Using APPS To Measure Shutter Speeds

[ReginaldSMith]

In another post, where I discovered my KONICA III shutter was running slow, I used my second had on a watch to measure the slow speeds of 1 and 2 on the dial. I wanted something more accurate. I discovered the use of oscilloscope apps on my Android phone. After testing a few, I settled on "Sound Oscilloscope" for it's ease of use. It's free. This approach measures the SOUND of the shutter opening and closing.

Method

I open the app and select a X-axis (time) range of about 500ms per division. This is for measuring the slow speeds. I place the app in RUN mode. I place the camera near the MIC input of the phone, set the speed to 1 and cock the shutter. Let all the noises of shuffling die down (you need a somewhat quiet environment). I depress the shutter halfway to get the slack out of the button. When I see the trace begin at the left side of the X-axis I depress and hold the button to allow the shutter to open and close. Here's the tricky part - - right after the bressing the camera shutter, you must press PAUSE on the scope. You just need to have your other hand hover over the PAUSE button.

You now have a scope trace with a big vertical spike as the shutter open followed by another spike as it closes. You then look at the X-axis scale and measure how many mS (milliseconds) are between the two spikes. That's your shutter time.

Decipering the trace. Because you have extra noises after the shutter push, such as releasing the button, moving your nad to the PAUSE and so on, there will be extraneous "spikes" in the trace. You need to practice on the slower speeds to be able to discern the opening spike, the closing spike and the noise afterwords. It's fairly easy.

Here are my first try results:

On speed 1 = 1,750mS or 1.75 seconds. When I used my second hand on watch, I saw this as 2 seconds.
On speed 2 = 1,000mS or 1 second. About the same as I saw on my watch.
On speed 10 = 200mS or 0.2 seconds. I was not able to measure that with watch.
On speed 100 = 22mS or about .02 seconds.

I tried several attempts to measure 250, but was unable to separate noise from the shutter spikes.

What this showed was that my shutter is running at about half speed up to 100. It's a more precise method than using a watch. And, you can measure up to at least 100 speed.

All this was what I would regard as minimal research on techniques. I am sure there are better APPS, better methods. But the concept I was chasing was: How NOT to spend a lot of money for as "shutter speed timing device." Yes, you could build a photodiode circuit and use other apps to achieve a light-based in place of sound based result, but I'm in the summer cabin and have no solding gear with me.

End result? I am going to use my camera now by stopping down 1 full stop. Easy.

If you have tried any of these APPs, please report!

 

[paul ron]

the sound of a shutter firing is not an accurate way to measure speeds.

the actual light passing through the shutter is the only proper way to measure a shutter's speeds.

not all shutters sound the same and considering all the moving parts in them, what exactly are you hearing?

find an app that uses your camera in the phone.


.

 

[ReginaldSMith]

Right. It's not precise in the sense of using it as a shutter adjustment tool. But, once you see the X-Y trace it's fairly clear to discover OPEN and CLOSE spikes. The two slowest speeds nearly matched my measurements using a second hand and visually watching the open/close.

 

[shutterfinger]

And the second hand check is just a rough estimate when one has no other means available.

 

[GRHazelton]

I seem to recall folks using a phonograph turntable to measure shutter speeds in the film era. IIRC a cardboard disc with a radial line was photographed. The angle described by the line as it rotated could be measured and the shutter speed then calculated. Given that most TTs had 33 1/3 rpm, perhaps 45 and 78.26 (and rarely 16 2/3) speeds also, the slow speeds couldn't really be measured, but the slower speeds could be "guesstimated" by looking through the camera. I'd imagine that using a stopwatch to time the light flash might do well, assuming that your reaction time was more or less constant.

 

[Luckless]

Audio isn't a great check of accuracy, but it is a useful check of consistency
- If a dozen trips of a shutter speed give you the same duration, then you can be confident that each time you trip the shutter you'll get the same exposure as per the last time you did it. If not, then something has worn, is dirty, or something else that needs addressing.
- If the difference from each shutter speed is dropping/rising roughly as expected from one setting to the next, then you can be a little more sure that things labelled a stop apart are likely to double or half the exposure.

For actual duration testing, I prefer using a light based timing circuit. (If you're going to make your own, be careful on photo diodes and stuff from cheap kid's kits. I've found many of them to float or have poor response times, making accurate measurement sketchy. Double check on forums before ordering parts.)

 

[GRHazelton]

Out of curiosity how do you measure the shutter speeds of a vintage Leica or the clones, with no easy access to the film plane? Hard for me to imagine a probe with light sensitive cells that could be used without risk to the shutter curtains....

 

[GRHazelton]

I ran across this just now. https://www.lomography.com/magazine/233990-shutter-speed-tester-for-your-iphone It uses the sound as described above, but includes a cheap circuit using a photoresistor to measure the light coming through the shutter. Looks good to me!

 

[GRHazelton]

Just mooching around the 'net I found this: https://www.ebay.com/itm/Camera-shu...033609?hash=item25f6c71ac9:g:6N0AAOSwNgxWEWrJ This version is for PCs, the seller has another for Macs. At the price, and given good user reviews, I decided to give it a try, it will be interesting to see how my various cameras test out.

 

[paul ron]

that circuit is probably similar to the one you posted for a light detector system... much better and ill bet more relable.

 

[ReginaldSMith]

I suspect that if you haven't used an oscilloscope much, and/or viewed the capture traces of the shutter release on one, it might seem improbably inaccurate. However, once you see the very explicit detail of the trace and understand the peaks and decays, you might change your mind. Again, this is not an adjustment tool, it's an assessment tool, that unlike all the visual stop watch methods allows for testing up to 1/250 with an accuracy I would say might be 20%. And, well, it's essentially free and requires no tools, circuits, batteries etc. Just a phone.

Right now that camera I used (Konica III) is loaded with film. But when empty, I will do the test again and take photos of the scope trace so people can see how it works. I might also try it with my new Hassey. I think I can lock up the mirror and back curtains to where only the lens shutter will fire.

When back home with my workbench and soldering tools, I will build the photo-electric timer and digital display that will read with uS resolution - - essentially 100% accurate.

I love the description of the turntable device!

 

[Luckless]

As someone who spent awhile playing with sound based solutions to measure shutter speeds before building custom light based units, I still have to say that sound alone was not great at accurately measuring duration. Picking out where the shutter actually opened is difficult, as sound on its own doesn't tell you anything about ramp up or ramp down within the mechanisms of the shutter, but merely "Here is where something started making noise, and here is where it stopped." - Where the light starts hitting the film takes more measuring.

Out of curiosity how do you measure the shutter speeds of a vintage Leica or the clones, with no easy access to the film plane? Hard for me to imagine a probe with light sensitive cells that could be used without risk to the shutter curtains....

Some shutter setups get a little more 'interesting' to test than others. Which is why I love working with cameras that have nice easily accessed leaf shutters. But for cameras that don't make poking around on either side of the shutter that easy, then one way to tackle it is to slip white paper or a reflective target into the film plane, and use a sensor that will measure the difference in reflected light. You'll get a spike on the reading while the shutter is open. As long as there isn't anything else in the lens/camera assembly that could cause weird reflections, back at the sensor, then you should be able to pick out a clean rise and fall point to go off of.

But oddly enough I've seen a few people rigging together high speed cameras for this sort of thing now rather than simple optical on/off or level sensors, simply because they've become something that is cheap enough that people are buying for other projects, and therefore have on hand. Technology gets weird at times.

 

[ReginaldSMith]

As someone who spent awhile playing with sound based solutions to measure shutter speeds before building custom light based units, I still have to say that sound alone was not great at accurately measuring duration. Picking out where the shutter actually opened is difficult, as sound on its own doesn't tell you anything about ramp up or ramp down within the mechanisms of the shutter, but merely "Here is where something started making noise, and here is where it stopped." - Where the light starts hitting the film takes more measuring.



Some shutter setups get a little more 'interesting' to test than others. Which is why I love working with cameras that have nice easily accessed leaf shutters. But for cameras that don't make poking around on either side of the shutter that easy, then one way to tackle it is to slip white paper or a reflective target into the film plane, and use a sensor that will measure the difference in reflected light. You'll get a spike on the reading while the shutter is open. As long as there isn't anything else in the lens/camera assembly that could cause weird reflections, back at the sensor, then you should be able to pick out a clean rise and fall point to go off of.

But oddly enough I've seen a few people rigging together high speed cameras for this sort of thing now rather than simple optical on/off or level sensors, simply because they've become something that is cheap enough that people are buying for other projects, and therefore have on hand. Technology gets weird at times.

Actually, the captured wave form tells everything about ramp up and ramp down that is happening with the shutter, from the beginning to the end. There is a very rich and detailed fingerprint of the entire operation start to finish. Hard to put into words, so I will wait until I can demonstrate with as 'scope photo. There is 20dB or more of differential amplitude that reflects the different moving parts and operations. Very easy to see.

 

[shutterfinger]

Just think if you had to rely on one of these:

 

[M Carter]

I have my own DIY/Ghetto speed test - one of my video cameras shoots at 120FPS@1080p; these are full frames, not interlaced. So I can check speeds up to about 1/100th at least. Shoot a few actuations, count the frames, do the math. Really handy for old single-speed box cameras, and since most of my leaf shutters seem to get weak on the slow end, it's better than nothin'. And it's a nice way to test flash sync, if you've altered an old camera's flash contacts to hit earlier for electronic flash. Actuate the shutter while attached to a strobe and check for fully-open when the light spikes.

One interesting thing when you actually watch shutter actuations at 1/120th of a second - is the shutter speed counted only when the shutter is fully open? You can clearly see in some cases the very beginnings of the shutter opening, followed by almost-open, than many frames fully open, and the inverse - a half-open and then barely opened or fully closed.

I'd guess that for accuracy's sake, if the beginning and ending frames are half-open, that counts as 1 frame's worth of light hitting the film. But generally I find that counting only the fully-open frames of a healthy shutter add up to the proper time. I'll have to try it with a more modern electronically controlled shutter (Nikon N90s, etc) and see how that plays out.

Anyway, it's kind of a fun way to test shutters.

 

[ReginaldSMith]

Great system! Thanks.

The partial stages raises an interesting question for determining the actual measurement. On my sound/scope system, it's easy to see that about 10% of the time between full open and full close is taken up by partial open/close. I think therefore the calculus is needed to integrate the area under the curve to find what the equivalent time would be if the open and close were infinitely fast. Thus, a 1/100 exposure is probably intended to mean 100% fully open for 100mS and no partials! I'm not sure.

 

[Eric Rose]

Back in the 70's I made one that used a light source and a photo cell plus timer circuit. At the time the fastest shutter was 1/2000 and it had no problem testing them. Should be easier to fabricate something like that these days.

 

[Chan Tran]

Just think if you had to rely on one of these:
View attachment 201238

Actually it's a good idea. May be I can build one to test shutter speed of digital camera.

 

[shutterfinger]

The partials of the constantly changing aperture from full closed to full open and the reverse is only a small amount of the total open time. The best I have been able to time is a leaf shutter goes from full closed to full open in 0.0002 to 0.0003 seconds. I have not timed full open to full closed but its likely the same or faster as the closing is done by the cocking spring tension.
Using this optical tester with Audacity audio program https://graflex.org/helpboard/viewtopic.php?t=6105 , a AA Mini Mag set to bright center beam set on the shutter, a Compur 00 shutter from a Rolleicord II freshly CLAed we get

for the shutter at 1/100. The vertical line is at the point where the sensor first sees light.

for the shutter at 1/300 , again the index line is at the start of the sensor seeing light.
The sensor is at the center of the shutter opening. For better testing of the open and close one would need two sensors, one at the center and one at the outer edge of the shutter's opening recording in stereo then measure the time lag between the sensors.

I recently put a Rolleiflex Automat back into service. I did a CLA on its shutter and it tested in tolerance at all speeds with the linked tester in this post. A roll of TMax 400 and out of date Portra 160 exposures were spot on as shown on page 2 of
https://www.photrio.com/forum/threads/rolleiflex-question-s.158678/ . I made sunny 16 and metered exposures of the same scren on the T Max and the densities were near identical.
I have tested shutter speeds at 1/1000 with this tester and its accurate.
The time from full closed to full open and full open to full closed is around 1% to 2% of total exposure at wide open so they have an insignificant effect on total exposure as wet printing needs a 1/3 stop difference to see in the print.
I'm 45pss at graflex.org.

 

[GRHazelton]

Speaking of the fastest speeds for leaf shutters, I seem to recall a 35mm camera some time in the late 50s or early 60s which offered perhaps 1/000 sec at smaller apertures. IIRC the shutter blades rotated 360, thus they didn't have to open, stop, and then accelerate to close. I'd like to hear from members who can tell us more, since I likely have some errors in my memory. I do have a WerraMatic with a leaf shutter which offers 1/750 through some sort of auxiliary shutter blades. I look forward to testing it when my Gizmo arrives!

 

[shutterfinger]

The Graflex 1000 uses 4 shutter blades mounted on eccentrics with a continuous circular drive gear. It was introduced around 1960.

 

[btaylor]

That’s one of vfmoto’s testers on eBay. I have 2 of his stand alone testers, I have found them to be really indispensable with all the old leaf shutters I have around. I like the stand alone versions because they are so easy to use, I don’t need to hook anything up. I recently got a bunch of weird density variations with a rolleiflex tlr. Five minutes with the tester and I confirmed an inconsistent shutter.
Also, we all know a perfectly consistent leaf shutter is usually off significantly at the higher speeds, an easy to use tester will tell you how far, and the appropriate marked vs real compensations can be made.

 

[ReginaldSMith]

Shutterfinger,
Nice information. I could not open the link. How does Audacity convert the optical to audio? Maybe I misunderstood.

 

[shutterfinger]

Audacity was not linked previously so here it is: Dead Link Removed
When light hits the photo LED (formerly photo transistor) it starts conducting current which shows up in Audacity as a negative dip from the center line. Think of the transistor/led output as a 0 hz signal. Not holding the light source results in a jagged response that sounds like static if played back through speakers or headphones. Human hearing is 20hz to 20,000hz unless there is damage to one's hearing.
Tester discussion: https://graflex.org/helpboard/viewtopic.php?t=6105

 

[ReginaldSMith]

Audacity was not linked previously so here it is Dead Link Removed
When light hits the photo LED (formerly photo transistor) it starts conducting current which shows up in Audacity as a negative dip from the center line. Think of the transistor/led output as a 0 hz signal. Not holding the light source results in a jagged response that sounds like static if played back through speakers or headphones. Human hearing is 20hz to 20,000hz unless there is damage to one's hearing.

Ok, thanks. Got it.