Got into accident...file corrupt?!

[Gibson99]

When I was the front car in a 3 car crash like yours, I heard the crash behind me and looked in the mirror right as the middle car got pushed into me. That happened way before I got into dashcams so reviewing footage wasn't even an option for me.

If the front car's driver will swear (under oath if necessary) that he heard the first crash before feeling the second one, then that is enough evidence to clear you of any fault.

 

[SawMaster]

State Laws vary, anyone here know how New York's work?

Generally in at-fault states your insurance initially pays for the car in front which you hit, the car behind you pays for you, all the way down the line. The insurance companies work out the details later and may sue the last car in line for the whole accident since they were the proximate cause of it all, and some liability may be assessed againstcthe car in front if they didn't have a legitimate need to stop or slow like they did. That won't involve you. And generally, no fault is given to the car(s) in the middle unless there is clear and strong evidence that they hit the car in front of them before they were themselves hit which isn't normally the case with chain-reaction crashes like this. Keep trying to salvage the video but don't worry too much just now- you're probably going to be in the clear.

If there are any charges made against you get a lawyer immediately and fight them if they recommend that. Don't count on your insurance company to offer decent legal assistance; they generally don't. All they want is to minimize their own losses any way they can even if that means throwing their loyal customers under the bus.

Phil

 

[SergioV]

Thank you for the advice regarding a 3 car accident. I will definitely keep this thread updated as to the outcome of this ordeal. I really appreciate everyone's help through this.

 

[Arcas]

Could some other A119 owners try some of the tests mentioned earlier in this thread? The remove-power-at-various-stages and remove-the-SD-card-at-various stages tests would provide useful data. Another test would be to toggle power very briefly. Also, it would be instructive to run some impact tests. I'm not suggesting that you throw your camera against a wall to see if it stops recording but maybe attach it to a table and drop a heavy book next to it to simulate a serious vehicle impact. Might need to repeat the test at different mounting angles.

Perhaps @SergioV simply got unlucky but it's a little worrisome that the one video file that he needed the most turned out to be unusable.

 

[Sunny]

Could some other A119 owners try some of the tests mentioned earlier in this thread? The remove-power-at-various-stages and remove-the-SD-card-at-various stages tests would provide useful data. Another test would be to toggle power very briefly. Also, it would be instructive to run some impact tests. I'm not suggesting that you throw your camera against a wall to see if it stops recording but maybe attach it to a table and drop a heavy book next to it to simulate a serious vehicle impact. Might need to repeat the test at different mounting angles.

Perhaps @SergioV simply got unlucky but it's a little worrisome that the one video file that he needed the most turned out to be unusable.

Taking SD card out while recording will definitely corrupt video regardless of what dashcam.

 

[Arcas]

Taking SD card out while recording will definitely corrupt video regardless of what dashcam.

Right. So is it possible to guess what happened based on the type of data corruption that @flip9 found? Given that this happened during a collision sufficiently violent that the vehicle was declared a total loss, it seems three possible culprits would be:
1) the SD interface's pins momentarily losing contact with the card at the moment of collision
2) some sort of power glitch (if power drops momentarily, what happens? what if it drops momentarily multiple times in rapid succession?)
3) the capacitors not having enough capacity to complete the shutdown sequence (though if this were the case, it seems @SergioV would have seen video corruption anytime he turns the vehicle 'off')

Does the nature of the corruption shed any light on what may have happened?

 

[2000rpm]

Right. So is it possible to guess what happened based on the type of data corruption that @flip9 found? Given that this happened during a collision sufficiently violent that the vehicle was declared a total loss, it seems three possible culprits would be:
1) the SD interface's pins momentarily losing contact with the card at the moment of collision
2) some sort of power glitch (if power drops momentarily, what happens? what if it drops momentarily multiple times in rapid succession?)
3) the capacitors not having enough capacity to complete the shutdown sequence (though if this were the case, it seems @SergioV would have seen video corruption anytime he turns the vehicle 'off')

Does the nature of the corruption shed any light on what may have happened?

4. The capacitors not having enough time to charge sufficiently, especially after a few days of not using the vehicle, possibly made worse by recording longer clip lengths, possibly made worse by extra power demands of a GPS module.

 

[flip9]

I've repaired the headers and managed to get the full 55 seconds out of the 146mb file. This is shortly before the impact and i cannot recover anymore after that since theres no data to work with. The following frames might have been stored on a buffer before writing to the card.

Last frame i could grab:

 

[SergioV]

That frame is just a second before impact. Shocking that the dash cam and SD card could have lost contact during the impact. Seems like that's the only reasonable excuse for such corruption. Definitely has to be addressed by @viofo for future iterations of this dash cam.

 

[2000rpm]

That frame is just a second before impact. Shocking that the dash cam and SD card could have lost contact during the impact. Seems like that's the only reasonable excuse for such corruption. Definitely has to be addressed by @viofo for future iterations of this dash cam.

I think there was an instance talked about on here a while back where, during a crash, a dashcam detatched from the windscreen, struck the dashboard, memory card ejected itself and last file was lost or damaged.

I would just add memory card loss of contact to the already substantial list of things that can cause a dashcam to fail to record an important moment. I recommend that people use two different makes of cam to greatly reduce the chance of not capturing an important moment.

 

[jokiin]

He actually recovered the files though from memory

 

[2000rpm]

Thinking a bit more about memory cards becoming detatched, I would expect, during a sudden deceleration, that the card would nudge towards the part of the vehicle which suffered the impact.

I notice a lot of dashcams have the micro sd with its contacts facing the user, so most cams might allow their card to briefly partially lose contact.

My A118C in my bits box (spare cam) has the contacts facing rearwards so I expect the A119 would, too, as do most of the cams in our cars - Transcend, Mio, Blacksys, Cobra, Panorama.

Our SGZC12RC, JooVuu and Mobius have horizontal card slots which might have more resistance to the card suffering a brief loss of contact in a heavy impact.

I also wonder whether a solidly-anchored mount (adhesive) would risk the card moving more than a suction cup which would buffer the force of the impact. However, the suction cup is probably more likely to detatch.

Also the way the cam attached to the mount could affect how easily the cam itself might come off. The small, light SG-RC would be unlikely to detatch (and the electronics with the card are tucked elsewhere anyway). The Mobius would rely on two plastic prongs in the front of its case and some friction of the cradle. The JooVuu has the screw going into the camera body so is probably very secure.

 

[flip9]

That frame is just a second before impact. Shocking that the dash cam and SD card could have lost contact during the impact

We should have at least seen a few frames of the impact on the video, which makes me think some frames might be stored on a temporary buffer before writing to the card (but in your case the card may have been fubared before it could write it). Perhaps viofo can shed some light.

It is concerning nonetheless.

 

[Sunny]

I've repaired the headers and managed to get the full 55 seconds out of the 146mb file. This is shortly before the impact and i cannot recover anymore after that since theres no data to work with. The following frames might have been stored on a buffer before writing to the card.

Last frame i could grab:

Good job.
I could only repair 28 seconds using various tools.
Your procedure can help recover corrupt files better.

 

[SergioV]

@flip9 @Sunny I appreciate all of the help you guys have offered me. You have gone above and beyond in helping me recover this footage. As well as everyone else's input in this thread. It's a serious issue that needs to be worked out and hopefully something like this doesn't happen to someone else.

 

[knesz]

My A119 suddenly corrupted every last file. I noticed that when I take of the power, camera shuts of immediately. Also settings wasn't saved. I opened it and found one of the capacitor to be blown. It could be the same reason for the OP? I'm now waiting new cam, warranty replacement from @viofo and using old trusted A118

Sent from my SM-G930F using Tapatalk

 

[2000rpm]

This is news to me. I've just now tested it for myself. Connected Mobius (capacitor-converted) to USB power bank, recorded for 30 seconds, then disconnected power. The 30 second file saved perfectly.

Mobius C2 (capacitor) last used in the afternoon of the 10th (exactly 5 days ago - manual says 7-10 days to retain date/time).
Switched on, took 5s to start recording, switched off after another 5s; no file present on card.
Repeated tested with a second identical C2c but allowed the 5s startup and 10s recording before switching off; still no file present on card.
Alternated cams and increased recording time by 5s but even after two minutes there's still no file being saved by either cam after power is disconnected - and that's with two cams 'climbing the ladder' of time intervals; if it was one cam I could believe it was a fault.
The cams are set for 'power-off-disconnect' of 10s because otherwise they don't cope with cars which have electrical power fluctuations when cranking (they switch off every time stop-start is activated or at certain other times).

 

[Sunny]

Mobius C2 (capacitor) last used in the afternoon of the 10th (exactly 5 days ago - manual says 7-10 days to retain date/time).
Switched on, took 5s to start recording, switched off after another 5s; no file present on card.
Repeated tested with a second identical C2c but allowed the 5s startup and 10s recording before switching off; still no file present on card.
Alternated cams and increased recording time by 5s but even after two minutes there's still no file being saved by either cam after power is disconnected - and that's with two cams 'climbing the ladder' of time intervals; if it was one cam I could believe it was a fault.
The cams are set for 'power-off-disconnect' of 10s because otherwise they don't cope with cars which have electrical power fluctuations when cranking (they switch off every time stop-start is activated or at certain other times).

That's very bad then.
What if accident happened right after you drive?
This makes capacitor based dashcams useless.

 

[2000rpm]

That's very bad then.
What if accident happened right after you drive?
This makes capacitor based dashcams useless.

I think it's the 10s power-off-disconnect that probably requires a fully charged cap to complete. But if I don't use a power-off-disconnect it rarely records anything in my car and sometimes fails in my wife's car.
On the other hand, CF-100 has a near-30s boot-up time which is also a weakness!
JooVuu doesn't take anywhere near that long - I haven't noticed any problems with saving files after very short periods of recording.
Not sure about the likes of Street Guardian or Panorama but I have a SG-RC and a Panorama and haven't noticed a major problem with short record times.

But just because I haven't noticed doesn't mean there isn't a brief window of time when a cap-cam might not save the file so I intend to look a lot more closely at this phenomenon in the next few days if time permits.

 

[russ331]

The cams are set for 'power-off-disconnect' of 10s

Mobius cams using the capacitor kit MUST be set to IMMEDIATE Power-Off Disconnect, not 10 seconds. Hence the issue you describe.

A timer delay device can resolve the problem created by your car's electrical system.