On May 13, 2005, at 6:03 AM, Cory Papenfuss wrote:
... Interesting. My understanding of TTL is basically the same... that there's a sensor in the camera that can read the light *during* the shot. A single additional pin from the body to the external flash triggers the external flash to stop firing when operating in TTL. I thought that pin was a binary threshold, though... not direct sensor reading.
Different cameras from the various manufacturers implement the communications and control in various ways...
The part that confused me was that for the flash, P-TTL is a superset of the requirements of TTL... they both have to be able to quench the flash at less than maximum. I guess the difference is that under P-TTL they body can use the *regular* meter to determine exposure, whereas for TTL it must use the TTL (mirror up) sensor.
I'm not sure about this, but I think what's happening is that the body evaluates the ambient and flash exposure on preflash, does the integration, and then just sets a time for the flash to operate during exposure. In other words, the flash sensor is inactive with the internal flash during the actual exposure. With an external flash, and TTL mode, the flash itself is reading the state of the sensor during the exposure and quenches when it arrives at its cut off value.
A quick clarification here... not only one that transmits aperture information, but one that is also set to use it?... with the DS body:
- The built-in flash unit will default to P-TTL flash metering whenever a lens that transmits aperture information is fitted to the camera. Whenever a lens which does not support aperture information is fitted, the built-in flash unit will operate at full power in non-metered mode.
Hmm. Perhaps my wording was unclear: the lens fitted must be both capable of sending aperture information AND set to the A position on the aperture ring to enable it for P-TTL to function.
In other words, an 'A' lens with the ring set to other than 'A' still sends min/max aperture via the sens pins. You said the preflash operation hasn't yet stopped down the lens, though, so without knowing what the aperture is *currently* set to, it cannot work P-TTL.
(I'm thinking of modifying an old K-lens to add the min/max aperture holes... can't make a true 'A' lens because of linear/log movement, but can fake an 'A' lens in non-'A' mode)
I don't know if that's true. When the lens is not set to the A position, an A series lens operates with all the restrictions of an M/K bayonet lens ... only full power manual on flash, no aperture operation except in Manual exposure mode, all metering modes default to Av other than Manual, etc. Whether it is sending min/max aperture information or not I don't know; it doesn't enable anything on the body that I can determine. So I'm not sure that a modified A lens will net you any additional function on the DS; I don't have one to fool with so I can't tell for sure.
I just did a test: on the DS with an A/F/FA lens set to any aperture other than A, the built-in flash fires at full power, unmetered. AF continues to work with the autofocus lenses. The aperture is NOT stopped down except in Manual mode; all modes except for M revert to Av operation at full aperture, and the meter pattern becomes CW Averaging. Exposures made with the mode selector in any autoexposure mode record full aperture in the EXIF data, on Manual mode no aperture value is recorded...
- All A, F and FA lenses operate like M/K bayonet lenses when their aperture ring is set to anything other than "A".Strictly speaking, there's more information to be had, since the camera knows the absolute aperture of what it's looking at due to the sens pins. The only thing it's missing is what it will be stopped down to (unless you preview).
I agree, strictly speaking, but the behavior doesn't seem to use that theoretically available data.
It will also *not* fire an external (dumb) flash if the shutter speed is set above the sync speed of 1/180. Protect you from yourself, I guess.
Good point. I'd forgotten about that.
Thanks a lot for the clarifications. The engineer in me sees opportunities to make improvements and workarounds for lots of these limitations. I'm sure that marketing had something to do with removing some of them.
Marketing and cost reduction measures, I imagine. Although how much cost is reduced by not adding a trace or two to a circuit board I don't know. ;-)
Godfrey
