Accuracy of leaf shutters

[Alan Gage]

With my foray back into 4x5 I'm finding myself with different lenses filtering into my hands and I've been testing the shutter speeds with a homemade tester, a light sensitive diode powered by a 1.5v battery and connected to a labscope. With a light source on one side of the lens and the diode on the other I can measure how long it was exposed to the light.

Testing these leaf shutters I'm finding the same thing I found when I built this thing and checked all my cameras/shutters a few years back, leaf shutters are just nowhere near their advertised highest speeds.

They all seem ok up to 1/60 but above that things start slowing down in a hurry. Of the six lenses I have to test the marked speed of 1/125 is actually around 1/80 on nearly all of them. The fastest speed I can get out of any of them is 1/200, which is marked on the shutter as 1/500. The others top out about 1/150 with marked speeds between 1/300 and 1/400.

I'd blame the tester for being inaccurate except this is pretty much what I saw before and at that time I had some focal plane shutters I could check and those were nearly all accurate up to 1/1000 on this same tester.

Has anyone else here tested their shutter speeds? What were your findings? Is it just me? How accurate are they after a fresh CLA?

I'm not really complaining, just curious. As long as I know the actual shutter speed it's not that big of a deal.

FWIW, three of the shutters I tested were Copals, probably from the mid-80's (two Copal 1 and one Copal 0). One is a Prontor SVS in a land camera that was CLA'd in the recent past, and the other two were the standard press camera affairs from the 50's that have been sitting around a long time.

Alan

[Julio1fer]

I've tested a Compur that did get close to advertised speed at 1/500, and a Prontor that was off on 1/100 but on the mark at 1/200.

[Mike Kovacs]

Where is the tester shooting through the shutter?  The problem is you are supposed to measure from about the halfway open position.

The problem is in the very centre there is a lag time as the blades begin to start their movement.  If you think hard, you can imagine why this does not affect the exposure time or cause unevenness in the exposure on film.

[Hoosier_Rich]

Maybe so on them being slow, but I've shot my Kowa on everything from long exposures in Antelope Canyon to 1/500 on sunny beaches all with mostly slide film and it's bang on, first try most of the time. So I can't complain much, works well enough for me!  Probably one of the reasons I can't bring myself to get rid of it.  :rolleyes:

[Alan Gage]

Where is the tester shooting through the shutter?  The problem is you are supposed to measure from about the halfway open position.

The problem is in the very centre there is a lag time as the blades begin to start their movement.  If you think hard, you can imagine why this does not affect the exposure time or cause unevenness in the exposure on film.

Maybe that does have something to do with it, it would make sense. In most cases the sensor was sitting directly under the lens, though in one case, the converted polaroid, the bellows were extended and the sensor was on the film plane, which I'd think would tend to even things out a bit.

Maybe I should just pick one of the lenses that was consistent with the others and shoot 4 sheets at the same EV but different shutter speeds and see how they look after development.

Do you have a good way of accurately testing speeds?

Alan

[Julio1fer]

Alan, I know of two ways for the amateur tinkerer.

1- You can test approximately if you operate the shutter while watching an image of a whole TV (analog type) screen. The vertical sweep rate is 60 times a second in alternate sweeps (even lines, then odd lines). So a whole screen is 1/60. 1/125 gives you half the area, 1/250 one quarter of screen, and so on. The images you'll see are diagonals, look at the fraction of the whole screen that you see illuminated when operating the shutter.

2 -If you like the Radio Shack way, you could make a shutter tester with a phototransistor, a load resistor and a battery. Dave Richter has a diagram in his site. Google "shutter testing". If you make a cable with a PC audio connection, you can use Audacity free software as an oscilloscope, to see the waveforms and compute the shutter times. Look at the capture angle of the phototransistor, but those are usually wide enough.

The TV test is best performed at full aperture, if you use the shutter tester you should set a mid-aperture, for instance f/5.6 in a 35 mm camera.

[Mike Kovacs]

I used technique #2 above.  If you used a highly focused source like a laser I suppose you could aim it quick carefully, provided you had the lens elements out for the test.

[epatsellis]

IMO, a purely academic problem, as most of my exposures tend to the 1 sec and beyond range.

[Alan Gage]

I did a little more tinkering today and learned that the phototransistor I'm using isn't a simple on/off, but that the signal level varies with light intensity (up to a point of course). So when viewing the signal on a scope at the time bases required for 1/125 and up it looks like /---\

I thought it was just slow to switch for some reason so I'd been measuring from the initial rise to the initial fall, but today I found if I varied the light intensity the voltage would vary as well. To prove the transistor was indeed capable of switching low to high much faster I shot it with a TV remote and got |---|.

So it would appear that the slope I'm getting in the signal is in relation to the amount of light actually being transmitted through the shutter. If I take that into account in my measuring then measured speeds improve and aren't too far off. I don't really know at just what point I should start/stop measuring though.

So I set up for a series of test shots out the back yard. 4 sheets at 1/60, 1/125, 1/250, and 1/500. They're drying now so I guess I'll see what I got in a couple hours. I had to shoot HP5 (ISO 400) to get the higher shutter speeds and unfortunately screwed up processing. Being used to shooting FP4 I went on autopilot and developed them for two minutes too long.  :rolleyes:

Highlights might be about gone but shadows should still tell the tale. I'll report back.

Alan

[Julio1fer]

Alan, if you have a scope things are easier. It is best not to saturate the phototransistor in order to get fast response, so if you see the rising and falling edges, and your signal level is varying with light level, then you are all set to measure.

Blade shutters open and close with a gradual movement: aperture starts in the center, then progresses to full aperture, where it stays for a time, then closes in the same manner. This is why you get rising and falling edges in your scope. Since what matters is total illumination on film, i.e. the area below the waveform, you can safely assume that the rising and falling edges are linear, and measure time between the midpoints of each edge to find true shutter speed.

There is a second effect to consider, i.e. that the waveform is affected by the aperture setting. It takes a bit longer for the shutter to reach higher apertures, and the waveform height goes higher with higher apertures. So you'd get slighty lower time measurements (between midpoints of waveform edges) with higher apertures. You should therefore choose an aperture for your tests and stick to it.

A technician once told me that he tested shutters in 35mm cameras with 50mm f/2 lenses at f/5.6, this being some kind of average manufacturer's recommendation. I don't know how to translate this into LF but you would surely figure it out.

Please let us know how it all comes out...

[Alan Gage]

This is why you get rising and falling edges in your scope. Since what matters is total illumination on film, i.e. the area below the waveform, you can safely assume that the rising and falling edges are linear, and measure time between the midpoints of each edge to find true shutter speed.

I guess I'd also need to experiment on the intensity of the light I'm using. A very bright light might reach full saturation before the shutter is fully open while a dimmer light would need the blades to be fully open before reaching the maximum 1.5 volts. I'd think light intensity would also affect the angle of the slope.

The negatives are dried and sleeved. I don't have a very good way to view them (no light table) but they look pretty consistent to my eye so far. On Monday I'll try and get to the college and do contact prints of each of them. Seems like it would be more consistent than trying to keep everything identical between scans.

Geez, I'm about ready to reach my saturation point with this little project. I don't think I have the ambition to start testing at different apertures too. Just so long as I can come up with "for practical purposes" readings I'll be happy.

Thanks for the help,

Alan

[Julio1fer]

I guess I'd also need to experiment on the intensity of the light I'm using. A very bright light might reach full saturation before the shutter is fully open while a dimmer light would need the blades to be fully open before reaching the maximum 1.5 volts. I'd think light intensity would also affect the angle of the slope.

You are of course right. I was assuming that the sensor was never saturated. This would be easy to test or adjust, using B and full aperture.

[Ron G]

This is a very interesting thread for me.I do not have a digital scope so cannot capture the trace as much as I would like to so I have to use the sound card as others have been doing.
I have been working on my Bessamatics and the highest speeds that I have witnessed have been in the low 300s and it is pretty tough to get that in my estimation.I finished one yesterday,that is I may have finished it,the light meter seems to be reading a little low so may open it back up to sort that before I run some film through it but I got a 322 out of it and am quite pleased with that.The other speeds are within tolerance except 1/15 and 1/30 are both the same at 32.I cannot account for that and hope some shutter activity will resolve it.Time will tell.
I have some measure of faith in my setup as it seems consistant but it may be consistantly wrong because I have no standard to compare it to.
I am using an 3 cell LED flashlight as a signal source which I place on the film plane and the sensor is placed in front of the shutter.
If I might someday acquire a newer scope that has the storage capability I could verify my measurements.
I actually have the attachment for a polaroid camera that will capture the single sweep trace on my SC504 if I was so inclined and could find whatever it uses for film but it feels like I am going backwards considering what is on the market today.I also have a newer TEK 2236 but is is analog as well.
I have read many articles written on the subject and the general consensus seems to be that I should not expect to achieve anything close to the advertised speeds on the leaf shutters that I am working on.That Synchro Compur is a great shutter they say but is limited on the high end.Was it ever that fast?Ron G

[br1078lum]

Jurgen Kreckel (certo6.com) is pretty adamant about leaf shutters not working as advertised.  Some of it has to do with the springs that drive the movements getting stretched over time.  Some of it is they never did match the settings in the first place. 

Two other things that you might consider are the consistancy of your light source, and the responce time of the photodiode you are using.

[Ron G]

I have pretty much concluded after reading many posts about it is that I do not have the equipment to accurately measure the higher speeds because rise and fall times are part of the equation and to get an accurate figure they need to be adjusted to truely reflect the amount of light that passes through the shutter at the higher speeds.I think that I am good to about 1/200 or 1/250 and after that I am guessing.
If 1/500 reads faster than 1/250 for example based on what I have observed I am going to assume that it is probably ok,maybe not within tolerance but functional.
One day I hope to get a little deeper into this but my time is pretty well spoken for at the moment.Ron G