My M9 is eating SD cards [MERGED}

[wlaidlaw]

Advertisement (gone after registration)

I did contact Andreas about this issue some time ago and asked if he could seek some guidance from Leica about what info from us could help Leica but I have heard nothing back. I sometimes wonder if there is an element of NIH (Not Invented Here) about Leica.

 

I sent a previous detailed proposal to Stefan Daniel about setting up a worldwide network of approved rangefinder service stations, with Leica trained but dealer employed technicians; at least one per country that has significant M sales. The analogy I quoted was that I would not expect to take my Porsche back to Stuttgart to have the front wheel alignment checked, so why should my Leica have to back to Solms for a minor rangefinder adjustment. He thanked me for the proposal and passed it on to a couple of senior Leica people for them to assess and let me know what they thought about the feasibility of the proposal. That was a year ago and I have heard nothing since.

 

Wilson

[Advertisement]

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

[john_newell]

Isn't the issue noted in your post below a significant limitation on the testing? I'm not trying to pick an argument, just wondering how much probative value the admirably detailed procedure in your earlier post will have when all is said and done?

 

Hi there,

 

BTW, it's K-H for Karl-Heinz.

I agree. Without the issues, the M9 is absolutely wonderful. No doubt about it.

 

Everybody has to be mindful that we are interested only in the intermittent problem.

This has to be very carefully separated from everyday malfunctions, like hardware parts or memory cards really going bad.

 

Best, K-H.

[john_newell]

Isn't the issue noted in your post below a significant limitation on the testing? I'm not trying to pick an argument, just wondering how much probative value the admirably detailed procedure in your earlier post will have when all is said and done?

 

Hi there,

 

BTW, it's K-H for Karl-Heinz.

I agree. Without the issues, the M9 is absolutely wonderful. No doubt about it.

 

Everybody has to be mindful that we are interested only in the intermittent problem.

This has to be very carefully separated from everyday malfunctions, like hardware parts or memory cards really going bad.

 

Best, K-H.

[k-hawinkler]

Isn't the issue noted in your post below a significant limitation on the testing? I'm not trying to pick an argument, just wondering how much probative value the admirably detailed procedure in your earlier post will have when all is said and done?

 

 

Hi John,

 

Thanks for your feedback. I really welcome it. Others please join in. This involves all of us.

 

Now to your point.

 

No, not really. It is part of a standard operating procedure for debugging intermittent problems in the worlds fastest computers to exclude all regular failures, e.g. components failing, etc.

 

One collects all sorts of statistics but keeps track of different issues in different bins.

 

In preparing a detailed procedure (help is always welcome) I am thinking along these lines.

 

At the beginning of the testing, we need to follow a certain protocol to ensure - to the best of our ability - that the memory cards used work correctly and don't have bad spots. I welcome forum members help to propose the detailed procedure for that. When during M9 testing then a card gets corrupted - but otherwise works fine (after reformatting) in a card reader or another non-M9 camera - that to me is an important relevant finding that ought to be properly flagged. I would then exclude that card from further participation in the test and use another one.

 

When testing, one needs to keep a complete record including positive and negative outcomes, how many tries, etc. In this case, positive means an error occurred and was recorded, negative means, everything worked fine.

 

If as part of this testing, it is recognized that there is a failure, that can be fixed as part of the regular repair process, then it will go in that appropriate bin and has no significance in terms of characterization of the intermittent issue.

 

We will also need a complete checklist in which all the M9 settings are listed and the active ones can be marked as a function of time.

 

I hope I could make this issue a little clearer.

 

If I failed in doing so, Howard, could you please help.

You have a certain way with words that I don't seem to possess.

 

Thanks again.

 

Best, K-H.

[k-hawinkler]

Isn't the issue noted in your post below a significant limitation on the testing? I'm not trying to pick an argument, just wondering how much probative value the admirably detailed procedure in your earlier post will have when all is said and done?

 

 

Hi John,

 

Thanks for your feedback. I really welcome it. Others please join in. This involves all of us.

 

Now to your point.

 

No, not really. It is part of a standard operating procedure for debugging intermittent problems in the worlds fastest computers to exclude all regular failures, e.g. components failing, etc.

 

One collects all sorts of statistics but keeps track of different issues in different bins.

 

In preparing a detailed procedure (help is always welcome) I am thinking along these lines.

 

At the beginning of the testing, we need to follow a certain protocol to ensure - to the best of our ability - that the memory cards used work correctly and don't have bad spots. I welcome forum members help to propose the detailed procedure for that. When during M9 testing then a card gets corrupted - but otherwise works fine (after reformatting) in a card reader or another non-M9 camera - that to me is an important relevant finding that ought to be properly flagged. I would then exclude that card from further participation in the test and use another one.

 

When testing, one needs to keep a complete record including positive and negative outcomes, how many tries, etc. In this case, positive means an error occurred and was recorded, negative means, everything worked fine.

 

If as part of this testing, it is recognized that there is a failure, that can be fixed as part of the regular repair process, then it will go in that appropriate bin and has no significance in terms of characterization of the intermittent issue.

 

We will also need a complete checklist in which all the M9 settings are listed and the active ones can be marked as a function of time.

 

I hope I could make this issue a little clearer.

 

If I failed in doing so, Howard, could you please help.

You have a certain way with words that I don't seem to possess.

 

Thanks again.

 

Best, K-H.

[Jamie Roberts]

I thought it was good practice to use more, small cards than fewer large ones, especially since the technology in SD cards is somewhat "flakey".

 

Don't you have to change the battery every 4 gig anyway?

 

Andy--

 

I think you're ok with 8GB--though I mainly use 4GB. But 4GB are hard to come by here... I needed more cards and my local store only had 8GB so I bought a couple and used them on the weekend.

 

I've had great luck with these Panasonic's HD SC class 10 "gold" cards meant for video capture, even up 8GB. They have good chip quality (they're guaranteed for 10 years) and they're very fast. Yes, they're expensive, but not compared with anything branded "Leica"

 

Anyway--I plugged them in another thread...

 

To the point on this thread: I'd be happy to test the M9 if someone can give me a definitive list of things to do to test it for failure.

 

I just shot over 1000 shots in 12+ hours on the M9, 2+ batteries and 8GB and 4GB cards without a single hiccup. Yes, that's way more than I normally shoot, but I'm pretty sick right now and took more than usual at last week's wedding!

 

FWIW--the settings were:

 

--manual

--single shot with continuous once for about 100 shots (by accident; I wish there was a way to lock out or lock in a shooting setting)

--no discreet mode

--DNG only

 

Since I was sick I chimped more than usual too.

[Robert_M]

Advertisement (gone after registration)

Here is my list for discussion:

 

• Your M9 is switched off. If you press down the shutter button, does the red LED blink. Please, record Yes or No.

• Firmware: Please record version. 1.162, 1.138, 1.116, 1.002. Is this correct, are there others?

• We should probably only test the newest two versions. Emphasis on the latest version.

 

• Please identify your memory card.

 

M9 parameters that may be involved:

 

• Advance: Standard, Soft, Discreet, Discr & Soft.

• Bracketing setup: No. of exposures: 3,5,7,Sequence: 0/+/-, -,0,+, EV increments: 0.5, 1, 1.5, 2 EV. <= Maybe only need to vary No. of exposures.

• Histogram: Standard, RGB, Clipping setup: Off, Overexposure only, Over / underexposure.

• Auto review: Duration: Off, 1s, 3s, 5s, Permanent, Rel.button pressed, Histogram: On, Off.

• Auto power off: 1 min., 2 min., 5 min., 10 min., Off.

• DNG setup: Uncompressed, Compressed.

 

• Compression: DNG, DNG & JPG fine (Probably less interesting to test: JPG fine, JPG basic, DNG & JPG basic)

• Exp. bracketing: Off, On.

 

Operator actions:

 

• Shutter switch set to S or C.

• How many shots fired?

 

While images were being saved to the memory card. Did you:

 

• switch M9 off?

• Press on Play button?

• Turn the wheel, flip through images, zoom in, reposition the zoom area?

• Press any of the directional buttons? Which one(s)?

 

• Also, after approximately how many shots did something go wrong?

• Any other relevant noteworthy observations?

 

Please, record your observations carefully.

 

I would like to suggest that folks, who have encountered problems, repeat and carefully record this time the scenario in which the M9 malfunctioned before.

 

Best, K-H.

 

 

 

Have been distracted with other photo/camera projects over the last few days. Will try to fill in details on K-H's list for the SD card problems I had.

 

In all instances I have been shooting with the following parameters. S mode, DNG only (not compressed), discreet shutter advance mode, no bracketing, 1.138 firmware, auto power off 1 min, auto review 5s, histogram rgb (no clipping). I've had three cases of spurious problems. ALL the SD cards I use have been purchased from large, major retailers. I check the cards and there is no evidence of "fake" cards. The fake card discussion is nonsense in my opinion.

 

Case 1: Sandisk 8 GB Ultra card, latest edition card. This was a new card and never used before. Formatted in the M9 (short formatting). Problem occured in warm, humid tropical conditions. It was while walking around on vacation taking photos, one at a time. The camera was not turned off between shots and there was not a series of rapid shots. The memory card was within the last 10% of capacity but not full. A series of pictures were not recorded and then I noticed the repeated blinking. Finally stopped it by pulling out the battery. A further shot also had problems so I exchanged cards and continued. That last shot on the card has file problems. I can see the shot in a M9 review but it has a file error which needs file reconstruction to retrieve.

 

Case 2: Another Sandisk 8 GB Ultra card. Also a new card bought at the same time as the above card. Formatted the same way. This second problem happened in similar conditions and a similar scenario as case 1. Also, it was within the last 10% of the card, but not full. A 0 size file was written. Rather than play around, I exchanged cards and continued to shoot. There is nothing in that 0 size file. I've inspected carefully with some computer utilities.

 

Case 3: Many months later. A new Sandisk 16 GB Ultra card (latest version). I formatted the card with the overwrite mode in the M9. This takes an enormous amount of time. I thought I'd try this just to see if it had anything to do with the previous card problems. In this case, the climate conditions were moderate and dry. I was using the card with some lens testing shots and within the first 10 frames recorded on the card, it stopped writing images to the card and just kept blinking for a very long time. It was clear that the buffer was full and the camera was stuck. I had to remove the battery again to reboot the camera. Since they were just test shots of little value, I reformatted the card in the M9 with the short format method and everything worked fine after. I was not taking rapid shots in succession at the time.

 

 

I have not had any of these problems with older Sandisk Ultra II cards (several years old) which were used in other cameras and formatted in a PC at various times. They were also formatted in the M9 before use.

 

The similarity of case 1 and case 2 makes me think there is some interaction between that version of the Ultra card and the M9 reading software/firmware. It is also striking to me that my problem with the new 16GB Ultra card is similar to another poster on this long thread.

 

Well, that is my data of problems (as near as I can reconstruct). I am doing some extensive testing of various SD cards I have on hand. I'll summarize and post later.

 

These experiences have trained me to check the preview images more often (not less) to make sure that the data is being recorded. The other parameter of use I will probably change is to increase the auto power off to about 5min instead of 1.

 

RM

[scrubs]

Hi Robert,

 

In case it is any help, I had exactly the same experience as case #3

with this generic class 4 card that came free with the camera from

the Dealer.

 

MyMemory 4GB SD Card (SDHC) - Class 4 £4.29 - Free Delivery

[Jamie Roberts]

{snipped} The fake card discussion is nonsense in my opinion.

{snipped}

 

RM

 

First, I'm sorry you're losing images. The only commonality I can see in all these threads of lost images is 16GB Sandisk cards, then, secondarily, 16GB cards.

 

Just another reason not to use 16GB cards right now.

 

But to call the counterfeiting problem nonsense is just wrong. When marginally performing chips get passed off as real, all it will take is a camera like the M9 to show them up. Just consider the amount of data you're trying to pass to clear a DNG + JPEG buffer to the card. It's not insignificant, and even a card that tests as "good" out of the box could get messed up by that kind of demand.

 

It's not the "bad" counterfeits that are the problem--you will recognize those right away. But the "back of the factory door chips that didn't quite make the quality cut" slipping out into the general public are the problem.

 

And that's why I don't buy Sandisk anymore.

[scrubs]

First, I'm sorry you're losing images. The only commonality I can see in all these threads of lost images is 16GB Sandisk cards, then, secondarily, 16GB cards.

 

Just another reason not to use 16GB cards right now.

 

But to call the counterfeiting problem nonsense is just wrong. When marginally performing chips get passed off as real, all it will take is a camera like the M9 to show them up. Just consider the amount of data you're trying to pass to clear a DNG + JPEG buffer to the card. It's not insignificant, and even a card that tests as "good" out of the box could get messed up by that kind of demand.

 

It's not the "bad" counterfeits that are the problem--you will recognize those right away. But the "back of the factory door chips that didn't quite make the quality cut" slipping out into the general public are the problem.

 

And that's why I don't buy Sandisk anymore.

 

 

Jamie the problem's with the San-disk cards have nothing to do with capacity

There are problems across the board relating to the "UHS" standard and M9

Firmware. This much Leica have already admitted, aknowledged and are working

on - having nothing to do with fake cards or capacity according to them.

 

If you do some digging you will find there are also many other users using other

brands of cards, not just San-disk and without that standard, having problems also.

[Guest WPalank]

Actually, my most reliable cards in the M9 right now are two SandDisk 16 GB SDHC cards.

[mjm6]

I have had these problems (as I stated in the early parts of the tread), and I can specifically date the first time I had the problem, and the cards that I was using. I'm almost positive it was right after the new firmware, but I need to check the dates tonight.

 

 

 

Just checked the dates. The first time I observed this problem was May 28, which was BEFORE the newest firmware. It was the first time using the newest 16GB Sandisk cards (Ultra SDHC) that I recently purchased, which is the same product as the older cards (about 1.5 years old) that I have, which have not given me any problems.

 

The new ones have a different layout on the card, but the specs seem to be the same in all respects.

 

I would like to test the process with both old and new cards and see if the old cards fail now, but as of today, they haven't given me any troubles, whereas the new ones do give me problems.

 

This was the reason earlier in the thread that I suggested that it may not be due to the firmware.

 

 

---Michael

[Robert_M]

First, I'm sorry you're losing images. The only commonality I can see in all these threads of lost images is 16GB Sandisk cards, then, secondarily, 16GB cards.

 

Just another reason not to use 16GB cards right now.

 

But to call the counterfeiting problem nonsense is just wrong. When marginally performing chips get passed off as real, all it will take is a camera like the M9 to show them up. Just consider the amount of data you're trying to pass to clear a DNG + JPEG buffer to the card. It's not insignificant, and even a card that tests as "good" out of the box could get messed up by that kind of demand.

 

It's not the "bad" counterfeits that are the problem--you will recognize those right away. But the "back of the factory door chips that didn't quite make the quality cut" slipping out into the general public are the problem.

 

And that's why I don't buy Sandisk anymore.

 

 

You are selectively reading what I just posted. I have had far more problems with the 8 GB cards. There is no way that you can take my data and say the problem is 16GB cards! The ONLY commonality in my card problems is that they are the latest SanDisk offerings of Ultra cards. And, this does not necessarily have commonality with the reports of others. Many people have reported problems from all card sizes and many different manufacturers.

 

The key fact that all the "nonsense" (yes, my opinion) discussion about fakes fails to address is: Why does the M9 have problems with these cards and the X1 or S2 does not? Why do we not hear much about "fake card problems" with other cameras, phones, mp3 players, etc which use the same memory cards?

 

The latest reports we hear are that Leica recognizes that there is a legitimate, M9 SD card read/write problem that needs to be addressed. If that is the case, why is there a subset of users posting here who want to still deny that there is a problem? I don't get it.

 

RM

[john_newell]

With sincerity, good luck - if we were talking about anything other than an M9 it would be a dizzying project. Lower numbers for this particular product will hopefully make it more manageable...

 

Hi John,

 

Thanks for your feedback. I really welcome it. Others please join in. This involves all of us.

 

Now to your point.

 

No, not really. It is part of a standard operating procedure for debugging intermittent problems in the worlds fastest computers to exclude all regular failures, e.g. components failing, etc.

 

One collects all sorts of statistics but keeps track of different issues in different bins.

 

In preparing a detailed procedure (help is always welcome) I am thinking along these lines.

 

At the beginning of the testing, we need to follow a certain protocol to ensure - to the best of our ability - that the memory cards used work correctly and don't have bad spots. I welcome forum members help to propose the detailed procedure for that. When during M9 testing then a card gets corrupted - but otherwise works fine (after reformatting) in a card reader or another non-M9 camera - that to me is an important relevant finding that ought to be properly flagged. I would then exclude that card from further participation in the test and use another one.

 

When testing, one needs to keep a complete record including positive and negative outcomes, how many tries, etc. In this case, positive means an error occurred and was recorded, negative means, everything worked fine.

 

If as part of this testing, it is recognized that there is a failure, that can be fixed as part of the regular repair process, then it will go in that appropriate bin and has no significance in terms of characterization of the intermittent issue.

 

We will also need a complete checklist in which all the M9 settings are listed and the active ones can be marked as a function of time.

 

I hope I could make this issue a little clearer.

 

If I failed in doing so, Howard, could you please help.

You have a certain way with words that I don't seem to possess.

 

Thanks again.

 

Best, K-H.

[k-hawinkler]

With sincerity, good luck - if we were talking about anything other than an M9 it would be a dizzying project. Lower numbers for this particular product will hopefully make it more manageable...

 

 

Hi John,

 

Thanks. Good luck is always welcome.

Lower numbers also mean less expensive if something had to be replaced.

Hopefully, the problem can be corrected with just a firmware update.

 

Thanks again.

 

Best, K-H.

[ho_co]

Jamie and Robert--

With respect, it seems to me you're both right.

 

Robert is looking for the specifically M9-originated problem with memory cards, and therefore saying, "Let's eliminate bad cards from the research to concentrate on what causes the M9 to act up."

 

Jamie is saying that some of the problems reported probably stem from bad cards, so we have to be sure that bad cards are eliminated--a difficult task.

 

I think that's the reason for Karl-Heinz's rigorous procedure: An intermittent problem is a mess to reproduce; therefore, all results must be tallied specifically, but properly categorized.

 

Let's not get buried in differences of approach at this stage of the game.

 

I don't think it has been proven that the M9 has a card problem, though the fact that there are no other threads discussing similar problems with the D-Lux 5 or the S2 or the X1 is a very strong indicator to the contrary.

 

Something is going on, and only a strict procedure is going to uncover it. Throwing brickbats won't help.

[k-hawinkler]

Less than a day ago I posted asking for feedback for 48 hours on a rigorous procedure.

There already was some great feedback. More is welcome.

I plan to collect all this feedback, consider it, and then post a concise description.

 

Thanks, K-H.

[Robert_M]

Jamie and Robert--

With respect, it seems to me you're both right.

 

Robert is looking for the specifically M9-originated problem with memory cards, and therefore saying, "Let's eliminate bad cards from the research to concentrate on what causes the M9 to act up."

 

Jamie is saying that some of the problems reported probably stem from bad cards, so we have to be sure that bad cards are eliminated--a difficult task.

 

I think that's the reason for Karl-Heinz's rigorous procedure: An intermittent problem is a mess to reproduce; therefore, all results must be tallied specifically, but properly categorized.

 

Let's not get buried in differences of approach at this stage of the game.

 

I don't think it has been proven that the M9 has a card problem, though the fact that there are no other threads discussing similar problems with the D-Lux 5 or the S2 or the X1 is a very strong indicator to the contrary.

 

Something is going on, and only a strict procedure is going to uncover it. Throwing brickbats won't help.

 

 

Howard,

 

Has Leica acknowledged there is an issue with the M9 and SD cards or not? Maybe someone can confirm that. (1) If not, then yes we can provide more evidence.. (2) if yes, then what is the point of your post and what is the point of debating interpretation of factual observations?

 

Sorry for this rather gruff response. I am disappointed with other responses that there are significant numbers or posters still in the denial phase. Again, I am reminded that the first phase of addiction is denial. And, I understand from psychologists (actually, I'll confess to not being a big fan of the field) is that the first phase of any crisis is denial. The next phase, they tell me, is bargaining. Have we moved onto that yet?

 

RM

[Advertisement]

Hallo,

Look here: Leica M

[k-hawinkler]

RM,

 

I would like to suggest the discussion on this forum effectively has already moved beyond this point.

There is no need to revisit the arguments. Anybody can reread this thread and find out.

Everybody following or posting in this thread cares deeply - I am sure.

No offense intended by my post. Please, let's move on.

 

Best, K-H.

[Robert_M]

RM,

 

I would like to suggest the discussion on this forum effectively has already moved beyond this point.

There is no need to revisit the arguments. Anybody can reread this thread and find out.

Everybody following or posting in this thread cares deeply - I am sure.

No offense intended by my post. Please, let's move on.

 

Best, K-H.

 

 

Yes, I hope so. I know that you and others feel that way. I just get frustrated logging on and reading so many post, and new threads, that revisit the same discussion and look for new excuses. OK, I'll stop reacting and ignore such posts. That is probably healthier....

 

BTW, I'm happy to help with testing protocol and testing. I think we need a simpler set of actions if you want significant numbers of participants. Maybe we should attack one set of topics at a time? Am I trying to make it too simple???

 

RM