snapsy wrote:
Good post. Another important point is some sensors are designed to increase MTF by reducing fill factor and thus reduce sensitivity to light, ie reducing the angles of light allowed to reach the sensor in order to increase acuity. Jack Hogan did a great write-up about this regarding his observations with the Z7:
In the thread you linked to, that theory was completely debunked via experiment by 'Just Tony'. The Z7 accepts light from wider incident angles than the D810 (used as comparator) - as it should - given that they put the pixel on the top.
amci4 wrote:
Nope, looked them up. The A1 is slower, but I am not sure if that one is accurate and CINED has the numbers for Video. The point is, if you want the speed there are more options out there.
I've been spoiled - speed to me means 50mp RAW @30fps, 240p 5mp EVF AND exceptionally fast focus acquisition. I'll leave it to you to figure out what makes it onto that list.
sjms wrote:
Priority Set in AF-C set to ‘Release’. The default is ‘Balanced Emphasis’. With it set to ‘Release’ the shutter will be released even if the subject is out of focus. With ‘Balanced Emphasis’ the camera tries to achieve a balance between focus and shutter release, however this reduces the maximum FPS slightly.
AF-C Priority set to AF or Balanced Emphasis does not reduce the maximum achievable FPS to less than 30FPS.
If you mean to say that AF or Balanced Emphasis MAY reduce the total number of images in a continuous sequence to below what one might get with Release if the camera loses focus on the subject then that would be correct.
I find that in practice in almost every sequence using AF priority the images are all 0.0333 seconds apart, and if the bird is not in focus then usually something else is (water, background...).
Steve Spencer wrote:
These sensor scan speeds do not need to be corrected for the size of the sensor. They are simply how fast the sensor is scanned when taking an image. Of course smaller sensors can be scanned faster than bigger sensors.
I think for determining if a bird's wings will distort into a funky (artistic?) mess does warrant looking at how fast it scans a certain height of sensor. If the same bird at the same distance is shot with two cameras with the same focal length then it really depends on how much of the sensor height the wings cover during an exposure. Let's say if the bird fills the APS-C frame and the APS-C camera can scan its entire height in 1/180s but the FF camera can scan its FF height in 1/270s. Well then the FF camera must scan the APS-C height that the bird's wingbeat occupies 1.5x faster than that. That is what I was thinking at least when I started doing maths on the speeds based on sensor size. But I could be looking at this wrong??
NonDecaf wrote:
In the thread you linked to, that theory was completely debunked via experiment by 'Just Tony'. The Z7 accepts light from wider incident angles than the D810 (used as comparator) - as it should - given that they put the pixel on the top.
What Tony's experiment demonstrated was that the Z's wider throat allows wider angles of incident light to reach the sensor. Notice that the F-Mount graph (D810) only starts at around 60 degrees vs 45 degrees on the Z. This was done by directing a distant laser through the mount onto the sensor. That does not demonstrate how the sensors relative full factors differ at such angles since only one sensor is represented, and what's not measured is potential angles with an actual mounted lens that isn't directing light at the laser's 40 feet.
The Z7's BSI properties are reflected in the higher light acceptance across the entire range of angles tested. For the Z-specific high angle measurements (45 to 60 degrees), notice the steep the light-acceptance falloff. That's indicative of high fill-factor attenuation of wider light angles.
Btw, I'm "horshack" on dpreview - I'm the one who suggested the experiments that lead to Tony's post.
duncang wrote:
AF-C Priority set to AF or Balanced Emphasis does not reduce the maximum achievable FPS to less than 30FPS.
If you mean to say that AF or Balanced Emphasis MAY reduce the total number of images in a continuous sequence to below what one might get with Release if the camera loses focus on the subject then that would be correct.
I find that in practice in almost every sequence using AF priority the images are all 0.0333 seconds apart, and if the bird is not in focus then usually something else is (water, background...).
then you live in a perfect world and even sony is wrong.
Steve Spencer wrote:
I don't know where you looked them up, but the number you got for the A1 is wrong as Geoff (arbitrage) noted on the previous page. The numbers are:
A1 - 1/260
X-H2S - 1/180
Z8/9 - 1/270
And the XH2S is about the same as the A9. These would all allow in almost all instances the use of an electronic shutter without any movement distortion. Higher sync speeds are still useful for fringe cases and flash synchonization, but all of these cameras do well.
These sensor scan speeds do not need to be corrected for the size of the sensor. They are simply how fast the sensor is scanned when taking an image. Of course smaller sensors can be scanned faster than bigger sensors....Show more →
Okay so I found a bad link on the Sony.
As for the speeds, you do have to account for the resolution, because the rest of the camera has to be built to move the data to the storage medium at that speed. That’s where the X-H2S appears to make up the difference. All 3 are fantastic cameras and I’m merely making the point, that if $4K to $7K isn’t in your budget, then $2500 is an easier pill to swallow with that feature set.
ChrisMak wrote:
What exactly are you trying to do here by getting the A1 sensor readout speed wrong by factor 1/2 and misrepresenting it as one of the slowest bodies in this regard?
Literally every review since its release quotes a readout speed of around 250 fps at 14 bit
Relax, I just did a quick search and that was the only one I found on short notice. It wasn’t intentional, but I do know the Sony Rolling shutter is not as good as the other two, so it made sense to me that it would be slower. It wasn’t intentional as and you will be okay.
In relation to the current thread and the OP, the R5 does have movement distortion for fast moving targets in electronic shutter mode. In my view it is enough to be concerned about for the types of things I shoot. This can be fixed by using the mechanical shutter, but that isn't silent and substantially drops the frames per second. The Z8, in my view, has a clear advantage in being able to shoot 20 fps without worries about movement distortion and it can do so silently. It is after all, things that tend to move fast for which many people want to shoot at high frames per second and in my experience shooting silently can really be useful for things like wildlife. The R5 is quite capable with it mechanical shutter at shooting things that move fast, but the Z8 with its electronic shutter that captures even quite fast movements without distortion is in my view substantial more capable....Show more →
The R5 is I believe the only one in the list without a stacked shutter if I’m not mistaken. Also, people love the X-H2S and OM-1 for shooting fast moving wildlife, so that 1/125 readout speed appears to be plenty fast enough for the camera. I personally love ES and never using anything but that in my X-H2S and I never really get distortion either, so maybe we have found the sweet spot for finally eliminating the Medhanical Shutter.
duncang wrote:
What Sony says and I quote "The continuous shooting speed may become slower and/or the shooting interval may become uneven, depending on the scene."
and what I experience are not mutually exclusive.
Note that they say "may... depending on the scene". They don't say "will become slower".
funny thing about the real world they wrote it for a reason. could it plausible deniability and knowing it is reality? you know shoulda, woulda, coulda.
now, sony also says that to shoot at 30 fps requires the use of compressed raw which i do believe is 12bit
"Just like Sony says, only Compressed RAW, JPEG or HEIF file formats are capable of achieving the maximum frame rate of 30 FPS."
amci4 wrote:
Relax, I just did a quick search and that was the only one I found on short notice. It wasn’t intentional, but I do know the Sony Rolling shutter is not as good as the other two, so it made sense to me that it would be slower. It wasn’t intentional as and you will be okay.
It does not matter to me, but I cannot make much sense of how you present the comparison, and how you support that with incorrect info.
Personally, I would dig a bit deeper to make sure I have at least the facts right before I make my statement.
I don't know either how you got to the conclusion that the A1 has worse rolling shutter than the Z9 or fuji XH2S, the latter being a crop camera on top of that, but it would be much more convincing if it were backed up with correct data.
Jun 03, 2023 at 07:59 AM
Steve Spencer Online • • • • • • • Upload & Sell: On
arbitrage wrote:
I think for determining if a bird's wings will distort into a funky (artistic?) mess does warrant looking at how fast it scans a certain height of sensor. If the same bird at the same distance is shot with two cameras with the same focal length then it really depends on how much of the sensor height the wings cover during an exposure. Let's say if the bird fills the APS-C frame and the APS-C camera can scan its entire height in 1/180s but the FF camera can scan its FF height in 1/270s. Well then the FF camera must scan the APS-C height that the bird's wingbeat occupies 1.5x faster than that. That is what I was thinking at least when I started doing maths on the speeds based on sensor size. But I could be looking at this wrong??...Show more →
That is true, but I think if you changed lenses on the full frame camera so the bird fills the frame the same amount then it is the sensor scan speed not correcting for sensor size that matters. Concretely if you shoot the same bird at the same distance with a 400mm lens on an APS-C camera and a 600mm lens on a full frame camera all that matters is the sensor scan speed for the camera. Said another way, if the image is equally magnified all that matters is total scan speed of the camera. Higher magnification, however, requires faster scan speed as the electronic shutter needs to move across more of the frame and that is true regardless of sensor size.
Jun 03, 2023 at 08:04 AM
Steve Spencer Online • • • • • • • Upload & Sell: On
amci4 wrote:
The R5 is I believe the only one in the list without a stacked shutter if I’m not mistaken. Also, people love the X-H2S and OM-1 for shooting fast moving wildlife, so that 1/125 readout speed appears to be plenty fast enough for the camera. I personally love ES and never using anything but that in my X-H2S and I never really get distortion either, so maybe we have found the sweet spot for finally eliminating the Medhanical Shutter.
Well, 1/125 will work for a lot of things and many shooters might never see a problem. I would want to test it out for really tough stuff like humming bird wings and golf balls flying off the club, if I shot those things regularly. That said it is a good sensor scan speed, but would never allow eliminating the mechanical shutter because of flash. It would create a camera with a very low flash sync speed (i.e., I don't think 1/125 would be a fast enough flash sync and ever the most basic consumer cameras do a lot better than that).
Jun 03, 2023 at 08:09 AM
Steve Spencer Online • • • • • • • Upload & Sell: On
amci4 wrote:
Okay so I found a bad link on the Sony.
As for the speeds, you do have to account for the resolution, because the rest of the camera has to be built to move the data to the storage medium at that speed. That’s where the X-H2S appears to make up the difference. All 3 are fantastic cameras and I’m merely making the point, that if $4K to $7K isn’t in your budget, then $2500 is an easier pill to swallow with that feature set.
I think what you are getting at is buffer depth. Yes, all of these cameras have some limit on how long they can shoot that fast, but the A1 and the Z8/Z9 both have way more buffer depth than I would ever need. Of course, YMMV, and you may need more buffer depth that either camera provides.
Steve Spencer wrote:
I think what you are getting at is buffer depth. Yes, all of these cameras have some limit on how long they can shoot that fast, but the A1 and the Z8/Z9 both have way more buffer depth than I would ever need. Of course, YMMV, and you may need more buffer depth that either camera provides.
No I’m saying the whole camera matters, Fuji’s image processing chip is ridiculously powerful and that’s not to say the others aren’t, but if the Fuji chip is faster than the output might be quicker regardless of the sensor dump speed. Making up a fraction of a second in RAM or in camera calculations might make the whole thing a wash for performance.
I still think the Z8 might be the best full frame camera on the market, but the Fuji is also an excellent budget option for those who can’t afford the Z8.
arbitrage wrote:
I see you've planted your flag and aren't going to budge. Here is just one more reference among many out there showing the A1 is around 1/260 scan in stills.
"The super-fast readout that enables 30 fps shooting is also key to the camera's electronic shutter being usable for almost all applications. At a measured 1/260s, the e-shutter is just as fast as many mechanical shutters, fast enough that it can sync with flashes at up to 1/200 sec exposures. You'll need to use a Sony TTL-compatible flash or flash trigger to reach this maximum speed, but it shows how quick the sensor is able to operate. If you crop in to APS-C mode, it takes even less time to read out the active area of the sensor, allowing a sync speed of 1/250 sec."
I agree that all the stacked sensor cameras you've listed should be in the conversation when looking to shoot fast action in ES.
The A1 and Z9/Z8 lead the pack for scan speed.
The A9II from 2017 can scan a FF sensor height as fast as the 2022 H2S can scan an APS-C height. So let's not get too excited about that camera's scan speed. That said, 1/180 on an APS-C sensor is probably enough for most things and is still scanning 2x faster than a Canon R5....Show more →
No I said I might be wrong and acknowledged I read a bad source. I have nothing to gain from intentionally making any claim.
ChrisMak wrote:
It does not matter to me, but I cannot make much sense of how you present the comparison, and how you support that with incorrect info.
Personally, I would dig a bit deeper to make sure I have at least the facts right before I make my statement.
I don't know either how you got to the conclusion that the A1 has worse rolling shutter than the Z9 or fuji XH2S, the latter being a crop camera on top of that, but it would be much more convincing if it were backed up with correct data.
CINED measured them and the Fuji was a hair faster than the fastest full frame which was the Z9. Rolling shutter is extremely impressive on those two cameras.
arbitrage wrote:
Video scan speeds have no relation to stills scan speeds in these high end stacked sensors. The stills speed is always significantly faster than the video scan.
6K Open Gate is 24 MP, it very much has relevance to the speed of the camera with only a 26 MP sensor. It tells you how you can get the best performance out of the body under certain circumstances.
Jun 03, 2023 at 02:34 PM
Steve Spencer Online • • • • • • • Upload & Sell: On
amci4 wrote:
No I’m saying the whole camera matters, Fuji’s image processing chip is ridiculously powerful and that’s not to say the others aren’t, but if the Fuji chip is faster than the output might be quicker regardless of the sensor dump speed. Making up a fraction of a second in RAM or in camera calculations might make the whole thing a wash for performance.
I still think the Z8 might be the best full frame camera on the market, but the Fuji is also an excellent budget option for those who can’t afford the Z8.
A lot of speculation there, I would want to see some sort of evidence that the camera is actually quicker.. I think the Fuji is a solid option if you want a smaller sensor and it rounds out their offerings nicely. Their medium format cameras are excellent for things that do not move or move slowly and It is great that they have a compelling option for those who want to shoot things that move quickly. Personally, I am still not very impressed by the lens lineup for shooting action, but that is just me and my preferences, YMMV.
