I experienced an issue with the 480 GB OWC Atlas Pro card and my A1 II today. I was shooting burst after burst, usually even if the buffer fills up the camera keeps shooting at a reduced rate more like couple of frames a second at least....but today the camera stopped shooting and I had to let the shutter go for a second or two to empty the buffer... It was annoying and I missed a couple of killer frames because of it... has anyone else notice that ? I have never seen this happen with Sony or Delkin cards
On a desktop SSD, TLC achieves speed by writing in SLC mode, but this also means it only write at full speed for at most 1/3 of the rated capacity
Then the SSD needs to do house keeping, which means unpacking the nand used in SLC mode back into TLC to recover space. However when this happens the SSD greatly slows down!
speedmaster20d wrote:
I experienced an issue with the 480 GB OWC Atlas Pro card and my A1 II today. I was shooting burst after burst, usually even if the buffer fills up the camera keeps shooting at a reduced rate more like couple of frames a second at least....but today the camera stopped shooting and I had to let the shutter go for a second or two to empty the buffer... It was annoying and I missed a couple of killer frames because of it... has anyone else notice that ? I have never seen this happen with Sony or Delkin cards...Show more →
I had a member contact me via PM a couple weeks ago. Same card and his was only shooting to a 67 shot buffer and then was taking over a minute to clear it. He had a friend with the same issue.
Not sure if this is similar to your issue?
OWC took his card and fixed it. He thought he'd be getting a new card but when he got the card back it had same serial number. It is working now so maybe they did FW update to the card
The same thing just happened with 2 480 cards and one 960 owc cards with three cards I thought it had to be my A1ii , talked to Sony and they blamed the cards, I tried a prograde ,and Sony and had no problems . If you try black magic speed test you will find they just stop writing than pickup speed again.I checked the firmware and all three have the latest firmware. One of the cards hasn't been used since April and that one developed the same problem. If you have the owc card reader and innerize program sanitize the cards seem to work ok.
On a desktop SSD, TLC achieves speed by writing in SLC mode, but this also means it only write at full speed for at most 1/3 of the rated capacity
Then the SSD needs to do house keeping, which means unpacking the nand used in SLC mode back into TLC to recover space. However when this happens the SSD greatly slows down!
A CF card is just a miniaturized SSD isn’t it
that's right, most of the memory out there is either QLC or TLC. they all use SLC cache to speed things up and then fold the data from the SLC cache into the TLC/QLC blocks. This normally does not reduce the capacity of the SSD but as the drive fills up the SLC cache gets smaller and smaller and eventually the drive writes directly to TLC / QLC blocks, that's why drives often slow down as they fill up
for the CFexpress cards the SLC cache is permanent, i.e. they reserve some blocks as SLC and that's why the capacity is a bit funny like 480Gb instead of there more standard 512 GB
in this case I think the issue could be the memory controller chip or the FW that has hiccups. for example they card may be performing "garbage collection" which refers to a term SSD's use for moving data around to avoid having partially programmed blocks.... if the garbage collection collides with SLC to TLS folding it can cause this kind of hiccups
at the end of the day this seems to be an issue with this brand of card, that happens to be one of the cheapest on the market so maybe something had to give.... it could just fine for other usage but shooting heavy bursts back to back with A1 / A9 series could be exposing a corner case that was dropped in the design of the card.
arbitrage wrote:
I had a member contact me via PM a couple weeks ago. Same card and his was only shooting to a 67 shot buffer and then was taking over a minute to clear it. He had a friend with the same issue.
Not sure if this is similar to your issue?
OWC took his card and fixed it. He thought he'd be getting a new card but when he got the card back it had same serial number. It is working now so maybe they did FW update to the card
that is bizarre, mine cleared after seconds and it did it only once but as you know by Murphy's law that one time happens right at the middle of food exchange!
I think I am going to stick to more trustable brands, it seems there are quite a few anecdotes like mine reported online....
johnsigs wrote:
The same thing just happened with 2 480 cards and one 960 owc cards with three cards I thought it had to be my A1ii , talked to Sony and they blamed the cards, I tried a prograde ,and Sony and had no problems . If you try black magic speed test you will find they just stop writing than pickup speed again.I checked the firmware and all three have the latest firmware. One of the cards hasn't been used since April and that one developed the same problem. If you have the owc card reader and innerize program sanitize the cards seem to work ok....Show more →
thanks, I did try black magic and indeed observed a hiccup during stress test for 4GB file size. It happened after about a minute of stress .... I don't have the OWC card reader unfortunately
speedmaster20d wrote:
thanks, I did try black magic and indeed observed a hiccup during stress test for 4GB file size. It happened after about a minute of stress .... I don't have the OWC card reader unfortunately
It's too bad manufacturers so far require OEM Readers to gain access to their card maintenance app, at least for those manufacturers that bother to offer such an app.
Many ended up with 50% discount on the OWC Reader when that was being offered, others decided not to take advantage of that offer and go with third party or some other Reader.
If you keep the OWC cards might be worth re-considering the Reader to have access to their health check and/or sanitize/refresh card service which might have ( I don't know) solved the card issue you described. Otherwise you can send the cards to OWC for analysis. I'm guessing they will probably run the same health check/sanitize process.
I also have the 480GB OWC Type A CFe and went through all that nonsense. The cost was not worth the aggravation. My Prograde and Delkins Black Type A cards work fine in multiple Sonys.
speedmaster20d wrote:
that's right, most of the memory out there is either QLC or TLC. they all use SLC cache to speed things up and then fold the data from the SLC cache into the TLC/QLC blocks. This normally does not reduce the capacity of the SSD but as the drive fills up the SLC cache gets smaller and smaller and eventually the drive writes directly to TLC / QLC blocks, that's why drives often slow down as they fill up
for the CFexpress cards the SLC cache is permanent, i.e. they reserve some blocks as SLC and that's why the capacity is a bit funny like 480Gb instead of there more standard 512 GB
in this case I think the issue could be the memory controller chip or the FW that has hiccups. for example they card may be performing "garbage collection" which refers to a term SSD's use for moving data around to avoid having partially programmed blocks.... if the garbage collection collides with SLC to TLS folding it can cause this kind of hiccups
at the end of the day this seems to be an issue with this brand of card, that happens to be one of the cheapest on the market so maybe something had to give.... it could just fine for other usage but shooting heavy bursts back to back with A1 / A9 series could be exposing a corner case that was dropped in the design of the card. ...Show more →
I got the OWC card reader - it's a type B reader (there's no type A). You'll need to A-B adapter which comes with their type A card. Since I don't have the problem OP described, so don't know if it's something that'll fix the issue
Add me to the list of people that have had minor issues with the OWC 480 card. I got some weird error messages twice, and it didn’t save some images once when those error messages popped up. It was not actively clearing a large buffer at the time.
Swapped out cards for the rest of the day and went home. I was able to download from the card and then do a full reformat in camera with no further issues. I don’t have the OWC card reader. This only happened recently.
It's not quite simple like that, the size of cache is not simply the difference and the content in cache is flushed to the main blocks in parallel at the same time, so it doesn't just sit there to fill up.... But really what enables the deep burst is the buffer inside the camera (DRAM), not the cache in the card.... once the camera buffer is full the card will be limiting the burst rate quite significantly, as Sony cameras only support NVME gen 3 protocol between the card and the camera which is obsolete and quite slow compared to today's gen 4
liggy wrote:
Wasn’t there a recall or a firmware update for a subset of the OWC cards?
I seem to recall contacting them to make sure mine weren’t part of the batch that needed the update since I don’t have their card reader.
Yes, end of last year IIRC and the issue it purported to fix was a "lock-up" of the card under extreme circumstances, although OWC at the time (again IIRC) said it was related more to video use than stills.
I'd be curious to know if those having issues checked their card for the potential need for this update. If updated cards are still locking up then clearly OWC didn't fix the issue.
aCuria wrote:
I am suspicious about cards using TLC NAND.
Most Type-A cards use TLC but the problem is probably more to do with the controllers and the fact that Type-A cards are inherently more bandwidth-challenged because they have fewer PCIe lanes than Type B cards.
Have multiple OWC 480GB cards. Haven't yet experienced an issue. I'll hammer them after my next shoot and see how they do. I have their reader so hopefully I can update firmware or whatever their fix is if it's needed.
compared to today's gen 4