Home · Register · Join Upload & Sell

Moderated by: Fred Miranda

New fredmiranda.com Mobile Site

New Feature: SMS Notification alert

New Feature: Buy & Sell Watchlist

FM Forums | Canon Forum | Join Upload & Sell

Archive 2023 · Q: record to multiple vs record separately to dual cards

rscheffler
Offline
• • • • • •
Upload & Sell: Off

p.1 #1 · p.1 #1 · Q: record to multiple vs record separately to dual cards

Hi all, I'm familiar with the basic differences when the camera writes images to the card if record to multiple or record separately is selected (multiple mirrors the cards, separately allows writing a different file type to each, or the same file type), but have a question...

First let me explain the situation that I experienced:

R6II with two identical make/model SDXC UHS-II V60 cards set to record to multiple in standard RAW file format. Several hours into an event, the camera locked up and I had to pull the battery to reset. Afterwards the camera could not recognize one of the two cards (the card in slot 1). I've had random incidents with the R6II where it would not read a card on start up, and I'd just restart it or remove and reinsert the card and all would be good. But in this case, after several tries, the camera still could not read the card. I swapped that card out and continued shooting. At the time I noticed that the last sequence I shot prior to the lockup was lost and not on the second card that continued to work correctly. This was likely because those images were 'stuck' in the camera's buffer and then lost when I turned it off to clear the lock up. Fortunately the lost images were not a super critical unique moment of the event.

Once back to the computer, the card that the camera no longer recognized would not mount when inserted in the computer's built-in SD slot, or any other external card readers I tried. It appears to be dead and has since been sent back to the manufacturer for warranty claim.

Whenever recording to dual cards (which I do most of the time), I've always used the record to multiple option which mirrors the two cards with the same file format. But thinking about this, has anyone had experience with the 'record separately' option during a failure of one of two cards? Might this setting instead allow the buffer to clear to one card even if the other card has suddenly hung up/failed/frozen? I can't think of a way to test this because it's dependent on one card failing, which I can't think of a way to simulate.

Thanks!

Sep 26, 2023 at 08:30 AM

deepbluejh
Offline
• • • • •
Upload & Sell: On

p.1 #2 · p.1 #2 · Q: record to multiple vs record separately to dual cards

SD card failures (while writing in camera no less) are so rare that I doubt you'll find much data here unfortunately.

It really depends on how the camera is programmed to operate. It sounds like if the camera encounters a fault to one of the cards while writing, then ALL writing is stopped. In which case, that series of images that is currently writing to the cards is likely lost. Writing to a single card would side step this, but with the obvious drawbacks.

FWIW, I'm about 600k frames into my original R6 bodies and have never had an SD card issue.

Sep 26, 2023 at 09:38 AM

artsupreme
Offline
• • • •
Upload & Sell: On

p.1 #3 · p.1 #3 · Q: record to multiple vs record separately to dual cards

It sounds similar to the situation when you are recording to two cards (different sizes) and the smaller card fills, it stops writing to the bigger card. This is lame. If you have a big CFe in slot one, and a smaller SD card in slot 2, the camera should keep writing to the CFe card when the SD card fills. This is not the case. The camera freezes when the smaller card is full. I learned this the hard way.

Sep 26, 2023 at 10:02 AM

snapsy
Offline
• • • • • •
Upload & Sell: On

p.1 #4 · p.1 #4 · Q: record to multiple vs record separately to dual cards

rscheffler wrote:
Whenever recording to dual cards (which I do most of the time), I've always used the record to multiple option which mirrors the two cards with the same file format. But thinking about this, has anyone had experience with the 'record separately' option during a failure of one of two cards? Might this setting instead allow the buffer to clear to one card even if the other card has suddenly hung up/failed/frozen? I can't think of a way to test this because it's dependent on one card failing, which I can't think of a way to simulate.
Thanks!

I troubleshooted this very issue with a Nikon MILC user about a month ago. He was shooting dual cards and the camera locked up during shooting with the card access light on. After restarting the camera (which required a battery pull), the images in the buffer were missing from both cards. In his case the failing card was still later mountable but the failing-session scenario appears to be the same as what you experienced.

I was able to reproduce his loss of images by simulating a card failure. I know from experience that Nikon's firmware locks up whenever it encounters a malfunctioning card, specifically one that is still mounted/detected but failing to complete outstanding writes. In this scenario the camera essentially waits forever, I presume hoping the card will start responding again. During this time it was theorized that the camera serializes writes of each image between cards, such that if one card locks up then the remaining buffer will not be written to the still-functioning second card.

To prove this theory I shot two cards with grossly-mismatched performance characteristics - one was an old 10 MB/s SD card and the other an 1800 MB/s CFE card (this was on a Nikon Z8). I shot a 20 image raw burst, waited a few seconds, then removed the CFE card while the camera was still attempting to clear the buffer. I let the camera finish writing to the SD card, which took over a minute. In examining the CFE card I saw it contained only the first image of the burst. This proved the camera serializes writes of images between cards.

I suspect the same occurs on other brands as well, and would explain what you experienced. It would of course be preferable if the camera would write to each card independently as fast as possible, then release the frame buffer after a given image has been written to both cards. This would still tie up the camera on the malfunctioning card (ie, not allow new images to be shot due to insufficient buffer space) but at least allow it to save a copy of the buffered images on the functioning card. An alternate solution would be for the camera to tell you which card it's waiting on, so that you can remove that card and induce the camera to resume buffer clearing on the alternate card.

My hunch is the situation will not improve if you try splitting the roles of the cards as raw/jpeg, as the resource being tied up is the original frame buffer and the serialization of writes is likely tied to that for implementation simplicity.

Sep 26, 2023 at 10:40 AM

rscheffler
Offline
• • • • • •
Upload & Sell: Off

p.1 #5 · p.1 #5 · Q: record to multiple vs record separately to dual cards

deepbluejh wrote:
SD card failures (while writing in camera no less) are so rare that I doubt you'll find much data here unfortunately.

It really depends on how the camera is programmed to operate. It sounds like if the camera encounters a fault to one of the cards while writing, then ALL writing is stopped. In which case, that series of images that is currently writing to the cards is likely lost. Writing to a single card would side step this, but with the obvious drawbacks.

FWIW, I'm about 600k frames into my original R6 bodies and have never had an SD card issue.

Thanks, that's probably how it works but was wondering if anyone may have experienced differently. I also have an R6 and have only had one problem with it that I remember, but I think it wasn't really the camera's fault. I'll explain it more in the reply to snapsy. The R6 is more or less my back up camera and is somewhat less used than the Mark II, and was purchased after the Mark II.

This is now the second problem I've had with cards used in the R6II. Back in February shooting to a different brand SDXC UHS-II card, probably about 1/3 of the images on the card were corrupt. Tried different card readers then too, because in the past my experience has been it was usually the card reader, but with the same results (second card in the camera was fine and 'rescued' the shoot). And now this incident with the bricked card. As mentioned, on occasion I've had the R6II not be able to read a card on the first try, but has always been able to on subsequent tries. IIRC it has been an assortment of cards where this has happened. It's got me wondering if it might actually be a problem with the camera's card slots. Just really difficult to pin it down on the camera, but will definitely be paying much closer attention.

Sep 26, 2023 at 01:25 PM

rscheffler
Offline
• • • • • •
Upload & Sell: Off

p.1 #6 · p.1 #6 · Q: record to multiple vs record separately to dual cards

snapsy wrote:
I troubleshooted this very issue with a Nikon MILC user about a month ago. He was shooting dual cards and the camera locked up during shooting with the card access light on. After restarting the camera (which required a battery pull), the images in the buffer were missing from both cards. In his case the failing card was still later mountable but the failing-session scenario appears to be the same as what you experienced.

I was able to reproduce his loss of images by simulating a card failure. I know from experience that Nikon's firmware locks up whenever it encounters a...Show more →

Thanks so much for taking the time to share that information/experience! It does sound very similar and you've given me some thoughts about how to simulate a card failure.

It also reminded me that I ran into a problem with the R6 (not Mark II) back in March shooting youth hockey teams at a tournament where the camera appeared to not save images to the card. At these youth tournaments I hand off cards to computer techs and really don't want to potentially lose costly high capacity, high performance, high quality cards when things get hectic and they're not paying attention to what's happening around them. So I bought a bunch of very cheap, slow, low capacity cards to hand over to the techs, but also mirrored to a high performance card in the second slot for back up. It appeared the problem was due to the massive write performance difference between the two cards because once I put two identical and cheap cards from a reputable brand in both slots, the camera worked normally for the rest of the tournament. So now I generally try to avoid too much of a performance mismatch between cards when writing to both. At least shooting stills though, I have used a UHS-II v90 in slot one and a regular SDXC v30 high capacity card in slot two without problems. But the camera didn't like this and displayed a warning when in video mode, which makes sense because it's continuously writing during a clip.

Sep 26, 2023 at 01:45 PM

rscheffler
Offline
• • • • • •
Upload & Sell: Off

p.1 #7 · p.1 #7 · Q: record to multiple vs record separately to dual cards

OK, I ran a test with the R6II. Slot 1 had a slow class 10 card and slot 2 had a UHS-II v90 card. Camera was set to record to multiple (mirror) and then repeated with it set to record separately to both cards. Shot sequences to fill the buffer and then opened the card door after a bit of time but before the buffer was cleared. On opening, the camera's rear display showed a message with the number of files still to be saved and to close the door. As the number of files to be saved counted down, I'd pull the slow card from slot 1. No matter whether set to record to multiple (mirror) or record separately, the camera would shut down immediately on pulling the card. Reinserting the card and reviewing images in camera always resulted in there were never any images to play back on either card, even though it had counted down the remaining number significantly and the v90 card could have cleared the buffer well before I pulled the slow card in slot 1. On reformatting the cards both did show that there was something on the cards taking up storage space, and both cards showed the exact same amount of storage use. Perhaps in theory those files might be recoverable with recovery software.

My immediate conclusion is that the R6II clears the buffer to both cards at the speed of the slowest card, given that the amount of storage shown to be used was identical on both, even though the faster card could cleared the buffer each time well before I pulled the slower card. If there was a write failure with one of the cards while clearing the buffer, it appears the camera probably would not continue writing to the second card. Pulling a card while the camera is writing to both immediately shuts it down and is not a solution to force it to resume writing to the good second card. Even if that were possible, as the camera is currently configured, there doesn't appear to be any way to know which of the two cards malfunctioned.

As an interesting aside, while doing this a bunch of times, there were a couple instances where the buffer countdown, while the card door was open, suddenly sped up really quickly, as though the camera suddenly ignored the class 10 card and was only writing to the v90 card. After this happened 2-3 times, the camera displayed an error message: "Err 02 Card cannot be accessed. Reinsert/change card or format card with camera." But there was no indication which card couldn't be accessed. It ended up being the class 10 card in slot 1. And same happened if moved to slot 2. Attempting to reformat the card in either slot also failed. Inserting the card in the computer's SD reader (a MacBook Pro) however did result in a 'do you wish to initialize card?' pop-up and I was able to erase and restore it in the computer. After inserting it back into the camera, I was able to reformat it there and it subsequently appeared to be working normally in both slots. Interestingly the 'failed' card was in slot 1, which was the same slot as the bricked card a few days ago. That was also the same slot as the card with the corrupt images back in February...

Either the R6II is more finicky about cards than other Canon cameras I've used and owned, or perhaps there's a problem with my camera's card slot(s). It could also be that it just doesn't like certain things happening to it, or being 'pushed' to its limit. This very much reminds me of my experience with a couple Leica M series digital cameras. In particular, the M9 very much disliked certain things happening while clearing the buffer, such as reviewing images and especially reviewing zoomed in images. That was really rolling the dice with respect to inviting a camera lock-up and loss of buffer contents.

Sep 26, 2023 at 02:52 PM

robstein
Offline
• • •
Upload & Sell: Off

p.1 #8 · p.1 #8 · Q: record to multiple vs record separately to dual cards

So if I read this correctly... dual writes do not actually help at the point of failure? If so that's really pathetic. Not what you would expect.

It would help if the image set completes and THEN one of the cards fail I guess.

Sep 26, 2023 at 06:24 PM

rscheffler
Offline
• • • • • •
Upload & Sell: Off

p.1 #9 · p.1 #9 · Q: record to multiple vs record separately to dual cards

robstein wrote:
So if I read this correctly... dual writes do not actually help at the point of failure? If so that's really pathetic. Not what you would expect.

It would help if the image set completes and THEN one of the cards fail I guess.

Yes, that's correct in respect to whatever is currently in the buffer will be lost. (It would be good to confirm this is also the case with Canon's higher-end R5 and R3 cameras, compared to my R6II.) If it's a critical moment, then poof, it's gone. In this respect, yes, it would be useful if Canon could program the system to write to the other card if one times out after a certain period (like 10 seconds) and then displays a warning to that effect.

But dual-write is still great to have, such as with my recent card failure experience. It meant I didn't lose 2-3 hours worth of non-repeatable photos from a client's reunion event for which some guests travelled great distances to attend. It could have easily been a wedding, major sports or news event, etc. I also dual-write to hedge against errors on my part, such as accidentally losing a card, accidentally formatting one, etc., though it hasn't yet come to that. After a shoot I transport each set separately, with one set on me, until I'm able to download and back them up.

I'd like to see Canon incorporate fast internal memory in cameras, say perhaps 128 or 256GB in addition to the ability to record to dual cards. It would be another hedge against card failure and if it's fast, could enhance buffer performance.

Sep 26, 2023 at 07:03 PM

snapsy
Offline
• • • • • •
Upload & Sell: On

p.1 #10 · p.1 #10 · Q: record to multiple vs record separately to dual cards

robstein wrote:
So if I read this correctly... dual writes do not actually help at the point of failure? If so that's really pathetic. Not what you would expect.

It would help if the image set completes and THEN one of the cards fail I guess.

Yep. As implemented, cameras' dual-card implementations only protect against data loss/corruption for data at rest. It does not protect against data loss for data in-flight, at least for most card malfunctions.

Sep 26, 2023 at 07:16 PM

artsupreme
Offline
• • • •
Upload & Sell: On

p.1 #11 · p.1 #11 · Q: record to multiple vs record separately to dual cards

snapsy wrote:
Yep. As implemented, cameras' dual-card implementations only protect against data loss/corruption for data at rest. It does not protect against data loss for data in-flight, at least for most card malfunctions.

Why would it be difficult to write the software to keep writing to a large CFe in slot 1 when a smaller SD card fills in slot 2? When card 2 fills it should just give you a warning that card 2 is full and continue writing to Card 1.

Sep 26, 2023 at 08:13 PM

rscheffler
Offline
• • • • • •
Upload & Sell: Off

p.1 #12 · p.1 #12 · Q: record to multiple vs record separately to dual cards

robstein wrote:
So if I read this correctly... dual writes do not actually help at the point of failure? If so that's really pathetic. Not what you would expect.

It would help if the image set completes and THEN one of the cards fail I guess.

snapsy wrote:
Yep. As implemented, cameras' dual-card implementations only protect against data loss/corruption for data at rest. It does not protect against data loss for data in-flight, at least for most card malfunctions.

artsupreme wrote:
Why would it be difficult to write the software to keep writing to a large CFe in slot 1 when a smaller SD card fills in slot 2? When card 2 fills it should just give you a warning that card 2 is full and continue writing to Card 1.

Presumably from the camera manufacturer's point of view, the point of mirroring to two cards is that you're mirroring everything. So if one card fills, continuing to write to the second is no longer mirroring.

At the least, they should provide you with the option of choosing whether or not the camera stops or continues to one card only. This reminds me of back in the days with the original 1D, and presumably other early Canon DSLRs - when the frame counter reached 9999 it locked the camera, rather than rolling over to 0001 in a new folder, which is now the norm. That burned me a few times in the middle of games/events until I got into the habit of resetting the file number for each event.

Sep 26, 2023 at 08:33 PM

artsupreme
Offline
• • • •
Upload & Sell: On

p.1 #13 · p.1 #13 · Q: record to multiple vs record separately to dual cards

rscheffler wrote:
Presumably from the camera manufacturer's point of view, the point of mirroring to two cards is that you're mirroring everything. So if one card fills, continuing to write to the second is no longer mirroring.

At the least, they should provide you with the option of choosing whether or not the camera stops or continues to one card only. This reminds me of back in the days with the original 1D, and presumably other early Canon DSLRs - when the frame counter reached 9999 it locked the camera, rather than rolling over to 0001 in a new folder, which is...Show more →

Yes, there should definitely be an option to "continue writing after one card is full".

Sep 26, 2023 at 08:40 PM

Ferrophot
Offline
• • •
Upload & Sell: Off

p.1 #14 · p.1 #14 · Q: record to multiple vs record separately to dual cards

The idea that the camera should contain a small memory internally so that it could operate without any cards inserted, or if a card malfunctioned, would be good. I haven't done it yet, but I know of photographers who have left home with no cards.

Sep 26, 2023 at 11:29 PM

vbnut
Offline
• •
Upload & Sell: Off

p.1 #15 · p.1 #15 · Q: record to multiple vs record separately to dual cards

The note at end of the "Recording Method with Two Cards Inserted" section (https://cam.start.canon/en/C010/manual/html/UG-07_Set-up_0030.html) of the R3 manual makes the behavior pretty clear (I assume it is similar for other models, but I don't know for sure). The 3rd bullet item says
[Card* full] is displayed when one of the cards becomes full, and shooting is no longer possible. To continue shooting, either replace the card or set [Stills Rec options] to [Standard] and select the card with free space.

It's not relevant to me as I write RAW to CF-E and JPEG to SD, erase the CF-E after copying images to my computer (3 locations), and erase the SD at the end of every month, and I've only filled the SD card once. I'm also not a pro, so there's nothing critical riding on me not losing a few shots. I agree it would be nice to have an option to continue writing to the other card when one is full.

Sep 26, 2023 at 11:38 PM

snapsy
Offline
• • • • • •
Upload & Sell: On

p.1 #16 · p.1 #16 · Q: record to multiple vs record separately to dual cards

I'm confused as to why the discussion has turned to card-full situations. The OP didn't mention one of the cards being full. Am I missing something?

Sep 27, 2023 at 03:25 AM

artsupreme
Offline
• • • •
Upload & Sell: On

p.1 #17 · p.1 #17 · Q: record to multiple vs record separately to dual cards

snapsy wrote:
I'm confused as to why the discussion has turned to card-full situations. The OP didn't mention one of the cards being full. Am I missing something?

Is it against the law to discuss two scenarios in one thread? Full card, bad card, similar behavior. This forum is absolutely dead and maybe there's some people who are interested in knowing if their second card fills their camera will freeze.

Sep 27, 2023 at 12:08 PM

snapsy
Offline
• • • • • •
Upload & Sell: On

p.1 #18 · p.1 #18 · Q: record to multiple vs record separately to dual cards

artsupreme wrote:
Is it against the law to discuss two scenarios in one thread? Full card, bad card, similar behavior. This forum is absolutely dead and maybe there's some people who are interested in knowing if their second card fills their camera will freeze.

No worries, I'm a big proponent of open discussions and letting conversations veer into whatever direction they may. I just wanted to make sure I wasn't overlooking something regarding the original problem and how it might relate to exceeding a card's capacity.

Sep 27, 2023 at 12:26 PM

rscheffler
Offline
• • • • • •
Upload & Sell: Off

p.1 #19 · p.1 #19 · Q: record to multiple vs record separately to dual cards

[Card* full] is displayed when one of the cards becomes full, and shooting is no longer possible. To continue shooting, either replace the card or set [Stills Rec options] to [Standard] and select the card with free space.

vbnut wrote:
The note at end of the "Recording Method with Two Cards Inserted" section (https://cam.start.canon/en/C010/manual/html/UG-07_Set-up_0030.html) of the R3 manual makes the behavior pretty clear (I assume it is similar for other models, but I don't know for sure). The 3rd bullet item says
It's not relevant to me as I write RAW to CF-E and JPEG to SD, erase the CF-E after copying images to my computer (3 locations), and erase the SD at the end of every month, and I've only filled the SD card once. I'm also not a pro, so there's nothing critical riding on me not losing a...Show more →

Well, the camera will write to the other card when one is full, but only if originally writing to only one card. It's the auto switch card option. It's a setting I used extensively in the past, but now I just buy massive capacity cards and never worry about switching them. I just mirror them in case of a mishap, such as the recent one.

What artsupreme is after is having the camera continue shooting to only one card when set to 'record to multiple' if the smaller capacity card fills. But this break in logic would probably cause massive confusion in the minds of Canon's engineers as to why anyone would want the camera to automatically switch away from mirroring if they purposefully selected mirroring in the first place.

Sep 27, 2023 at 08:50 PM


FM Forums \| Canon Forum \| Join Upload & Sell

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.