Shutter Speed Accuracy...

[dangie]

Purely as a matter of interest:

Back in the old film days of mechanical and then electro-mechanical shutters, AP and other magazines when testing a camera used to give the accuracy of the shutter. At faster shutter speeds this could often be up to half a stop out!

I know modern shutters can usually be set to at least a third of a stop, but how accurate are they to the indicated speed?

[Malcolm_Stewart]

I was browsing through some old APs a few days ago, and the same question crossed my mind. All I can say is that when my Canon DSLRs have used high shutter speeds, I've not noticed any problems with regard to exposure evenness or overall tone compared to say, 1/125s. That's with my 5D, 30D and 1D MkIII. I did knock-up a simple shutter speed tester, but that was for a film SLR. I suppose a moving and rapidly flashing LED could be used as a suitable test source.

[TimF]

I seem to remember that the Nikon F5 had a kind of self-diagnostic mechanism for maintaining accurate shutter speeds. The camera's cpu would monitor the shutter and apply correction if it started to drift.

I'd assume that such systems have spread to pretty much all cameras in the intervening decade and a half.

As for older cameras having somewhat inaccurate speeds, how much effect did it have in reality? A 1/3rd stop variation is quite subtle in its visual difference. For those shooting negative film the small extra exposure was likely to be beneficial anyway.

[beejaybee]

I know modern shutters can usually be set to at least a third of a stop, but how accurate are they to the indicated speed?

It really doesn't matter, provided they're consistent.

Back in the days of mechanical shutters, the real issue was the slow speeds. With electronic shutters, even 15 sec is very consistent, and probably accurate to 3 or 4 significant figures.

[Roger_Provins]

My experience, after repairing vintage and classic cameras over the years, is once the electronic shutters matured the accuracy became extremely good. They either work more or less spot on or they're broken and don't work at all.

I would guess modern dSLRs (I've never tested) will be at least as good as the last film shutters - nigh on perfect.

[dangie]

Thanks for the replies everyone. As I said it was purely out of interest. While trawling for some answers I came upon this:

"...Tolerances
For older, mechanical shutters, a plus or minus half stop tolerance along the whole range is deemed to be acceptable by most technicians. A top quality mechanical shutter should be within plus or minus quarter of a stop. The best electronic metal square shutters are often accurate within plus or minus 10% of a stop....".

If I recall correctly from my old Zenit/Spotmatic days there used to be a British Standard for shutter speeds (There was a BS for everything back then). A 1/500 shutter speed could be within tolerance anywhere between approx 1/350 and 1/700 second. With modern shutters being as accurate as they are (we hope) is there any need for the odd test, or is it a "given" that they are as almost spot on as what we think they are. Pure interest that's all.

[RogerMac]

With modern shutters being as accurate as they are (we hope) is there any need for the odd test

Exactly what I as thinking - perhaps the occasional test to make sure that standards are being kept up.

Incidentally I understand (no I have not looked up the references) that quoted ISO speeds are way off on some cameras, and this could cause a lot more bad exposure than inaccurate shutter speeds.

Roger

Edit: It is quite easy to check the slower speeds on your camera. Just set up an old fashioned turntable with a single bright point on it and take a photo - then measure the angle that the blur has moved. At 1/30 second and 45 rpm the blur should measure 9 degrees.
You should be able to measure speeds in the 1 to 1/30 sec. range using this method. Unfortunately not the most useful range to test