PDA

View Full Version : Leaf Shutter Speedometer Project



Neal Chaves
14-Mar-2016, 19:37
Yesterday member Drew Bedo sent me a PM asking if the circuitry in my Ghostmaster device might be of use in making a shutter speed checker. I wrote back that I did not think so, but at 4AM I awoke with the following idea.

This strategy is limited to x-synch leaf shutters. I am sure I could write up a patent application for this in few hours and be successful, but my previous experience with patents is they offer little advantage in a niche market. Once we start discussing it here, the invention is "open source" and cannot be patented.

The blades in a correctly functioning leaf shutter take only a few milliseconds to open or close(If not it needs a CLA). The electronic flash contact contact points close when the blades have flown fully open and open when they start to close. Therefore, the time of contact closure is very close to the exposure time. There should be some very useful information about the actual shutter speed Vs. that indicated if we connect an appropriate circuit to the electronic flash contacts. I know that this could be easily measured with an oscilloscope.

The scope will also show that the contacts bounce open again for a few micro seconds after closing when the shutter opens. Any circuit for measuring contact dwell time must be able to deal with this. Accurate measurements at very high speeds 1/250-1/500 may not be possible.
However, at the speeds at which most LF photographers work, less than 1/125, this approach could be very useful.

One advantage is that the measurement could be easily made with the lens in place on the camera, just before or even during actual film exposure. Perhaps the device could be incorporated in a cable release which also had an X-Synch connection.

Drew mentioned a possible smart phone interface. This could be wireless I suppose.

Any interest?

Jody_S
14-Mar-2016, 21:14
The electronic flash contact contact points close when the blades have flown fully open and open when they start to close. Therefore, the time of contact closure is very close to the exposure time.

I'm not convinced this is 100% true, especially with shutters that have M/X syncs. Also, using a cheap piece of photo-voltaic material, it is possible to make a 1-piece shutter speed tester that plugs into the jack of a smartphone (using the microphone contact), that can also be used on roller-blind shutters including early Leicas. It's just the photo-voltaic material with a lead and the correct jack soldered on. I'm not sure what the advantage is, as the x-sync tester would at a minimum require a power source (battery) whereas the photo-cell version does not (the photo-voltaic element generates enough voltage to register as microphone output).

Randy Moe
14-Mar-2016, 21:19
Sounds doable and useful.

Since I just ruined 8 sheets because my DIY Packard sync failed at f32, I would also like a way of correlating full open shutter to sync which I guess is just the other side of your coin?

Dual trace Ocilloscope App?

barnacle
15-Mar-2016, 02:00
Neal, isn't your widget just testing the duration of the flash sync signal? Or have I misunderstood your post? I think you'd really want to know the duration of the actual opening (and to some extent the profile of the light curve), and for that I suspect you're better off with a light sensor rather than looking at the trigger signals - which after all may not, on a lens in need of servicing, occur quite when you expect them.

Neil

Randy Moe
15-Mar-2016, 04:22
Anybody use one of these?

SainSmart NEW Nano ARM DSO Note II Portable Mini Handheld Touch Screen Digital Storage Oscilloscope

Drew Bedo
15-Mar-2016, 05:59
Neal: Thanks for kicking this off with a serious discussion and a sophisticated approach.. I brought this subject up maybe two years ago and got some dismissive responces on another website.

I have no engineering or electronic project experience, so I deffer to about anyone else on theory or technical practicalities. But I would like to expand on my idea:

My concept is much like Barnicle's—a light sensor (photo-voltaic or CdS) senses the light and generates a signal. This is turned into a tone whih is analysed by an audio program. The curve of light intensity over time is the data set, but the curve doesn't need to be displayed only a usable shutter speed. The application on the phone can be tweeked to display a useful shutter speed from this curve.

I envision a "widget" that mounts to the front of the lens like a clip-on or slip-over lens cap. The signal could be sent to the smart phone or tablet via cable or Bluetooth. This will allow determining variability in shutter speeds when in deep cold or high heat/humidity as conditions change. A measurement could be made in the field prior to taking a shot and that info could be used in making exposure decisions.

It is my hope that someone will publish a circuit diagram and parts list for the widget, and perhaps someone else will develop the app or apps that do the processing. This could be used with any film camera I think. There may be some processing trick for focal plane shutters.

rbultman
15-Mar-2016, 06:50
Anybody use one of these?

SainSmart NEW Nano ARM DSO Note II Portable Mini Handheld Touch Screen Digital Storage Oscilloscope

Randy, I'm getting "invalid URL" from the link you sent.

Tim Meisburger
15-Mar-2016, 07:24
Neal, I prefer your idea. Setting up the photovoltaic tester takes time, and is not really something you could do int he field. On the other hand, the iphone tester that works off sound is easy to use in the field, and maybe accurate enough for our needs?

Randy, what kind of sync are you using on your Packard? All I have is a piece of bare stranded copper wire that the piston hits at full extension. It took about ten minutes to make with hot glue, and looks it, but it fires reliably.

MartinP
15-Mar-2016, 08:01
Randy, I'm getting "invalid URL" from the link you sent.

The 'correct' link appears to be this (http://www.sainsmart.com/sainsmart-2015-new-ds202-nano-arm-pocket-mini-nano-touchscreen-oscilloscope-dso202.html).

A shutter timer can be a very, very simple thing. Here (http://www.baytan.org/prak/shutter.html) is a simple option, and another (http://www.mraggett.co.uk/shuttertester/shuttertester.htm), that plugs in to the microphone port on a computer and can be 'displayed' using, for example, a simple audio-editor such as Audacity. The signal from the device is a voltage under 1Volt, just like that from a microphone, so the audio-editor is effectively a pseudo-oscilloscope in this case.

Edit: Don't bother messing around with a laser pointer as mentioned in the links. Just use a decent LED torch or bike-light -- the wider light source is MUCH quicker and easier to set up.

Neal Chaves
15-Mar-2016, 08:58
I'm pleased that there has been such interest and that some members seem to have a good grasp of what I am proposing. The next step is to test a shutter with an o-scope and determine how closely X contact dwell time relates to shutter speed. Shutters are calibrated to account for exposure produced at opening and closing, but at slower speeds using the low speed gear train this effect is a negligible percentage of total exposure.

Drew Bedo
15-Mar-2016, 09:06
What about building a photo-voltaic detector and electronics for the shutter tester into an old Kodak film pack holder? Would there be enough volume there? It would make a durable and convenient package for field use.

Another question: I am not clear on the x-synch approach. What plugs into the PC fitting on the shutter? Is there a blue box with electronic components or does the smart phone plug right in?

Neal Chaves
15-Mar-2016, 09:10
That little scope is very cool. I'm going to buy one myself.
http://www.sainsmart.com/sainsmart-2015-new-ds202-nano-arm-pocket-mini-nano-touchscreen-oscilloscope-dso202.html

Neal Chaves
15-Mar-2016, 09:13
Drew, as I suggested to you privately, if you must measure the light transmission through the lens, a digital body on an adapter for the view camera and a gray card will do it. Using the X-synch approach, the circuit would essentially be a digital stop watch that starts on contact make and stops on contact break.

rbultman
15-Mar-2016, 09:19
What about building a photo-voltaic detector and electronics for the shutter tester into an old Kodak film pack holder? Would there be enough volume there? It would make a durable and convenient package for field use.

I like this idea.

el french
15-Mar-2016, 11:47
Put one of these (https://www.sparkfun.com/products/10266) in back of the shutter and one of these (https://www.sparkfun.com/products/9349)in front. Program an Arduino (or other micro controller) to capture the interval.

Randy Moe
15-Mar-2016, 12:50
That little scope is very cool. I'm going to buy one myself.
http://www.sainsmart.com/sainsmart-2015-new-ds202-nano-arm-pocket-mini-nano-touchscreen-oscilloscope-dso202.html

I agree, finally something affordable, tiny and it seems we can add a monitor.

Randy Moe
15-Mar-2016, 12:55
Neal, I prefer your idea. Setting up the photovoltaic tester takes time, and is not really something you could do int he field. On the other hand, the iphone tester that works off sound is easy to use in the field, and maybe accurate enough for our needs?

Randy, what kind of sync are you using on your Packard? All I have is a piece of bare stranded copper wire that the piston

hits at full extension. It took about ten minutes to make with hot glue, and looks it, but it fires reliably.

Mine is similar and has been working well but put a beautiful model in front and it failed to trigger as expected. It did flash my 4 strobes, but maybe late. I need to find the problem. I will. Thanks for suggestions. The same strobes worked perfectly with exact same DSLR settings. Check my tiny format portrait I just posted.

domaz
15-Mar-2016, 13:46
Put one of these (https://www.sparkfun.com/products/10266) in back of the shutter and one of these (https://www.sparkfun.com/products/9349)in front. Program an Arduino (or other micro controller) to capture the interval.

Or just use an Arduino and a simple photocell (https://www.sparkfun.com/products/9088) it should pick up ambient light enough to get a response, especially if yout put a flashlight in front of the shutter. You may need to find a photocell with a high enough response though.

jp
15-Mar-2016, 14:51
Ive got an o scope like that. Handy but battery life is poor. Keep a phone charger nearby. Put a small inductor inline and that should provide a good trigger pulse.

Jim Jones
15-Mar-2016, 18:06
Sain Smart and others have produced a variety of such devices, some even less expensive than the model linked to in post 9. Many of these are used in conjunction with a laptop. I haven't seen enough information or reviews yet to even guess if they would be the basis of an effective shutter speed tester. My choice would be a triggered sweep oscilloscope with a light sensor. This can check between-the-lens and most focal plane shutters for effective speed, flash sync delay, shutter bounce, and other characteristics. 43 years ago, while in a remote overseas location, I used a basic Heathkit X-Y scope for most of these tests. Accuracy was excellent.

Drew Bedo
15-Mar-2016, 18:26
OK; so it seems a digital mini oscilloscope is the key component. Does it plug into the PC socket? How does this interface with a smart phone application to give a shutter speed?

Neal Chaves
15-Mar-2016, 19:42
That little o-scope has hundreds of functions we don't need for this application. We just need a digital stop watch and circuitry to start and stop with X-synch contacts.

Randy Moe
15-Mar-2016, 19:58
That may be, however I dislike buying specialty tools, when a multi use tool costs about the same and does 100's of other tasks I will have for it.

How much have many of us already spent on the last shutter speed app and dongle? I think about $30+ all in with APP, sensor and necessary extension cable.

I also want to measure flash bulb duration. For starters...

I'm not suggesting the little O Scope is for everyone, but others need to step up and examine all possibilities. Or not.

MacGyver this as little clock kit, same company, they have oodles of cool stuff. (http://www.sainsmart.com/sainsmart-new-electronic-crystal-table-watches-diy-kit-for-scm-led-watch-digital-watches.html)

barnacle
16-Mar-2016, 02:05
To get a little technical here - the scope, and a light sensor, will tell you everything you need to know about the shutter opening times and the flash sync *if* you know how to interpret what you're seeing. You will also need a light source and some sort of *fast* light sensor; an opto-diode would probably be ideal though I haven't considered the matter in detail or done any tests.

Anything which uses a microphone to listen to the shutter opening and closing has the disadvantage that it's listening to what's happening and trying to deduce what's going on; the algorithm used may be more or less successful but it is certainly unlikely to be accurate at high speeds.

Anything which uses a light source, a sensor, and a microphone input on a phone or computer has a possible issue: most of these microphone inputs are AC-coupled and so produce a differentiated version of the signal seen. This is not insurmountable, but it should be borne in mind if you're considering this context. Calibration may also be an issue; see below.

Anything which is looking at the flash contacts is telling you about the flash contacts only.

Remember we are concerned with the integrated light passing through the lens: how much, and for how long. Iris-type shutters do not instantaneously open but expand from the centre, pause, and close to the centre. On a long exposure the time taken for this is effectively immaterial, but on a fast shutter - my fastest shutter claims 1/500th of a second, 2 milliseconds - if it takes a millisecond to open and another millisecond to close, then this is significant. Here's an idealised drawing of the light intensity as the shutter opens and closes:

148338

The area under the 'curve' is how much light is arriving at the film - the height times the length - so in this case, if we assume that the rise and fall times are both 1ms, our nominal 2ms opening is actually 3ms, and occurs over 4ms. This is the curve we need to visualise with any measuring device. To make life even more complicated, you might consider that different iris settings will have different rise and fall times, since the shutter has to clear a smaller space.

With a scope, light source, and sensor, you can make these measurements directly. You need to ensure that the fully exposed level does not clip as this will give an incorrect estimate of the rise and fall times, and you will need to take repeated measurements to check consistency. And finally, you need to know how to use the oscilloscope... The bonus is that the flash sync can be directly visualised on a second channel of the scope.

A suggestion was made upthread to integrate this measurement equipment into a film back. That intrigues me; perhaps something that slides in the place of the film back could perform the measurements automatically and transmit the results to a mobile phone. I need to think about that...

Neil

Drew Bedo
16-Mar-2016, 06:26
Sinar had (has?) focal plane spot probes that integrated with light meters from Gossen, Minolta and Broncolor. Is there anything to be gleaned rom how these instruments work?

plaubel
16-Mar-2016, 07:10
My one, a Gossen Spot for screen measuring only, is slow working.
It wouldn't work as a fast shutter measuring instrument.
The other problem could be a general delay between measuring and "showing" the value, - the Gossen works ultraprecise, but analog with a moving coin, so he is moving like a slowt, especially with low light.

Maybe a flash meter would work ?

barnacle
16-Mar-2016, 09:48
I don't think so, Drew. They are just measuring the intensity at a given point, which you'd have to do, but not timing. To measure a 1/500th shutter with any accuracy, you'd need to be counting microseconds or so, or at least tens of microseconds, and that's a job of work for the analog-digital converted fitted to most microcontrollers. To test across the range of say f5.6-f45 is seven stops, so there's an equivalent gain change required too - over a range of 1:128.

As I said in an earlier post, this is easy to do with an oscilloscope, if you know how to use it. It's not quite so easy to automate it, and both would certainly require manual input as part of the test.

Neil

Drew Bedo
16-Mar-2016, 12:46
Thanks Neal,

The state of my knowledge base and intellectual tool kit is obvious.

Yet, I have faith that something effective can be put together. Knowing what to do and how to do it is the thing.

Jac@stafford.net
16-Mar-2016, 13:36
[... snip good stuff ...]I also want to measure flash bulb duration. For starters...]

The literature for bulbs is good enough for that.

I don't understand what this whole 'speedometer' project is about. All one needs to know is that the flash is triggered at the right time, which for electronic flash is when the shutter is full open. We can determine this if the flash shows on the ground glass (or without) when triggered. It's a nobrainer, no? Or do I need a new brain? Yeah, mine is seventy years-old but it still works.

For flashbulbs, meter with a conventional flash meter using a PC cable to the meter. It works.

Jac@stafford.net
16-Mar-2016, 14:08
[...] Shutters are calibrated to account for exposure produced at opening and closing, but at slower speeds using the low speed gear train this effect is a negligible percentage of total exposure.

When using X-sync the flash is triggered when the shutter is wide full-open. No other calibration is necessary. When not using flash all you need is a stone simple shutter tester.
.

Randy Moe
16-Mar-2016, 14:18
Sometimes we want to climb the mountain already topped.

I had a problem with flash sync this weekend. I want to check my systems, simple as that.

I have a basic rule, never stop people from working, chasing a dream nor attacking windmills. Life is not a computer program, Mr Spock.

barnacle
16-Mar-2016, 14:21
OF course, the traditional way to test shutter speeds is to paint a dot on a 78rpm gramophone record (fortunately most of us here are just about old enough to know what one is!) and photograph it while it's turning. The length of the trace tells you how long the shutter was open, after a bit of maths - a full circle is 1/78 of a minute; 769ms. One degree of arc is 2.2ms, so you can use this down to about 1/500th of a second. Moving the camera slightly between shots means you can test a whole bunch of shutter speeds on one negative.

I have instructions to do this from a 1923 copy of Amateur Photographer!

Neil

Randy Moe
16-Mar-2016, 14:23
Wasn't there a way to use a CRT TV also? Perhaps just to check 1/60th?

Jac@stafford.net
16-Mar-2016, 14:26
I had a problem with flash sync this weekend. I want to check my systems, simple as that.


If you are using electronic flash, then look at the lens from the film side. Trigger the flash. If you see it you are good to go. If you are using flash bulbs, turn the shutter speed to 1/25.

Randy Moe
16-Mar-2016, 14:59
If you are using electronic flash, then look at the lens from the film side. Trigger the flash. If you see it you are good to go. If you are using flash bulbs, turn the shutter speed to 1/25.

I know that Jac.

Thanks for the advice.

barnacle
16-Mar-2016, 16:12
Yes, Randy: CRT TVs draw half the lines in 1/60th of a second for US/NTSC systems, and in 1/50th of a second for Euro/PAL systems, so you can calculate what proportion of the image is seen. The resolution is 64 microseconds per line for PAL, 63.5 microseconds for NTSC. The problem is that in both cases approximately twenty-five lines aren't shown so you need to adjust for the actual visible lines, and it's difficult to measure times longer than the 1/50th / 1/60th.

None of this works with modern LCD displays.

Jac@stafford.net
16-Mar-2016, 16:34
What problem is being solved by this thread? Is it not really an exercise in making conversation while sitting on our butts rather than making pictures?
.

Drew Bedo
16-Mar-2016, 18:03
Hi Jac@


The original concept was to come up with a electronic project that could be used as a shutter tester for use in view cameras while shooting in the field.

One version of this idea is components built into a film pack holder. Suggestions have been made to incorporate a digital mini-oscilloscope to make the measurement with input from the X-synch contacts or from a light sensor. Other suggestions include having a connection (Cable or Bluetooth) to a smart phone or tablet and an application to render the technical measurement as a shutter speed instead of displaying a curve.

The fundamental concept is a shutter tester that is convenient enough and rugged ehough to be used in the field prior to making decisions about exposure. Some combination of these and other ideas can come together to bring forward an effective, convenient and affordable shutter tester I am cure.

While I do not have the technical knowledge or skill set to make this happen, I am sure that there are others who can do the electrical engineering, project bench work and yet others who can creat an application for it.

Randy Moe
16-Mar-2016, 18:17
I got a PM to elucidate. Perhaps this explanations fails, but it's what i am thinking about.


The old iPhone app works fine, sort of. Very crude. What I want to see is actual shutter opening dimensions and exact sync time with duration compared to the shutter opening, resting or closing. 2 dwell times. 2 frozen scope traces.

Particularly a Packard has an odd set of leaves that obstruct light differently than leaf or FP shutters.



Additionally studio strobes have very short duration, perhaps 1/8000th second. Flash bulbs designed for FP shutters such as Speed Graphics, have long light production duration in seconds to accommodate synchronization with slow or fast moving slit shutters. The area opened to the film is a moving target. So is a leaf shutter, but they worked/perfected those longer, i.e. M, X settings and solenoids. Beat the Reaper methods. I also need that method.

Some Packards are tri-leaf shutters, others have 4 moving leaves, that have something in common with moving slits, however exposed Packard shutter light area changes radically in shape and speed. I see this shape on my incorrectly exposed film.


Maybe what I want is high speed video of whats happening. :)

Jim Jones
16-Mar-2016, 22:01
OF course, the traditional way to test shutter speeds is to paint a dot on a 78rpm gramophone record (fortunately most of us here are just about old enough to know what one is!) and photograph it while it's turning. The length of the trace tells you how long the shutter was open, after a bit of maths - a full circle is 1/78 of a minute; 769ms. One degree of arc is 2.2ms, so you can use this down to about 1/500th of a second. Moving the camera slightly between shots means you can test a whole bunch of shutter speeds on one negative.

I have instructions to do this from a 1923 copy of Amateur Photographer!

Neil

I've used this for between-the-lens shutters, but not for focal plane shutters. Since both the image and the shutter gap in a FP shutter are moving, there can be an error. Think of the photos of curved helicopter blades.

barnacle
17-Mar-2016, 00:07
Yes, for an FP shutter you'd need something moving linearly at right angles to the shutter opening, and you'd get a diagonal trace which you'd have to measure the length of in the shutter movement direction. Doable, but not so easy.

Neil

Neal Chaves
17-Mar-2016, 07:09
Your eye will retain an image of the shutter blades frozen by the electronic flash burst for 1/10 sec. Look through the shutter (cells removed) as the flash illuminates a wall. You will see the position of the blades as the synch contacts close. If you have a Ghostmaster you can see their position when the contacts open. Many large studio strobes can have a flash duration of close to 1/300 sec. You can easily cut off part of the burst if you use a high shutter speed like 1/250 or 1/500. The tail end of the EF burst is where most of the red component is, so by cutting it off you can end up with a colder color rendition.


I got a PM to elucidate. Perhaps this explanations fails, but it's what i am thinking about.


The old iPhone app works fine, sort of. Very crude. What I want to see is actual shutter opening dimensions and exact sync time with duration compared to the shutter opening, resting or closing. 2 dwell times. 2 frozen scope traces.

Particularly a Packard has an odd set of leaves that obstruct light differently than leaf or FP shutters.



Additionally studio strobes have very short duration, perhaps 1/8000th second. Flash bulbs designed for FP shutters such as Speed Graphics, have long light production duration in seconds to accommodate synchronization with slow or fast moving slit shutters. The area opened to the film is a moving target. So is a leaf shutter, but they worked/perfected those longer, i.e. M, X settings and solenoids. Beat the Reaper methods. I also need that method.

Some Packards are tri-leaf shutters, others have 4 moving leaves, that have something in common with moving slits, however exposed Packard shutter light area changes radically in shape and speed. I see this shape on my incorrectly exposed film.


Maybe what I want is high speed video of whats happening. :)

Jac@stafford.net
17-Mar-2016, 08:36
Thank you for the cogent explanation, Drew! It is an intriguing project. By gosh, I might even have a use for my Busch film pack adapter!

Randy Moe
17-Mar-2016, 09:29
Peter De Smidt suggests I use my D750 at 60P and record the action from the rear.

Good idea.

Perhaps tonight.

Video at 11.

wombat2go
17-Mar-2016, 11:13
I use a photodiode in photoconductive mode and an old Tek466 storage oscilliscope.
the p/d is arranged to receive a narrow beam, of scatter from the ground glass, usually.

This link should be Dir of scope traces and a summary sheet of
RB67 50mm shutter tests including variation over frame
https://app.box.com/s/6d06x1j0n6dm5et3jmypxpijpscdym8a

Here is Graflex Graftar f/4.5 103mm. The narrow spike of light is the electronic flash firing
http://www.largeformatphotography.info/forum/attachment.php?attachmentid=128540&d=1422319890

barnacle
17-Mar-2016, 12:28
Lovely demonstration, Wombat. The traces clearly show the opening and closing delay, the fact that it's faster on a small aperture (though I did not expect it to be quite so large a ratio), and the difference in signal height with iris.

It's clear that an automatic system will have to do some thinking.

Thanks.

Neil

Randy Moe
17-Mar-2016, 20:17
My video below of my problem.

During the video notice the flash is a quick white line across the DSLR field or lens, but not coming from the View Camera taking lenses, despite the strobe pointing at the view camera. The orange glow is the modeling lamp.

I deliberately operated the Packard slowly, also notice the PCB radio is on upper left and the red led always blinks late, just as it does when syncing with a DSLR. So ignore that.

I think I need to move the upper right micro switch lower to contact the piston sooner. This used to work, but perhaps has worn...

https://youtu.be/jzeLQES4XUE

Neal Chaves
19-Mar-2016, 08:05
My video below of my problem.

During the video notice the flash is a quick white line across the DSLR field or lens, but not coming from the View Camera taking lenses, despite the strobe pointing at the view camera. The orange glow is the modeling lamp.

I deliberately operated the Packard slowly, also notice the PCB radio is on upper left and the red led always blinks late, just as it does when syncing with a DSLR. So ignore that.

I think I need to move the upper right micro switch lower to contact the piston sooner. This used to work, but perhaps has worn...

https://youtu.be/jzeLQES4XUE

We are getting pretty far afield from the original idea I proposed, but it is good to have a complete understanding of X-synch challenges. One that I have encountered recently is that the new Norman power packs, 24/24 and newer may not have sufficient synch voltage to reliably operate older radios and IR sets. The guys at Holly Enterprises are aware of this and can change some resistors to increase synch voltage enough to operate these older remote sets but not damage other digital do-dads that may be attached.

Randy Moe
19-Mar-2016, 08:40
I know we are pushing this thread, but we may as well have one conversation. imho

I am going to make a better video that demos a few options.

Need a couple days for that.