Canon R7 rolling shutter

[BobGilbody]

I tried to make sense of a lot of information I found on the net about rolling shutter.
I found that the sensor read out speed for the R7 is 31.3 ms.
I hope this is not non-sense.
That means to me that it takes 1/300 [ sorry 1/30 ] of a second [ near enough ] for the frame to be read.
On the R7 I have a choice of mechanical focal plane shutter, electronic front curtain, or electronic shutter. I have been using electronic front curtain but am quite happy to use mechanical focal plane if it matters.
The following photo is not much good but it is a good example of rolling shutter.

Untitled by Bob Gilbody, on Flickr

205mm f5.6 1/800sec ISO800

I don't think I was panning.

The longer bit of twig in the bird's beak is about 1/10 of the vertical frame.

My understanding is therefore that the time between recording the top of the stick and the bottom is 1/3000 sec [ should be 1/300 sec ].

Am I making sense this far?

It seems to me that if I crop the image so that the bird fills the frame it takes about 1/1000 sec [ 1/100 sec ] to read the frame as cropped.

Maybe that is why first curtain electronic is ok most of the time, mechanical is ok and if there were global on normal type cameras that would be better.

I hope this is confusing because I would not wish to be out of step with the rest of the net experts.


I got it wrong. Not 1/300 sec but 1/30th. Getting old.

Forgot to say in mechanical shutter mode it is about 4 ms to read the full frame. Is it? Seems strange to me.


I am becoming very confused. How is it possible to take photos using a speedlite? Does the little pixel thingies get the light and hold on to it?

[pnodrog]

4ms readout mechanical shutter vs 31.3ms electronic shutter readout is interesting. For the mechanical shutter the pixel wells will all have been cleared of charge prior to the exposure. With electronic shutter the removal of well charge probably has to be synchronized with the exposure time and readout rate so it all takes longer. I should try and read up on it before making such speculations....interesting homework for me.

[BobGilbody]

Hello Paul,
I'm begining to think that maybe ms is millionth of a second rather than one thousandth. I got the info off https://www.youtube.com/watch?v=HrT93f5yq9E
at about 9min 20seconds in.
I tried to find more info but failed. I have taken photos using flash to freeze motion but that was with the 5Ds mechanical shutter. All the same the flash duration hits all pixels at the same time so read out of 1/250 sec just does not make sense to me.

PS: why would using a focal plane mechanical shutter make a change to 'read out time' ? Again it does not make sense to me.

A photo of nos. 1 daughter moving the speedlite manual as fast as she could, and a flowing tap, taken using 600EX-RT and Canon 5Ds:

Untitled by Bob Gilbody, on Flickr

Untitled by Bob Gilbody, on Flickr

[pnodrog]

A millisecond (ms) is 1/1000 of a second. Due to the fact that there seems to be no problem with electronic front shutter it seems the time to empty the wells of charge is not a major issue. So far on a quick google I have only found simple explanations of electronic shutters rather than full technical details of the timing. I will do a bit more research.

P.S. Using electronic front curtain should not cause any significantly greater rolling shutter effect than a mechanical shutter.

[pnodrog]

Just found some info. Not finished reading it yet but I think it is worthwhile drawing it to your attention.

I have just realised the 4 ms is a typical shutter travel time not the readout rate. It is the fastest "freeze time" that a such a mechanical shutter can achieve. The read of the sensor starts once the shutter is closed (or closing?) and will take a similar time to the electronic shutter.

[BobGilbody]

Thanks for the 'info' Paul. I must be missing something because I don't understand 'pseudo global shutter'. It looks similar to diagrams used to explain 'flash synch shutter time'.
Another dumb question: how does the gizmo setup 'know' the difference between bright sunlight and the flash from a speedlite? Maybe the answer is that sunlight is sort of continuous but the flash discharge is of limited duration. I don't understand a lot of the tech stuff that is around these days.
I believe Duade, the bloke that does the youtube stuff, is ok. However I'm not so sure about the intensions of camera manufacturers. The Canon R7 is an amazing bit of tech. If I were to be ultra cynical I would say that Canon shot themselves in the foot. The engineering in cameras is complicated but the engineers don't determine corporate strategy. In order to make money it makes sense to get people to buy the mk 2, then the mk 3, then the mk whatever. I remember in the late 60's Texas Instruments sold calculators that had different capabilities for different prices. The guts of the calculator were the same the only difference was the number of buttons. I would hope this is not the case with cameras.
BTW I knew that ms is millisec or 1/1000 of a second. I was thinking maybe Duade got it wrong or was fed bs.

Hopefully 'the truth is out there'

PS I just remembered reading about using two polarizing screens one of which could be flipped by 90 degrees using electrical charge. This was back in the early 90's when I was trying to understand how a photo of a bullet going through an apple could be obtained. If that 'simple' system was used as a shutter in a modern camera body it would be the same as using sunlight to mimic a speedlite, only better.

[DanK]

here's an explanation. the key part is in the section on electronic shutters.

https://johnplatt.com.au/electronic-...ter-explained/

[BobGilbody]

Thanks for the info Dan.

It seems I can have 30 frames per second if I accept rolling shutter effect.
I can have 15 fps if I want to minimise rolling shutter.

The R7 has a 'shutter life' of about 200000 actuations.
If I use mechanical shutter either as full mechanical or part mechanical in electronic front curtain mode and use the machine gun technique to chase the illusive good shot, I might average out at 15 frames per click. If that ends up correct then on average the shutter will fail after about 13500 'actuations'. That figure is not high. 200000 I can die happy, 13500 and I might end up with a camera with no shutter. In all probability the Canon R7 Mk 2 will be available before the shutter fails me and I can toss the R7 I have with its worn out shutter into the bin.
I hope I have got my sums wrong.

[BobGilbody]

I just came across this in the net:
https://www.youtube.com/watch?v=fVRyHBXIwqQ

I found it very interesting.
At 9min 52 seconds in it mentions the Rapatronic Camera. This uses the two polarized screen system I remember from thirty years ago.

[BobGilbody]

Hello Paul,
This morning I found this:
https://www.google.com.au/search?q=c...id:wt9d-cI4Ujk

It is a better explanation than the one I tried to do in my first post. However the two cameras compared are the R6 and the R7.
The R6 is full frame 24.2 MP.

It seems to me that the 'answer' lies in the characteristics of the pixels and how they transfer information. In your first post you mention the pixels. I know nothing about how they work, but I think you are on the right track. I don't want to chase that line of thought because I would have to learn too much to start. If you chose to I would be most interested in anything you find.

Bob

[Astro]

The link below takes you to a Canon released explanation of shutter modes

https://www.eos-magazine.com/article...ter-modes.html

[BobGilbody]

Thanks James.
Excellent presentation of how to choose which shutter mode to use. The presentation explains a great deal; I found the section on bokeh interesting.
However it did not address the problem fully.
Rolling shutter is not a problem that was introduced by the move from focal plane shutter to electronic shutter. The problem arose from leaf shutter to focal plane shutter. Back when I was using the Pentax 645 it was possible to obtain a Pentax lens that had a leaf shutter. This was of importance for high flash synch shutter speeds.
Focal plane shutters allow high shutter speeds but do have the introduced problem of rolling shutter and flash synch limitation.
Electronic shutters allow for high frame rate capture but have rolling shutter and flash synch speed problems.
All this leads to inherent problems associated with how pixels functions. As I mentioned above I am not keen to 'learn' all that would be required about pixel functions.
It seems to me that the problem is 'leaf shutter' to 'focal plane shutter'. Whether the focal plane shutter is achieved mechanically or electrically the result is the same. Mechanical might not be as bad as electronic but the problem of rolling shutter is still there.
This is why I am interested in the Rapatronic Camera shutter system. It appears to be an electronic leaf shutter.

Bob

[dje]

Hi Bob

The key parameter with rolling shutter is the curtain transit time (ie the time taken by the curtain to cross from one long side of the sensor to the other). For mechanical shutters, this is typically in the range 3.3 ms to 4 ms (1/300 s to 1/250 s). Electronic first curtain is much the same. All the first curtain has to do is activate the pixels row by row and this can be done very quickly.

With full electronic shutter, transit times are somewhat higher. (Ie slower). It varies considerably between camera models but I’ve seen figures quoted from 1/25 s to 1/150th sec. The latter is the exception at present and was measured by Jim Kasson on a Sony A1. The transit times are slower because the 2nd curtain has to read out the pixel values line by line as it goes. The readout speed depends on the sensor technology and the MPixels. With a mechanical shutter, this time constraint is not placed on the 2nd curtain.

The result is that at present, full electronic shutters are more prone to rolling shutter distortions.

Dave

[BobGilbody]

Hello Dave,
What you say is correct. If you consider what happens when electronic flash is employed to 'freeze the moment' it is seen that the 'empty pixel' gets its light then there is no more light to give problems. I realise that the 'flash' is not instantaneous. First or second curtain 'flash' gives different end images. Car lights are a secondary source of light to the flash. As you say, electronic shutters are at present prone to 'rolling' shutter, as to a lesser extent is a focal plane shutter. Leaf shutters are different in their operation but limited in their minimal duration.
My question now is simply this:
The rapatronic shutter technology is near enough 80 years old. It gives the benefit of the leaf shutter with very short exposure times. Is it possible to build a DSLR with a 'rapatronic shutter' between the mirror and the sensor small enough to give a hand held DSLR camera?

All the best Bob

[pnodrog]

Hi Bob

The key parameter with rolling shutter is the curtain transit time (ie the time taken by the curtain to cross from one long side of the sensor to the other). For mechanical shutters, this is typically in the range 3.3 ms to 4 ms (1/300 s to 1/250 s). Electronic first curtain is much the same. All the first curtain has to do is activate the pixels row by row and this can be done very quickly.

With full electronic shutter, transit times are somewhat higher. (Ie slower). It varies considerably between camera models but I’ve seen figures quoted from 1/25 s to 1/150th sec. The latter is the exception at present and was measured by Jim Kasson on a Sony A1. The transit times are slower because the 2nd curtain has to read out the pixel values line by line as it goes. The readout speed depends on the sensor technology and the MPixels. With a mechanical shutter, this time constraint is not placed on the 2nd curtain.

The result is that at present, full electronic shutters are more prone to rolling shutter distortions.

Dave

I think you mean "discharge, clear or empty" the pixel wells. If we could "activate" and therefore deactivate the pixels quickly then there would not be a problem. It's the reading time that mucks it all up.

[dje]

I think you mean "discharge, clear or empty" the pixel wells. If we could "activate" and therefore deactivate the pixels quickly then there would not be a problem. It's the reading time that mucks it all up.

Yes Paul I think you are correct.

Dave

[dje]

Hello Dave,
What you say is correct. If you consider what happens when electronic flash is employed to 'freeze the moment' it is seen that the 'empty pixel' gets its light then there is no more light to give problems. I realise that the 'flash' is not instantaneous. First or second curtain 'flash' gives different end images. Car lights are a secondary source of light to the flash. As you say, electronic shutters are at present prone to 'rolling' shutter, as to a lesser extent is a focal plane shutter. Leaf shutters are different in their operation but limited in their minimal duration.
My question now is simply this:
The rapatronic shutter technology is near enough 80 years old. It gives the benefit of the leaf shutter with very short exposure times. Is it possible to build a DSLR with a 'rapatronic shutter' between the mirror and the sensor small enough to give a hand held DSLR camera?

All the best Bob

Sorry Bob I know nothing about the rapatronic shutter.

Dave

[pnodrog]

The rapatronic shutter is exceptionally fast but I doubt could be easily employed in portable battery powered cameras. As it is based on two polarizing filters with their polarization set at right angles to each other and by their very nature even when aligned they attenuate the light significantly. At the minimum at least a stop, probably much more. At low light levels they would be a huge hindrance. I'm not sure how much power the intervening Kerr cell takes but for the exposure times used in normal photography it's probably well beyond what could be battery powered.

I think eventually the sensors themselves or an element just above them (LCD technology?) will have some way of stopping the light sufficiently that it's effect on the exposure is negligible. No doubt there are people working on it now.

[BobGilbody]

Thanks Paul,
I'm sure they are working on it.

All the best Bob