Any Leica users who own a camera not already in my rolling shutter database and would like to contribute their camera's measurement to my site please send me a PM Here's a link describing the project and process:
So, ... I'm looking at the database and the table info.
I notice that the fields for:
# bands in measurement
total # bands in image
As one who tries to understand where the quid pro quo of "there's no such thing as a free lunch" exists, the differences in these values are strikingly large. However, I have no understanding of what they actually refer to or the meaningfulness behind them.
So, while the readout speeds for these cameras seem correlated to the # of bands ... I realized that I have no clue as to what these bands are, and thus why they would / wouldn't matter to me. I understand that more info takes longer to process, less info takes less time to process, pretty simple stuff there. But, I'm not sure I understand what the influence of # of bands is. My gut tells me that it relates to increments of refinement in some regard, but again ... no clue on my end.
If someone could explain what this is, I'd appreciate that.
But, I found it interesting as the tester pushed the file. Had me wondering how that relates back to the # bands (and thus the readout speed).
Admittedly, I don't know what I don't know ... so clue me on the impact / influence on the # bands involved as to the pro / con vs. quid pro quo of "no free lunch", particularly as it pertains to IQ vs. readout speed vs.
But, just a cursory correlation of the models mentioned above and their respective # bands, it does seem to correlate as a quid pro quo of IQ vs. speed (no surprise that the two would be inversely correlated).
But, here again ... ummm, I need some education regarding the matter of bands, thanks.
rscheffler wrote:
The S system also had leaf shutter options. I don't think it's the ideal solution. Leaf shutter adds size, complexity and expense to each lens. IMO the better technological solution is what Sony and Nikon have done with their stacked sensor cameras. I believe the Nikon Z8/Z9 don't even have mechanical shutter options available, proving the technology exists now to virtually eliminate need for a mechanical shutter. It would be interesting to know the reasons why Leica didn't opt for a stacked sensor in the SL3 as unfortunately every non-stacked Sony FF sensor option appears to have significant challenges with rolling shutter....Show more →
Because the SL3 IQ is measurably better. More resolution and more DR. Right now the SL3 and M11 are as close to miniMF as 24x36 can get.
If people don't like it then go buy an A1 for fucks sake. You don't see this whinging about the A7R5 or A7CR. The SL3 has issues. This isn't one of them.
I have an A1 and I;m glad my SL3 doesn't have that sensor.
RustyBug wrote:
So, ... I'm looking at the database and the table info.
I notice that the fields for:
# bands in measurement
total # bands in image
As one who tries to understand where the quid pro quo of "there's no such thing as a free lunch" exists, the differences in these values are strikingly large. However, I have no understanding of what they actually refer to or the meaningfulness behind them.
So, while the readout speeds for these cameras seem correlated to the # of bands ... I realized that I have no clue as to what these bands are, and thus why they would / wouldn't matter to me. I understand that more info takes longer to process, less info takes less time to process, pretty simple stuff there. But, I'm not sure I understand what the influence of # of bands is. My gut tells me that it relates to increments of refinement in some regard, but again ... no clue on my end.
If someone could explain what this is, I'd appreciate that.
But, I found it interesting as the tester pushed the file. Had me wondering how that relates back to the # bands (and thus the readout speed).
Admittedly, I don't know what I don't know ... so clue me on the impact / influence on the # bands involved as to the pro / con vs. quid pro quo of "no free lunch", particularly as it pertains to IQ vs. readout speed vs.
But, just a cursory correlation of the models mentioned above and their respective # bands, it does seem to correlate as a quid pro quo of IQ vs. speed (no surprise that the two would be inversely correlated).
But, here again ... ummm, I need some education regarding the matter of bands, thanks.
flash wrote:
Because the SL3 IQ is measurably better. More resolution and more DR. Right now the SL3 and M11 are as close to miniMF as 24x36 can get.
If people don't like it then go buy an A1 for fucks sake. You don't see this whinging about the A7R5 or A7CR. The SL3 has issues. This isn't one of them.
I have an A1 and I;m glad my SL3 doesn't have that sensor.
Gordon
I agree that the SL3 has better IQ, resolution and DR, but the SL3 represents a significant investment for some of us. If I was looking for the medium format look, I would buy the Fuji or Hasselblad. With the SL3, for the price, I expect fairly modern camera. And it is not unreasonable in 2024 to expect things like BSI, decent AF, and a better EVF than the one from the previous generation. It feels very "off the shelf" to me, with little to no innovation. I have set aside the money for the SL3, but I find its limitations are currently stopping me from buying. I will need to personally shoot one to see the impact of these issues.
1/10 readout is a significant problem, as I photograph a lot of aviation. I had hoped that the SL3 would have AF capable of tracking fast moving aircraft at close range, head on. It does not. The readout speed will make the camera unusable for propeller driven aircraft. So this is an issue for me. I will continue to use my R3 and future R1 for this purpose.
For me, this then relegates the SL3 to travel/street cam, but the above portrait is an illustrates the rolling shutter in mildly dynamic situations. It makes me wonder if the camera has a use case for anything with movement. As you said more MF like, but then this is a 35mm format camera, not medium format.
I can't help but wonder if the SL3 is basically an EVF M11, and no longer what the SL & SL2 were - which was very functional usable camera's.
snapsy wrote:
Those are just the internal calculations to produce the values you see in the live table. It's documented in my GitHub landing page readme:
Okay, so I get the number of bands ... sort of like size / # of shutter slats on window blinds to cover the window.
Begging the question of the A7RV vs. the A1 or A9 (sticking within a given brand) ... what would be the determining factor whether the engineering decision was made to have 39 bands by design vs. 1 band by design?
Is there any quid pro quo for having more / fewer bands ... rows being read. Is this a computational limit, power consideration, heat dissipation constraint, bit depth, well depth, or
It strikes me that there has to be some kind of quid pro quo involved between something as significantly different as the A7RV vs. the A1 ... on an engineering design level, that then translates into pragmatics limits ... whether that influence be on IQ vs. Performance vs.
Short
Why make the A7RV and the A1 so radically different? What's the benefit / trade-off that's under the hood / behind the scenes as the driver for such engineering decisions?
Okay, so I get the number of bands ... sort of like size / # of shutter slats on window blinds to cover the window.
Begging the question of the A7RV vs. the A1 or A9 (sticking within a given brand) ... what would be the determining factor whether the engineering decision was made to have 39 bands by design vs. 1 band by design?
Is there any quid pro quo for having more / fewer bands ... rows being read. Is this a computational limit, power consideration, heat dissipation constraint, bit depth, well depth, or
It strikes me that there has to be some kind of quid pro quo involved between something as significantly different as the A7RV vs. the A1 ... on an engineering design level, that then translates into pragmatics limits ... whether that influence be on IQ vs. Performance vs.
Short
Why make the A7RV and the A1 so radically different? What's the benefit / trade-off that's under the hood / behind the scenes as the driver for such engineering decisions?...Show more →
Faster sensor readouts require reading multiple rows in parallel at a time and/or having a stacked design with memory close to the sensor to support moving that data off the sensor. The drawbacks are cost, complexity, and increased sensor read noise, which can translate into marginally lower dynamic range.
Mar 30, 2024 at 08:41 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
Planetwide wrote:
I agree that the SL3 has better IQ, resolution and DR, but the SL3 represents a significant investment for some of us. If I was looking for the medium format look, I would buy the Fuji or Hasselblad. With the SL3, for the price, I expect fairly modern camera. And it is not unreasonable in 2024 to expect things like BSI, decent AF, and a better EVF than the one from the previous generation. It feels very "off the shelf" to me, with little to no innovation. I have set aside the money for the SL3, but I find its limitations are currently stopping me from buying. I will need to personally shoot one to see the impact of these issues.
1/10 readout is a significant problem, as I photograph a lot of aviation. I had hoped that the SL3 would have AF capable of tracking fast moving aircraft at close range, head on. It does not. The readout speed will make the camera unusable for propeller driven aircraft. So this is an issue for me. I will continue to use my R3 and future R1 for this purpose.
For me, this then relegates the SL3 to travel/street cam, but the above portrait is an illustrates the rolling shutter in mildly dynamic situations. It makes me wonder if the camera has a use case for anything with movement. As you said more MF like, but then this is a 35mm format camera, not medium format.
I can't help but wonder if the SL3 is basically an EVF M11, and no longer what the SL & SL2 were - which was very functional usable camera's.
Keep in mind that you only get rolling shutter effects with the electronic shutter. When you use the mechanical shutter you don't get these effects. So the SL3 will be totally useable for moving subjects with the mechanical shutter. Further it is a BSI sensor, just not a stacked BSI sensor and it has phase detect auto focus (PDAF), which should improve AF considerably over the SL2 that did not have PDAF. Except for electronic first curtain mode this camera has a sensor very similar to the Sony A7r V, which is a very modern sensor and one of the most recent cameras introduced by Sony. The A7r V also has a 1/10 sensor read out speed. So it is not as if the SL3 doesn't have a modern sensor. It does. It just is a sensor that is better suited for high IQ for static or slow moving subjects and less well suited for action but it can be used for action in mechanical shutter mode and you will really only get movement distortion in electronic shutter mode.
Planetwide wrote:
I agree that the SL3 has better IQ, resolution and DR, but the SL3 represents a significant investment for some of us. If I was looking for the medium format look, I would buy the Fuji or Hasselblad. With the SL3, for the price, I expect fairly modern camera. And it is not unreasonable in 2024 to expect things like BSI, decent AF, and a better EVF than the one from the previous generation. It feels very "off the shelf" to me, with little to no innovation. I have set aside the money for the SL3, but I find its limitations are currently stopping me from buying. I will need to personally shoot one to see the impact of these issues.
1/10 readout is a significant problem, as I photograph a lot of aviation. I had hoped that the SL3 would have AF capable of tracking fast moving aircraft at close range, head on. It does not. The readout speed will make the camera unusable for propeller driven aircraft. So this is an issue for me. I will continue to use my R3 and future R1 for this purpose.
For me, this then relegates the SL3 to travel/street cam, but the above portrait is an illustrates the rolling shutter in mildly dynamic situations. It makes me wonder if the camera has a use case for anything with movement. As you said more MF like, but then this is a 35mm format camera, not medium format.
I can't help but wonder if the SL3 is basically an EVF M11, and no longer what the SL & SL2 were - which was very functional usable camera's.
I thought it was an odd "line" I heard from Leica management SL3 announcements; something along the lines of...not much more can be added to modern cameras so we decided less in more in some circumstances... Again, not exact wording.
Probably only Leica can get away with that type of announcement for their latest mirrorless camera in 2024. Ha Ha
Sans the stacked sensor, the SL3 carries on with the SL system's ability to shoot to 1/8000s in mech shutter. At a minimum to 1/8000s, is helpful for day time Nocitlux 0,95 shooting and other very fast moving subjects in bright light conditions too.
However the very low FPS and related FPS AF limitations on the high resolution SL3 doesn't lend itself to performance shooting anyway. Never mind the current capabilities of freshly minted Panasonic hybrid AF system with PDAF. I think Leica's AF it's going to take some time for the new hybrid AF tech to mature, hopefully via firmware updates.
Otherwise, IF there is something like a FF 24 MP SL2-S coming e.g. SL3-x, perhaps Leica can squeak out a bit more FPS and PDAF performance with a lower resolution FF BSI sensor.
snapsy wrote:
Faster sensor readouts require reading multiple rows in parallel at a time and/or having a stacked design with memory close to the sensor to support moving that data off the sensor. The drawbacks are cost, complexity, and increased sensor read noise, which can translate into marginally lower dynamic range.
Steve Spencer wrote:
Keep in mind that you only get rolling shutter effects with the electronic shutter. When you use the mechanical shutter you don't get these effects. So the SL3 will be totally useable for moving subjects with the mechanical shutter. Further it is a BSI sensor, just not a stacked BSI sensor and it has phase detect auto focus (PDAF), which should improve AF considerably over the SL2 that did not have PDAF. Except for electronic first curtain mode this camera has a sensor very similar to the Sony A7r V, which is a very modern sensor and one of the most recent cameras introduced by Sony. The A7r V also has a 1/10 sensor read out speed. So it is not as if the SL3 doesn't have a modern sensor. It does. It just is a sensor that is better suited for high IQ for static or slow moving subjects and less well suited for action but it can be used for action in mechanical shutter mode and you will really only get movement distortion in electronic shutter mode....Show more →
I had the A7R4 for a while, and the sensor is very good. Rolling shutter can become evident even with mechanical shutter on things like propellors etc... which are typically shot at lower shutter speeds for partial prop blur. To be honest, I did not expect the Leica to be much more than it is. The issue is, that the that the competition has caught up lens wise for my particular use case.
Planetwide wrote:
I agree that the SL3 has better IQ, resolution and DR, but the SL3 represents a significant investment for some of us. If I was looking for the medium format look, I would buy the Fuji or Hasselblad. With the SL3, for the price, I expect fairly modern camera. And it is not unreasonable in 2024 to expect things like BSI, decent AF, and a better EVF than the one from the previous generation. It feels very "off the shelf" to me, with little to no innovation. I have set aside the money for the SL3, but I find its limitations are currently stopping me from buying. I will need to personally shoot one to see the impact of these issues.
1/10 readout is a significant problem, as I photograph a lot of aviation. I had hoped that the SL3 would have AF capable of tracking fast moving aircraft at close range, head on. It does not. The readout speed will make the camera unusable for propeller driven aircraft. So this is an issue for me. I will continue to use my R3 and future R1 for this purpose.
For me, this then relegates the SL3 to travel/street cam, but the above portrait is an illustrates the rolling shutter in mildly dynamic situations. It makes me wonder if the camera has a use case for anything with movement. As you said more MF like, but then this is a 35mm format camera, not medium format.
I can't help but wonder if the SL3 is basically an EVF M11, and no longer what the SL & SL2 were - which was very functional usable camera's.
Then you buy one of the three excellent high res fast read out sensor cameras instead. ie: A1, Z8 or Z9.
Oh, and stop relying just on specs for the rest. The Sl3's IBIS is at least as good as the A7R5 and everyone I've shown both cameras to so far thinks the EVF's are the same. Don't forget the Nikons are using lower res EVF's and they're considerd *modern cameras*. And the A1 sensor is as old as the one in the SL3. You can't just compare on paper specs. DpReview tried this in another recent piece of clickbait bullshit. Pick them up and compare side by side in a decent store. Or rent them. That's the only way.
Personally, I have no idea why an aviation photographer would be shoooting L mount when the Sony A1 exists.
Not only that but to acheive the shooting speeds advertised the bit depth needs to drop below 14bit. Usually to 12 bit, further widening the IQ gap. It's small but significant for how pliable a file becomes in post. Leica is chasing IQ. Sony and Nikon are obviously chasing the sports and bird photographer. Canon are throwing shit at a wall to see what sticks....
stgrove wrote:
Wonder if on the SL3S they will opt for stacked sensor technology?
Unlikely. Especially since their longest fast prime is 90mm. Leica is clearly in the IQ camp rather than the performance one. Not exactly chasing the sports market with that one. If anything it might be Panasonic but they seem tied to Leica pretty hard for now.
flash wrote:
Not only that but to acheive the shooting speeds advertised the bit depth needs to drop below 14bit. Usually to 12 bit, further widening the IQ gap. It's small but significant for how pliable a file becomes in post. Leica is chasing IQ. Sony and Nikon are obviously chasing the sports and bird photographer. Canon are throwing shit at a wall to see what sticks....
Gordon
+1 that a reduction in bit depth would enable quicker readout times ... or inversely, greater bit depth increases readout times (or increases quantity processed > heat > noise). A few options how to approach, but just wanted to understand who was paying for "lunch".
flash wrote:
Unlikely. Especially since their longest fast prime is 90mm. Leica is clearly in the IQ camp rather than the performance one. Not exactly chasing the sports market with that one. If anything it might be Panasonic but they seem tied to Leica pretty hard for now.
I wouldn't be holding my breath.
Gordon
Frankly I could care as I am happy with SL3 and all other Leica gear. Some files from the Fuji just have a special look to me and it is hard to put a finger on , but I like them.
rscheffler wrote:
The S system also had leaf shutter options. I don't think it's the ideal solution. Leaf shutter adds size, complexity and expense to each lens. IMO the better technological solution is what Sony and Nikon have done with their stacked sensor cameras. I believe the Nikon Z8/Z9 don't even have mechanical shutter options available, proving the technology exists now to virtually eliminate need for a mechanical shutter. It would be interesting to know the reasons why Leica didn't opt for a stacked sensor in the SL3 as unfortunately every non-stacked Sony FF sensor option appears to have significant challenges with rolling shutter. ...Show more → flash wrote:
Because the SL3 IQ is measurably better. More resolution and more DR. Right now the SL3 and M11 are as close to miniMF as 24x36 can get.
If people don't like it then go buy an A1 for fucks sake. You don't see this whinging about the A7R5 or A7CR. The SL3 has issues. This isn't one of them.
I have an A1 and I;m glad my SL3 doesn't have that sensor.
Gordon
Like you said elsewhere, it must be a use-it-to-see-it situation because that measurably better sensor in the a7R5/SL3 compared to the a1, in respect to DR, sure doesn't look like much of an advantage on paper (at Photons to Photos).
Here's a link describing the project and process:
