Sony introduces the a9

[Les Olson]

I see where you are coming from, but the measurement was done at the hot shoe. It should be done at the sensor.

I would be interested to see more measurements on recent cameras.
I don't follow your reasoning here.
It is not the mirror that is the primary vibrating mass. The mirror hits other stuff and causes waves to travel all over the camera, lens, and tripod legs (put your ear on a tripod leg while the mirror moves and you will hear a rather loud sound). The shutter by itself does not cause such a loud sound.
The mirror slap must have been recognized by Nikon as a source of vibrations. See Section 3 on this page:
http://www.nikon.com/news/2014/0626_dslr_01.htm

You have contradicted yourself.  If "The mirror hits other stuff and causes waves to travel all over the camera" then measuring them at the hot shoe is perfectly reasonable.  And, in the D200, the shutter does cause a louder sound than the mirror. 

It is not in question that the mirror mechanism makes a noise, or that when you are doing macro photography or using long lenses and (in both cases) shutter speeds in the 1/15 - 1/30 sec range, there may be a benefit from using mirror lock-up.  I say "may" because I have never seen an actual example, but I have only tested it with medium format film. 

[Les Olson]

According to the review of Imaging Resource, the shutter lag of D200 is 57msec and that of D800 is 44msec.  So, the shutter shock may occur before the vibration caused by the mirror settles.  Nikon deliberately slowed the shutter lag of D810 (54-56msec, according to Imaging Resource) to slow down the mirror movement sequence and thus to reduce the shock.

Also, on D800, the direction of the vibration initiated by the shutter could be the same as that of the remaining vibration caused by the mirror, which could recover the vibration "effectively".

Reducing the shock by slowing the mirror is also called making the camera quieter, which has social advantages that are more often a reason to value it than the effects of vibration on images in unusual conditions. 

And, sure, mirror and shutter vibrations could be additive, but if they were out of phase they could equally well interfere destructively.

[simsurace]

Well, you said that mirror-induced blur is an urban myth. If you are trying to argue that this is true based on a single experiment and a (overly) big dose of extrapolation, you've lost me right there. Particularly since it is so easy to reproduce.

If I'm sitting here with my camera trying to get rid of mirror-induced shake, it does not help me that someone 'disproved' it on his camera and specific conditions he was testing in.

On a general note, in science it is very hard to be successful proving that certain phenomena don't exist. What you can do is falsify theories. You can take one of the specific examples where people report blur and where they think that it is because of the mirror movement, reproduce it, and then track down why the blur occurs. Maybe the theory about mirror-induced shake is wrong (in that example). You can prove that by showing measurements that exclude the mirror as a source of vibrations. But then you still have to offer alternative explanations of why the blur occurs. The linked D200 experiment does not achieve that because the conditions are very different from those in which people usually report mirror-induced problems.

Another way to put it: if you're saying that all mirror-induced vibrations will always die down just before the shutter opens, then this requires extraordinary evidence and a very detailed theoretical calculation to have merit, and you still haven't accounted for all the reported cases of camera shake where the mirror is a likely source.

It is nice that Ansel Adams did not have these problems, but we have progressed quite a bit in terms of how much resolution we pack onto relatively small sensors and demands for stability have increased as we are able to examine our shots as closely as never before.

[David H. Hartman]

Ansel Adams had to contend with wind and an 8x10" camera's ability to catch that wind or so I surmise.

Dave

[Bjørn Rørslett]

A comment from the side line: the first time I used a "field camera", a rather light-weight Toyo 4x5" in the mountains I was amazed how susceptible it was to  being buffeted by wind. Not one of the first 10 sheets came out adequately sharp. I later dropped the cherry-wood Toyo for an Arca F-line metal camera that was much more robust and the problems diminished, but did not vanish entirely.

[Ilkka Nissilä]

If one fills the buffer in a D500 how long does it take the D500 to write all to memory? Is shooting suspended for the duration or can one start shooting when there is enough buffer available to hold a file?

I thought the D5 and D500 can write to G series Sony XQD cards faster than they can generate new data (14-bit lossless compressed NEF), so you should be able to shoot a 200-image burst at 12fps, take off your finger and immediately fire another.  The 200 image limit is the last defence set in firmware, to prevent the camera from accidentally firing the shutter forever if the shutter button is squeezed in the bag, for example. With uncompressed NEFs or use of slower cards (CF or older XQD) you can run into a real memory limit sooner than reaching 200.

Edit: Correction: I found an imaging resource test which shows the D5 took only 183 14-bit lossless compressed NEFs in a burst before buffer full (the D500 managed 200 images):

http://www.imaging-resource.com/PRODS/nikon-d5/nikon-d5A6.HTM
http://www.imaging-resource.com/PRODS/nikon-d500/nikon-d500A6.HTM

They report a 5-second clearing time for the D5 buffer and 3 seconds for the D500. I wonder why this result differs from Nikon specifications in the manual for the D5; perhaps it is due to a different card type or the test subject. Imaging Resource used a Lexar Pro 2933x XQD 2.0 400MB/s flash card. This would suggest the camera isn't quite able to write as fast as it generates data on those cards tested.

I read practically realized write speeds around 270-290MB/s on the D5 (297MB/s for the D500):

https://www.cameramemoryspeed.com/nikon-d5/fastest-xqd-cards/
https://www.cameramemoryspeed.com/nikon-d500/sd-and-xqd-card-speed-test/

12 fps * 25MB typical lossless compressed 14-bi NEF size = 300MB/s so it is a little bit more data than the write speed realized. Today the fastest XQD cards have manufacturer reported write speeds of 390 or 400MB/s; the Sony G 2 series being a few percent faster than the Lexar tested above.

Anyway, I suspect the different results reported may be due to the differences in the image content which affect the file size a bit. So if one buys the fastest cards today one can expect 3-5 s clearing time after a 200-ish image burst of 14-bit lossless compressed NEFs with these cameras, or perhaps if one is lucky with the compression of the image and can get a full buffer with virtually no clearing time.

I'm sorry for writing the initial post's assumptions - I tried to be more comprehensive in my correction. I should probably now test this with my camera and cards. At least I can take a look at how much the compression affects file size for my typical subject matter.

[Les Olson]

Well, you said that mirror-induced blur is an urban myth. If you are trying to argue that this is true based on a single experiment and a (overly) big dose of extrapolation, you've lost me right there. Particularly since it is so easy to reproduce.

It is nice that Ansel Adams did not have these problems, but we have progressed quite a bit in terms of how much resolution we pack onto relatively small sensors and demands for stability have increased as we are able to examine our shots as closely as never before.

But it is not easy to reproduce.  Here, eg, is a clever set-up which shows no effect of mirror lock-up on camera vibration (https://petapixel.com/2011/05/02/dslr-mirror-vibration-shown-using-a-laser-pointer/ - one interesting aspect is the title given to the piece, and the number of people making comments who assert there was an effect of mirror lock-up: people see what they expect, not what is in front of their eyes). 

Of course the phenomenon of mirror vibration affecting images is real: here is an example where the image is sharper with mirror lock-up (http://www.discoverdigitalphotography.com/2015/what-are-mirror-slap-and-shutter-shock/), but it used a 300mm lens and 1/25 shutter speed - exactly when Ansel Adams said mirror lock-up made a difference. 

Using a 300mm lens at 1/25 sec is hardly routine.  What is an urban myth is the idea that mirror slap is routinely blurring images.  Even where the effect can be seen, it is small - trivial, compared to the effect of pressing the shutter with your finger, let alone compared to hand-holding. 

[Bjørn Rørslett]

...

Using a 300mm lens at 1/25 sec is hardly routine.

Very routine if you are a nature photographer in the Nordic countries

What is an urban myth is the idea that mirror slap is routinely blurring images.  Even where the effect can be seen, it is small - trivial, compared to the effect of pressing the shutter with your finger, let alone compared to hand-holding.

Agree that user-induced movement is a major cause of unsharpness, in particular with long lenses and slow shutter speeds. The so-called "long lens technique" (LLT) which advocates pushing down hard on your long lens in fact is well capable of destroying image sharpness, unless shutter speed is very fast. And in the latter case, even a poor method such as LLT has little impact on the final outcome.

[Ilkka Nissilä]

Very routine if you are a nature photographer in the Nordic countries

I can confirm this. In the winter, in particular, if I'm shooting some landscape details just before sunrise, 1/25s sounds about right. 

I was shooting with the 80-400 AF-S for a few months before becoming frustrated and trading it in. I could not achieve consistently sharp results until the shutter speed was about 1/250s or 1/320s even using the Kirk collar. This meant waiting for the sun to be above the horizon in practice, which gives a harder light. Of course one can rise the ISO but that won't give the best tonal quality.

EFCS helps quite a bit but at that time it wasn't available on any Nikon camera; the D810 came the next year. Wind can still ruin slow speed shots with long focal lengths of course, but at least it isn't constantly present.

If one uses long focal lengths for action, then the mirror and shutter vibrations are largely mitigated by the fast shutter speeds typically required to freeze the subject movement. However, I think long focal lengths can be very useful for landscape as well, if one can get around the vibration issue. I do believe wildlife photographers run into vibration issues from time to time with long focal lengths as many animals are out there to be photographed in quite dim light. Again shooting before sunrise and just after can give an interesting variety of lighting for the photographs.

[pluton]

Using a 300mm lens at 1/25 sec is hardly routine. 

It is if you're trying to make shots like this:

[Frank Fremerey]

A comment from the side line: the first time I used a "field camera", a rather light-weight Toyo 4x5" in the mountains I was amazed how susceptible it was to  being buffeted by wind. Not one of the first 10 sheets came out adequately sharp. I later dropped the cherry-wood Toyo for an Arca F-line metal camera that was much more robust and the problems diminished, but did not vanish entirely.

Outdoors wind is the main problem for me. I can IR release and MUP but I cannot stop the object movement...

[David H. Hartman]

Outdoors wind is the main problem for me. I can IR release and MUP but I cannot stop the object movement...

If a low angle works and a tripod legs can be spread wide then not extending the lower legs of a tripod can help. I use a 7kg Bogen video tripod that way.

Dave Hartman

[Bjørn Rørslett]

Tripod weight per se has no bearing at all on the issue of movement prevention. The mechanical coupling of the gear to the tripod does, however.

[MFloyd]

Interesting, but this has not much to do with the initial topic .... 😏

[Bjørn Rørslett]

Yes and no. The discussion veered into the domain of vibration and stability.

Besides, the flow of a forum thread always has a certain Brownian aspect to it.