My M9 is eating SD cards [MERGED}

[scrubs]

Advertisement (gone after registration)

Could a mod please do me a favour: When I got my reply from Leica i asked if they

would be ok with me posting it here in this thread and before waiting for a reply - made

the assumption this would be fine. Just got a reply today and they are fine with it but,

requested I remove the employers name from the email..

 

So, could a mod please remove the name at the bottom of post #114 and quoted on post #115.

Both on page 6. It will no longer let me edit.

 

 

Thanks!

[Advertisement]

The Leica Q3 is available for pre-order at B&H Photo Video and Adorama

[M9reno]

My question, I guess, was to what extent one could further isolate the problem by avoiding any manipulation of the card. Going USB would certainly discount possible corruption of the card through ejection from M9 or computer, or even the slightest physical damage to which the M9 might be particularly sensitive. It might also make the card less prone (though I imagine not altogether immune) to corruption by a computer writing or deleting information directly to or from it.

 

The only practical "workaround" that I can see at the moment is to place no trust on higher capacity cards, and go for the 2 and 4 GB versions. It seems to me almost certainly to be a problem with the camera's operating system.

[andybarton]

Could a mod please do me a favour:

 

I have done that, but you are lucky that I saw your request. Incredibly, the Mods don't read every post...

 

If you want to attract the attention of a Mod regarding a post, please use the Warning triangle button on the left hand side of the post. That way, an email will be sent to all Mods and you are much more likely to get something done.

[scrubs]

Great stuff. Thanks for that Andy and Cheers for the heads up!

[scrubs]

Oh... Andy I just noticed the whole post (letter from Leica) was deleted - I just wanted the

employee name deleted from the bottom of letter (name beginning with 'M').

Maybe I wasent clear enough. Is it too late to re-paste the letter - just leaving out the employee name?

[jaapv]

I'm afraid it is not possible to restore the post. If you just repost the mail without the names.

[andybarton]

Advertisement (gone after registration)

Oh... Andy I just noticed the whole post (letter from Leica) was deleted - I just wanted the

employee name deleted from the bottom of letter (name beginning with 'M').

Maybe I wasent clear enough. Is it too late to re-paste the letter - just leaving out the employee name?

 

My apologies - a slip of the mouse. I will see if it's still in my cache somewhere.

 

Edit: No, sorry.

[scrubs]

I will PM it

[scrubs]

I'm afraid it is not possible to restore the post. If you just repost the mail without the names.

 

Thanks Jaap,

 

 

Just PM'd it to Andy and explained that it was too important of a message to lose

and would probably only serve to confuse the thread further by reposting out of

context several pages later, so probably better being placed back in original post.

[andybarton]

Sorted. Thanks for your patience.

[Venkman]

YES!! I've managed to provoke the behaviour. The following way:

Fill the buffer on C. When the camera starts writing hit "play"and then switch off the camera. All files written after pushing play will be empty.The next shots will be normal again.I have Autoreview switched off.

Now I only need to figure out a real-world shooting situation where I would do this...

 

Jaap,

let's theoretically say that some Leicas are faster writing than others - I can remember that some people had such a feeling, now let's just take that as granted. Then let's say one shoots jpg and dng compressed, which might take a bit longer to process. Under ose circumstances I wouldn't be too surprised that I get that error after just 2 frames in the buffer, would you?

 

You reproduced a strange behaviour, which probably has not much impact on every day shooting, but the question here is: are there probably more scenarios where this could happen to a small number of cameras and if so, why should those affected tolerate it?

 

I for my part will run a couple of tests once I am back home and will contact Leica with the results, because it affected me a couple of times and in one case so much, that I was really annoyed with this otherwise great camera.

 

Dirk

[jaapv]

Strangely enough the camera wrote 4.5 Mb jpgs for four of the seven files. It was set to DNG uncompressed and JPG fine.

[UliWer]

YES!! I've managed to provoke the behaviour. The following way:

Fill the buffer on C. When the camera starts writing hit "play"and then switch off the camera. All files written after pushing play will be empty.The next shots will be normal again.I have Autoreview switched off.

Now I only need to figure out a real-world shooting situation where I would do this...

 

My M9 has a bug!

 

DNG+JPG fine, C mode, 7 exposures, pushing the play button, switching the camera off - the red light goes on signalling for ages when the camera is off....

 

Switching it on again - seems to be dead, then "wait until...", 7 exposures are there (DNG 18.3 MB; JPG 3.4MB).

 

Tried two times: same result.

 

Do I have to send it Solms to ensure that I get empty files?

[Venkman]

Funny. Wanna switch cameras?

[k-hawinkler]

Funny. Wanna switch cameras?

 

He is not trying hard enough.

 

K-H.

[Robert_M]

Since forum readers want to continue to bring fake SD cards into the discussion, let's try to define what a "Fake" memory card is and let the reader decide if it is relevant to their observed issue or problem.

 

First, I hope we can agree that the motivation for someone putting out a fake memory cards is financial. The faster the money the better. Remember, the expenses involved in running the scam are an important consideration.

 

Second, I'd like to dispel the notion that a fake cards are junk that is put together in some backyard, or shack in the third world. Yes, one can make milling machines in garage type settings. I've even seen complex welding fabrication "shops" on the sidewalks of crowded Asian cities. One can do small, precision optics fabrication in a minimal garage setting. I've even seen high end carbon-fiber bicycle frame manufacturing (for a major name brand) being done in shack-like conditions in a developing country. BUT, one can NOT make memory chips in similar circumstances. Semiconductor, in particular memory chip, fabrication takes millions of dollars of capital equipment and involves sophisticated technology (ie, processing under ultra-high vacuum , cryopumps, 1micron scale photo lithography, class 100 clean rooms, dicing machines, etc). So drop the image of some phoney factory putting out product.

 

So what are fake memory cards? I'd divide them into several different categories.

 

1) First, what I'll call the Stolen IP (Intellectual Property) version. The production of these memory chips (and cards, thumb drives, etc) are subcontracted to factories in China. I don't mean to cast nasty dispersions, but it is an unfortunate fact that contracts and IP are not treated with great respect in Chinese culture. I've seen this firsthand since I've had extensive dealing with organizations and business in Taiwan and China. You wouldn't believe how many times I've heard the phrase "a contract is just a piece of paper". Anyway, mainland China can be particular bad since there are state entities involved in some of the factory ownerships. So, there is a category of goods where the factory produces product A for a same firm (say memory chips for a big, international brand) and there is an extra production which slips out the back door and onto another distribution channel without the contracting firm knowing. Well, the big international firm surely knows but can't stop it. This is the source of imitation goods which is of greatest concern to major firms. There is noting wrong with the cards of this "fake" version. But, the whole scheme is stealing (by another name). I suspect that this is the biggest source of "fake" cards but I can't back that up with any facts.

 

2) The Scam cards is another version. These are cards which are inexpensive, small capacity memory cards which are then re-programed in the card firmware to appear like a higher capacity card. So, one takes a lower priced 2 or 4 GB card, hacks the firmware (which a cheap machine), and relabels it as a highend high capacity card. And, makes a nice profit in the distribution. The card will work for the initial 2 (or 4) GB of use but then the card tries to write to non-existent memory and in some cases rewrites over what is already used. A real mess. These are the cards which the ebay scams links talk about. There are a few youtube videos out there if you want more info. Do a google search if you want more info. It is possible to test for these card and I'll supply more on that in another posting.

 

3) Another category I'll call the Revived Reject version. On the production line, some memory chips don't pass the tests and are rejected. Someone goes and saves these from the trash bin and makes cards from them. They work, but not to full potential in memory space or other spec. This category seems to have been a problem 5 or more years ago and it appears to be less of a problem now. (Again, I'm going by anecdotal reports and can not verify the magnitude by statistics). I can see that big firms can come in and cleanup quality control to minimize this problem. And, the factory management would be more attracted to making money off version 1 above than letting some renegade scam off their trash bin.

 

 

So, one will have to decide for oneself if the problem they see is due to the card or the device. If one suspects a problem is the memory card, it is possible to run some checks on the card to verify or dispel that thesis. I'll discuss in another post. This is already too long.

 

RM

[scrubs]

Thanks you said what I wanted to regarding OEM products but could not be bothered.

small example: I recently bought a stylus gauge for a turntable - a reputable company

sells a 'branded' version of this product for over £250 I went direct to the factory

in China bought the 'Exact' same product - same factory, same machinery, 'un-branded'

for under £20. This can be applied to pretty much most products on the market, and not

just in the technology sector.

[Robert_M]

I will note that there are now 3 different posts in this thread that have had trouble with the SanDisk Ultra 16GB class 4 card (apparently new cards). All the problems seen are early on in the card use (within first dozen frames). I was one who also had problems with such a card. It is worth investigating further since this is specifically a card recommended by Leica in their FAQ. To be precise, Leica actually recommends the Sandisk Ultra II 16 GB class 4", however this seems to be replaced by cards marked "Ultra" rather than Ultra II". Whether there is a technical change or not, I don't know. Could just be marketing; I don't know.

 

Anyway, I've now thoroughly tested such a card with specific test software and can find no flaws in the card nor can I find any evidence of a fake. BTW, my card was bought from Newegg.com which is one of the largest retailers of computer related products in the USA and is extremely cautious as an internet sales retailer. In my experience, B&H photo and Newegg are the two most careful, secure internet sales sources I've encountered.

 

To test Memory Cards and USB drives for internal problems, I used the program

H2testw 1.4

This program is free and available for download at dozens of web sites. I used the site H2testw Download - Softpedia The program has instructions in German and English. If you do a google search you can find information and videos with example results of bad cards. This program writes to every memory location and then verifies the location by reading it. The output contains speeds of read and writes, bad blocks, and information on false labeling (if any) for the card's specifications.

 

I tested one of the two Sandisk Ultra cards which I own (both purchased at the same time and same shipment). Be prepared that this testing takes a long time. At 6 MB/s write speed and 16GB data that is ~45min to do that part of the test procedure alone. You do the math for your own card.... Here are the results for my card's test:

 

Test finished without errors.

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

Writing speed: 6.05 MByte/s

Reading speed: 16.7 MByte/s

H2testw v1.4

 

I note that the tests exceed the quoted specifications for the card by SanDisk (one of the reasons techies view them as a reliable source). Not a very interesting output when the card is fine....

 

The H2testw v1.4 program is written for MSWindows but it is very simple and does not require an installation into the OS. There are routines in Linux (Ubuntu) which also do some of this investigation. However, I haven't found one with the directed purpose that H2testw has. I recommend trying this program on suspected flawed cards and suspected flawed USB drives. Certainly worth understanding if a card has errors before tossing it in the trash bin.

 

Hope this information helps someone.

 

RM

[Advertisement]

Hallo,

Look here: Leica M

[ho_co]

I... Whether there is a technical change or not, I don't know....

That's an important point, Robert. The card designations may change, or the card itself may change.

 

There were reports on this forum of two identically-named cards formatting to different available space--something like 40 bytes different, IIRC, IOW not falling into one of your fakes categories. (Those may have been SD cards, though, not SDHC.)

 

Which is just to say that not all legitimate cards with the same specification from the same manufacturer will be the same.

 

Also, isn't a "Class 4" or any "class X" a minimum specification? In other words, a Class 4 is defined as a card that meets a minimum transfer rate. Thus, having a card that "beats" the minimum specified is merely fitting the defined speed category.

[k-hawinkler]

Hi Howard,

 

I have only bought memory cards from Leica's recommended list from reputable sources, like B&H.

I have not bought newer or older versions. Just from their list.

Wouldn't you think it's up to Leica to make sure they work on the M9?

 

The way I look at the M9 is as a cutting bleeding edge piece of technology where problems are to be expected.

As they have worked on their other issues, I am expecting they will find their more important firmware bugs soon, and issue a new firmware update.

I don't think their problems are just with the memory cards.

From my observations, they also have to address how the various camera controls interact with the firmware and the storing of images.

After all their M9 has to be an interrupt driven system.

 

Best, K-H.