Sound Card Shutter speed tester - improving accuracy at high speeds?

[Peakbagger]

I built a shutter speed tester based on the really simple circuit that's floating out there (like this one - http://static.photo.net/attachments/bboard/004/0044cW-10288684.pdf). For the phototransistor, I used the Radio Shack 276-145. Testing with the Supermatic shutter, I am getting values up to 1/50 and then after that, the values are the same. like 1/125th, approximately, so I don't believe that 200 and 500 are at all accurate.

I had heard that people had similar experiences with that circuit and I was wondering if anyone had improved faster shutter performance by using a different phototransistor. I've seen mention of different phototransistors like the BPW40 and others. Will this make a difference, or are there other factors that I can change that will affect this?

[RichSBV]

No answers at all????

Well, curiosity got to me so I looked into that circuit. Checked the rise time of the photo transistor and the RC constant of the resistor/capacitor. The circuit should be plenty fast enough for any shutter ever made!

So, first thing would be to double check your components to make sure you have the right values. We could take it for granted you have the right photo transistor. But wrong values for the resistor or capacitor could affect the timing of the circuit. Make sure you are using a 4300 PICOfarad capacitor!

Beyond that, the method of shutter testing comes into play big time. You can NOT check the shutter through the center of light travel. Your reading will be way off the higher the shutter speed tested and similar to what you observe. Try checking at the half way point from center to edge. This will be much more accurate. Two thirds out towards the edge would be more accurate. It all gets into the physics of leaf shutters and how they are rated, but the center will never give you an accurate reading... I got into this many years ago when I first purchased a shutter tester, got bad results and had to find out why.

And now that I haven't fully explained things, maybe more people will chime in ;)

[Jim Graves]

I have tried two different versions of the sound card version type of tester ....and, as I have adequately chronicled elsewhere on this forum, I had the same observations that they work quite well at low shutter speeds and fail miserably at high speeds.

I remain very skeptical that measuring the period between sounds ... or the duration of sound ... will ever give a reasonably accurate light exposure (shutter speed) reading. It simply is measuring the wrong thing.

I would also guess that different makes of shutters ... and, even different sized shutters of the same type from the same maker ... will vary in the time sequence of sounds when compared to the sequence of actual light transmission through the iris. If this is true, it greatly limits the use of sound board shutters.

I do understand that shutter time tolerances ... i.e. the variation off labeled speeds that is "acceptable" is pretty loose ... but this even supports the argument that sound ranges should not be used ... as it only adds even more variation into an already inexact measurement.

[RichSBV]

" It simply is measuring the wrong thing. "? It's measuring the period between shutter open and shutter close. Same principal as any other shutter tester.

Without doubt, the electrical components have the capability of performing within the range of any shutter speed. Leaving out human error... That would leave two options (that I can think of at the moment).

1: The width of the pulses produced by the circuit/sound card. The higher the shutter speed, the closer these pulses would be together. If the pulses were too wide, than an accurate reading would be difficult.

2: What software is doing the actual measuring??? I haven't seen this mentioned and it's important as sound is not being measured or recorded. The two pulses must be recorded and then the time distance between them measured. It might be easy to measure one second and difficult to measure 1/250 of a second. All depends on the programmer I suppose...

So to satisfy my curiosity, what software is doing the measuring??

[Peakbagger]

" It simply is measuring the wrong thing. "? It's measuring the period between shutter open and shutter close. Same principal as any other shutter tester.

Without doubt, the electrical components have the capability of performing within the range of any shutter speed. Leaving out human error... That would leave two options (that I can think of at the moment).

1: The width of the pulses produced by the circuit/sound card. The higher the shutter speed, the closer these pulses would be together. If the pulses were too wide, than an accurate reading would be difficult.

2: What software is doing the actual measuring??? I haven't seen this mentioned and it's important as sound is not being measured or recorded. The two pulses must be recorded and then the time distance between them measured. It might be easy to measure one second and difficult to measure 1/250 of a second. All depends on the programmer I suppose...

So to satisfy my curiosity, what software is doing the measuring??

Thanks for answering, I'm using Audacity, but I've also used Adobe Audition to measure the electrical pulses with similar results. I also tried 3 different soundcards, although they were all integrated chipsets, not dedicated cards. The resolution is pretty good as far as I can tell, although the new version of audacity allows for fewer decimal places than the old version.

It never occurred to me that the position of the phototransistor relative to the center of the shutter would have an effect.
Right now I have the circuit prototyped on a breadboard, with the phototransistor fully exposed. I just put a tube over it and the shutter on top of that. I wonder if I enclose the transistor so that the only light coming through is through the top portion that I will get a better reading? I have used a headlamp and my flashlight on my smartphone as light providers.

Also, I reconfigured my circuit, to more closely match this one, but with only one phototransistor. http://www.kyphoto.com/classics/combinationtester.html, but results are the same.

I am using a 4.7K resistor here as in the schematic.


Yes, I was thinking of your observations about the high speed accuracy failures. Still, I have seen that some have been able to report successful tests at higher speeds so, I am spurred on by these reports.

[Jim Graves]

Another possible variable is the intensity of the light.

If I recall correctly, one of the commercial testers specified the wattage and distance of the light from the tester.

[BetterSense]

With leaf shutters the intensity of light is not constant, so in principle, a temporal measurement can never tell you everything about your exposure. If exposure is what you are mainly interested in, you could use a piece of photo paper in a film holder with a step wedge taped over it. It's easy to see how the equally-density patches shift -- hopefully by one stop.

[mmerig]

Although the laser pointer is a simple light source, it makes measuring more complicated, primarily because the system relies on a binary voltage signal. An improvement is to use a pulsed LED with a wide field of illumination (wider than a laser, anyway). This wider beam alleviates the "center of the lens" problem mentioned above, and also can indicate opening and closing dynamics when it is pulsed on and off very fast.

The photo transistor behind the lens and shutter records the pulses. Sound cards can record up to 20 kHz, so the resolution is good. In the system I built, I set the pulse for 10K.

The circuit design that I used is available on the internet. If interested, and you are not sure about it, I'll post what I did and some results. I mentioned this shutter speed tester in a post on APUG a while back, and there is a good discussion and examples on that post. I did not post my results there as requested later, because I did not follow the thread closely enough.

http://www.apug.org/forum/index.php?threads/cable-shutter-test-results.126932/#post-1677407

There seems to be some confusion about using a sound card to measure shutter speeds. Some testers do use sound (like the smart-phone based ones), but the phototransistor converts light to a voltage, and that is what the sound card is reading (it is not sound converted to a voltage, like a microphone in a smart phone would do).

Some prefer oscilloscopes instead of a sound card. A sound card measuring voltage amplitude over time is essentially a recording digital oscilloscope with a bandwidth of 20,000 Hz. Way slow for what scopes are typically used for, but good enough for most shutter speed testing, and way, way cheaper than a digital storage scope.

[Jody_S]

This wider beam alleviates the "center of the lens" problem mentioned above, and also can indicate opening and closing dynamics when it is pulsed on and off very fast.


Can't the simple $3 circuit give the same info simply by positioning it in the center and again at a couple of points on the perimeter? And since we mostly shoot at f16 or smaller, at speeds slower than 1/60, I don't see the edge effect as being that much of a problem in practice.

[mmerig]

Can't the simple $3 circuit give the same info simply by positioning it in the center and again at a couple of points on the perimeter? And since we mostly shoot at f16 or smaller, at speeds slower than 1/60, I don't see the edge effect as being that much of a problem in practice.

The pulsing circuit is not much more expensive, (the timing chip is about 50 cents) and the nice thing about it is that the signal for opening, open, and closing is easy to see. You quickly can take say three measurements at each speed to check consistency. The wider beam also means you don't have to be so precise in aiming the laser at the phototransistor.

Moving the laser around, as you suggest would take more measurement time, and if there was a big difference in shutter speed time from edge to center, you'd have to account for it somehow. I never used the laser-pointer version, so I don't know how problematic the laser really is.

The first shutter-speed tester I built was based on a resistance-capacitance time constant, where the longer the shutter speed, the higher the voltage retained in a capacitor which was charged by a phototransistor. Quite simple and cheap, not counting the costs of a voltmeter, but it was unreliable and tricky to use. That is why I built the pulsed-LED, phototransistor system. It's a little bit over-kill, as we don't really need to know the details of the opening and closing time when it is fast, as it should be, but if it slows up too much then you know for sure.

By the way, on leaf shutters in large format lenses, I see very little difference in shutter speed time for each marked speed between F8 and F22.

My camera equipment is old and needs checking once in a while. Ruining say 5 sheets of B & W film with bad exposures costs much more than the shutter tester that I built.

[Jim Jones]

. . . By the way, on leaf shutters in large format lenses, I see very little difference in shutter speed time for each marked speed between F8 and F22. . . .

At the highest shutter speeds, you might see a difference in the effective shutter speeds. The time needed for the shutter blades to fully open and to fully close becomes an appreciable part of the total shutter speed. A shutter tester that measures the time from the beginning of the shutter opening to the last moment before it is closed will indicate a slower shutter speed than a tester that measures only the time that the shutter is nearly wide open. Your system of charging a capacitor through a phototransister avoids this problem if the circuit values and light source are chosen so the charge on the capacitor is proportional to the light reaching the film. Those choices can be difficult to balance. With long experience in electronics, I prefer to use an oscilloscope. It can also check for shutter bounce and measure sync delay.

[mmerig]

The pulsed-LED phototransistor method, which records the voltage signal on a sound file, allows a time measurement for the fully open period, (effective shutter speed), or you can include the opening and closing parts that show up in the signal. If I get a little time, I'll go back to some of my files and see what the difference is and report them.

One of the problems I had with the capacitor-based system was that the voltmeter loaded the circuit too much, (it has a 7.5 M ohm buffer), so that the charge on the capacitor discharged through the meter as I was reading it. I could have used a meter with a larger buffer, but I would have had to buy one, or built a buffer for the capacitor-based circuit. I liked the idea behind the pulsed LED system, so I built that instead. The other problem with the cap-based circuit was that the voltages did not agree for the same shutter speed when measured in the high-speed setting and the low-speed setting.

This shutter speed stuff is like that old saying in electronics: "The man with one voltmeter thinks he knows the voltage, the man with two voltmeters knows he does not know the voltage".

[tonyowen]

To check the speeds on my xenar f4.7, 135mm I used a simple microphone attached to my computer and used the Audacity program.

I set the 'recording mode of the program and fired the shutter.

The replay showed the peaks of the 'click-clunk' and with some judicial ignorance of some peaks gave me readings for 1/500th thru 1sec.

The replays inferred that all speed settings were off typically about 50% ie 1/500 read as 1/250th, 1/30 as 1/20th and 1/2s as 1 sec

I'm note certain of my results, but the lens is old and has not had a cla for years.

regards
Tony