Shot this image to test for rolling shutter. ISO 3200, f11, 1/50.
Shot this image to test for rolling shutter. ISO 3200, f11, 1/50.
2 Member(s) liked this post
It would be better to call this 'SB' (for scanning back) rather than 'ES' - but obviously that's a harder feature to sell.
This isn't really an electronic shutter in the sense that I imagine it - i.e. one that can read all the data from the back in something like 1/250" or less.
As a scanning back, the IQ3-100 is a bit quicker than Betterlight's 80MP model - though the Betterlight records true RGB data - and a bit slower than Seitz's 6x17 Digital Panoramic Camera, which can pull a 160MP image composed of true RGB data in less than a second.
This "Electronic Shutter" is a great feature to have and will be suitable for many situations. However, anything with movements especially along the X-axis and a global shutter is the right choice. ES is another convenience tool for the photographers and offers something that Copal shutter can not. However, we have to understand when its usage can result in the "rolling shutter effect" and make informed choice accordingly.
IQ3 100 H • Cambo 1600 • Rodenstock 23,32,50,90 • Zeiss 350SA
3 Member(s) liked this post
I can't really see the comparison to a Seitz 6 x 17, as the normal exposure is between 2 and 3 seconds in normal light as I understand it. A true scanning back.
The ES on the IQ100 operates between any selected shutter speed, i.e. 1 hour to 1/4000 at any ISO selected on the camera, more similar to the ES on a Fuji X-T1.
No doubt the images from a 40K or so Seitz are going to be excellent, but any movement at all, will be blurred, as you are limited to around 2 seconds for any exposure, where as the ES on the Phase easily stops motion of leaves, grass, as it can be used in a full range of shutter speeds. The issue of rolling shutter is nothing really that new, shows up in many DSLR's that shoot video even the latest from Canon still have the issue.
You can hand hold the XF and use the ES also, but shutter speeds are limited to beyond 1/250 and results vary. But when on tripod I have not had any rolling shutter problems within the range of normal landscape images for me, even yesterday with a pretty strong wing blowing shutter speeds of 1/250 and 1/500 worked fine. But the motion of small objects, such as leaves, grass, plants etc don't seem to have issues, obviously a car or any larger object does. Just like with Fuji's X-T1 or now X-T2.
For the tech camera users, this is a huge move forward, as you are now free of the Copal shutter, and it's totally non-accurate shutter speeds, and now can also gain 1/3 or 1/2 stops which to me is a nice feature, but just being able to get true 1/30, 1/60 1/125 and 1/250 is a big plus. Not to mention you are free from the other side of the equation, an electronic shutter for each tech camera lens, that needs it's own battery, and controller and most are limited to 1/250th as the fastest shutter speed. The the cost and additional weight, and mass of an ES on a tech lens has always made such a solution non-tenable for me.
I also agree with what Paul has said...... this is a very welcome feature for me. It will probably taker the next iteration of sensors for the incorporation of a complete on/off shutter type effect. In the mean time this step is very welcome.....
The electronic shutter is simply an electronic version of the focal plane shutter. Because it is limited by readout speed it travels slowly.
I don't think a global shutter is possible with today's technology, so I don't think we will see that feature on any large sensors any soon.
But, I may be wrong, of course.
Portfolio: http://echophoto.smugmug.com1 Member(s) liked this post
As a tech cam user shooting mostly architecture I welcome the move, as I don't care much about the jelly effect. To me the real drawback is that 1" flash sync, meaning I couldn't use the ES for interiors, unless relying on available or continuous light.
For you guys shooting mostly landscapes this is a great feature.
It's actually closely analogous. The fundamental difference between the Seitz 6x17 and the IQ3 100 electronic shutter is that the IQ3 100 is exposing different pixels over the duration of the "scan" of the electronic shutter, whereas the D3 scanning back in the Seitz is moving a block of pixels across the frame.
Exposure time at the pixel level on the Seitz can range from 1/2000th of a second up to 10 seconds. It was an impressive piece of kit in its day, but I think it's been left behind somewhat (claimed DR is only 11 stops, for example).
1 Member(s) thanked for this post
Obviously, the technology behind this is way over my head, and I can learn more from Erik and others, with the engineering background and always welcome the knowledge.
From my own perspective, as a photographer, I see this feature (ES) as a huge advantage, and not sure why Phase choose to hold it away from Kina.
Simple example, of where I have immediately found 100% improvement. Phase/Schneider 240LS lens, an amazing piece of glass, but heavy and it has not tripod collar. With the XF, I was finding that even with a remote release, shutter speeds of 1/8 to 1/30 could possible contain blur, slight blur, but easily noticed especially on distant subject matter (1 mile to 6 miles out). I was able to get a better result by adding in the RRS large lens stabilization tool. But that device is both heavy, and takes time to setup. Plus, it totally limits me to a horizontal capture and many times I want to go vertical. XF and Vertical just don't mix well as you can visually see the camera move when the mirror goes up, so vibration is there (note I was always in the Vibration reduction mode on the XF). Just touching the back to fire the camera adds movement and I have had mixed results with the 2 sec delay.
Yesterday, I took the same setup out and shot at 1/15 to 1/30 with the ES, you will get a full shutter action on the first frame and after than all you hear is the beep. I did use the remote release. The images were tack sharp, none with the slight blur, none showed any rolling movement problems (subject matter was outdoors, trees, leaves plants). Personally I can't see any delay either between firing the normal shutter and using the ES, as 1/15 and 1/30 and 1/60 all seem to take the same amount of time. I understand that the ES is reading by lines thus slower capturing the data slower than if fired by focal, but I am not seeing any issues for my normal photography and the benefits for long glass are 100% improved.
Overall a very positive step forward, and as Erik states, I am sure the next round will be a global ES, (not sure what Fuji has implemented on their X cameras, if it's a global or not, but it will shoot up to 1/32000, but has immediate problem with indoor lighting creating banding so I am assuming it's recording similar to how Phase One is?)
Wow, using the ES seems to be complicated especially with respect to the Dark Frame.
Thank you guys for bringing clarity to this wonderful new tool.
Although I am not a Phase One owner, I find the technology fascinating
www.howardcubell.com1 Member(s) liked this post
I thought about this some more and I did some more testing today, with about a 5 to 10mph wind.
Tree leaves at 1/250 appear fine, but if I slow down to a 1/40 to 1/25, I will see blur but that is standard motion blur. Clouds so far are not an issue but on a very windy day, where you can clearly see movement that might be a different issue, but again a cloud is so non uniform, not sure it will matter.
If anything it adds to water work for me, as I prefer a slow shutter, say 1 to 5 sec. I have not tried to stop water at a faster shutter speed, curious what that will look like.
The other way to look at it for me is, tree leaves/grass/plants that are stationary are moving but not moving across the shutter, thus their slight motion doesn't seem to cross the boundary and create the rolling shutter issue. If a entire tree limb is moving in high wind, it's much more possible to have an issue.
Larger objects, cars, planes, trains, animals etc. more than likely won't work since they are moving faster and their motion is not from a static base, (like a tree leaf), thus they are going to have the classic rolling shutter effect. And I just realized during this test, that a falling leaf, can create a bit of problem if it happens to cross the sensor close enough up as it's rate of fall is moving it over a much greater portion of the sensor again creating the rolling look.
It's a good problem to have I guess, but you do need to check your shots if you are seeing a lot of wind movement.
It should work just fine, but limits DR to 12 bits. Readout is something like 10X faster than the 100 MP Phase One back.
There is little reason to use silent shutter on the A7rII as electronic first shutter curtain takes care of the vibration issues anyway.
On the other hand, it is totally silent and has no mechanical wear.
www.howardcubell.com2 Member(s) liked this post
Using it on the XF also has some advantages (as mentioned by Paul), and on that body camera handles the dark frame. I’ve been thinking of buying a used 300, which most of us have given up on because it is nearly impossible to get a sharp shot because of shutter vibration. I’ve heard the EFC has helped that, but now it seems very viable in some circumstances using ES.
- - - Updated - - -
Images of, say, 20-60 secs each for star trails would be interesting with ES. Depending on focal length and the angle of movement across the frame, could the star movement interact with the sweep of the sensor read-out?
Ed Hurst, www.spiffingpics.com
The scan time will be around 0.6 seconds whatever the exposure time. So, if you expose for 20-40 seconds the back will do a sweeping reset that takes about 0.6, after that all of the sensor collects light, the electronic shutter is fully open. After 19-59 seconds a sweeping readout starts. So in essence that the ES will work just as a focal plane shutter.
With Electronic First Curtain there is a a second curtain that terminates exposure, that can sweep reasonably fast. Reset can be very fast. So it will still operate as an FP shutter, but sweep time will be decided by the travel time of the second curtain.
It is also possible to do a global reset. This is used on the Hasselblad X1D. The exposure starts with the leaf shutter open and the sensor just does a global reset. Exposure is terminated by the closure of the leaf shutter.
Reset is easy, it is just about connect the photodiode to signal ground. Readout is a different process. A reference voltage is ramped up and compared with photodiode voltage. The time to reach equal voltage is the digital output. This ramping is a bit slow and it is done for each row of the sensor. So momentous global readout cannot be made with the technology used now. It was possible on interline CCDs, that could pop the charge into a storage position and readout in post.
I am no expert in this area, it is just my understanding how those things work.
Currently Sony's Exmor technology is up to the point where they can utilize BSI to push all on-sensor circuitry behind the photosensitive area, which means the wiring no longer occupies the same area as the sensor wells, allowing for larger pixels. But it may also mean that a global electronic shutter can now be implemented without significant drawbacks, as previously it came at the cost of reduced DR and higher noise. This is going to be the development to watch out for.
Yes that is correct, but wiring occupies a very small area on modern sensors anyway. Sony now also has stacked sensors but only in small sizes. The RX10 can do a lot of magic with new sensor technology.
Regarding global shutters, I guess it is something like 2-3 sensor generations ahead. I would suggest that things take time.
We had a fantastic journey with digital technology and it is not over yet, but things are slowing down a bit…
I was impressed with the testing I did. Before ES, I pretty much had to limit my exposures with the 240LS to 1/45 or greater (on tripod), as the slower speeds, many times would show slight blur. If the camera was vertical (which I prefer many times) problems were worse. Use of the RRS long lens support fixed the problem most of the time, but that setup:
a: takes time to setup
b: won't easily allow for a vertical orientation of the camera
My use always had the Vibration Reduction ON, so I assume that the EFC was in effect, but personally I never really understood the EFC on the IQ100. On my K1 or D810, it's very clear when I am using the EFC as you can hear a difference, not with the XF and IQ100, at least for me.
Once I figured out ES on the XF, my first thoughts were just how it might improve the performance of the 240LS with the IQ100, and the results were impressive. 1/8 to 1/30 shutter speeds are now quite useable. I briefly tried the 300 Mamiya but gave up on it due to the same vibration issues, (but worse than the 240LS), but I figure that the ES would make use of the 300 possible, with slower shutter speeds, note I had to use 1/125 or higher with the 300, again on tripod.
I have also tried the older 210mm and 75-150 and both seem work very well with the ES.
I have to say again, the ES is a huge advantage for Phase One, for both long lens work and tech camera work.
I wanted to find out if the ES would distort moving water, so I did a series of identical exposures using the FP shutter and then the ES. I did short exposures at high ISOs and long exposures with an ND filter, both in landscape and portrait orientation. (I didn't think the long exposures would show any difference but....)
Camera on firm tripod, VR turned on, triggered with my iPhone. As I suspected, there are no distinguishable differences in the files at all. Of course, I was using the SK55 mm which would not be prone to vibration like, say, the 240 mm, but that wasn't the purpose of my test. Tomorrow I'll "do" the 240 mm.
www.billcaulfeild-browne.ca2 Member(s) thanked for this post2 Member(s) liked this post
Thanks, Paul. It's in the little gorge at the base of Inglis Falls in Owen Sound, Ontario. Very very slippery trail to get there!
A test of the ES versus the LS on the XF camera using the SK240 mm lens.
First shot to set the scene. Tripod, cube, 1/4 sec at f11, vibration delay, Capture Pilot.
Next is the LS shutter version at 100%
Next is the ES version.
It's hard to see the difference on a dumbed-down 1200 px jpg, but the resolution and micro contrast is very obvious in the RAW file. The ES is definitely worth using at slow shutter speeds. I also did a series at <100th/sec but could not see any difference.
www.billcaulfeild-browne.ca1 Member(s) thanked for this post3 Member(s) liked this post
Thanks for sharing. On most natural subjects I would not expect a lot of distortion from the slow moving electronic shutter. Shoot some racing cars from tripod without panning and there will be some distortion…
Nice images, by the way!
Portfolio: http://echophoto.smugmug.com1 Member(s) thanked for this post
Impressive. Really obvious if you load the two JPEGs into a stack and then flip the second layer on and off.
Were these processed identically though? It's odd that the exposure on the leaves in the two images is practically identical, but on the moss/ferns there is a significant difference.
1 Member(s) thanked for this post
They were processed identically, C1 defaults, but I had to alter the exposure very slightly on one of them to match the other. This, I suspect, is because the ES is more accurate than the LS.
Indeed, I saw this in many shots - though it is less evident when comparing the ES with FPS. As you know, LS effective speed is slightly affected by the aperture at which you're shooting. And probably very cold weather might slow the LS, too.
www.billcaulfeild-browne.ca1 Member(s) thanked for this post