Rolling shutter full frame vs crop camera

Maybe someone can help me if I've got something wrong.

Do we have to adjust the read out speed for the crop factor to get a real comparison?

For example the new S1RII needs 20 ms in 12 bit to read the hole sensor in lines. The OM1 does need about 8 Ms for it's m43 sensor. But it's about half the heights. If I crop out a half heights image of the S1RII, that part should have took half the time of the read out, that means 10 Ms. So the rolling shutter should be not much more noticeable.

Of course, you only would get 11 Mpix. But even a APS-C Crop would take only about 13-14 Ms, what would be about the same as a GH7 or G9II, and would get you about 20 Mpix.

Or did I got something wrong?

I think you've got it right.

An additional factor is the lens. If an image detail is in motion on the sensor while the sensor is scanning, it will not be rendered properly.

So, a long telephoto will have a lot of details that are almost impossible to keep still while the sensor is scanning, while a wide angle lens will be easy.

Practically speaking, I have a Sigma fpL (60mp, electronic shutter only) and seldom notice rolling shutter. But, I don't shoot sports or other action.

I think that rolling shutter is not related to sensor size but only to number of vertical pixels, bit depth, sensor type and generation / optimization.

For example Panasonic S1R, S1RII, Fujifilm X-T5 are all non-stacked sensors in the 40-47 MP range and their rolling shutter at 14 bit is 36-39ms very similar.

olympus OM-1 is quicker due to stacked sensor, not due to smaller sensor

grasmuc said:

I think that rolling shutter is not related to sensor size but only to number of vertical pixels, bit depth, sensor type and generation / optimization.

For example Panasonic S1R, S1RII, Fujifilm X-T5 are all non-stacked sensors in the 40-47 MP range and their rolling shutter at 14 bit is 36-39ms very similar.

olympus OM-1 is quicker due to stacked sensor, not due to smaller sensor

Click to expand...

But the point is, a OM-1 isn't much quicker at all in practice. Only a little bit quicker. Despite it's a stacked Sensor. The OM1 is only 20 Mpix. If I crop the S1RII to about 20 Mpix, thet crops rolling shutter is not that much worse compared to the 20 Mpix image out of a OK1 and about the same compared to the 25 Mpix image of a G9II or GH7.

So I asked my self if a OM-1 or G9II would even make any sense for me as a addition for sports and wildlife, if I also could use a wider field of view in a S1RII and crop later and would get similar results in terms of rolling shutter. Even the lenses wouldn't be necessary smaller as a 100-400 Sigma or 150-600 sigma are basically the same size as the OM 100-400 or 150-600.

QM said:

grasmuc said:

I think that rolling shutter is not related to sensor size but only to number of vertical pixels, bit depth, sensor type and generation / optimization.

For example Panasonic S1R, S1RII, Fujifilm X-T5 are all non-stacked sensors in the 40-47 MP range and their rolling shutter at 14 bit is 36-39ms very similar.

olympus OM-1 is quicker due to stacked sensor, not due to smaller sensor

Click to expand...

But the point is, a OM-1 isn't much quicker at all in practice. Only a little bit quicker. Despite it's a stacked Sensor. The OM1 is only 20 Mpix. If I crop the S1RII to about 20 Mpix, thet crops rolling shutter is not that much worse compared to the 20 Mpix image out of a OK1 and about the same compared to the 25 Mpix image of a G9II or GH7.

So I asked my self if a OM-1 or G9II would even make any sense for me as a addition for sports and wildlife, if I also could use a wider field of view in a S1RII and crop later and would get similar results in terms of rolling shutter. Even the lenses wouldn't be necessary smaller as a 100-400 Sigma or 150-600 sigma are basically the same size as the OM 100-400 or 150-600.

Click to expand...

Maybe I'm misunderstanding the scenario, but I don't think cropping after the shot changes anything. Suppose you have a fence post that appears at an angle due to RS. If you crop that image, the post will still be distorted at the same angle.

If you actually shoot in a crop mode then yes, the scan of that area on the sensor will be quicker.

Tim200 said:

QM said:

grasmuc said:

I think that rolling shutter is not related to sensor size but only to number of vertical pixels, bit depth, sensor type and generation / optimization.

For example Panasonic S1R, S1RII, Fujifilm X-T5 are all non-stacked sensors in the 40-47 MP range and their rolling shutter at 14 bit is 36-39ms very similar.

olympus OM-1 is quicker due to stacked sensor, not due to smaller sensor

Click to expand...

But the point is, a OM-1 isn't much quicker at all in practice. Only a little bit quicker. Despite it's a stacked Sensor. The OM1 is only 20 Mpix. If I crop the S1RII to about 20 Mpix, thet crops rolling shutter is not that much worse compared to the 20 Mpix image out of a OK1 and about the same compared to the 25 Mpix image of a G9II or GH7.

So I asked my self if a OM-1 or G9II would even make any sense for me as a addition for sports and wildlife, if I also could use a wider field of view in a S1RII and crop later and would get similar results in terms of rolling shutter. Even the lenses wouldn't be necessary smaller as a 100-400 Sigma or 150-600 sigma are basically the same size as the OM 100-400 or 150-600.

Click to expand...

Maybe I'm misunderstanding the scenario, but I don't think cropping after the shot changes anything. Suppose you have a fence post that appears at an angle due to RS. If you crop that image, the post will still be distorted at the same angle.

If you actually shoot in a crop mode then yes, the scan of that area on the sensor will be quicker.

Click to expand...

It shouldn't make a difference to crop in post, shooting in crop mode or shooting with a crop camera. The read out time for the fraction of the frame, the crop, is the same. Therefore the distortion should be the same.

QM said:

Tim200 said:

QM said:

grasmuc said:

I think that rolling shutter is not related to sensor size but only to number of vertical pixels, bit depth, sensor type and generation / optimization.

For example Panasonic S1R, S1RII, Fujifilm X-T5 are all non-stacked sensors in the 40-47 MP range and their rolling shutter at 14 bit is 36-39ms very similar.

olympus OM-1 is quicker due to stacked sensor, not due to smaller sensor

Click to expand...

But the point is, a OM-1 isn't much quicker at all in practice. Only a little bit quicker. Despite it's a stacked Sensor. The OM1 is only 20 Mpix. If I crop the S1RII to about 20 Mpix, thet crops rolling shutter is not that much worse compared to the 20 Mpix image out of a OK1 and about the same compared to the 25 Mpix image of a G9II or GH7.

So I asked my self if a OM-1 or G9II would even make any sense for me as a addition for sports and wildlife, if I also could use a wider field of view in a S1RII and crop later and would get similar results in terms of rolling shutter. Even the lenses wouldn't be necessary smaller as a 100-400 Sigma or 150-600 sigma are basically the same size as the OM 100-400 or 150-600.

Click to expand...

Maybe I'm misunderstanding the scenario, but I don't think cropping after the shot changes anything. Suppose you have a fence post that appears at an angle due to RS. If you crop that image, the post will still be distorted at the same angle.

If you actually shoot in a crop mode then yes, the scan of that area on the sensor will be quicker.

Click to expand...

It shouldn't make a difference to crop in post, shooting in crop mode or shooting with a crop camera. The read out time for the fraction of the frame, the crop, is the same. Therefore the distortion should be the same.

Click to expand...

Ok, I see in theory what you're saying. The smaller the subjects of interest in the frame, the less they are impacted.

Tim200 said:

QM said:

Tim200 said:

QM said:

grasmuc said:

I think that rolling shutter is not related to sensor size but only to number of vertical pixels, bit depth, sensor type and generation / optimization.

For example Panasonic S1R, S1RII, Fujifilm X-T5 are all non-stacked sensors in the 40-47 MP range and their rolling shutter at 14 bit is 36-39ms very similar.

olympus OM-1 is quicker due to stacked sensor, not due to smaller sensor

Click to expand...

But the point is, a OM-1 isn't much quicker at all in practice. Only a little bit quicker. Despite it's a stacked Sensor. The OM1 is only 20 Mpix. If I crop the S1RII to about 20 Mpix, thet crops rolling shutter is not that much worse compared to the 20 Mpix image out of a OK1 and about the same compared to the 25 Mpix image of a G9II or GH7.

So I asked my self if a OM-1 or G9II would even make any sense for me as a addition for sports and wildlife, if I also could use a wider field of view in a S1RII and crop later and would get similar results in terms of rolling shutter. Even the lenses wouldn't be necessary smaller as a 100-400 Sigma or 150-600 sigma are basically the same size as the OM 100-400 or 150-600.

Click to expand...

Maybe I'm misunderstanding the scenario, but I don't think cropping after the shot changes anything. Suppose you have a fence post that appears at an angle due to RS. If you crop that image, the post will still be distorted at the same angle.

If you actually shoot in a crop mode then yes, the scan of that area on the sensor will be quicker.

Click to expand...

It shouldn't make a difference to crop in post, shooting in crop mode or shooting with a crop camera. The read out time for the fraction of the frame, the crop, is the same. Therefore the distortion should be the same.

Click to expand...

Ok, I see in theory what you're saying. The smaller the subjects of interest in the frame, the less they are impacted.

Click to expand...

Yes. You can determine the scan time from the height of the object in pixels. Then you can then determine the amount of skew or warp if you know how fast that object is moving in pix/sec (easier said than done).

So, all you really need to do is to zoom out a bit. No need to use crop mode specifically. But again, before worrying about that, just spend some time understanding how the camera responds naturally.

--
http://georgehudetzphotography.smugmug.com/
Capture One LUMIX FF feature request thread: https://www.dpreview.com/forums/thread/4658107#forum-post-66298057

QM said:

grasmuc said:

I think that rolling shutter is not related to sensor size but only to number of vertical pixels, bit depth, sensor type and generation / optimization.

For example Panasonic S1R, S1RII, Fujifilm X-T5 are all non-stacked sensors in the 40-47 MP range and their rolling shutter at 14 bit is 36-39ms very similar.

olympus OM-1 is quicker due to stacked sensor, not due to smaller sensor

Click to expand...

But the point is, a OM-1 isn't much quicker at all in practice. Only a little bit quicker. Despite it's a stacked Sensor. The OM1 is only 20 Mpix. If I crop the S1RII to about 20 Mpix, thet crops rolling shutter is not that much worse compared to the 20 Mpix image out of a OK1 and about the same compared to the 25 Mpix image of a G9II or GH7.

So I asked my self if a OM-1 or G9II would even make any sense for me as a addition for sports and wildlife, if I also could use a wider field of view in a S1RII and crop later and would get similar results in terms of rolling shutter. Even the lenses wouldn't be necessary smaller as a 100-400 Sigma or 150-600 sigma are basically the same size as the OM 100-400 or 150-600.

Click to expand...

Just a heads up, the OM-1/II and OM-3 is using a quad bayer sensor, the final 20MP output comes from 80 million photosites being read out. That is quite a lot of data to parse, and it does come around ~1/125 on the electronic shutter. Unfortunately no plain, just stacked regular bayer pattern m4/3 exists.


Please ignore the last marketing point of 35mm equiv. performance- they were raked with hot coals on that and there was really no reason to say it. It was at end of day them saying they improved the noise characteristics of the JPEGs- but even then not equal.

I haven't seen numbers for the GH7 and G9 II on their rolling shutter yet (are we talking about stills?), but it's somewhere inbetween the now "old" fast (a slower version exists, mainly for the non high performance bodies) 20MP sensor which is about 1/60s and the newer OM sensor.

You check out this blog which explores "small sensor" cameras, and the one I linked below is specific for the Olympus Pro Capture. You can see if it bothers you or not

https://smallsensorphotography.com/category/pro-capture

--
I like cameras, they're fun.

Ainisru said:

QM said:

grasmuc said:

I think that rolling shutter is not related to sensor size but only to number of vertical pixels, bit depth, sensor type and generation / optimization.

For example Panasonic S1R, S1RII, Fujifilm X-T5 are all non-stacked sensors in the 40-47 MP range and their rolling shutter at 14 bit is 36-39ms very similar.

olympus OM-1 is quicker due to stacked sensor, not due to smaller sensor

Click to expand...

But the point is, a OM-1 isn't much quicker at all in practice. Only a little bit quicker. Despite it's a stacked Sensor. The OM1 is only 20 Mpix. If I crop the S1RII to about 20 Mpix, thet crops rolling shutter is not that much worse compared to the 20 Mpix image out of a OK1 and about the same compared to the 25 Mpix image of a G9II or GH7.

So I asked my self if a OM-1 or G9II would even make any sense for me as a addition for sports and wildlife, if I also could use a wider field of view in a S1RII and crop later and would get similar results in terms of rolling shutter. Even the lenses wouldn't be necessary smaller as a 100-400 Sigma or 150-600 sigma are basically the same size as the OM 100-400 or 150-600.

Click to expand...

Just a heads up, the OM-1/II and OM-3 is using a quad bayer sensor, the final 20MP output comes from 80 million photosites being read out. That is quite a lot of data to parse, and it does come around ~1/125 on the electronic shutter. Unfortunately no plain, just stacked regular bayer pattern m4/3 exists.


Please ignore the last marketing point of 35mm equiv. performance- they were raked with hot coals on that and there was really no reason to say it. It was at end of day them saying they improved the noise characteristics of the JPEGs- but even then not equal.

Click to expand...

Yes, I got that.

Ainisru said:

Ainisru said:

I haven't seen numbers for the GH7 and G9 II on their rolling shutter yet (are we talking about stills?), but it's somewhere inbetween the now "old" fast (a slower version exists, mainly for the non high performance bodies) 20MP sensor which is about 1/60s and the newer OM sensor.

Click to expand...

Click to expand...

I also haven't seen numbers in stills mode. But it should be around 1/80s. An about 20 Mpix 3:2 crop from the S1RII should be at about 1/75s. So not much difference.

Ainisru said:

You check out this blog which explores "small sensor" cameras, and the one I linked below is specific for the Olympus Pro Capture. You can see if it bothers you or not

https://smallsensorphotography.com/category/pro-capture

--
I like cameras, they're fun.

Click to expand...