It depends on the flash situation that you're facing, obviously. Fill flash vs. a truly dim scene. The exposure always starts as soon as the first shutter starts opening. But you know that.
Not necessary in all the cases. High speed flash photography is an example.you are Right in case of flash photography, where the first curtain Needs to fully expose the sensor.Well, you must be pretty slow, because the difference can be a portraits with eyes shut or open... of course, one can always machine gun the model and hope for the best...The difference between the electronic shutter and the mechanical is very large, but you are right, I have to take a few more pics similarly as I tested the electronic shutter but with the EFCS and mechanical shutter so that I get some reference results, after all, the mechanical and the EFCS is accurately measured.
The difference is just 6ms, and this is understandable since the sensor needs to close the mechanical shutter first, before it can start opening the first curtain. My guess is that this 6ms difference is what it takes to close the shutter.
Since your technique for the electronic shutter (ES) involves taking a photo, that measurement will include the sensor readout time (at least up to the sensor rows where your LCD display is positioned in the frame), whereas your mechanical/ES technique do not. That readout time must then be subtracted from your calculation otherwise the ES lag measurement will appear slower than it actually is.
that the travel time of fully electronic shutter is quite Long compared to travel time of electronic first curtain or mechanical first curtain is clear and well known - due to the line by line readout process for fully electronic shutter (see also Horshacks post below)
So you are doing the same measurement that I am, with two differences.
That's plenty for me to see what you're doing.
I start my measurements with the shutter half-pressed.
I start my measurements with the shutter half-pressed.
I no longer have a D800. I measure the D5 at 45 ms, if I remember right.
--
By the way, I have to repeat my measurements even without a lens. I don't know if it makes a difference, I don't think so, but never the less, it must be tested.
blog.kasson.com
Actually, both your and my techniques don't know when the shutter is fully open, but only when the camera says it is. That's plenty good enough for this exercise, though.
Well explained up above...
Travel time for the mechanical shutter is about 4 msec. If it were 6 msec, it couldn't offer 1/200 sec X-Synch.Not necessary in all the cases. High speed flash photography is an example.you are Right in case of flash photography, where the first curtain Needs to fully expose the sensor.Well, you must be pretty slow, because the difference can be a portraits with eyes shut or open... of course, one can always machine gun the model and hope for the best...
The difference is just 6ms, and this is understandable since the sensor needs to close the mechanical shutter first, before it can start opening the first curtain. My guess is that this 6ms difference is what it takes to close the shutter.
By fully open in this concept I mean as open as it is going to be, not at the end of movement.
any idea which technical processes cause the difference in resulting lag between
In AF, the camera will have focused if you have done a half-press, where it will have to focus if you have not.
Yes.
Metering is being done continuously and is reported in the finder.
The half-press versus none does not explain an EFCS vs mechanical shutter difference, should one exist.
but you are using the x-contact for all measurements ... does this x-contact close at "higher" (you name it "after") shutter Speeds earlier than the first curtain having accomplised full travel?
With mechanical and EFCS, it's pretty easy to correct results if the start of exposure is the desired metric. In the Z7, just subtract 4 msec from the OP's reported result.
Yes, basically we are doing the same thing, but using different means to measure.JimKasson said:
Half press is simulated by my devices because that is necessary for the camera, otherwise it will not trigger at all. However, the function of half press (AF and metering) is not relevant in my case here, since I have the camera set to manual. My device can also measure AF speed and in that case the half press is more important.Member said:
If I had not made my own dedicated instruments for the tasks of camera testing, I could use a scope as well, but I prefer it this way. I have many instruments, even a very accurate timer/counter/signal generator. If I did not have two dedicated, completely independent devices to measure, I would have used my signal generator/counter to verify my measurements.Member said:
I have a 300ms delay between the "half press" and the actual shutter release signal, but the time measurement is not not taking this into account and starts only after the "shutter release" signal. You can actually visually see this 300ms in my video, the white LED is turned on 300ms before the red, the red is actually the shutter release. Note, before anyone mentions the long time those LEDs are lit: none of that is part of the measurements, the LEDs are handled outside and they are just for visual indication of the signals.Member said:
I never had the D5 so I don't know, but you see, the 22ms is EXTREMELY short time, and I'd actually expect the D5 to be faster than the Sony you have in the table below.Member said:
I have now repeated the tests, and it makes no difference if I have a lens on or not. My measurements show the same values, so I guess this ends my shutter lag testing, and the next time it will be for final test of my second instrument.Member said:Member said:
By the way, I have to repeat my measurements even without a lens. I don't know if it makes a difference, I don't think so, but never the less, it must be tested.
There is a flash signal even at higher shutter speeds, but unless you switch to high speed flash use, the image will be vignetted if you try to use a flash above the sync speed.mcabato said:
no, thanks, this is what I meant and Jim had it already adressed.
4 msec fo Z7, more like 2.5 or 3 Msec for D850 and D5.
Resetting the sensor can be very fast, and even with a rolling reset when not using EFCS, it only needs to be ahead of the mechanical shutter. So the reset time is not going to affect the latency.
I don't see why that needs to be the case. I think it is more likely a difference in the way the mechanical shutter works.
blog.kasson.com
