Just when needed -- complete failure of Nextbase 512G

The speed rating I mentioned is what sdcard.org says for Class 10 cards, not what Kingston, Transcend, Samsung or any other manufacturer says. If you're saying that's how sdcard.org achieve those numbers, then no card is compatible with dashcams, right?

Just out of curiosity, yesterday I tested my Kingston card with h2testw against the Transcend cards supplied with the SG9665XS. The Kingston's write speed was 10~11MB/s, while the Transcends all tested between 13 and 14MB/s. I don't know where you got the idea that Kingston cards are faster. If this test counts for something, the Kingston card would be an ultra marathon runner, not a sprinter. In any case, what matters to this discussion is that a Class 10 card's write speed is high enough for the demands of any dashcam. I can't even imagine what the camera's bitrate would have to be for its write speed to exceed even a Class 10 Kingston's "low" write speed.

Correct: no memory card has been designed with dashcams in mind. They were designed for phone and computer-type use which isn't continuous writing. That's why some card companies explicitly mention that their warranty is not valid if the card is used in a dashcam.

In some tests I saw in a magazine a couple of years ago, Kingston were highly rated for speed as well as being a bargain price. I haven't looked at their recent products.

The bottom line is that some memory cards don't cope with the heavy-duty requirements of a dashcam and from what I've seen Kingston are among the most likely to struggle or fail prematurely with dashcam use whereas Transcend and Samsung are among the least likely to struggle.
 
Everything is part of a system. If a card can keep pace with the cam's data writing under all circumstances then you're OK. Simple. Given that video recording is probably the toughest job imposed on these cards, using a best quality card rated well above the expected tasks is simply a smart idea- that is the very first thing I learned about dashcam quirks and it has been my mantra ever since :cool: My fave is the Transcend 64GB class 10 400x. Of my several, one has positively seen well beyond 600X (around 3 years 24/7 use), others have seen 3-400x or more. And all just tested 100% on h2testw :D

Phil
 
The bottom line is that some memory cards don't cope with the heavy-duty requirements of a dashcam and from what I've seen Kingston are among the most likely to struggle or fail prematurely with dashcam use whereas Transcend and Samsung are among the least likely to struggle.
I can only speak for the one I have but it's been working non-stop for 2 years and 7 months now in several dashcams and tested 100% (again) on h2testw.
 
If a card can keep pace with the cam's data writing under all circumstances then you're OK. Simple.
Exactly the point I've been trying to get through. What's the point of buying a card with 20+MB/s write speed when the camera hardly writes at 3MB/s?
 
What's the point of buying a card with 20+MB/s write speed when the camera hardly writes at 3MB/s?

Most likely it can only sustain 20+MB/s for short periods, after which its speed drops.
And because a fragmented card will run much slower.

The figure printed for a card's performance is virtually meaningless for dashcam. Faster cards have a better chance of being adequate for a dashcam than a slower card but there will be so-called fast cards that can't keep up and there may well be occasional so-called slow cards that sometimes manage to keep up.
 
Last edited:
Because it can only sustain 20+MB/s for short periods, after which its speed drops.
And because a fragmented card will run much slower.

The figure printed for a card's performance is virtually meaningless for dashcam. Faster cards have a better chance of being adequate for a dashcam than a slower card but there will be so-called fast cards that can't keep up and there may well be occasional so-called slow cards that also manage to keep up.
Again, I don't know if the h2testw test counts for anything, but at the beginning of the test the writing speed is always higher and then stabilizes at 10~11MB/s until the end of test, which usually lasts between 45 and 50 minutes. Can this be considered a "short period"?
My dashcam writes at an average of 1.6MB/s. I don't think the card's write speed will drop that far, otherwise it would've started giving me problems a long time ago.
 
When I bought my SD card I, blindly at the time, went for one advertised as being for video use. It's a 64Gb Sandisk Ultra Plus, formatted FAT, and it's never let me down in a 402G. The longest non stop journeys it's done is just over 3 hours, cam set to 3 minute chunks and they've always been there.
 
Might be a silly question but you know the nextbase stores protected videos in the DCIM/PROTECTED folder? A video from a crash should have been protected.
 
Might be a silly question but you know the nextbase stores protected videos in the DCIM/PROTECTED folder? A video from a crash should have been protected.

Not if the internal battery wasn't holding enough charge to complete a shutdown after an interruption to power (or if G-sensor had been turned off due to the UK's potholed roads that trigger G-sensors a couple of times per hour or more).
Failure of internal batteries is a major problem with dashcams - and Nextbase cams seem especially prone.
 
So a user replaceable button battery is a reasonable request, then we could replace them annually, just to be sure.
 
So a user replaceable button battery is a reasonable request, then we could replace them annually, just to be sure.

The battery on my NB402G lasted less than three months.
 
Sorry to drag up an older thread as my first post. I am looking at getting a dashcam and like the look of the Nextbase cameras (specifically the 512GW), which is how I ended u here, doing some research.

Correct: no memory card has been designed with dashcams in mind. They were designed for phone and computer-type use which isn't continuous writing. That's why some card companies explicitly mention that their warranty is not valid if the card is used in a dashcam.

What about the likes of the SanDisk SDSDQQ-064G-G46A? Just wondered if anyone else was using one of these with their dashcam.

Also regarding the formatting of the card, surely it would be a simple firmware addition to add a reminder if the card hasn't been formatted for a set period.

Regards,

Les
 
There are a few SD cards that are advertised as being for video use.

These cards are typically Class 10 U3 cards which allow read/write up to 98Mbps.

I got a Samsung 128Gb one from Scan (search my posts as I did provide a link to it). I chose Scan because they are listed on the Samsung website as retailers. There are a lot of fakes around.
 
Mine - and it's never caused any issues:

0DxEy2e.jpg
 
Faster cards have a better chance of being adequate for a dashcam than a slower card but there will be so-called fast cards that can't keep up and there may well be occasional so-called slow cards that sometimes manage to keep up.
So, why does a dashcam-dedicated card, like the SanDisk @Pointy indicated, only reads and writes up to 20MB/s and not the 45 or 98MB/s people claim to be better for dashcams?
 
So, why does a dashcam-dedicated card, like the SanDisk @Pointy indicated, only reads and writes up to 20MB/s and not the 45 or 98MB/s people claim to be better for dashcams?

Manufacturer speed tests are performed by writing a file (or a short burst of files) onto an unfragmented card or reading a file from an unfragmented card.

Dashcams constantly write/delete/overwrite files and the card could soon become fragmented.

I think what happens is that a card becomes badly fragmented and the camera wants to write the next file so it (and the card) start searching for somewhere to put the next file but because the card is so fragmented the card hesitates for a sufficient amount of time that the card isn't ready to write the next file as the camera starts sending the data. So the backlog causes the card and camera to stumble and stop recording. A few minutes later the camera may well start recording again only to stumble again later.

Fragmented performance often bears no resemblance to unfragmented performance. Cards that appear to be slow in the official tests can actually turn out to be very good in dashcams while cards that appear fast can actually slow down severely once fragmented.
A better test for dashcam use would be the read/write speed of a badly fragmented card.
 
Back
Top