Dodgy SD Cards

[Gein07]

So last year when I thought I might be getting a Pandora for Christmas I went ahead and purchased some SD Cards.... from E-bay. I was young and stupid back then and I didn't realize just how prolific the knock-off SD market really was. Long story short, I'm about 95% percent sure I got a pair of class 4 16gb knock-offs since the physical housing of the cards are slightly different despite having the same label. Both cards are in fact 16gb cards, so I'm lucky in that respect, but here are a few questions I have for those of you with some experience on the matter:


Is there any utility I can use to truly test and see if these are knock-offs?


What risks do I run in using them as my primaries for my Pandora? (Using them in my DSLR now so far with no issues.)


Should I just pony up and buy a retail card and be done with all the anxiety over card failure?


Gein07

 

[Prometheus]

Looks like the Prometheus-signal...


I doubt they are genuine cards, or genuinely 16GB - if you've come to that conclusion from putting files on them, you've most likely been fooled. The controller chips in such cards are hacked in order to report a false capacity, which your OS will report as being whatever it says it is, and which you will seemingly be allowed to fill, but in reality nothing is getting written past the real capacity of the flash chip, and some controllers will cause corruption in the actual genuine space because of how they think they're handling writes and erases.


To test for sure, you'll need to use H2testw (an application for Microsoft Windows, but which also runs under WINE on x86 versions of Linux and Mac OS X), or the native Linux equivalent F3 (which also has a Pandora port by slaeshjag). This will take some time, and it will take even longer if the cards are fake, so you would be best to leave this testing in the background.


You run a lot of risks by using them as primary cards in your Pandora, as far as I can see. After all, there are a few genuine retailed cards that, due to certain hardware in them, behave oddly and cause random file-corruption and long boot-up times even if you are running your Pandora from the internal NAND space (I just ditched a Verbatim card that did these things, and they're quite a well-known brand). Fake-capacity cards are usually made from parts surreptitiously rescued from the reject bin, so I would imagine that these will be much more given to causing such problems.


In short, I would suggest picking up a genuine card (SanDisk come highly recommended, as unlike many other manufacturers they make all of their own parts in-house, and their cards tend to be very reliable as a result), and save yourself the hassle of running into strange problems caused by problematic cards.

 

[Gein07]

Looks like the Prometheus-signal...


I doubt they are genuine cards, or genuinely 16GB - if you've come to that conclusion from putting files on them, you've most likely been fooled. The controller chips in such cards are hacked in order to report a false capacity, which your OS will report as being whatever it says it is, and which you will seemingly be allowed to fill, but in reality nothing is getting written past the real capacity of the flash chip, and some controllers will cause corruption in the actual genuine space because of how they think they're handling writes and erases.


To test for sure, you'll need to use H2testw (an application for Microsoft Windows, but which also runs under WINE on x86 versions of Linux and Mac OS X), or the native Linux equivalent F3 (which also has a Pandora port by slaeshjag). This will take some time, and it will take even longer if the cards are fake, so you would be best to leave this testing in the background.


You run a lot of risks by using them as primary cards in your Pandora, as far as I can see. After all, there are a few genuine retailed cards that, due to certain hardware in them, behave oddly and cause random file-corruption and long boot-up times even if you are running your Pandora from the internal NAND space (I just ditched a Verbatim card that did these things, and they're quite a well-known brand). Fake-capacity cards are usually made from parts surreptitiously rescued from the reject bin, so I would imagine that these will be much more given to causing such problems.


In short, I would suggest picking up a genuine card (SanDisk come highly recommended, as unlike many other manufacturers they make all of their own parts in-house, and their cards tend to be very reliable as a result), and save yourself the hassle of running into strange problems caused by problematic cards.

Oh wow, that's worse than I originally thought then. Most pictures I'm taking are of my infant son so I could actually be losing pics and not know it. I'll discontinue use of them immediately until I can fully test them. Thanks for posting those resources as well.


Is there a utility I can download that will tell me where these fake card distributors live so I know whose house I should be burning to the ground? =D

 

[Prometheus]

Oh wow, that's worse than I originally thought then. Most pictures I'm taking are of my infant son so I could actually be losing pics and not know it. I'll discontinue use of them immediately until I can fully test them. Thanks for posting those resources as well.

Ack, that is the worst use case for such cards. Try to rescue all you can from them before testing, though sadly you will most likely find that you won't be able to get them all.

Is there a utility I can download that will tell me where these fake card distributors live so I know whose house I should be burning to the ground? =D

No, no there is not. The best thing you can do is keep hold of those links, and raise awareness among others if you can - at least that way more people will know what to avoid.

 

[Granitehead]

I also suggest SanDisk if you want to be sure, but I just got a 64GB Transcend one which works nice so far too (partially tested it with F3, but I'm quite sure it's no fake as I bought it from a trusted seller, if it were other problems they had probably showed up in the first few GB. Going to test the full 64GB over night though, I want to be sure before using it).

 

[Gein07]

I also suggest SanDisk if you want to be sure, but I just got a 64GB Transcend one which works nice so far too (partially tested it with F3, but I'm quite sure it's no fake as I bought it from a trusted seller, if it were other problems they had probably showed up in the first few GB. Going to test the full 64GB over night though, I want to be sure before using it).

Yeah I think I will be purchasing a SanDisk card or two tonight. Fortunately I spoke to my wife and I had forgotten I made her switch out to a retail purchased 32gb card for our DSLR camera. No lost memories due to this issue, just hurt pride if the cards do turn out to be phoney.


Anyway, since I bought these suspect cards through e-bay to save some dough, anyone got any good leads on inexpensive sd cards from a reliable source?

 

[Prometheus]

Anyway, since I bought these suspect cards through e-bay to save some dough, anyone got any good leads on inexpensive sd cards from a reliable source?

Honestly, I hate to say it, but this one of those things where you get what you pay for, due to the slim profit-margins on genuine-capacity flash media unless moved in great volume.


The best I can suggest is keeping an eye on Amazon, Newegg (if you're in the US), or Play.com (if you're in the UK), for good deals offered by these retailers. (I would suggest avoiding third-parties, just in case any fakes pop up with them.)


I'm glad to hear that the card in your camera had already been switched for a retail-purchased one, by the way.

 

[Gein07]

So, I tested both suspect SD cards when I got home today and to my surprise they seem to be legit cards after all. Both 16gb cards stored 14.8gb worth of data via the tests. The read and write speeds are a little off between the cards, but way better than what class 4 should be rated at according to google. I tested with h2testw 1.4 as Pometheus recommended. Here are the test results for both cards:


Test finished without errors.


You can now delete the test files *.h2w or verify them again.


Writing speed: 11.4 MByte/s


Reading speed: 18.1 MByte/s


H2testw v1.4


Test finished without errors.


You can now delete the test files *.h2w or verify them again.


Writing speed: 14.4 MByte/s


Reading speed: 18.6 MByte/s


H2testw v1.4


Everything I read today said class 4 should be writing around 4mb/second, so I'm wondering why this benchmark showed results as high as 14.4. Should I not be so concerned about these cards now if they passed the test? Thanks.

 

[Prometheus]

Both 16gb cards stored 14.8gb worth of data via the tests.

14.8GB suggests a lower grade of flash chip than you might want to trust with certain things (i.e., I still wouldn't trust them with photographs), but wow, you were lucky. :blink:

Everything I read today said class 4 should be writing around 4mb/second, so I'm wondering why this benchmark showed results as high as 14.4.

The reason for this is that the write speeds indicated by the class are the minimum that they are guaranteed to maintain (it's meant as an indication for usage scenarios such as digital camcorders, where even a moment's drop in the minimum writing speed can be catastrophic). Many cards will tend to exceed this.

Should I not be so concerned about these cards now if they passed the test? Thanks.

Given the apparent chip-quality, as I noted above, I still wouldn't trust them completely, and certainly not with lone copies of data, but they should be fine for things for which you have backups.

 

[Gein07]

Both 16gb cards stored 14.8gb worth of data via the tests.

14.8GB suggests a lower grade of flash chip than you might want to trust with certain things (i.e., I still wouldn't trust them with photographs), but wow, you were lucky. :blink:

Everything I read today said class 4 should be writing around 4mb/second, so I'm wondering why this benchmark showed results as high as 14.4.

The reason for this is that the write speeds indicated by the class are the minimum that they are guaranteed to maintain (it's meant as an indication for usage scenarios such as digital camcorders, where even a moment's drop in the minimum writing speed can be catastrophic). Many cards will tend to exceed this.

Should I not be so concerned about these cards now if they passed the test? Thanks.

Given the apparent chip-quality, as I noted above, I still wouldn't trust them completely, and certainly not with lone copies of data, but they should be fine for things for which you have backups.

Awesome, thanks for all your help Prometheus. I feel like I dodged a bullet, but I know now just how shady the SD market can be.

 

[Prometheus]

Glad to be of assistance - and glad that you not only dodged a bullet, but also got an insight into that whole fake SD Cards mess before something dire happened.

 

[slaeshjag]

Sometimes it's silly how far they go (in case you missed it, just look at this)

 

[Prometheus]

^ Wow. :blink: That's a new one, even to me.

 

[DaMummy]

you gotta admit, those nuts make for some good grounding of the drive though, plus it looks like a nice cozy case for that 128MB flash drive, which btw, is frickin impossible to find these days, could be worth some money, i say no hdmi kills the deal though