DIY shutter tester design contemplation

[onre]

Greetings all,

Last night I could not sleep and so I was thinking about testing a curtain shutter. I don't know anything about how this is REALLY done, but thinking of it, I concluded that the relevant thing would be even exposure on all parts of the film gate. Most DIY testers seem to only measure one point, so they don't help that much in matching curtain speeds.

So, first focusing on a 135 film format, let's make a 36 mm wide array of, say, eight phototransistors. Install them in recessed slots on a frame. The idea is to construct a device that can be put right against the film gate. This is basically just eight "cheap DIY testers" next to each other, light-isolated from each other so that the first one "sees" leftmost 1/8 of the film gate, next one the second-leftmost 1/8 etc.

Then, let's take a computing device that is equipped to do analog-to-digital conversion and has a provision for a display device. Now, connecting the analog signals from the phototransistors to the A/D converter, we can sample the inputs and essentially reconstruct in software what happened inside the shutter. We could display average time for all sensors, separate times for each of sensors and also the captured waveform itself for each sensor so that detecting things such as difference in curtain speed or shutter bounce would be rather trivial just looking at the data. Also, if we equip the device with a servo and a trigger cable, we could make it to do, say, a 100-shot series of exposures and display a histogram of results. Basically, I'm under the illusion that this system could do all measurements related to a shutter.

So, the questions.

Has someone done this already, so I don't have to?

Is there something fundamental that I'm completely overlooking?

What measurements can you do with a professional shutter tester that I've not mentioned above? (flash sync comes to mind, it would just mean one more A/D line)

 

[paul_c5x4]

No need for A/D conversion, all you are interested in is when the curtain has uncovered the sensor. A couple of more points to consider is some curtains operate from side to side, others go up and down. Then there are leaf shutters that have multiple blades opening in an arc.

To be honest, having multiple sensors to measure shutter times is over-thinking & over-engineering the problem - Just one detector in the centre is adequate for checking the actual shutter speed matches the dial. Checking that the shutter is fully open for flash sync would need two, perhaps three sensors.

 

[onre]

To be honest, having multiple sensors to measure shutter times is over-thinking & over-engineering the problem - Just one detector in the centre is adequate for checking the actual shutter speed matches the dial. Checking that the shutter is fully open for flash sync would need two, perhaps three sensors.

But how about shutter curtain speeds? If they're mismatched, you get different exposure on different parts of the image. I've understood that professional shutter test equipment measures this.

 

[wombat2go]

Hi Onre,
I recently tested an RB67 shutter and compared the "on- axis" and "frame corner" timing and brightness variations. Thread here:
(there was a url link here which no longer exists)
Data:
https://app.box.com/s/6d06x1j0n6dm5et3jmypxpijpscdym8a

However this was a leaf shutter, quite different from the curtain shutter you mention.

 

[onre]

Thank you, this is interesting as well.

The main idea to do A/D conversion and then measure things in software domain would be price. Suitable hardware for this costs less than $100. The amount of eight sensors... well, I just pulled that off a hat. Probably three would suffice for curtain shutter. Left edge, center, right edge.

Also, given that phototransistors cost next to nothing, different sized/shaped "heads" for different measurements wouldn't be that expensive. I'd imagine some sort of cardboard-epoxy construction would be good enough.

 

[shutterfinger]

I use a single infrared LED with deep red filter material over it at the center of leaf shutters and Audacity audio recording software. Any white light source works if its bright enough. I use a 2AA cell Mini mag for leaf shutters.
On Focal plane shutter large format SLR's I use a 2 infrared LEDs. All sensors are mounted in foam core. For the FPS I mounted each sensor in a small square of foam core which I attach to a second piece of foam core cut to fit the film holder opening of the camera. I place one sensor at the top right about 3/8 inch from the edge of the film opening and the other in the lower left about 3/8 inch from the lower edge of the film frame. On FPS that have very good tensioning spring the sensors are identical to off .0005 ms at all speeds. On worn shutters the difference in speeds is usually less than .003ms.
Some of my ideas for testing came from this site: http://www.kyphoto.com/classics/combinationtester.html

 

[Steve Smith]

You just need an oscillator with its output gated by an opto sensor going into a digital counter. Put a light on one side of the shutter and the sensor on the other. Reset the counter, fire the shutter and work out the shutter speed from the count depending on the oscillator frequency.


Steve.

 

[John Koehrer]

Commercial testers use three sensors. Two in diagonally opposite corners and one in the center.
The corners read travel times and because of their location read either vertical or horizontal travel.
The center reads exposure and works with either focal plane or leaf shutters.

There's someone on ebay that sells a couple of types from a simple single sensor to one that also reads travel times.
I think(!) he's got the user instructions on his site.

 

[Philip Taylor]

I happen to work for a high speed camera manufacturer, so I well...photograph my shutters.
An interesting issue arises though. The abilities of the high speed cameras are such that I can record the slightest shutter movement (I can record 1.4 million FPS if wanted...there are some things film can't do)... And this has me frustrated because I can calculate the start to finish of shutter movement (centre on a leaf shutter) whereas the edge is very different. What do professional testers record?

 

[Jeff Bradford]

You just need an oscillator with its output gated by an opto sensor going into a digital counter. Put a light on one side of the shutter and the sensor on the other. Reset the counter, fire the shutter and work out the shutter speed from the count depending on the oscillator frequency.


Steve.

Exactly right.

 

[wombat2go]

Exactly right.

https://app.box.com/s/0w6q2wbrzhsd3stwiav3w1mshr9upf8v

Well, it would be OK at slow shutter, but with this old RB67 shutter at 250th, the speed with that method could be anywhere between 1/333th and 1/166th depending on the f/- and the threshold of the gate.

 

[Vitoret]

i've built one just for fun using an arduino + 1 light sensor (total cost around 20,00 euros)

camera open back with a little torchlight, arduino+sensor on the other side

I've to do some fine tunings to the code yet but it's working.. maybe i'll make a specific post when it's done

 

[mjork]

I happen to work for a high speed camera manufacturer, so I well...photograph my shutters.
An interesting issue arises though. The abilities of the high speed cameras are such that I can record the slightest shutter movement (I can record 1.4 million FPS if wanted...there are some things film can't do)... And this has me frustrated because I can calculate the start to finish of shutter movement (centre on a leaf shutter) whereas the edge is very different. What do professional testers record?

According to the text book "Basic Photographic Materials and Processes" by Leslie D. Stroebel et al, you should be measuring from the time when the shutter is half-way open to when it is half-way closed (with aperture set to max opening).

 

[onre]

i've built one just for fun using an arduino + 1 light sensor (total cost around 20,00 euros)

camera open back with a little torchlight, arduino+sensor on the other side

I've to do some fine tunings to the code yet but it's working.. maybe i'll make a specific post when it's done

The ridiculously low cost of embedded computing devices like this is exactly the reason why I thought of doing this digitally. Even with something totally over-powered like a Raspberry PI the total parts cost would be under $100. Writing the software should be fairly trivial, I'm a programmer.

I know this can be accomplished with an oscillator and simple digital counter(s), but that's really not the design I'm after. If I wanted to go in the "make it simpler" direction, I could just get a synchro-motor to rotate a disc in front of a light at a constant known speed and look at the stripes that I see through the shutter when shooting at it.

 

[paul ron]

Why bother DIY when a Calumat shutter tester used sells for around $50?

I made a simple one just using a photo diode directly into the sound input of my computer. I used a free sound program, Cool Edit or was it Cool Cut, to record the waves. Its very good up to 1/500th compared to my Calumat. Total cost.. $1.50!

My light source was an LED camping head light.

 

[sagai]

Why bother DIY when a Calumat shutter tester used sells for around $50?

I made a simple one just using a photo diode directly into the sound input of my computer. I used a free sound program, Cool Edit or was it Cool Cut, to record the waves. Its very good up to 1/500th compared to my Calumat. Total cost.. $1.50!

My light source was an LED camping head light.

It is reassuring to see these posts those telling me that common sense is still alive!

 

[onre]

Why bother DIY when a Calumat shutter tester used sells for around $50?

If they were $50 - or available at all - around here, I wouldn't spend a single second considering DIY.

 

[onre]

It is reassuring to see these posts those telling me that common sense is still alive!

Thank you, sir.

 

[paul ron]

http://www.ebay.com/sch/i.html?geo_...oapp=true&_nkw=shutter+tester&device=t&crdt=0

fancy digital readout diy tester
http://www.instructables.com/id/Build-a-Camera-Shutter-Speed-Timer/?ALLSTEPS

hack a mouse for parts...
http://makezine.com/2010/06/15/dead-simple-shutter-speed-tester-fr/

and another lije mine.. just a photodiode into your pc
http://www.photoplug.de

heres another very common one from the dark ages

http://www.baytan.org/prak/shutter.html

oooooh finlanddddddd.

sheesh the other side of the planet.

get a photo diode or photo transistor. just orient its polarity accross 2 wires n a jack into your pc. no fancy electronics needed. on the net, download a free recording program n you're all set to go.

if you'd like it fancy, mount it in a box for a few bucks more.


btw the calumat tester was $80 new.

 

[John Koehrer]

Calumet testers have commonly been going for well over the original price.

 

[Philip Taylor]

According to the text book "Basic Photographic Materials and Processes" by Leslie D. Stroebel et al, you should be measuring from the time when the shutter is half-way open to when it is half-way closed (with aperture set to max opening).

I'm not sure if you're still around, but thank you for replying! I've been out of the LF world for a while, and am now getting back into it, and vaguely remembered this thread. I should really do my own research since "basic Photographic Materials & Processes" was our main theory text at uni, and I have a copy readily available

Thanks for your reply...now to start testing my shutters again...

 

[radiant]

You can measure shutter with for example iPhone slo-mo video (240fps) up to about 1/125 speed. For sure you can measure 1/60 and slower with nice accuracy.

240fps = 4.2 milliseconds per frame

1/125 shutter = 8,6milliseconds