Double flash?

scott657

New member
Ive had my D3100 since May. Up until now I did not realize the pop up flash double flashes even when not in red eye mode. Is this normal. I have reset to the factory settings.
Thanks for your help
 

WayneF

Senior Member
Ive had my D3100 since May. Up until now I did not realize the pop up flash double flashes even when not in red eye mode. Is this normal. I have reset to the factory settings.
Thanks for your help

In TTL flash mode, there is always a preflash immediately before the shutter opens, and then the final flash when the shutter does open (maybe 100 msec separation). Normally, they are indistinguishable to humans, looks like one flash.

(it is NOT related to Red eye, which delays about one second while flashing a couple of times, trying to reduce the diameter of the eye pupil - this is very distinguishable.)

If you set the camera to rear curtain sync mode, then the final flash is at the tail end of the shutter duration. If a slow shutter speed, then this can make the two flashes appear more separate. You won't realize it for normal front curtain sync.

Preflash is absolutely necessary for digital cameras.. The preflash is what is metered, to set the final flash power level. All compact and DSLR cameras must do it (for automatic TTL flash). You probably only see it if Rear Curtain Sync mode.

You can select Manual flash mode to eliminate the prefash. Then, instead of the TTL automation metering the preflash and setting the final flash level, then it is up to you to directly set the final flash level manually, yourself.
 
Last edited:

Bob Blaylock

Senior Member
In TTL flash mode, there is always a preflash immediately before the shutter opens, and then the final flash when the shutter does open (maybe 100 msec separation). Normally, they are indistinguishable to humans, looks like one flash.
·
·
·​
Preflash is absolutely necessary for digital cameras.. The preflash is what is metered, to set the final flash power level. All compact and DSLR cameras must do it (for automatic TTL flash).

I've got some faint memory of reading of a system that Olympus implemented in some of its 35mm cameras, where it metered the flash by measuring light reflected off of the film during the actual exposure. The concept seems dubious to me, because I have to doubt that the albedo would be consistent enough from one variety of film to another for this to work reliably. It does seem, now, thinking of it in the context of a digital camera, that since you always have the same sensor, that the albedo of that sensor could be assumed to be consistent enough for this idea to work. Then again, why use light bounced off of the image sensor to be metered with an additional sensor, rather than doing the metering directly from the image sensor itself?
 

WayneF

Senior Member
I've got some faint memory of reading of a system that Olympus implemented in some of its 35mm cameras, where it metered the flash by measuring light reflected off of the film during the actual exposure. The concept seems dubious to me, because I have to doubt that the albedo would be consistent enough from one variety of film to another for this to work reliably. It does seem, now, thinking of it in the context of a digital camera, that since you always have the same sensor, that the albedo of that sensor could be assumed to be consistent enough for this idea to work. Then again, why use light bounced off of the image sensor to be metered with an additional sensor, rather than doing the metering directly from the image sensor itself?

Nikon film TTL did exactly that, metered the flash reflection directly from the film surface, in real time of the final exposure. Worked really well too as to overall exposure, but it could not meter individual flashes. It was real time, and when the light was sufficient, it cut off all the flashes, regardless of their contribution.

But digital anti-alias filters put an end to that - reflection no longer worked for digital.

So early D-TTL (early digital) painted a light gray strip on the front of the shutter curtain, and metered a preflash reflected from that (after the mirror was raised). There was lots wrong with that system.

So Nikon brought out iTTL (about ten years ago), which meters preflfash in the viewfinder (not in the image sensor), similar to the ambient meter, and it doesn't really work quite as well, but well enough, and it does add significant features, like metering the individual flashes for their proper contribution.
 
Last edited:
I've got some faint memory of reading of a system that Olympus implemented in some of its 35mm cameras, where it metered the flash by measuring light reflected off of the film during the actual exposure. The concept seems dubious to me, because I have to doubt that the albedo would be consistent enough from one variety of film to another for this to work reliably. It does seem, now, thinking of it in the context of a digital camera, that since you always have the same sensor, that the albedo of that sensor could be assumed to be consistent enough for this idea to work. Then again, why use light bounced off of the image sensor to be metered with an additional sensor, rather than doing the metering directly from the image sensor itself?


I have an OM2n and it does have that system. I used it with 3 flashes that were wired to the camera and all fired. I could set the flashes to different powers or use them on different umbrellas and the exposure was always perfect. For 1970s tech it was pretty good.
 
Top