Reciprocity issues at exposures < 1/1000?

'lo all. I've just received a 2000FCW body to use as a backup to my 201f, the added bonus being the 1/1500 and 1/2000 top speeds. I won't need to shoot that fast very often but they're a welcome addition, and will come in use for those close up fisheye skate shots where 1/1000 isn't quite quick enough. Having tested the camera it appears to be in very good nick, but there seems to be issues with the two top speeds. You can see a contact sheet from the test here: Dead Link Removed, apertures were f8 at 1/500 down to f4 at 1/2000 (+1 and +2 were intentional over exposures, you can also see vignetting on the +2 shot down to shooting at f2)

If you look at the sheet you will notice how the higher speeds look a good stop under exposed. I suspected that the speeds might not be accurate given the age of the camera (1987) but then I realised that for the negs to be under then shutter would have to be a stop *faster* than stated. Pretty unlikely? Then I thought "well maybe I metered incorrectly?" Again, pretty unlikely 4 times in a row. The only other thing I suspect is that it could be down to reciprocity issues. Also notice how the 1/1000 2000FCW shot is darker than the 1/1000 201f shot - I suspect my 201f may be running slow.

So, does anyone agree this could be the case? The film was Tri-x 400 (fresh), developed in Ilfosol S (10 mins @ 20 degrees C). Having looked at the data sheet I see Kodak suggest +10% development times for exposures less than or equal to 1/1000 so that may be slight confirmation. Like I said I won't be shooting at those speeds often; Regardless I can open up a stop when I do anyway. I'll probably repeat the test with HP5, but I can't think of any other reason for the under exposed frames. It is the first time I've come across this issue though, although I admit I've never done a test like this before.

Very fast shutter speeds are often inaccurate in my experience (except probably with high-tech electronic gear where it ought to be rather more accurate).

I suggest some controlled test shots with various shutter speeds and the corresponding aperture adjustment which ought to give negatives nearly identical to each other. This can be very revealing of shutter consistency!

The reciprocity failure at 1/2000 is probably not going to be much. At 1/10,000 Tri-X needs +1/2 stop exposure and +15% more development, and 1/2000 is closer to 1/1000 where no exposure adjustment is needed (but +10% dev increase is recommended).

My data is from my copy of Kodak's pdf which is not the current one, however.

Since a focal plane shutter's open time is a factor of both its speed of travel and it slit with, it can very well be fast even at 1/10000 second due to improper calibration of the slit width. Any slight error at longer indicated shutter times that is due to slit width will be much greater at the shortest time, either longer or shorter, as the same clockwork setting is used.

Some focal plane error sources can be identified by photographing a TV screen at shutter speeds shorter than 1/60 seconds. You will not see the full screen. You should see a diagonal stripe whose width and shape may tell you what you want to know.

gainer said:

Some focal plane error sources can be identified by photographing a TV screen at shutter speeds shorter than 1/60 seconds. You will not see the full screen. You should see a diagonal stripe whose width and shape may tell you what you want to know.

Click to expand...

Ah, smashing idea. I was trying to think what I could use to determine the speeds and that might do it.

I forgot to mention that if the shutter travel is horizontal, hold the camera in landscape and the other way round if it travels vertically.

A follow up for those interested. I tested the camera using my TV and found the shutter speeds to be around 1/2 a stop quick:

1/250 -> 1/400
1/500 -> 1/750
1/1000 -> 1/1500
1/2000 -> 1/2500 ish (margin of error here is greater than the above)

I don't think this would account for the full stop underexposure, but taking into consideration Kodak's own advice to increase development times by 10% for shutter speeds <= 1/1000 I think I can attribute the other 1/2 stop to reciprocity failure.