New Delkin Black claims no buffer on Z9...

If you would like to post, you'll need to register. Note that if you have a BCG store account, you'll need a new, separate account here (we keep the two sites separate for security purposes).

Both Dave and I have tickets in with Delkin. I suggest everyone else who sees the problem does the same.

That said, do note that I'm not seeing the SAME performance from my old and new card. The old card drops to like 1 fps when it hits the buffer, the new card seems to be able to maintain 10-15 fps after hitting the buffer and it clears much faster.
The extremely slow performance is similar to what happened when I formatted a card in my D850 and then tested in a Z camera. It had starts and stops as long as 2 seconds with an average write speed of 1.5 fps. Fresh formatting in the camera being used resulted in normal speed. But you have already formatted in the camera being used, so my guess is it needs a deeper formatting which can be provided by Delkin. Just guessing at the solution, but the type of performance you are getting suggests a formatting issue where file structure is not currently optimized for the Z9 camera.
 
I don't see how that would help with a card that has never been used before.
I found an article from Delkin Devices dated 7/9/20. In this article they state: "Now that you understand the importance of formatting, another thing to know is the proper way of formatting, specifically only in a camera and not a computer. A memory card can be formatted in a computer, but if it’s trying to be used in a camera, there’s a chance the camera may not be able to read the card’s file structure properly. As each host device’s operating system formats differently from one another, it’s important to only use memory cards that have been formatted in the camera you are planning to use."
They also go on to talk about Deep/Full formatting or Low Level formatting in a computer to fully erase all data on card. A long process. But this would only correct problems arising because of previous use of the card leaving buildup of data from old files. So Delkin only recommends this after extended use of a card.

So I can only conclude that the statements from Delkin, that this card is much faster and allows unlimited buffer, is just not an accurate statement.
There are different levels of formatting. At a very basic level, one just erases the information in the table stating where data is stored so that the space it occupied can be written to again, often referred to as a quick format. Eric is referring to a format which takes that a step farther and is essentially recreating the table instead of just erasing its contents. As to why a new, unused card would require this is difficult to say, but there were a few reported speed issues when people received their Z9 and new Delkin Black cards at the beginning of the year and this is the procedure Delkin recommended and resolved the issue.

I have the new new 150GB card, but didn’t try a buffer test yet. I was skeptical of the claim of unlimited buffer from the beginning since the previous Delkin Black and ProGrade Cobalt cards already have a maximum sustained write speed higher than the technical requirements needed. Obviously other factors could come into play, so it’s a possibility it could result in unlimited buffer.
 
I don't see how that would help with a card that has never been used before.
why a new, unused card would require this is difficult to say, but there were a few reported speed issues when people received their Z9 and new Delkin Black card
It seems odd that Delkin would ship a card that is wrongly formatted for a Z9 when they are specifically touting Z9 performance - but obviously within the realm of possibility, especially for a new product.

The extremely slow performance is similar to what happened when I formatted a card in my D850 and then tested in a Z camera.
Interesting - I was using my 128gb in my D850 for about 1.5 months before getting the Z9. I did immediately do a format in the Z9 before using it though. John - could this perhaps be your issue? 1fps is much too low...
 
Last edited:
Ok, for completeness I just did another test - the "simple" test of the 512gb with a lenscap, fixed ISO, MF to see if it performed as advertised under those non-real world conditions (the 128gb did to a bit better in my previous tests). It eeked out a few more frames 83 before switching to 16fps - so a tiny performance improvement, but still does not meet the marketing claims.

Side note: actually, I hadn't realized that I was still getting ~15fps after hitting the buffer with these cards - that's pretty good! I do wish it didn't stutter though, which is distracting - rather it was a steady 15fps, but now I know to ignore it in those rare cases I hit the buffer.

Cheers!
 
Interesting - I was using my 128gb in my D850 for about 1.5 months before getting the Z9. I did immediately a format in the Z9 before using it though. John - could this perhaps be your issue? 1fps is much too low...
In my test it was specifically BECAUSE the card was formatted in a different camera and the formatting specs were apparently different. If you format in the camera being used, it normally is fine - but not always. I could not explain the technical details for the required file structure in one vs. the other, but the point is Delkin's initial formatting might not be aligned with Nikon's Z9 file structure. It could be a firmware update for the card or just deep formatting with a program from Delkin. It could also be a firmware update, but Nikon is not likely to do that unless there is something longer term happening - especially since the earlier Delkin Black cards were fine.

There was a similar issue with early Sandisk CFExpress B cards. Sandisk is the Canon partner and the cards worked fine in Canon cameras. But for Nikon, while on paper they should work, there were issues with the 64 GB cards delaying certification by Nikon until Sandisk replaced the card with a new model that was compatible with both Canon and Nikon cameras. The Sandisk cards are fine now.
 
Ok, for completeness I just did another test - the "simple" test of the 512gb with a lenscap, fixed ISO, MF to see if it performed as advertised under those non-real world conditions (the 128gb did to a bit better in my previous tests). It eeked out a few more frames 83 before switching to 16fps - so a tiny performance improvement, but still does not meet the marketing claims.

Side note: actually, I hadn't realized that I was still getting ~15fps after hitting the buffer with these cards - that's pretty good! I do wish it didn't stutter though, which is distracting - rather it was a steady 15fps, but now I know to ignore it in those rare cases I hit the buffer.

Cheers!
15 fps with a full buffer is pretty remarkable. That's where the fast card really pays off.

The stutter can be meaningful. When I tested other cards and looked at frame rate by counting frames in full second increments, I would get some mild stuttering on some cards, but on one card it was particularly bad. There were full seconds when I did not get a single frame, and then it resumed with normal variations and stutters. Temporarily clearing the buffer partially did not help and at times the Remaining Frames count was actually increasing even though the camera performed as though the buffer was full.

I don't expect the Delkin card to have that problem, but it's the kind of thing that reference to "stuttering" brings to my mind. You'd hate to need a high frame rate and go a full second or two at a critical time with 1-2 fps because of stuttering. On the other hand, if stuttering means you still get 10-15 fps, I could live with that.
 
another unscientific test. seems to maintain a smidge higher frame rate than the 650GB Prograde Cobalt (but similar).

also, writing to both the new Black and the Cobalt in backup mode, maintains a steady, but (obviously) reduced frame rate. maybe 4-5 fps?
 
There were full seconds when I did not get a single frame, and then it resumed with normal variations and stutters. Temporarily clearing the buffer partially did not help and at times the Remaining Frames count was actually increasing even though the camera performed as though the buffer was full.
That sounds really weird - the card and camera must not have been communicating properly, or the card was poorly designed to throttle the I/O under severe loading situations. In either case, I'd blame the card, and perhaps a firmware update could fix that, but if it's a bad bus design, then can't be fixed by software/firmware.
 
My assessment is the lineup is now more consistent with their minimum write speeds, but it does appear that the OLD 128 GB card were superior to all the new cards for some reason.

One other thing that I love about these cards that is not often mentioned is their cool temperatures, which will lead to lower probability of errors and better longevity. I, especially appreciate this as I often shoot sporting events in extremely hot temperatures with the sun beating down on the camera. Better for offloading of images where the card is working for 5-10 straight minutes as well.

1652319625772.png
 
So Delkin support got back to me with a very generic "see this page" upon which has a note about Z9 performance posted yesterday:


It doubles down on saying it can shoot indefinitely at 20fps lossless compressed RAW, however it does have some weasel words to the effect of YMMV.

I wrote them back saying effectively "yah, i'm not getting that". I guess we'll see what they say.
 
So Delkin support got back to me with a very generic "see this page" upon which has a note about Z9 performance posted yesterday:


It doubles down on saying it can shoot indefinitely at 20fps lossless compressed RAW, however it does have some weasel words to the effect of YMMV.

I wrote them back saying effectively "yah, i'm not getting that". I guess we'll see what they say.
So it looks like they clarified to say no stops, not a slight drop in fps. Since it seems like it will only drop to 15fps based on other comments, I’m more than ok with the performance. I agree the marketing, or at least original statement, is/was slightly misleading.
 
So it looks like they clarified to say no stops, not a slight drop in fps. Since it seems like it will only drop to 15fps based on other comments, I’m more than ok with the performance. I agree the marketing, or at least original statement, is/was slightly misleading.
they also say that the buffer never maxes out, and that's not true of my testing. you can see the remaining value goes to zero and stays there which is not what they are implying.
 
How does this compare to ProGrade Cobalts?
I have both the 325GB Cobalt cards and the "new" 325GB Dellkin cards and in my testing shooting 20FPS Lossless Compressed Raw, Z9 set to manual, ISO 400 shutter speed 1/2500 C-AF, this is how I normally shot in the real world and I don't see really any difference in performance between the 2 cards, They both shoot for about 4 seconds before they start to stutter... Bottom line, I paid $799 for 2 325GB Prograde Cobalt cards and $860 for 2 of the new Delkins.. Not the end of the world.
I also agree with others posts, the limitation is the Z9 and not the these "high" end cards
 
So Delkin support got back to me with a very generic "see this page" upon which has a note about Z9 performance posted yesterday:


It doubles down on saying it can shoot indefinitely at 20fps lossless compressed RAW, however it does have some weasel words to the effect of YMMV.

I wrote them back saying effectively "yah, i'm not getting that". I guess we'll see what they say.
I'd ask them to escalate the issue to one of their managers for more clarification. I don't think their original statement would be completely made up, so they should be able to outline the parameters under which the buffer never fills.

I would not agree that the limit is the camera at this point. We thought that was the case with the Z6/7 and the Z6ii/Z7ii, but the Delkin Black and ProGrade Cobalt cards produced higher write speeds in those cameras. The increases were only 15-20%, but still meaningful when you look at the impact on maximum burst and buffer clearing. The camera is a limit, but only until a faster card is released. The boundary remains minimum sustained write speed and there are new cards on the horizon with faster speeds than today.
 
I have both the 325GB Cobalt cards and the "new" 325GB Dellkin cards and in my testing shooting 20FPS Lossless Compressed Raw, Z9 set to manual, ISO 400 shutter speed 1/2500 C-AF, this is how I normally shot in the real world and I don't see really any difference in performance between the 2 cards, They both shoot for about 4 seconds before they start to stutter... Bottom line, I paid $799 for 2 325GB Prograde Cobalt cards and $860 for 2 of the new Delkins.. Not the end of the world.
I also agree with others posts, the limitation is the Z9 and not the these "high" end cards
Thanks. I'll stick with my Prograde Cobalt cards. I shoot in HE* - I have not done a comparison between formats, IQ and noise yet, based upon everything I have heard /read HE* is basically equivalent to lossless Raw
 
This is what got back from Delkin
I also spoke to a support representative and he said in “lab” conditions they are getting unlimited 20fps using lossless raw with the Z9, I was told by the rep that this was achieved using “P” mode, and that shooting at higher shutter speeds would be an issue.. I think they “somewhat” over represented what these new cards can do.
 
This is what got back from Delkin
I also spoke to a support representative and he said in “lab” conditions they are getting unlimited 20fps using lossless raw with the Z9, I was told by the rep that this was achieved using “P” mode, and that shooting at higher shutter speeds would be an issue.. I think they “somewhat” over represented what these new cards can do.
i'm curious what "high shutter speed" has to do with anything
 
they also say that the buffer never maxes out, and that's not true of my testing. you can see the remaining value goes to zero and stays there which is not what they are implying.
Yeah, I thought about that afterwards and to me that part would be incorrect in their statement. I have the new 150GB and haven’t bothered to test it with the buffer. I’ll have to do that.
 
I suspect the customer service reps involved are completely clueless. It's too bad they can't escalate to someone with some level of knowledge. I can't believe Delkin would run an ad for a new card based on a customer service rep's test in P mode - obviously without knowing what they were doing. This just does not add up.
 
I suspect the customer service reps involved are completely clueless. It's too bad they can't escalate to someone with some level of knowledge. I can't believe Delkin would run an ad for a new card based on a customer service rep's test in P mode - obviously without knowing what they were doing. This just does not add up.
Wonder if the claim in based upon real world testing, or just combining specs from testing different components????
 
Back
Top