CFexpress choice confusion- Z8

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).

Member's gallery
marler1957 said:
ALik Griffin has actual test data on most of the cards at https://alikgriffin.com/best-cfexpress-type-b-memory-cards/ Also look under the "Reviews" pulldown for test data on other card formats. You are considering good brand cards from my previous research for my Z8...also note the comment from Alik on some Nikon shooters finding the Angel Birds cards to be too thick (YMMV).

I am using a 1TB Lexar Gold for 8K video as well as 4K when not shooting stills. For still photos, I shoot RAW plus large-fine-jpeg and limited to my Sandisk Extreme Pro V90 at 268MB/s writes.
Click to expand...
Those updated evaluations are useful I had not seen them. From that it looks like if you went for the Delkin Power you would need the 650 to get a fast enough speed for rapid fire stills.

Be aware that early reports I saw also documented overheating. From what I recall the Lexar cards in that test hd an issue with overheating. The Dellkin Black and Prograde cobalt maintained the coolest temperatures.

Only the Delking Blacks offer 48 hour no questions asked replacement. You need to register your card to get the protection.

I tried to order the 650 gb Delkin Black that was on sale but they are out of stock at B H. Hopefully they get some in.
 
Member's gallery
wotan1 said:
I think this is clear to everyone but I just want to make the point just in case.

The card speed issue has to do with the time it takes to move an image from the camera buffer to the card.

When you shoot video or high frame rate stills the data initially goes to the buffer. The data has to get off the buffer and onto the card before the buffer gets full. If not the camera will slow down.

As I understand it test results with the Z8 and Z9 have shown that the sustained write speed is the bottleneck. For video you need 800 something sustained write speed to record continuous video until the card fills. For stills the biggest speed pressure comes from shooting 20 fps RAW. For that you need about 1400.

If you don’t have the video speed the camera as I understand will stop shooting. For the stills the frame rate slows down.

They have sustained write claims in the fine print. The question I have is whether those are hypothetical or based on real experience. The cards have made advances that are beyond the capacity of current cameras. Is the sustained speed an estimate of what can happen or based on actual tests with existing cameras.

My feeling is I paid a lot of money for these cameras and i use high frame rate burst shooting all the time. I don’t want to hamper my camera by putting slower cards in there.
Click to expand...
Actually the number for stills is 1300MB/s sustained to get the 82 frame buffer in lossless compressed. The Delkin Black at the current Gen is 1530MB/s still gets 82 frame buffer. I have a velocity CINE 128GB card that is made by Delkin and is the exact same specs as the Delkin Black previous Gen. Same 1710MB/s sustained write speeds and the same 48 hour replacement warranty. That card gets me 84 frames in lossless compressed.
 
Member's gallery
wotan1 said:
Be aware that early reports I saw also documented overheating. From what I recall the Lexar cards in that test hd an issue with overheating. The Dellkin Black and Prograde cobalt maintained the coolest temperatures.
Click to expand...

I definitely had overheating issues shooting long videos or multiple short ones close together in time when at the F1 race in Austin last October. Ambient temperature was 90F+ which definitely was not a help. Did not have an issue with shooting 10fps or 20fps stills but I was being careful not to have the camera exposed to the sun for more 20-30 minute periods.

I probably need to add a Delkin Black to my card set soon. Thanks for the reminder.
 
Member's gallery
Thread made me do some Z8/Z9 buffering research. Found some year old tests on youtube:
1) Nikon Z8 vs Z9 - Buffer, Speed and FPS testing (Matt Granger) ->
(yes Matt, SD card in my Z8 hurts)
2) Nikon Z8 Buffer Test - Is It Better Than the Nikon Z9? (The Results Are Surprising!) (Jack Beasley)->

Decided on rough estimates to use for card sizing for how I mostly shoot:
1) exFat formatting leaves just under 93% usable space on a card
2) one max size RAW photo = 60MB (typically mid 50s)
3) one max size fine JPEG photo = 30MB (typically high 20s)
 
Member's gallery
marler1957 said:
Thread made me do some Z8/Z9 buffering research. Found some year old tests on youtube:
1) Nikon Z8 vs Z9 - Buffer, Speed and FPS testing (Matt Granger) ->
(yes Matt, SD card in my Z8 hurts)
2) Nikon Z8 Buffer Test - Is It Better Than the Nikon Z9? (The Results Are Surprising!) (Jack Beasley)->

Decided on rough estimates to use for card sizing for how I mostly shoot:
1) exFat formatting leaves just under 93% usable space on a card
2) one max size RAW photo = 60MB (typically mid 50s)
3) one max size fine JPEG photo = 30MB (typically high 20s)
Click to expand...
I looked at that test.

With my Z9’s I never record to more than one card at a time. The tests he ran involve some sort of second card recording whether it was backup or jpeg or whatever. Obviously the SD card in the second slot of the Z8 is going to be slower than the CF B card in the Z9 second slot. I would have wanted a test with just recording to the CF B card in each camera to tell if there is a difference. Maybe I missed something but I wonder about that.
 
Member's gallery
Never mind the above I looked at the second video which tends to show the Z8 being slower is due to the use of a slower SD card in the second slot.

What would be interesting is to run these tests again with later firmware. In other words does the camera firmware affect buffer speed.
 
Member's gallery
wotan1 said:
Never mind the above I looked at the second video which tends to show the Z8 being slower is due to the use of a slower SD card in the second slot.

What would be interesting is to run these tests again with later firmware. In other words does the camera firmware affect buffer speed.
Click to expand...
No the firmware should have zero to do with the buffer depth. It's solely dependent on the card speed.

The hardware is the card slots are PCIe 3.0 2 Lane. This gives a max write speed of roughly 1.8GB/s

I've seen people, i believe one was Alex Phan, use M,2 1TB 2230 NVME drive and an adapter get unlimited buffer with lossless compressed on the Z9. This drives write at a sustained 3GB/s so they max out the PCIe speeds.

So the max buffer is what it is based on the hardware and the cards sustained write speeds
 
Last edited:
Member's gallery
marler1957 said:
Realizing now that I know too little about Expeed architecture and DRAM quantities in Nikon cameras, so I need to do some research...I thought some assigned amount of DRAM was involved in buffering?
Click to expand...
There is memory that the camera writes to not how fast the cameras hardware can wire to the card will determine if you hit the buffer limit or not. If they had put more memory in the camera that could have extended the buffer a bit but in the end it's still about hour fast it can write to the card to keep that buffer clear.
 
Member's gallery
EricKlees said:
No the firmware should have zero to do with the buffer depth. It's solely dependent on the card speed.

The hardware is the card slots are PCIe 3.0 2 Lane. This gives a max write speed of roughly 1.8GB/s

I've seen people, i believe one was Alex Phan, use M,2 1TB 2230 NVME drive and an adapter get unlimited buffer with lossless compressed on the Z9. This drives write at a sustained 3GB/s so they max out the PCIe speeds.

So the max buffer is what it is based on the hardware and the cards sustained write speeds
Click to expand...
I've tried M.2 NVME adapters, and I both have not been able to replicate Alex's results and I suspect there was some sort of flaw with the test.

At this time I think there is no way to sustain RAW at 20FPS
 
Member's gallery
John Navitsky said:
I've tried M.2 NVME adapters, and I both have not been able to replicate Alex's results and I suspect there was some sort of flaw with the test.

At this time I think there is no way to sustain RAW at 20FPS
Click to expand...
I have seen other video's showing the unlimited buffer but I can not remember who or where they were.

Though I can't find Alex's video on this anymore either but his video was pretty cut and dry with the rear LCD showing the settings. I have never tried making my own as I was concerned over the heat these DIY create and I shoot HE* anyway and never hit the buffer with the Cobalt 325GB card.

There were discusions on this that only certain types of drives could be used and not all M.2 NVME drives would produce the same performance
 
Member's gallery
I have settled on Delkin Power as my personal preference. I do have the 512Gb but I now have 2 x 650Gb Delkin Power. They are G4 the size shows as 1.03Tb on the computer. The speed of the cards seems to be very very good. I do use RAW HE* as my normal output. There is no buffer limit as far as I can tell.
For smaller cards the Delkin Power 128Gb is a very good performing card.
 
You must log in or register to reply here.
Back
Top