Interesting. It was kinda weird how slow it was at max capture speed in the first place. Could be a later firmware version having an effect. And/or temperature related, even though I recall dclarks measurements being replicated by others. You don't just happen to be in APSC mode, I assume?
Daran wrote:
Interesting. It was kinda weird how slow it was at max capture speed in the first place. Could be a later firmware version having an effect. And/or temperature related, even though I recall dclarks measurements being replicated by others. You don't just happen to be in APSC mode, I assume?
Slow? It exceeded 120fps? (122)
I double checked and the raw files are 24MP. FF mode, not apsc.
120fps was achieved by pressing down the C button, then holding the shutter button down.
aCuria wrote:
Slow? It exceeded 120fps? (122)
Slow when offloading the buffer. Either A1 can sustain >10 fps till heat exhaustion. The A9III has only 1/2 the data to move but sustains only 5 fps?
Daran wrote:
Slow when offloading the buffer. Either A1 can sustain >10 fps till heat exhaustion. The A9III has only 1/2 the data to move but sustains only 5 fps?
aCuria wrote:
Has anyone seen the 5fps slowdown happen?
No, mine slowed down to around 15fps when I tested it a long time ago. Some cards have a buffer and once that is filled the write speed becomes pretty slow. Get VPG400 cards that have a minimum write speed.
If you look at p.1 #6 of the referenced thread you will see that there was a question about why the buffer with capacity of 236 frames emptied at different rates when in 120fps or 60fps and 30fps or slower. The expected crossover for 120fps or 60fps if the rate of emptying the buffer is the same is is a bit less than 350 frames for 120fps and 60fps. It may be that a firmware update fixed that issue. I don't currently have access to the A9III so I can't repeat the measurements with current firmware.
The measurements shown in this thread look a bit strange, especially the second and third charts. Are you extracting msec frame timing? Are you operating the camera in focus priority rather than release priority mode? Can you show data for the msec frame intervals such as was shown in the 6th chart in the referenced thread?
It's worth noting that some of the cards used in the testing are VPG400 so that cannot be an explanation for any difference.
dclark wrote:
If you look at p.1 #6 of the referenced thread you will see that there was a question about why the buffer with capacity of 236 frames emptied at different rates when in 120fps or 60fps and 30fps or slower. The expected crossover for 120fps or 60fps if the rate of emptying the buffer is the same is is a bit less than 350 frames for 120fps and 60fps. It may be that a firmware update fixed that issue. I don't currently have access to the A9III so I can't repeat the measurements with current firmware.
The measurements shown in this thread look a bit strange, especially the second and third charts. Are you extracting msec frame timing? Are you operating the camera in focus priority rather than release priority mode? Can you show data for the msec frame intervals such as was shown in the 6th chart in the referenced thread?
It's worth noting that some of the cards used in the testing are VPG400 so that cannot be an explanation for any difference....Show more →
Second and third charts were using af priority. First Chart had AF off.
