurs whenever the front display is turned off. If I set this setting on the first two tested cameras, the reboot occurs again. With Front display set to Never, the reboot does not occur, even when the camera is very hot.

Do you have anything to do with it? I practically don't care for normal use of such a setting, it's just that the camera's consumption is a little higher, but it's still strange.

[dnewman-gpsw]

It is an interesting finding and it might be a clue, however I also typically have both screens timeout, and I'm not seems this. One caveat, I'm not doing much with HERO11 any more, so it might be relates to some of the enhancements in the later models.

[MomoBrut]

I am just confirming that after many shoots with 2 pieces of the naked version on a quad it works, which is de facto the most demanding use, the settings as I mentioned are fine including the original IMU trigger. It is simply necessary that the front display does not turn off, if it is active all the time, the camera works reliably without reboots.