Just when needed -- complete failure of Nextbase 512G

D

Denbigh Gabbitas

New Member
Joined
Dec 15, 2015
Messages
6
Reaction score
0
Country
France
Having fitted this camera to my camera and recorded interesting events from time to time - I decided to fit one to my wife's new car, new on 10th May. On the 3rd of June we had a full-frontal crash with a white van cutting the corner and on our side of the road. Our air-bags went off and the car filled with smoke. By the time we pushed down the air-bags, the other vehicle had disappeared ! I was so pleased we had this dashcam and when the police arrived, they were as pleased as me.
However, in the accident, the dashcam broke from the suction mount and struck the windscreen - the part holding the GPS chip simply split, thus letting it free - and when I picked it up off the floor, the screen was totally non-working and nothing could be seen. The police accompanied me home to see the recording on my laptop, direct from the SDHC chip. But there was nothing there from about 2 minutes before the accident - a long time in a car. The device records in 3 minute lumps, and if the recording is not properly closed it doesn't exist. The very thing for which it is most definitely required - an accident - was too much for the poor thing! There is a setting which is low medium and high, G-sensor, which is supposed to protect the current recording in the case of G-shock. Clearly non-functioning part of the spec.
When buying this dashcam, I followed reviews which showed the performance pottering along motorways and villages, day and night - I simply assumed that the fundamental requirement would be fulfuilled. I was very wrong.
Interestingly, NextBase are not in the slightest bit interested, as two emails to the support, wondering if they have any sophisticated software to retrieve the lost un-closed recording, have yielded zero response, they are too busy selling, I suppose.
 
A file segment that have not been proberly finalized by the camera can often be repaired, i have been able to do that a fjew times with the repair feature in registrator viewer.

But if you cant even see the broken files than you cant repair them, but if you can see them but just not play them with a player on the computer then it can often be fixed.

Registrator viewer need the broken file + a working file recorded by same camera, and then it should repair the broken file.
 
Just a thought, but the predecessor of the 512G (the 402G) has been showing signs of an issue with defective backup batteries - it's those batteries which are intended to complete the saving of the last file if power to the cam is cut in an accident.
So did your cam not save the last file properly due to defective battery or was it due to being damaged by hitting the windscreen?

Have a look back through the files that are still intact. If there are corrupted files at the end of each journey, that indicates the battery was failing.

Please let us know what you find because if the 512G has bad batteries/bad charging circuits I'm sure the folk here would like to know. :)
 
I can't help but feel dash cams should be writing data in a completely different way. They aren't PCs, and don't need to emulate PC file writing methods, just the file formats.

They should create an empty file in advance (1 minute, 5 minute or whatever), then as it records video it should dump data to the file one sector/cluster at a time.
That way the file itself will always be there, because it was made in advance, and from then on the directory structure for that file does not need to change. Only data sectors need to be written, and that is a continuous process. If done properly there is no need to lose more than one cluster of data in the event of catastrophic failure. I imagine there is some internal buffering going on, so a battery/capacitor may still be needed, it may also increase how much data can be lost. But to lose several minutes of recording just because writing never got finalised is pretty shameful. They aren't even trying, because it isn't the kind of thing that is immediately apparent. But if a manufacturer was to use an improved file writing method it would be a strong selling point.
 
Wouldent a larger memory buffer allso be good, maybe, but it would proberly allso make the camera more expensive.
 
@Rajagra - I like your thinking, but file structures are not quite as simple as that, with parity check and crc blocks, IF, repeat IF, they are to be compatible with computer file systems. Otherwise some special software would be needed to upload to a PC, but it could be done with a bit of gumption.

@kamkar - I have some pretty sophisticated software myself and was able to find the unfinished file, but unfortunately the actual fact of switching the camera on caused the camera to start recording and it recorded over the unfinished file space. I could have recovered the file if I had immediately taken the chip out of the camera {{{{ which I always would now if anything like this happened, just don't give the camera the opportunity to restart }}}}.

The annoying thing is that there is this parameter used to ensure that the file gets written in the event of a shock. The parameters are low medium and high - but the manual doesn't say if low means low sensitivity or low threshold, the two possibilities are opposites.
 
Putain de merde :(

Lets hope you will be more lucky with a new camera, or better yet never have any use for it.
 
Well, to be honest, I was quite satisfied with the 512G up to that accident, with one in our first car, that despite this freak result, I have bought a replacement the same. Delicate differences such as wider or lesser angle, etc. don't really seem to matter. I can read the number plates of cars which I pass in the night, I keep resolution as high as possible since tiny details can matter - I had a small incident on our other car where some road-works were improperly signed and a barrier stuck out into the road with a small inch-wide metal tag or nib - the road was narrow and I passed another car, touching the little nib which scratched my near-side door. I gave the video to the insurance and the contractors paid up for a resprayed door. I am pretty certain I would have been whistling in the wind without the dashcam evidence, and the high res. made the little nib stand out quite clearly.
 
No matter how good camera you have and how expensive a brand memory card you use there is a fjew things you must do.

1. once in a while format the sd card in the camera to restore FAT on the memory card, just like a HDD in a computer when it get defragmented it can hurt performance.
Depending on how much you drive, but i would do it at least 1 time every month.

2. when you have the card out of the dashcam and in the computer, look to see if the files recorded match what you have been driving lately.

I do a random test of files to see if windows player can launch the files, not all of them but 2 -3 files in every driving session on the memory card.
And keep close eye on all last files in driving sessions, you will often see problems with last file not shut down proberly and then beeing corrupt so your player cant read it.
If last files start to be corrupted it might be due to memoory card dying soon, or lipo battery or capasitors not working as they should.
So i check all last files when i inspect my SD cards, dont play all of the 3 minute file offcourse, just look if windows player can launch and read them.

I think this take me around 5 minutes for each of my 64 Gb memory cards, only problem i have 6 cameras in the car, and aside for one 32Gb they are all 64 Gb.
Good thing is i never inspect all my cameras at once.

The new Lukas dashcams is claimed to be "format free" i think this mean the camera have some form of way to keep the FAT of the memory card in good shape.
Maybe using somthing like the defragment tool in windows.
 
Last edited:
kamkar1 said:
.....I do a random test of files to see if windows player can launch the files, not all of them but 2 -3 files in every driving session on the memory card.
And keep close eye on all last files in driving sessions, you will often see problems with last file not shut down proberly and then beeing corrupt so your player cant read it.
If last files start to be corrupted it might be due to memoory card dying soon, or lipo battery or capasitors not working as they should.
So i check all last files when i inspect my SD cards, dont play all of the 3 minute file offcourse, just look if windows player can launch and read them.
Click to expand...

Yes, there's usually enough craziness on the roads that I pull the card out of each cam about once a week.
Like you, I always check the presence and playability of the first and last files of every journey on the memory card, plus a check that no files are missing by seeing whether the time labels on each file are three minutes (or whatever the cam is set for) - the time and date of each file is usually part of what the cam has named it so a skim down the file name list is adequate to see whether the cam stopped unexpectedly at any point.
 
Allso a good idea to keep a eye out for files not the size you want them to be and not beeing the last files.

If there is a 2 minute 11 seconds files in the middle of a session its worth to investigate why, i have seen it a fjew times, and the reason i assume was a glitch in the power supply.
Smaller than usual file still playable, but looking at times in following footage you can see camera have rebooted.
The wierd thing then is i have only seen this happen single times, and if the PSU in not seated right it really should happen again shortly.

Over time i have more or less seen this on all my cameras, but ensuring PSU is proberly seated have remooved this problem every time.
I have not seen this on cameras connected to my hardwired USB power supplies.
 
Here I am again, not sure whether to start a new thread or extend this one...
I considered my experience with the first 512G a freak occurence, which enabled a hit-and-run driver to get away with it, and still went ahead and bought two more. One front and one rear.
Yesterday, for the first time since last summer, I had a look to see what was being recorded. I was shocked to find that gaps in the recording trail were occurring, so that some minutes dotted around were not covered. Last night, I formatted the memory chip in the camera. Today, I find that among those recordings which are on the chip at least half are unreadable by VLC Media Player, Quicktime, Windows Media Player and Nextbase Replay. QT and WMP say that the file is not one readable by QT or WMP... VLC and NR both just behave as if I had not clicked on the file.
This is very worrying, as one must rely totally on the recording to provide evidence in case of an accident. I am not sure how to proceed now. My next step is to try a different SDHD chip, to eliminate one factor in the equation.
Incidentially, kamkar1 I see no reason to worry about slight itermittency on the power supply, since it takes about 3 mins to stop recording after the car is halted and the keys removed from the ignition.

To ALL READERS: the first and last test is not good enough. The random occurrence that I have found means that it could easily be that the first and last are OK. However, if the last is consistently OK, then that means the moment of crash should be caught. I dont think, on the evidence so far, that one can cound on the last always being OK.

Remember, I am now the proud owner of 4, not 1, 2 or 3, but 4 of this model of dashcam, so can be thought of as unusual and worth listening to.
 
Remember that if the last file is OK, that's only reassuring if it truly is (should have been) the last recording. If the final moments were being written to a file but that never got written into the directory structure, there will be no bad file to warn you of problems.
You could move your hand in front of the camera just before shutdown, or say something to mark the moment when the camera shuts down.

Sent from my SM-G903F using Tapatalk
 
Intermittent stop-start is usually a failing, under-spec or fake memory card. Occasionally loose or inadequate power supply.
 
I have had a similar issue this morning using a high quality Kingston microSD card in a Nextbase 212. I stopped at lunch time to (for the first time) take a recording off the card and guess what? About half of my journeys recently don't even show on the card and then I looked at today's to discover that my early morning journey isn't there and the one I needed only starts 10 minutes into my journey - and 5 minutes after I needed the footage. I used Sandisk Rescue Pro to try and recover footage, but there are loads of files that only record for a second, even though the camera is set to 5 minutes. Indeed, it is only intermittent ones where the full recording is there. This is shocking Nextbase and leaves those of us with a dashcam as part of our insurance cover liable to significant risk.
 
Last edited:
Would this suggest why out of 58 on the card only 13 are complete at 360 MB each, 14 are incomplete, but with between 130 and 360 MB and the rest are presumably information files at around 87K ? At no point are there more there more than 2 3 minute segment full videos available.
 
Yes - that would identify a card that isn't writing the information fast enough.

Do you have any other cards there @MikeyI ?

You can use a card between 8Gb and 128Gb and at least with a write speed of 48Mb/s.
 
Last edited:
I am using a Kingston microSDHC speed class 10 UHS-I speed class 1.
 
You must log in or register to reply here.
Back
Top