I've tested several memory cards in the Nikon Z6 and Z7II to understand performance of the different CFExpress and ZQD cards I own. I also wanted to understand how to achieve and maintain the highest frame rates and write the largest number of images to my card in a period of time.
I tested the Delkin Power 128 GB CFExpress, the Prograde Gold 128 GB CFExpress, and the older Lexar 64 GB XQD. The Delkin is a high performance CFExpress card and the Prograde is an economy CFExpress card.
My test methodology was to hold down the shutter for about 32 seconds, and then count the number of images shot in each second. I shot a complete series with 14 bit lossless compressed RAW and 12 bit lossless compressed RAW. Camera settings were set to otherwise minimize processor impact and support faster speed - no noise reduction, single AF, electronic shutter, Silent shutter off, etc. I used a 50mm f/1.8 S lens at base ISO producing a shutter speed of around 1/400 sec.
The vertical axis shows the number of frames I shot in each second. The horizontal axis shows the rank order of each one second period. I only used whole seconds based on the camera's time stamp, so the start of each test had 2-4 wasted images that were not in the counts. That means after filling the buffer, I might have gotten just a few more frames in the first second that was not shown at maximum.
The graphs clearly show that the cameras performed at the maximum frame rate until the buffer filled, then slowed to 3-4 fps for the duration of the burst. The Z7II (before the latest firmware update) showed more variation than the Z6 but that may be partially related to firmware not being tuned fully yet.
I had several surprises. On several occasions, the camera simply stopped firing for second or two while continuing to write images and raising the counter for number remaining. A few times the number remaining rose above 9 frames - but the actual frame rate continued at 3-4 fps. Even if I stopped shooting for a second or two, the camera did not regain the high frame rate. I had to stop for long enough to completely write all images and a second or two to resume a fast frame rate, and it resumed for a shorter period of time. Another surprise was that card choice did not make much difference. It did affect the length of the initial high speed burst, but once that ended, the cameras all slowed and did not regain a fast rate. I was especially surprised that the XQD card held its own, albeit it was more variable in the Z7II. For me, the big worry is an extreme slowdown - dropping to 2 or fewer frames in a second. As long as it was shooting 3-4 fps I could live with the results, but a 1-2 second break in the action without warning is a problem.
In terms of performance, my ranking was Delkin Power, Lexar XQD, and then the ProGrade Gold. There are a lot of other cards that could be tested.
My take is that Nikon has conservatively set the cameras to avoid heat buildup and cutoffs. I have not figured out an optimal workaround to improve speed. Stopping for a few seconds and then shooting a full burst produced around the same number of images in 30 seconds.
This is a point of reference and suggests you might want to try your own testing. Card size and card brand and model will make a difference.
I tested the Delkin Power 128 GB CFExpress, the Prograde Gold 128 GB CFExpress, and the older Lexar 64 GB XQD. The Delkin is a high performance CFExpress card and the Prograde is an economy CFExpress card.
My test methodology was to hold down the shutter for about 32 seconds, and then count the number of images shot in each second. I shot a complete series with 14 bit lossless compressed RAW and 12 bit lossless compressed RAW. Camera settings were set to otherwise minimize processor impact and support faster speed - no noise reduction, single AF, electronic shutter, Silent shutter off, etc. I used a 50mm f/1.8 S lens at base ISO producing a shutter speed of around 1/400 sec.
The vertical axis shows the number of frames I shot in each second. The horizontal axis shows the rank order of each one second period. I only used whole seconds based on the camera's time stamp, so the start of each test had 2-4 wasted images that were not in the counts. That means after filling the buffer, I might have gotten just a few more frames in the first second that was not shown at maximum.
You can only see EXIF info for this image if you are logged in.
You can only see EXIF info for this image if you are logged in.
You can only see EXIF info for this image if you are logged in.
You can only see EXIF info for this image if you are logged in.
The graphs clearly show that the cameras performed at the maximum frame rate until the buffer filled, then slowed to 3-4 fps for the duration of the burst. The Z7II (before the latest firmware update) showed more variation than the Z6 but that may be partially related to firmware not being tuned fully yet.
I had several surprises. On several occasions, the camera simply stopped firing for second or two while continuing to write images and raising the counter for number remaining. A few times the number remaining rose above 9 frames - but the actual frame rate continued at 3-4 fps. Even if I stopped shooting for a second or two, the camera did not regain the high frame rate. I had to stop for long enough to completely write all images and a second or two to resume a fast frame rate, and it resumed for a shorter period of time. Another surprise was that card choice did not make much difference. It did affect the length of the initial high speed burst, but once that ended, the cameras all slowed and did not regain a fast rate. I was especially surprised that the XQD card held its own, albeit it was more variable in the Z7II. For me, the big worry is an extreme slowdown - dropping to 2 or fewer frames in a second. As long as it was shooting 3-4 fps I could live with the results, but a 1-2 second break in the action without warning is a problem.
In terms of performance, my ranking was Delkin Power, Lexar XQD, and then the ProGrade Gold. There are a lot of other cards that could be tested.
My take is that Nikon has conservatively set the cameras to avoid heat buildup and cutoffs. I have not figured out an optimal workaround to improve speed. Stopping for a few seconds and then shooting a full burst produced around the same number of images in 30 seconds.
This is a point of reference and suggests you might want to try your own testing. Card size and card brand and model will make a difference.
