Why is there a difference between the shutter lag times for prefocus and for manual focus?

Question

In my lab I have two cameras, the Canon T2i and the Canon T3i, which are without their standard lenses. Instead they are mounted on a microscope. To capture the images of my samples, I set the camera to the manual mode and then I press capture button in the EOS Utility software which is installed on a nearby computer.

The images thus captured, as far as I can discern with my limited senses, seem to have a consistent lag between the mouse click and the actual capture. Now, I am trying to automate a process of capturing 100 images over a period of about 200 seconds (198 seconds to be more accurate). I will be using the technique detailed on this link. However, instead of the manual switch I will be using a micro-controller (the controller is extremely accurate — so, for this problem, please neglect the µs errors arising due to the controller).

Concentrating on the milliseconds, I want to figure out how much would be the shutter lag (i.e. the time gap between the image capture and the trigger command issued by the micro-controller). The shutter lags for the cameras are noted on the following links:
Canon T2i : http://www.imaging-resource.com/PRODS/T2I/T2IA6.HTM
Canon T3i : http://www.imaging-resource.com/PRODS/T3I/T3IA6.HTM

The information provided there has become a source of confusion. I hadn't read about prefocus before, but from what I have understood, prefocus is the shutter lag after half-pressing the trigger. Thus the photographer "prefocuses" on the target before capturing the image. Why is there a difference between the shutter lag times for prefocus and for manual focus (in the case of each camera)? Is there an underlying difference between the mechanisms of prefocus and manual focus? I would have assumed that after the prefocus operation is complete the camera will react the say way as if it was in manual focus. After all, in both cases, the object is already in focus when the camera receives the command for capturing the image.

Answer 1

The reason "prefocus" is faster even when it doesn't actually focus is that half-pressing the shutter does more than just focus, when you half-press the shutter the camera:

1. Powers up any component that may have been turned off to save power.
2. Focus (if in auto-focus mode)
3. Meters the exposure
4. Do all the preprocessing before taking the image, decide what the value of each automatically set value should be and set it
5. Display selected values and focus confirmation.

Even if you are taking a sequence of images (so everything should be powered up), in manual exposure mode and manual focus mode you still have the metering and pre-processing (even if you don't use the camera's metering results the camera still shows you at least how far you are from the metered values).

For what you are doing all of this is probably meaningless, you can keep the shutter half pressed while you shoot you entire sequence but without testing this yourself you don't know how this effects shutter lag.

Answer 2

The problems I see with with the "test" are:

1. Overly precise measurements. A 0.001 second difference is meaningless.

2. No disclose about the testing detail- e.g. what target for AF, what distance, what lens, definition of outliers, etc

3. No evidence of a properly conducted study, i.e. multiple cameras and multiple lenses, each camera tested with each lens multiple times.

4. No sort of any data analysis. We have no idea what time differences can be considered statistically significantly different- they don't even give confidence intervals to show us

Now for the really bad news: as published, those tests are completely irrelevant to your issue. For your needs you can ignore shutter lag as long as it's reasonably fast. What you actually need is consistency in the lag, or in statspeak, a low standard deviation with few outliers.

Since you aren't going to get those numbers anywhere, you're stuck with trial and error.

Answer 3

The time taken to pre-focus varies from lens to lens.. As you have a microscope, I can't calculate the the time to pre-focus. But if the lens is set to Auto focus and the camera is set to AI Servo, and then the image moves, it will try to refocus the subject while the shutter button is half pressed. But if you focus it manually, the the camera will neglect the AI Servo mode. Moreover the shutter lag is the time taken for the mirror or prism(in high end cameras) to rotate or change its position so that the light is directed towards the sensor. So after taking a photo it will revert back to the original position and the process goes on like that in burst mode. So you see the Auto focus plays no role in the shutter lag. As in the reference you gave, the time includes the auto focus time(that is the prefocus time) and that's why they have mentioned the lens model number. You do in manual focus and forget all worries as the photos will be taken via micro controller.