I don't think any of us, including me, really do care much about speed of memory cards - correct me if I'm wrong.
I just remember once we had a thread about the buffer size (of the D850) and I had a hard time to see the 51 frames it should be able to record before buffer slows down.
Now I just read a sidetone about a Nikon demonstration and a rep talking about XQD cards in his D850: He only uses them and no SD card next to it because they would slow down the record process (although not set to SD? weird). I wanted to see that for myself. Usually the camera ist set to SD card as I have readers for them in both Macs. The XQD reader I always need to plug in and one Mac is still on USB 2.0.
So first I used only the Lexar 128 GB XQD and not the Sony QD-G64E XQD which is mentioned in the D850 manual. This time right from the beginning I got the 51 shots (no AF, Manual mode, no Auto ISO, no Auto WB, lens defocused) on the Lexar card. And only 42 on the Sony... SD card inserted and XQD card formatted in camera - 30 frames on XQD :o
Again, this time without SD: 32 frames, same card and again formatted before I started to torture the shutter.
Long story short. A freshly "formatted" XQD card which was in use before will not be able to catch up with what the buffer can deliver. In other words, doing the burst over an over (counter went up to 500 frames), the frame number will increase to 51, in some burst even a bit more.
Bottom line: If you ever want to know the real burst rate of a given memory card in your camera, don't format the card after each burst, the results will be wrong or questionable.
Another bottom line: Usually we're told to empty batteries and don't charge them when they still have 50-90% capacity. Maybe I also need to think about if it's really clever to delete pictures I just downloaded form the card and never use the full capacity of the card. I know SSD are slower with overwriting "deleted" blocks than saving to a fresh SSD. I'm bit afraid now to use always only the "entry space" of the card while the higher registers just collect byte-dust.
After that I tried the SD card solo - that made the buffer slowing down after less than 40 frames. Also, the CF-card (Sandisk Extreme Pro 90MB/s, 16GB) in the D810 starts to stutter after 25 shots. 25 × 36 MP = 900 MP/burst. The D850maxes out at 51 × 45.7 MP = 2330 MP/burst. Meaning: Nikon, who can't make DSLRs ( © miro
) improved buffer size roughly by factor 2.6!
I hope Lexar gets back to XQD manufacturing with the old performance
If I knew 2 years ago what I know now, I would have bought another 128 GB card... Costly yes, but worth the price. While Sony... much more costly (CHF 5.47/GB or today CHF 2.70/GB) than Lexar (CHF 2.18/GB) and about 20% less fast.
I just remember once we had a thread about the buffer size (of the D850) and I had a hard time to see the 51 frames it should be able to record before buffer slows down.
Now I just read a sidetone about a Nikon demonstration and a rep talking about XQD cards in his D850: He only uses them and no SD card next to it because they would slow down the record process (although not set to SD? weird). I wanted to see that for myself. Usually the camera ist set to SD card as I have readers for them in both Macs. The XQD reader I always need to plug in and one Mac is still on USB 2.0.
So first I used only the Lexar 128 GB XQD and not the Sony QD-G64E XQD which is mentioned in the D850 manual. This time right from the beginning I got the 51 shots (no AF, Manual mode, no Auto ISO, no Auto WB, lens defocused) on the Lexar card. And only 42 on the Sony... SD card inserted and XQD card formatted in camera - 30 frames on XQD :o
Again, this time without SD: 32 frames, same card and again formatted before I started to torture the shutter.
Long story short. A freshly "formatted" XQD card which was in use before will not be able to catch up with what the buffer can deliver. In other words, doing the burst over an over (counter went up to 500 frames), the frame number will increase to 51, in some burst even a bit more.
Bottom line: If you ever want to know the real burst rate of a given memory card in your camera, don't format the card after each burst, the results will be wrong or questionable.
Another bottom line: Usually we're told to empty batteries and don't charge them when they still have 50-90% capacity. Maybe I also need to think about if it's really clever to delete pictures I just downloaded form the card and never use the full capacity of the card. I know SSD are slower with overwriting "deleted" blocks than saving to a fresh SSD. I'm bit afraid now to use always only the "entry space" of the card while the higher registers just collect byte-dust.
After that I tried the SD card solo - that made the buffer slowing down after less than 40 frames. Also, the CF-card (Sandisk Extreme Pro 90MB/s, 16GB) in the D810 starts to stutter after 25 shots. 25 × 36 MP = 900 MP/burst. The D850maxes out at 51 × 45.7 MP = 2330 MP/burst. Meaning: Nikon, who can't make DSLRs ( © miro
) improved buffer size roughly by factor 2.6!I hope Lexar gets back to XQD manufacturing with the old performance
If I knew 2 years ago what I know now, I would have bought another 128 GB card... Costly yes, but worth the price. While Sony... much more costly (CHF 5.47/GB or today CHF 2.70/GB) than Lexar (CHF 2.18/GB) and about 20% less fast.