How Fast Is the Z7's Silent Shutter

[makoto honda]

This is a highly technical question but a practical one as well. The silent (electronic) shutter is a relatively new concept and many people think of it as just an add-on feature for a special occasion (like shooting in the church, as the camera brochure advertises). Of course, many small cameras today use an electronics shutter exclusively.

The Z7 has a silent shutter mode - it means the camera just turns on and off the sensor output to the memory buffer, thus altogether forgoing the mechanical, focal-plane shutter. (The silent shutter has nothing to do with the mirrorless per se, and a DSLR has a silent shutter too.)

The focal-plane shutter is another complex mechanism of the past, like the complex and bulky reflex-mirror mechanism (that the mirrorless has gotten rid of). The silence is the wave of the future. As I talk about this, I even do not know how fast or slow the silent shutter traversal speed is - Does anyone know?

For a typical focal-plane shutter, it is 1/250 sec (which is the flash sync speed). The so-called JELLO effect is directly related to this speed.

In a typical focal-plane shutter, even if you use 1/8000, it takes 1/250 sec (much longer than 1/8000) for the camera to finish the exposure, because the front curtain must come down to open the shutter and then the rear curtain must come down to close the shutter. That's 1/250 sec. So, to make a truly "accurate" picture, your subject has to be still during that 1/250 sec. This is the shutter traversal speed I am talking about.

I wonder what the equivalent traversal speed is for the silent shutter. I suppose it is appreciably slower than 1/250; that's why the camera manufactures cannot forgo the mechanical shutter just yet. It must be a bit longer for the Z7 than the Z6 since more pixels have to be activated for the Z7. (But then, I do remember the days not long ago when a top-of-the-line SLR camera's flash sync speed is just 1/30-1/60 second!)

In practice, this is what I am curious about::: If you set your camera to the silent mode for a day or so and take pictures just like you normally do, and at the end of this test, look at all the pictures you have taken. Now, how many pictures can you find that you would say - "Oh gee, I should have used a normal mechanical shutter instead of this friggin silent shutter! My picture is totally messed up!"

With the continuous advancement of CMOS sensor technology, I would not be surprised that, in 5 years from now, no high-end single-lens camera comes with a mechanical shutter.

One thing to add is that the current Z7 only offers a "flash" of the EVF to indicate that a shot is taken during the silent mode. And, it offers NOTHING if you are doing the H+ shooting. Sony cameras offer an artificial sound for it in addition to the visual cue.

 

[scottinpollock]

Things you are most likely to notice with electronic shutter are banding with some frequencies of artificial light, and motion distortion when something is quickly moving across the frame. These things happen due to the readout speed (scanning) of the sensor, which can be much slower than the exposure time.

In video this is described as Rolling Shutter, but it affects stills as well. The rather high resolution and full frame size of the Z7's sensor makes it's readout speed rather slow (DPR estimated it at 1/15th of a second).

 

[acnomad]

Well since you asked...

I had to test it. Here are examples of rolling shutter effects in each camera:

1. Z7

Subscribe to see EXIF info for this image (if available)

2. Z6

Subscribe to see EXIF info for this image (if available)

FWIW, my distinctly non-expert observations and hastily-made conclusions:

• The blade at the 3 o'clock position is least affected. This confirms that the lines of pixels are read horizontally.
• The blade at the 1 o'clock position is definitely more affected in the Z7 image than the Z6. That is consistent with having more pixels and a similar read speed.
  
• Approximately 7.5 bands visible in the Z7 image, approximately 6 in the Z6. (How's that for appropriate numbers?!) AC power in the USA is 60 Hz, so that implies it took about 1/8" to read the frame on the Z7 and about 1/10" on the Z6.
  
• It is not very difficult to get lots of images without even being sure you shot them with the Z6 in H+ mode; it is blazing fast and truly silent. Every bit as good as the trusty E-M1.1 in this respect.

All of the above is subject to verification and correction by those more knowledgeable!

 

[scottinpollock]

The blade at the 3 o'clock position is least affected. This confirms that the lines of pixels are read horizontally.

To be clear, the sensor is scanned top to bottom, hence horizontal movement is more affected then vertical movement. Vertical movement tends to just get fattened or narrowed (depending on direction), while horizontal is bent as the top of the one o'clock blade is captured earlier than the areas below it.

 

[acnomad]

To be clear, the sensor is scanned top to bottom, hence horizontal movement is more affected then vertical movement. Vertical movement tends to just get fattened or narrowed (depending on direction), while horizontal is bent as the top of the one o'clock blade is captured earlier than the areas below it.

That’s a much better statement than mine. I was thinking of the pixels being arranged in horizontal lines, and that they are read all the way across the frame, then the next line is read, and so on. Is that the right concept?

 

[scottinpollock]

Is that the right concept?

Indeed. And that is what I thought you meant... I was just clarifying for others that might misconstrue your description to mean the sensor was scanned from side to side.

 

[makoto honda]

Thank you, guys, for the input!

The readout of 1/15 sec is rather slow for comfort, compared with a 1/250 speed for today's focal-plane "readout". But, as I said, people were using a 1/30-1/60 sec focal-plane "rolling shutter" for a long time. Just one more push of the CMOS technology, and we can get rid of the mechanical shutter for good!

Footnote: I am surprised that I was able to capture even a few shots with reasonable sharpness when I took photos of bees and flowers (posted here a few days ago), now that I know the silent readout is as slow as 1/15 sec!

 

[MNglass]

Thank you, guys, for the input!

The readout of 1/15 sec is rather slow for comfort, compared with a 1/250 speed for today's focal-plane "readout". But, as I said, people were using a 1/30-1/60 sec focal-plane "rolling shutter" for a long time. Just one more push of the CMOS technology, and we can get rid of the mechanical shutter for good!

Footnote: I am surprised that I was able to capture even a few shots with reasonable sharpness when I took photos of bees and flowers (posted here a few days ago), now that I know the silent readout is as slow as 1/15 sec!

Remember that the 1/15 sec is for the whole sensor - all 5504 lines - which (if my math is correct) is 0.000012 seconds per line. If the movement is within some small part of the sensor (say leaves in a landscape) then the 'scan' speed over that small portion is going to be relatively quick and the distortion may be acceptable.

 

[scottinpollock]

0.000012 seconds per line

And if the set shutter speed is fast enough, we're not talking about blur... every line captured will be sharp. It is just that if there is movement, it will be recorded in a slightly different position. Can sometimes make for some cool effects.

 

[makoto honda]

Yes, so in my bee pictures, a rolling distortion did not affect the result too negatively. I believe the shutter speed used was somewhere around 1/100 - 1/300 sec. I do see a "jello" effect on the bee's wing in flight on one picture (not a clear picture).

So, to me, a lesson to take away from this is that I should be consciously using the silent shutter more often in a normal situation, unless I foresee some issues of rolling distortion - because it is virtually free lunch.

 

[MNglass]

Yes, so in my bee pictures, a rolling distortion did not affect the result too negatively. I believe the shutter speed used was somewhere around 1/100 - 1/300 sec. I do see a "jello" effect on the bee's wing in flight on one picture (not a clear picture).

So, to me, a lesson to take away from this is that I should be consciously using the silent shutter more often in a normal situation, unless I foresee some issues of rolling distortion - because it is virtually free lunch.

When I had the Olympus OMD-EM1, I used the electronic/silent shutter often and it had about the same scan time and I'm not sure I ever saw a problem - though it was 16mpx. With the OMD-EM1.2 the scan time was closer to 1/60 on a 20mpx sensor and I actually used the silent shutter most of the time - (again for casual photography - not sports, not BIF, not fast moving things, not artificial light!).

The downside is that if there is resultant distortion, it is almost impossible to correct! So you have to be cognizant of the circumstances.

 

[scottinpollock]

a lesson to take away from this is that I should be consciously using the silent shutter more often in a normal situation

It certainly cuts down on the wear and tear of the mechanical shutter. My mirrorless of choice for the last year or so has been the Lumix G9, which I leave set to E-Shutter most of the time. But it has significantly faster readout time (granted with a sensor half the size and resolution).

You'll just have to experiment a bit to determine what level of movement you can tolerate with it. And of course you can't use it for flash and long exposures.

 

[acnomad]

Does anyone know if there’s an easy way to see whether mechanical or e-shutter was used when reviewing images in Lightroom or Photo Mechanic (in other words, without digging through raw EXIF data)?

 

[scottinpollock]

anyone know if there’s an easy way to see whether mechanical or e-shutter

Doubt it. The only time I have ever seen it is in ExifTool.