Added: See "Yes!!!" at end.
EFCS may cause large under-exposure at very high shutter speeds.
Sample of 1 so far. Mine !!!
Makes sense (see end) but exact basis as yet TBD.
Sony SLT & mirrorless cameras and many Canon EOS DSLrs offer EFCS.
I have a Sony A77 SLT - I use it in EFCS mode and have only used mechanical FCS to compare the modes. For me the main advantage is reduced noise. The SLT fixed mirror and single shutter operation vastly reduce noise compared to a mirror plus 2 shutter actions SLR.
It is implied by Sony that the EFCS may cause image blurring with moving objectcsat large apertures at larger focal lengths. The rationale for this is given below - basically the EFCS has a finite dot writing speed which causes it to be unable to provide a high enough "enable" signal to start sensor light-writing.
From this web discussion.
The Sony NEX5n manual says (lost in translation):
They note that Sony suggests EFCS not be used in some situations eg.
- When you shoot at high shutter speeds with a large diameter lens attached, the ghosting of a blurred area may occur, depending on the subject or shooting conditions. In such cases, set this item to [Off].
And then provides this plausible discussion:
The first electronic curtain is likely rate limited by the clock it shared with readout. So it can't clear rows faster than it could read them. The second curtain isn't limited by this clock. A mechanical shutter with a sync speed of 1/250 exposing at 1/8000 will move a slit 250:8000 of the height of the frame across it. If the height has 4000 rows, then the size of this slit is 4000*250/8000 = 125 rows. For an electronic shutter, which has a sync speed of say 1/12 (at still resolution; the A77 shoots at up to 12fps, right?) this becomes 4000*12/8000 = 6 rows. At 1/24 it would be 12 rows. So if you're panning with a fast lens wide open, tracking a subject, it's not hard to see how the background could smear.
What they're really saying is that at high shutter speeds and narrow depth of field, turn it off when shooting moving subjects. Sports, wildlife; that sort of thing.
Sony also say:
- When a Minolta/Konica Minolta lens is used, set this item to [Off]. If you set this item to [On], the correct exposure will not be set or the image brightness will be uneven.
See Stan's comment in this. I'd have thought that the camera could easily add delay if required - say a default delay if the lens is not in the camera's "library". And a user adjustable delay could easily be added.
And discussion says:
Not sure what the difference in lenses is; maybe newer Minolta and all Sony branded lenses have better defined aperture timing, or can go from one aperture to another without passing through an implicit reset state (wide open). Or some other practical difference that matters here.
My A77 often exposes incorrectly at very high shutter speeds - ay high ISO and bright day when subject matter swings from shadow into bright sun.
In aperture priority mode, as light increases shutter speed increases. As increaseing light moves shutter speed from 1/4000s to 1/6000s the image darkens appreciably and at 1/8000s the image is massively underexposed.
BUT I just tried this with mechanical front curtain shutter and the problem appears to vanish. This testing was done with CFL bulbs and a white ceiling as its 1am here but I'll be trying this in daylight asap.
In this case it may be that the very fine tolerances between electronic turn on and mechanical exposure removal are easily swamped by small error in mechanical timing.
Even at a full curtain traverse time of say 1/250s, to get 1/8000s effective the shutter needs to be open 250/8000 = 1/32 ~+ 3% of the width of the sensor. If shutter edge is "out" by 1% in position the exposure may be reduced by 1/3. A 2% position error gives only 1/3 exposure and a ~= 3% position error gives ~= darkness.
This may well be a calibration adjustment.
More anon when/if I find out.
I do not have time now to experiment further at present but:
Exposure with and without EFCS at 1/8000s was good with Sony 50mm, f/1.8 (DT 1.8/50 SAM) even when imaging a CFL light bulb and surroundings at f/2.4.
Exposure problems occurred at 1/8000s with EFCS but NOT with mechanical FCS with
Minolta 17-37 f/2.8-4 D -> usually an excellent performer.
Tamron 28-200mm, f/3.8-5.6 -> an "only OK" lens overall.