Cameras: Coming Full Circle

[charlie]

  The reason to have leaf shutters is, of course, to synchronise with flash at any shutter speed, and the reason to do that is so you can use flash to stop very fast action without turning off the lights. . 

Why would you do that? It makes much more sense to freeze action with short flash duration as opposed to shutter speed, then you can leave the lights on to boot. This is after all the point of making strobes with 1/80,000s flash duration.

In my opinion a better argument for leaf shutters is the ability to set your aperture to something in the f/2.8 - f/5.6 range and not being limited to f/11 or above when combining strobes and bright ambient light, with out the need for ND's. This is a great benefit for portrait work.

[JohnBrew]

Where can I read more about the sensor-heat issue. How serious is it?

Don't know. The people who have handled the camera at the intros remarked that the camera got hot thus raising the question of how long can you use it before it gets hot and long exposures, supposedly the camera is good for 60 minutes just like the H5D and H6D with the same sensor, could be a genuine concern.
getDPI has the most active MF forum I know of and several of the posters have handled the camera and actually used it (though no one has been able to put their own card in!) and have commented on it.
When I used an H5D-50c the longest exposure I tried was around 8 minutes and didn't notice any issues, but then the digital back has a much larger heat sink than is possible with the new mirrorless design.

[David H. Hartman]

There is a problem with very fast shutter speed and flash duration. With a shutterspeed of 1/2,000 and a T0.1 flash duration of 1/500 how much flash gets through? How efficient is the shutter at 1/2,000? I know about the issue of shutter efficiency at high shorter speeds but nothing specific. Certainly not with modern shutters.

Dave

[Michael Erlewine]

Don't know. The people who have handled the camera at the intros remarked that the camera got hot thus raising the question of how long can you use it before it gets hot and long exposures, supposedly the camera is good for 60 minutes just like the H5D and H6D with the same sensor, could be a genuine concern.
getDPI has the most active MF forum I know of and several of the posters have handled the camera and actually used it (though no one has been able to put their own card in!) and have commented on it.
When I used an H5D-50c the longest exposure I tried was around 8 minutes and didn't notice any issues, but then the digital back has a much larger heat sink than is possible with the new mirrorless design.

So, that's not even a rumor, but something you just thought? I can find no discussion of that and I asked someone who should know, and they never heard anything about the sensor being too hot. Thanks for clarifying where this came from.

[Les Olson]

Why would you do that? It makes much more sense to freeze action with short flash duration as opposed to shutter speed, then you can leave the lights on to boot. This is after all the point of making strobes with 1/80,000s flash duration.

When the flash is the only source of light it does not matter that the shutter is open for longer than the flash, because the exposure during the non-flash period is negligible.  But if the shutter is open for 1/200 sec and there is a significant amount of light for the whole period you get ghosts - partial images of the moving subject. 

[David H. Hartman]

The shutter can close faster than the flash duration so the full power of the flash is not used. The faster the shutter speed the greater the loss.

At full power the flash duration of an SB-800 is longer  than the shutter is open on a D800 at 1/320 and 1/250 so the tail of the flash is clipped. This causes a shading of the top of the frame. This problem should be gone by 1/200.

If a leaf shutter is too fast it will cut the flash exposure short but the loss will be uniform across the frame. If one is trying to "overpower the sun" this shutter v. flash duration can be a problem.

Dave

[Les Olson]

So, that's not even a rumor, but something you just thought? I can find no discussion of that and I asked someone who should know, and they never heard anything about the sensor being too hot. Thanks for clarifying where this came from.

Build-up of heat generated by the electronics is a problem for all sensors, because it creates noise.  The problem is greater for CMOS sensors, because, although their power consumption is less than CCD sensors, they have a lot of circuitry generating heat next to the photosites, so they have high dark noise and reset noise (you can read the technical details at http://micro.magnet.fsu.edu/primer/digitalimaging/cmosimagesensors.html).  Because heat is not lost from the surface but from the edges, the bigger the chip the harder the heat is to control, and, obviously, longer exposures  = more heat. 

In cameras where size and weight are not a consideration you can have heat sinks next to the sensor.  We don't know about the X1D.  So asking about heat management is perfectly reasonable.  The internet being what it is, a reasonable question may have been turned into a rumour that it is a problem, but it is not baseless.

[Michael Erlewine]

Build-up of heat generated by the electronics is a problem for all sensors, because it creates noise.  The problem is greater for CMOS sensors, because, although their power consumption is less than CCD sensors, they have a lot of circuitry generating heat next to the photosites, so they have high dark noise and reset noise (you can read the technical details at http://micro.magnet.fsu.edu/primer/digitalimaging/cmosimagesensors.html).  Because heat is not lost from the surface but from the edges, the bigger the chip the harder the heat is to control, and, obviously, longer exposures  = more heat. 

In cameras where size and weight are not a consideration you can have heat sinks next to the sensor.  We don't know about the X1D.  So asking about heat management is perfectly reasonable.  The internet being what it is, a reasonable question may have been turned into a rumour that it is a problem, but it is not baseless.

I see, so this is a generic problem for all CMOS sensors, rather than already a significant problem with the X1D, correct? I think this is how rumors start.

[Ilkka Nissilä]

If a leaf shutter is too fast it will cut the flash exposure short but the loss will be uniform across the frame. If one is trying to "overpower the sun" this shutter v. flash duration can be a problem.

Good point but there are flashes with fast durations and I don't use my speedlights at M1/1 and I'd rather use them at around 1/8 or at 1/4 at most, to avoid slow recycle times and overheating.

I have two battery powered Elinchrom Quadra Hybrids, with the action head at 100% flash energy from the main port you get t.5 time of 1/2800s (400Ws) and from the "fast" port 1/5800s. The fast port output by itself is 1/3 of 400Ws which is 133 Ws.  For outdoor photography with 1/200s sync, if it is a bright sunny day, and I'm shooting a group, using e.g. 200Ws or more, the flash can overheat and start missing shots after a while (there is no warning, it just won't fire until it has cooled down sufficiently). However, by using 1/500s or 1/1000s with a central shutter lens,  I could reduce the flash output significantly and probably avoid overheating, even if at 1/2000s it might not quite give me 400Ws. I don't usually need that much light but what I need/want is consistency.

With Nikon flashes,  At 1/4 maximum output if I am shooting outdoors with umbrella or softbox, 1/200s won't let me output enough light for a balanced exposure with sunlight; in fact I've used three speedlights (behind a shoot through umbrella) at around 1/2 and that was about right, but there can be overheating and missed shots. Again, by using a faster sync speed, I may be able to use just one speedlight, or run three at lower settings. SB-5000 flash duration (I assume t.5) at M1/2 is 1/1110s, at M1/4 it is 1/2580s, M1/8 1/5160s, so by using 1/4 or 1/8 I should be able to use 1/1000s which means each flash is quadrupled against the sun. The key is that by being able to use lower flash energy settings I get less heating and faster recycle times, and faster recycle means I can shoot at a fairly rapid rate, and get good expressions (since I don't want a forced smile, I need to get a number of shots of a group so that I can avoid closed eyes).  Often when photographing a family or a group of kids, the going can be wild, and so I want to get a number of shots before they disperse. There is no way I can monitor all their expressions and just fire once every five seconds, things change much faster than that. A single speedlight might not be able to light a big group shot against the sun with softbox or umbrella but it can do more with a leaf shutter than a fp shutter.

However, I can't afford to spend 17k€ to get a single camera and two lenses (Leica S would cost even more). So I will lug around my Quadra and sweat, and wait for it to cool if it overheats, and hope that there are clouds, because if there are clouds I might be able to get away with just a little light and the sun is not going to contribute so much to my personal heating or the flash's.

[Ilkka Nissilä]

I see, so this is a generic problem for all CMOS sensors, rather than already a significant problem with the X1D, correct? I think this is how rumors start.

I think sensor heat can be an issue if the body is closed and very compact. I believe that the sensor shaking image stabilization may be partly contributing to heating experienced in some cameras as if there is a mechanical system that moves the sensor, it can be more difficult to disperse of the heat. If the sensor is fixed then the heat should dissipate through fixed heat conducting structures to the body, assuming a good design. 

However, I've never used a camera which heated appreciably in use. I've had flashes heat though, on sunny days. Not so much a problem in Finland except during peak wedding season.

[Michael Erlewine]

I think sensor heat can be an issue if the body is closed and very compact. I believe that the sensor shaking image stabilization may be partly contributing to heating experienced in some cameras as if there is a mechanical system that moves the sensor, it can be more difficult to disperse of the heat. If the sensor is fixed then the heat should dissipate through fixed heat conducting structures to the body, assuming a good design. 

However, I've never used a camera which heated appreciably in use. I've had flashes heat though, on sunny days. Not so much a problem in Finland except during peak wedding season.

I appreciate your comments. They are helpful. I was commenting on those who suggest that we have a heat problem with this particular camera (X1D), when that has not been established. Obviously, Hasselblad is tweaking the firmware and hopefully nothing in hardware. As a system programmer who founded the second oldest software company still on the Internet (the first being Microsoft), I well know how difficult it is to bring a new software product to market. It gives you fits, for sure.

[JohnBrew]

Michael, I do not start rumors. I will repeat that the problem was mentioned more than once by people who have handled the camera. getDPI has a thread with 25 pages of comments on the X1D.
Do your own homework. I'm out of this conversation.

[Ilkka Nissilä]

Michael, I completely agree, it is best to see real-world reports of the final release of the camera before judging it. Given Hasselblad's heritage and experience with medium format sensors it would be quite surprising if there is some problem.

[Les Olson]

Given Hasselblad's heritage and experience with medium format sensors it would be quite surprising if there is some problem.

But the company producing the X1D has no connection whatever with Hasselblad's heritage.  Victor Hasselblad's company was sold in 1996, but Hasselblad management was involved in the buy-out.  In 2002 the development of a digital camera was cancelled and the team working on it was dispersed.  A majority stake in the company was sold in 2002, to Shriro, a Hong Kong-based firm which does mainly distribution and marketing.  The new owners decided to restart the development of a digital camera, and bought Imacon, a scanner company, for its expertise in sensors and image processing.  100% of the company was sold in 2011 to a venture capital fund run by a guy called Helmut Vorndran (he is relatively progressive, as venture capitalists go, but the core strategy is buy, restructure, and re-sell, not long term management).  These are the owners who made the Lunar and the Stellar, and have now made the X1D. 

Nikon has a much longer experience with mirror boxes than Hasselblad has with sensors, but that did not prevent the D600 fiasco - and Nikon has not had three changes of ownership in 20 years.

[D800Dominic]

Here is what I have been doing for a long time, this with the D810 and the Noct Nikkor. I like it, but I am ready for portraits and landscapes.

Wonderful capture, as a result, you may have 3 discrete systems in your 2017 future, Fujifilm, Hasselblad, and the D810 replacement, sounds like a grand adventure, happy shooting. When I listen to U2, "I still haven't found what I'm looking for..." I will fondly think of your quest. Sounds like fun. Thanks for sharing.