snapsy wrote:
Hmm, I don't have a Sony camera handy but I don't recall the whole FF area being captured in crop mode and undoable in post. I just searched online and see conflicting information. Is it different between various Sony models perhaps?
In crop mode on Sony, the Raw file is indeed saved as a cropped file to cover only the APS-C area, it is not the full sensor resolution. I think that is what you are asking, right?
chiron wrote:
@snapsy@@, help me out here. I am not a techie and I have no engineering background, so what I am saying here may just reflect my ignorance.
But I feel that something is missing from the discussion of sensor speed and crop factor. The reason the sensor readout speed seems shorter is because the area scanned is smaller--but the rate of scanning stays the same. So, for example, LED banding would occur as badly with a crop image as with a full-frame image, because the rate was the same even though the total time spent scanning was shorter. Similarly for rolling shutter effects: the same subject motion will cover a larger percentage of area on a crop image than it would on a full-frame image--so the degree of rolling shutter would be proportional to the area/movement and thus the same.
chiron wrote:
@snapsy@@, help me out here. I am not a techie and I have no engineering background, so what I am saying here may just reflect my ignorance.
But I feel that something is missing from the discussion of sensor speed and crop factor. The reason the sensor readout speed seems shorter is because the area scanned is smaller--but the rate of scanning stays the same. So, for example, LED banding would occur as badly with a crop image as with a full-frame image, because the rate was the same even though the total time spent scanning was shorter. Similarly for rolling shutter effects: the same subject motion will cover a larger percentage of area on a crop image than it would on a full-frame image--so the degree of rolling shutter would be proportional to the area/movement and thus the same.
You've got it correct - the perceptual effects of rolling shutter are not reduced when shooting a FF sensor in APS-C crop mode since the readout rate per row is the same, thus the reduction in readout time from shooting in crop only has potential implications for non-perceptual related metrics, for example burst rate fps.
I am not suggesting that a 5000 ppi display is a real thing — it was just a way to conceptualize the issue.
duncangr wrote:
A 5000ppi display is not analogous to a 218 ppi display. A 5000ppi display would result in blocks of 20x20 or greater pixels being seen as individual pixels - in effect downsampling the image by a huge amount. All pixel size fine details, such as hair would disappear.
Never mind that the entire 10k A7r5 image would be around 2 inches wide on this imaginary 5000ppi display of yours when viewed at 100%.
Remember with retina we are talking individual pixels not being discernible from each other - we are not talking about them not being visible. They are just small enough so that sharp rounded edges such as is found on text looks completely smooth to the naked eye - as if drawn by an ink pen with a smooth stroke rather than a dot matrix printer. ...Show more →
rob_ww wrote:
I suspect it will still scan the whole sensor when in APS-C mode, as usually the whole image is still captured and the crop can be undone in post.
I don't see how it is possible to undo the crop in post. In fact, taking back-to-back images of the same exact subject matter with the same settings results in roughly a 50% size difference, so I don't think those pixels are ever being recorded to the card. Maybe a feature in another brand?
You can't. It just means you have to be thoughtful when you use crop mode. For instance, I always use crop when shooting small birds because I know I'm always going to crop in to the image later. No reason to shoot full frame images and waste all that space when you just have a small bird in the center of the frame. For landscapes, I never use crop mode because I can choose an infinite number of crops in post later (if I choose). For portraits, I'll use crop mode frequently because it lets me see exactly the composition I want if I'm using a normal to wide lens and need more telephoto. It lets me get the exact distance I need to make the subject look good without distortion that might be present if I were viewing the entire frame.
snapsy wrote:
You've got it correct - the perceptual effects of rolling shutter are not reduced when shooting a FF sensor in APS-C crop mode since the readout rate per row is the same, thus the reduction in readout time from shooting in crop only has potential implications for non-perceptual related metrics, for example burst rate fps.
Snapsy, I don’t think you got it right.
Say I’m shooting with a 24GM and decide I want a tighter framing for the next shot. I can either:
A: switch the camera into APS-C mode
B: swap lenses and use the 35GM instead
Option A will have less rolling shutter, assuming APS-C mode has a faster readout speed.
aCuria wrote:
Snapsy, I don’t think you got it right.
Say I’m shooting with a 24GM and decide I want a tighter framing for the next shot. I can either:
A: switch the camera into APS-C mode
B: swap lenses and use the 35GM instead
Option A will have less rolling shutter, assuming APS-C mode has a faster readout speed.
A moving subject on a FF sensor shot with an APS-C crop will move the same number of horizontal pixels across each successive row as it would shot in FF mode, ie its skew from rolling shutter will be the same. This is because the increase in readout speed from the APS-C crop is only due to the fewer number of vertical rows read - it doesn't increase the readout speed along the horizontal axis. Cropping out a portion of the subject or a change of the subject's size as a percentage of the frame (from cropping) doesn't affect that - ie, cropping in-camera via APS-C mode will yield the same rolling-shutter result as shooting that same image in FF mode and cropping in post.
Using a 24GM in FF mode will reduce the size of the subject across the sensor (including vertical axis), and thus will reduce its rolling shutter skew vs shooting with the 35GM in FF mode.
duncangr wrote:
Which part of "5x faster readout than its predecessor" are you struggling to understand ? It's hardly Sony's fault that you and a few others got a bit over excited thinking it would be as fast as or faster than an A1.
And as for it must have a purpose - to quote Jan Wegener:
“If you are after breathtaking resolution, extraordinary cropping capability, and the added bonus of a genuine action camera all in one body then the A7r6 delivers like nothing else on the market”
The A7V has a partially stacked sensor, but a faster readout speed for images. Yeah, it has less pixels. Fully stacked sensors have all been under 10ms readout. So....yes...it is Sony's fault that a lot of people were wondering whether the A7RVI would be close to the A1 in readout speed when they heard "fully stacked sensor". And I wasn't questioning the purpose of the camera; I was asking about the purpose of the stacked layer.
I won't ask what part of that YOU don't understand because it's obviously all of it.
hatsubai wrote:
I don't see how it is possible to undo the crop in post. In fact, taking back-to-back images of the same exact subject matter with the same settings results in roughly a 50% size difference, so I don't think those pixels are ever being recorded to the card. Maybe a feature in another brand?
It can be done on the RX1R III, I own that camera and most certainly can do it on its cropped image files
evanhanded wrote:
The A7V has a partially stacked sensor, but a faster readout speed for images. Yeah, it has less pixels. Fully stacked sensors have all been under 10ms readout. So....yes...it is Sony's fault that a lot of people were wondering whether the A7RVI would be close to the A1 in readout speed when they heard "fully stacked sensor". And I wasn't questioning the purpose of the camera; I was asking about the purpose of the stacked layer.
I won't ask what part of that YOU don't understand because it's obviously all of it.
So it's your contention then that it is Sony's fault that a) partial information was leaked by someone, and b) someone else jumped to the wrong conclusion based on this partial information.
Jazzgear296 wrote:
It can be done on the RX1R III, I own that camera and most certainly can do it on its cropped image files
This is a specific feature of the RX1R III called Step Crop Shooting mode. Quote:
"RAW files shot using Step Crop still capture the full-frame image and simply embed the crop in the metadata. The camera displays the cropped view on playback, but in Lightroom, you can revert to the full-frame version if you prefer."
In other words, RAW shooting will always save the full 61MP image to storage. This is not a feature of MILC. Instead, MILC reduces the RAW image area for faster retrieval from the sensor, reduced in-flight processing, and a smaller volume on the storage medium. With MILC, the RAW crop is automatically selected when the attached lens is APS-C.
snapsy wrote:
A moving subject on a FF sensor shot with an APS-C crop will move the same number of horizontal pixels across each successive row as it would shot in FF mode, ie its skew from rolling shutter will be the same. This is because the increase in readout speed from the APS-C crop is only due to the fewer number of vertical rows read - it doesn't increase the readout speed along the horizontal axis. Cropping out a portion of the subject or a change of the subject's size as a percentage of the frame (from cropping) doesn't affect that - ie, cropping in-camera via APS-C mode will yield the same rolling-shutter result as shooting that same image in FF mode and cropping in post.
Using a 24GM in FF mode will reduce the size of the subject across the sensor (including vertical axis), and thus will reduce its rolling shutter skew vs shooting with the 35GM in FF mode....Show more →
snapsy wrote:
A moving subject on a FF sensor shot with an APS-C crop will move the same number of horizontal pixels across each successive row as it would shot in FF mode
If we make the very reasonable assumption that the APS-C shot is framed identically to the full-frame shot, meaning the focal length used on full frame is 1.5x longer than that used on APS-C, then this statement is incorrect.
This statement is only true if you use the same focal length in APS-C and Full Frame modes
snapsy wrote:
Using a 24GM in FF mode will reduce the size of the subject across the sensor (including vertical axis), and thus will reduce its rolling shutter skew vs shooting with the 35GM in FF mode.
Yes, but to be clear using the 24GM in APSC mode will reduce rolling shutter vs shooting with the 35GM in FF mode.
Assuming of course that the shots are framed the same, meaning distance from camera to subject is identical.
snapsy wrote:
You've got it correct - the perceptual effects of rolling shutter are not reduced when shooting a FF sensor in APS-C crop mode since the readout rate per row is the same, thus the reduction in readout time from shooting in crop only has potential implications for non-perceptual related metrics, for example burst rate fps.
This statement is unclear because its only sometimes true. For example when I use the A9iii then this is true... Or if the user takes a shot with the same distance to subject and same focal length in apsc and ff modes then this would be true.
aCuria wrote:
If we make the very reasonable assumption that the APS-C shot is framed identically to the full-frame shot, meaning the focal length used on full frame is 1.5x longer than that used on APS-C, then this statement is incorrect.
This statement is only true if you use the same focal length in APS-C and Full Frame modes
Yes, but to be clear using the 24GM in APSC mode will reduce rolling shutter vs shooting with the 35GM in FF mode.
Assuming of course that the shots are framed the same, meaning distance from camera to subject is identical.
This statement is unclear because its only sometimes true. For example when I use the A9iii then this is true... Or if the user takes a shot with the same distance to subject and same focal length in apsc and ff modes then this would be true....Show more →
You've restated what I said and no, I don't agree it's reasonable to assume equal framing (ie, a focal length change) when discussing a property of the sensor in this context.
evanhanded wrote:
The A7V has a partially stacked sensor, but a faster readout speed for images. Yeah, it has less pixels. Fully stacked sensors have all been under 10ms readout. So....yes...it is Sony's fault that a lot of people were wondering whether the A7RVI would be close to the A1 in readout speed when they heard "fully stacked sensor". And I wasn't questioning the purpose of the camera; I was asking about the purpose of the stacked layer.
I won't ask what part of that YOU don't understand because it's obviously all of it.
take it with a grain of salt
don't count your chickens
Wisdom.
Sony's upcoming A7RVI will give you A1II performance for a lower price
duncangr wrote:
So it's your contention then that it is Sony's fault that a) partial information was leaked by someone, and b) someone else jumped to the wrong conclusion based on this partial information.
You're right, I don't understand that.
The information that was "leaked" was "fully stacked sensor". That information was correct. My contention is that Sony used that terminology without explaining that it meant something out of the ordinary in this case. All the Youtubers "promoting" the camera used that terminology supplied by Sony...but again, with no explanation of what was different this time. Maybe if the leaker had seen a better explanation of the sensor stack from Sony, they could have leaked better. All that said, I highly doubt that you want to understand as YOU keep attempting to mischaracterize everything I write.
Choderboy wrote: take it with a grain of salt
don't count your chickens
Wisdom.
Sony's upcoming A7RVI will give you A1II performance for a lower price
Clickbait.
Yeah, it's a rumor site. Their track record is way less than 100%. And Sony has surprised us before. I was skeptical that it would match the A1ii, or that any of the rumored "facts" were accurate. I still contend that Sony marketing pulled a fast one. I was holding off from buying the A1ii until I saw what this new camera had to offer. Back to the A1ii.
May 17, 2026 at 10:15 AM
Steve Spencer Online • • • • • • • Upload & Sell: On
evanhanded wrote:
The information that was "leaked" was "fully stacked sensor". That information was correct. My contention is that Sony used that terminology without explaining that it meant something out of the ordinary in this case. All the Youtubers "promoting" the camera used that terminology supplied by Sony...but again, with no explanation of what was different this time. Maybe if the leaker had seen a better explanation of the sensor stack from Sony, they could have leaked better. All that said, I highly doubt that you want to understand as YOU keep attempting to mischaracterize everything I write.
There was also a leak that the sensor scan speed was 16ms (it turns out to be just a bit slower at 19.6ms at least with 14-bit output), and for people who were paying attention that was and I would say should have been a red flag this was a different sort of stacked sensor camera. The camera was obviously in a lot of people's hands, so the leaks could have been from someone violating an NDA, but they could have been orchestrated by Sony as well. Perhaps both sources. With leaks it is often, if not always, difficult to know from where they came. So, is Sony to blame? Perhaps, but even if they were responsbile for the leaks, in this case I think there was enough information out there to give anyone pause that this was going to be an A1 II replacement. Those who made that interpretation weren't even thinking about the leaked specs carefully enough, IMO.
