let's talk reciprocity failure here

[Bryce]

Noel-
Yes, I establish exposure with the meter at the subject location, and then set the flash off from the location it will be during the actual exposure.
The meter has proved quite reliable for studio shots under more normal conditions, and has always agreed with the exposure calcualtor wheel on the flashgun within a half stop.
As nearly as I can tell, the problem comes as a result of multiple flash exposures yielding less negative density than would a single pulse with the lens open wider.
Seems strange, no?

 

[Xmas]

Bryce

There is a high speed effect as well as a slow but a 285 in manual should be ok, i.e. in the linear range.

So the meter says you are four stops under with a single flash and you fire the flash 16 times?

Noel

 

[Bryce]

Yes that is what I'd have to give it... and 16 pulses is a good part of the life of the batteries!
Needless to say I spend a fair amount of time waiting for them to recharge, and have leaned to put new batteries in in the dark.
As far as I understand, the 285 has a relatively long pulse length when used at full power. It is long enough that it won't "freeze" a hummingbird's wings. So I don't think the problem is that the pulse is too short, and anyway I don't get underexposure when I can use a single pulse to light the scene.
Like I said, the real solution would be to test more scientifically and make a chart that allows me to predict the actual number of pulses needed for a given level of exposure.

 

[Xmas]

Bryce

Stumped, full power is about 1/1000 from memory so that is well inside linear. You could try the Fuji, if that is ok or better then you have discovered something.

The normal table for exposure assumes one 'shot' we both are assuming that integration is 100% efficient at the say 1/1000, but faint light level. Ive never got any traces of fogging from the perfectly visible flashes I get while loading 120 film and tearing off the backing paper, that could be why.

Noel

 

[Bryce]

Frank, Noel-
Thanks for bringing this rhread up. It caused me to search around again online and this time I found something, an old thread on P-Net:

http://photo.net/bboard/q-and-a-fetch-msg?msg_id=004mQP

The phenomenon I've been struggling with is real, has a name even. It is called intermittency failure. Unfortunately, it sounds like data sheets don't discuss it and give a chart because it depends on many variables. So testing is the only solution.
Poo.