Alik Griffin: Best SD Memory Cards Leica M11

[Ne314satel]

Advertisement (gone after registration)

On 1/9/2023 at 7:23 PM, SrMi said:

Delkin Black v90

but the camera also sometimes freezes with it.

[SrMi]

21 minutes ago, Ne314satel said:

but the camera also sometimes freezes with it.

Did your camera freeze with Delkin Black and stopped freezing when you used a different card or internal memory?

My M11s did not freeze yet with Delkin Black.

Edited January 12, 2023 by SrMi

[Advertisement]

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

[Ne314satel]

3 hours ago, SrMi said:

Did your camera freeze with Delkin Black and stopped freezing when you used a different card or internal memory?

My M11s did not freeze yet with Delkin Black.

I have Delkin Black 64gb - sometimes freezes with it. When working with internal memory, everything is ok

I'm tired of it) I just want to take pictures. Therefore, I turned on m11 for the priority of internal memory.

Edited January 12, 2023 by Ne314satel

[M11 for me]

and does this always work?

[SrMi]

9 hours ago, Ne314satel said:

I have Delkin Black 64gb - sometimes freezes with it. When working with internal memory, everything is ok

I'm tired of it) I just want to take pictures. Therefore, I turned on m11 for the priority of internal memory.

If it freezes with other cards as well, then the issue is not the type of SD card but some hardware issue. You could test that, and ask for a replacement.

[ivohula]

If you read the article carefully, you will find he says that the M11 camera is doing the following:

What I think is happening is many new memory cards are no longer compatible with the legacy hardware the Leica M11 is using. It looks like the Leica M11 is using very old protocols to communicate with the cards. This is a problem we were having when testing the Panasonic GH4 back in the day. The Leica M11 looks like it’s using similar hardware as that 9-year-old camera based on how it’s performing.

Since UHS-I memory cards were also performing very poorly, it seems like the Leica M11 is only using the UDMA 4 mode to communicate with the cards. This puts the cards at around a max write speed of 66.7MB/s per lane. I get around 50MB/s when testing UHS-I cards, so this seems correct.

My guess is that the latest v90 UHS-II cards might just be dropping this legacy communication protocol from the memory controllers and the cards drop to UHS-I speeds, or down to a single lane of data transfer instead of the two, which results in half the speed.

I’m not sure why Leica decided to use 10-year-old hardware for communication with memory cards in a $9,000 dollar camera. It would have been nice if they used modern hardware so we got a faster performance out of the cards."

 

It means that except for the cards he recommends, you will get the same performance with old UHS-I cards as with the non-reccomended UHS-II cards, except for downloading the faster cards to your computer.

In my experience, lock-ups happen when a card is not communicating with the camera properly, or not getting data at the proper speed within the parameters set by the camera and the camera shuts the information transfer down. This could be due to the protocols of the camera-card drivers or something like oil/grease on the card contacts.

[lct]

Advertisement (gone after registration)

54 minutes ago, ivohula said:

Since UHS-I memory cards were also performing very poorly [...]

My UHS-I cards did work perfectly before 1.6 firmware update.

[ivohula]

All the cards I use in my M11 are the same ones I used in my M9; they are Sandisk Extreme Pro 45MB/s or 95MB/s. I don't see a speed difference between the two on my camera at all. I have never had a lock up, but then my M11 has never seen a UHS-II card. I am using 1.6 firmware. Write speed is not an issue for me in the Leica, but lock-ups certainly would be. My Canons get CFexpress cards as that is where the speed counts (1200MB/s). 

If Griffin is correct, then we are wasting our money on UHS-II cards unless we buy his recommended ones and even then, you are only at 103MB/s at best.

[Photoworks]

If you look at his article you can see that he promotes cards from Lexar and Sony , and some other once.

He is promoting Lexar in many other posts, and always has affiliate link to buy the cards. It is smells little fishy!

 

Personally I stay from Lexar like the plague. And Sony tough cards are notorious not to work well on Leica cameras.

All the good cards where not even tested. LOL

[CDodkin]

The amount of FUD and conformational bias from certain people in this thread is just mind-blowing 

Edited January 15, 2023 by CDodkin

[Kwesi]

I disagree. 
The one thing we can all glean from the post on this thread is that there is no magic card that solves the freezing issue. 
In fact based on this thread we can’t even say with any certainty that it’s a card issue.
 

[CDodkin]

15 hours ago, Kwesi said:

I disagree. 
The one thing we can all glean from the post on this thread is that there is no magic card that solves the freezing issue. 
In fact based on this thread we can’t even say with any certainty that it’s a card issue.
 

Given that the linked testing at no point made any claims of solving anyone's freezing issue - that's purely people in this thread trying to draw a conclusion from non-related SD card speed testing.

One thing we can glean from this thread is that people clutch at straws and jump to their own conclusions.

We also have excellent data on SD card read/write performance with the M11 - which in itself is extremely useful.

[dscdsc]

On 1/16/2023 at 7:33 AM, Kwesi said:

I disagree. 
The one thing we can all glean from the post on this thread is that there is no magic card that solves the freezing issue. 
In fact based on this thread we can’t even say with any certainty that it’s a card issue.
 

yea I think the sd card is a red herring and people are jumping to conclusions.

[lct]

21 minutes ago, jqian6 said:

and it doesn't work perfectly anymore after 1.6 firmware update? or you simply just didn't update the firmware?

I did update to firmware 1.6 but i tried a UHS-II card (Sandisk 300 mb/s) to see how it worked with the new firmware. Got a couple of freezes to begin with, then i updated my user profiles under firmware 1.6 and i never got the least freeze with the same UHS-II card since then.

[Tailwagger]

7 hours ago, lct said:

I did update to firmware 1.6 but i tried a UHS-II card (Sandisk 300 mb/s) to see how it worked with the new firmware. 

Thats what I've been using since day one with a few freezes early and none of late since the 1.5 updates. 

As I've said in the past I seriously doubt that there's a single factor involved. If it was simple it would have been fixed long ago. That's not to say that the mem card isn't involved, but its not as simple as just buying the 'right' card.  I'd bet that there's a semaphore(s) or lock(s) somewhere thats not getting released under some hard to replicate, timing sensitive, conditions that could involve a write or initiating a write to the memory card.  ie. this is a software problem. 

It's likely that the internal software design uses a main monitor thread that's coordinating operations across various independent sub threads/routines that are performing specific tasks like dealing with menus, config, playback, file writes, evf refresh, evaluating metering, auto iso calculations, PC, etc, etc, etc. One duty of that monitor thread is to ensure that when certain critical operations are in flight, others are not allowed to run.  The canonical example for this is writing the memory card.  When thats taking place, you can't allow the user to turn the camera off until it has completed or the contents of the card could be corrupted which in the extreme could lose every shot on the card.  Thus the OS takes this very seriously.  If somehow the core logic misses the fact that the write operation completed, then the camera can lock up***.  Classic race condition and a likely explanation as to why you can't switch the camera off when things freeze... its a virtual switch, not a physical break on the electrons from the battery... the os is alive in there, it just wont let you turn the camera off until it thinks the write has completed. If it misses the message, it never will think that, so you must pull the battery to clear the fault by rebooting the system.

I also strongly suspect, having seen lockups in every M I've owned from the 240 onward (as have others), that this flaw(s) of which we speak is buried deep in the system and has been with us for a very long time.  It is an ancient evil that has re-awoken with the increased speed of the M11's hardware.  Anecdotally, its interesting to note that the M10-R was slower in its file writes than the M10 due to larger files being processed by the same hardware. This might explain why many seem to think that camera freezes less frequently than any of the recent Ms.  Long story short, AFAICT this seems to be a time sensitive problem, a deep and non-obvious one, but AFAIC is almost certainly a software, not a hardware problem per se. Slower cards might make the problem less likely to occur, but they're not solving anything, merely masking it. 

*** I'm obviously oversimplifying here.  I'm not an expert on what it takes to read the sensor, stuff it into the intermediate buffer, applying corrections or recasting down to 16MPx or doing demosaicing, jpg processing and conversion, dealing with SD memory controllers and writing out the bits, etc, etc, etc.  That the shutter doesn't respond suggests that the flag that's being missed is somewhere in that tangle of independent but co-reliant operations required to write out an image. Perhaps the sensor is incorrectly seen to be busy or the membuf it has to write into is locked or it could be a multi-level locking problem... dunno.  But without access to the code, while the details are all speculative, the symptoms are absolutely classic and quite familiar to anyone whose spent time working at the operating system level.

 

Edited January 21, 2023 by Tailwagger

[Photoworks]

5 minutes ago, Tailwagger said:

Thats what I've been using since day one with a few freezes early and none of late since the 1.5 updates. 

As I've said in the past I seriously doubt that there's a single factor involved. If it was simple it would have been fixed long ago. That's not to say that the mem card isn't involved, but its not as simple as just buying the 'right' card.  I'd bet that there's a semaphore(s) or lock(s) somewhere thats not getting released under some hard to replicate, timing sensitive, conditions that could involve a write or initiating a write to the memory card.  ie. this is a software problem. 

It's likely that the internal software design uses a main monitor thread that's coordinating operations across various independent sub threads/routines that are performing specific tasks like dealing with menus, config, playback, file writes, evf refresh, evaluating metering, auto iso calculations, PC, etc, etc, etc. One duty of that monitor thread is to ensure that when certain critical operations are in flight, others are not allowed to run.  The canonical example for this is writing the memory card.  When thats taking place, you can't allow the user to turn the camera off until it has completed or the contents of the card could be corrupted which in the extreme could lose every shot on the card.  Thus the OS takes this very seriously.  If somehow the core logic misses the fact that the write operation completed, then the camera can lock up***.  Classic race condition and a likely explanation as to why you can't switch the camera off when things freeze... its a virtual switch, not a physical break on the electrons from the battery... the os is alive in there, it just wont let you turn the camera off until it thinks the write has completed. If it misses the message, it never will think that, so you must pull the battery to clear the fault by rebooting the system.

I also strongly suspect, having seen lockups in every M I've owned from the 240 onward (as have others), that this flaw(s) of which we speak is buried deep in the system and has been with us for a very long time.  It is an ancient evil that has re-awoken with the increased speed of the M11's hardware.  Anecdotally, its interesting to note that the M10-R was slower in its file writes than the M10 due to larger files being processed by the same hardware. This might explain why many seem to think that camera freezes less frequently than any of the recent Ms.  Long story short, AFAICT this seems to be a time sensitive problem, a deep and non-obvious one, but AFAIC is almost certainly a software, not a hardware problem per se. Slower cards might make the problem less likely to occur, but they're not solving anything, merely masking it. 

*** I'm obviously oversimplifying here.  I'm not an expert on what it takes to read the sensor, stuff it into the intermediate buffer, applying corrections or recasting down to 16MPx or doing demosaicing, jpg processing and conversion, dealing with SD memory controllers and writing out the bits, etc, etc, etc.  That the shutter doesn't respond suggests that the flag that's being missed is somewhere in that tangle of independent but co-reliant operations required to write out an image. Perhaps the sensor is incorrectly seen to be busy or the membuf it has to write into is locked or it could be a multi-level locking problem... dunno.  But without access to the code, while the details are all speculative, the symptoms are absolutely classic and quite familiar to anyone whose spent time working on working on operating systems.

 

I totally agree with you.

I am not a programmer and can't imagine what is involved. It sure seams that the M11 is using a new database in the cards, this is supposition on my part because non of the cards coming from other cameras don't work until formatted for M11. This is probably to work on the internal and SD card, and be able to use the DB in FOTOS.
In the M11 there are many changes in hardware that require software. Leica is a small company and they have already a full load of work. Working on the FW is probably in the list, but it takes time to give us new option and fix bugs, get testing and fix some more. you can do this forever, or never come out with anything new anymore until it is perfect!

I have suggested Profiles been a condition where some bugs only where reproducible , now everyone is talking about profiles been the problem. I don't think that. The issues probably come from one of the many setting changed in the camera and save to a profile, it can be any of the setting in combination with specific condition to cause an issue. this is hard to pinpoint. So please don't stop using Profiles, or say profiles are the problem.

I am looking forward to collaborate in detail if they like to make the FW stable for everyone.

[lct]

43 minutes ago, Tailwagger said:

Thats what I've been using since day one with a few freezes early and none of late since the 1.5 updates.

Different experience here. My only way to avoid freezes was to use UHS-I cards before firmware 1.6. It is only firmware 1.6 which allowed me to use UHS-II cards and i had to update my user profiles for that. 

[Photoworks]

18 minutes ago, jqian6 said:

The whole writing speed theory related to SD card and amplified by profiles might be the true culprit. But as many have said here it's deeply buried in one or more of those settings. At this point of time, I incline to believe it's a software issue.

I know many of you have tried many things, such as not using profile, going back to slower card i.e. UHS-I, etc, has anyone tried to not to shoot raw at all? Is it an option (note I don't have M11 yet). If it's possible, just try to prevent the system writing DNG files, which are huge. This might give some clues?

Again, it's just a theory, for now it doesn't hurt to try

you can pick any of the combination flavors to save your photos in the M11, just like many other M before.

I am not sure I would give up DNG. in any case if I consider one freeze in 2 month and going, it is not an issue for me. in fact FW 1.6 is quite stable. I am sure other are different opinion.
Heck! my iPhone freezes more often for no reason.

[Tailwagger]

52 minutes ago, lct said:

Different experience here. My only way to avoid freezes was to use UHS-I cards before firmware 1.6. It is only firmware 1.6 which allowed me to use UHS-II cards and i had to update my user profiles for that. 

More evidence, since our hardware is the same and the SD card is the same, that some combo of options and software is at the heart of the issue.  I will say, anecdotally, that I don't bother using profiles on the M.  I find things simple enough that I don't need them. Perhaps they're involved, but I suspect if so,  its more about the parameters they contain rather than using them per se.

[Advertisement]

Hallo,

Look here: Interesting Leica gear

[Antonio Russell]

It would be good to get some definitive test resutls for different cards... I appreciate Griffin's efforts but measuring card speeds based on the blinking LED on the base of the camera doesn't, on the face of it, appear to be a particularly reliable method of doing so.

It would be nice if Leica themselves would chime in with this information and help support users of the M11 range of cameras....

 

Edited March 29 by Antonio Russell