Measuring your shutter speed without any new equipment

[KarlF]

I think that I might have found a way to measure the speed of a LF lens shutter without buying any equipment ( if you already own a digital slr). I tested it and it worked for me. I am submitting here for your review.

- Have a controlled light source.

- With your large format point to a fixed object lit by the controlled light.

- Open the LF shutter and leave it open

- put the digital camera inside the LF and shoot through the LF shutter and lens at 1/10th, 1/20th, 1/30th, 1/60th. You will see a round light spot. If you are measuring for a speed of 1/30th for example, make sure that at this speed, the light spot is greyish ( not too bright)

- Then close the LF shutter, put the digital camera in B and open it, trigger the shutter of the large frame.

The picture taken should be matched to the pictures taken with the digital camera speed.

So if the picture of the spot timed by the LF shutter most closely resembles the 1/20th speed of the digital camera shutter then your LF shutter is at about 1/20th. you can do the compare directly from your digital SLR lcd screen.

This is how I confirmed that my Packard shutter is shooting at 1/10th not 1/25th as it should ;(

I think that this method should be more accurate than anything based on sound because it measures the actual total light that goes through the shutter. It also takes minutes. Finally, no need for a computer so it can be done on the road.

Let me know your thoughts or if this is not clear enough.

Best Regards,
Karl

[Dirk Rösler]

This is great, thanks for sharing the idea.

[Bob Salomon]

How do you know your DSLR shutter speed is accurate? To be considered accurate a shutter speed can be ±30% of the marked speed.

[KarlF]

Hi Bob,

I knew that this question would come up so thanks for giving me the chance to answer it. The electronic shutter of a modern digital DSLR is partialy based on mechanics but mainly based on electronic timing and is typically extremely accurate. Certainly more accurate than I would ever need for my LF.

Best,
Karl

[Dirk Rösler]

Does anyone really think that in 2009 a DSLR's shutter speeds would be inaccurate? Quartz timing etc have been around for 20-30 years now, infinitely more accurate than the mechanical stuff (not that it has to be super precise anyway, see Leicas).

[Bob Salomon]

Does anyone really think that in 2009 a DSLR's shutter speeds would be inaccurate? Quartz timing etc have been around for 20-30 years now, infinitely more accurate than the mechanical stuff (not that it has to be super precise anyway, see Leicas).

Electronically timed and actuated but carried out mechanically by that shutter mechanism. Still can be ±30%.

[Steaphany]

A method I've used years ago was shooting a frame from the screen of an analog Television. This method works for anything faster than 1/30th. The video is interlaced so two scans make up a complete image.

If your test shot is between 1/60th and 1/30th you'll see an image across the whole screen, but part will be brighter than the rest which is where the second field is being captured.

If your test shot is less than 1/60th you will only capture part of the screen with an image.

To determine shutter speed, compare the vertical image size to the total screen height.

A 1/100th shutter would capture an image that covers 60% of the screens vertical
A 1/125th shutter = 48%
A 1/250th shutter = 24%
A 1/500th shutter = 12%

A 1/50th shutter = 120% - a Full Screen image with 20% showing the brighter interlaced area

To calculate the screen image percentage that a shutter speed corresponds to use:

100 * ( Shutter Speed ) / ( 1 / 60 )

Where Shutter Speed is the actual fractional time, i.e. 1/100, 1/125, etc.

A note, I haven't tried this with a new digital TV.

As an alternative, if you have or know someone with an oscilloscope, shoot a test image of the scope screen as a sweep of the screen is made. Here you can control the speed for very slow shutters, less than 1/30th, and knowing the time base of the scope, you'll be able to read the shutter speed right off the test image.

[BetterSense]

Electronically timed and actuated but carried out mechanically by that shutter mechanism.

I think if you understood the mechanism of most dual-curtain FP shutters you would agree that the shutter speed is entirely under control of the electronics, actually. The first curtain is released, an electronically determined amount of time is allowed to pass, and then the second curtain is released. A variable or malfunctioning shutter mechanism will generally cause uneven exposure, not evenly slow or fast exposure. And since digital has less latitude than color negative film, if the digital cameras is working at all the shutter must be within spec.

[Dan Fromm]

I think if you understood the mechanism of most dual-curtain FP shutters you would agree that the shutter speed is entirely under control of the electronics, actually. The first curtain is released, an electronically determined amount of time is allowed to pass, and then the second curtain is released. A variable or malfunctioning shutter mechanism will generally cause uneven exposure, not evenly slow or fast exposure. And since digital has less latitude than color negative film, if the digital cameras is working at all the shutter must be within spec.Interesting theory. Tell that to my oh-so-slow Ilex electronic shutter. Its timer energizes a solenoid to open it, after a predetermined interval has passed turns the power off and a spring closes it. Can't possibly go wrong, but it does, and consistently.

Interesting theory. Tell that to my former F135s' solenoid-actuated shutters. One solenoid to open, another opposing one to close. Bang! Bang! All it takes is a sticky solenoid.

Interesting theory. Timing circuits can be off, curtains or metal blades (think Copal Square Shutter and descendants) can move more slowly than expected, springs can be weak, solenoids/magnets can be sticky.

Many are the ways in which an electronically timed shutter can go wrong. Not that clockwork is necessarily any better, even in the cine cameras (and a few 70 mm aerial cameras too) whose shutters just go round and round.

[Tiziano]

I use my digital video camera for measuring 1/30 or longer times. This camera takes a digital video at a speed of 30 frames/second. I first record the shutter opening and closing, and than watch the video with a simple video editor software that will let you go forward one frame at a time.

[Ivan J. Eberle]

Can't say for sure that the newest Nikon DSLRs do this (expect that the pro models all do), but going back to the F5 the shutter was self-diagnostic. If the actual shutter time is off, the camera throws a code and reports the malfunction.

[Bob Salomon]

Can't say for sure that the newest Nikon DSLRs do this (expect that the pro models all do), but going back to the F5 the shutter was self-diagnostic. If the actual shutter time is off, the camera throws a code and reports the malfunction.

But that brings you back to a correct shutter speed is ±30%.

[Ivan J. Eberle]

But that brings you back to a correct shutter speed is ±30%.

Is it your assertion, Bob, that an electronically-timed DSLR shutter can be no more accurate than this commonly-agreed-upon acceptable 1/3 stop error in LF shutter timing?

Because I'm not buying that.

[Bob Salomon]

Is it your assertion, Bob, that an electronically-timed DSLR shutter can be no more accurate than this commonly-agreed-upon acceptable 1/3 stop error in LF shutter timing?

Because I'm not buying that.

I am stating that the industry standard for camera shutter speeds are ±30%.

[Marko]

How about non-slr digital cameras? As far as I know, they are completely electronic and do not have mechanical shutters at all.

[John MacManus]

Karl

This link may interest you:
http://decisivemomentum.blogspot.com/2008/07/using-dslr-as-shutter-tester.html

The fella uses the histogram to pinpoint the equivalent DSLR shutter speed. I've used this to estimate the speeds available on my 4X5 and 5X7 Speed Graphics; a bit tedious but effective.

Happy New Year ... J

[Ivan J. Eberle]

Or, instead of reinventing the wheel, there are numerous sites that show how for less than $20 and an hour's time, you can build your own precise photo transistor shutter tester which plugs into your sound card port on your PC or laptop and where to download a free piece of software to run it. Extremely accurate timing with an dead-simple, easy-to-build device. Results are graphically stored and interpreted via any of several sound editors ( though the "Sound Card Shutter Tester" is a misnomer as it doesn't sense sound; the sound editor software is just a very convenient way to quickly record and compare the results).

Also a bunch of recent threads discussing the topic here on the LF forum.

[KarlF]

Hi Ivan,

I am aware of the sound card test, but in all humility, I think that it is actually less accurate. Here is why:

At the risk of upseting some, we actualy don't really care about the shutter speed, we only care about how much light goes through the shutter.

A shutter that opens 125th of a second will let go different amounts of light depending on a lot of parameters. One of them is the speed at which it reaches full open and the speed at which it closes back. If it fully opens very fast, stays open for 125th/s then closes very fast, it will transfer a different amount of light than if it opens slowly and closes slowly. Is the wave measuring the light that goes through the shutter described more as a square or more as a semi-circle. For the same speed, the semi-circle will let less light through.

Put in another way, for the sound card test to be accurate, you would have to compute the 'surface area' under the wave described by the shutter, then do this in different positions, not just the middle of the shutter, finally find out how this surface area translates to the actual amount of light going through ( or calibrate your photocell).

The sound card test would be accurate if the shutters opened infinitely fast waited 125th/sec then closed infinitely fast. But that is never the case.

The test I describe, measure the total amount of light received, regardless of how the shutter behaves. Single curtain or double curtain, fast or slow. Thus making the test more accurate and a lot more simple.

I am of course open to rebuttal.

Thanks,
Karl

[Nathan Potter]

Karl, your point is well taken. We do care about the "dose" of exposure that the film receives. The "dose" is the light intensity X time (lux*sec.). A variable in leaf shutters clearly could be the velocity of opening and closing of the leafs (leaves?). The moving slits, at high speeds, of focal plane shutters are another special case of this intensity*time product. Thus it seems there may be a terminology glitch here in that we should be concerned about "dose" rather than shutter speed. We should leave the intensity variable "f/stop" out of the equation altogether in considering your comment.

Seems like you have a good point but I don't know how significant is the velocity of the shutter leafs as a function of the total open period of the shutter. Certainly irrelevant at long period shutter speeds. Is this phenomena taken into account at the factory when designing and adjusting the shutters?

Nate Potter, Austin TX.

[KarlF]

Hi Nathan,

I see what you are saying but I am not talking about the aperture at all. What I am talking about is the time that the shutter spends being partially opened or partially closed.

If a shutter opens instantly fast, stays fully open for 1/125 second and then closes instantly fast then we can say that there is full use of the 1/125 second.

On the other have is a shutter spends 1/400th opening ( period during which not al the light is going through) then another 1/400th closing, we can say that for 1/200th of the time, the film is not getting full light from the lens.

At the factory they will compensate for it and give the shutter a little more time to do these two operation. Let say 1/90th for the whole thing.

So for this particular shutter it needs 1/90th to produce the light of 1/125th. Calibrating this shutter to 125th with a sound board will actually hurt you by underexposing the film.

This hold true for all shutters, none of them moves infinitely fast. I would therefore say that sound boards are not good for calibration. The larger shutters the worse the problem.

Do the seniors agree ?

Happy New Year,
K

[Peter K]

The larger shutters the worse the problem.
What problem?

Normaly a shutter is used to expose a light sensitive material like film to the light formed with an optical system. And if you will see, or measure, no difference in density also if the shutter has a tolerance of 30 % there is no need to hold the tolercances below this value. It's much more important a shutter works well in different enviroments like different temperatures, humidity, dust etc.

And for other applications like for very short exposure-times there are shutters without any moving part aviable like "Kerr's cell". Or with very small moving parts like the "light valve" used for the sound of movies.

Peter

[BetterSense]

I agree that in principle the DSLR test is more accurate, since it actually is integrating the light coming in the lens. I wouldn't have thought of such a clever way to achieve this!

Practically, my opinion is that the sound card test is well within accuracy expectations most of the time. At high (200+) speeds, large apertures (which does change the 'effective' shutter speed), or large lenses, simply measuring the 'open time' could be inaccurate for exposure purposes. Probably not drastically, though.

[Ivan J. Eberle]

I just think you're making it harder on yourself than you need to, that's all.

How are you comparing multiple exposures to gauge shutter timing consistency? Is it fast and easy? (It sure is slick and dead simple to measure the mean shutter open to mean shutter closing duration via the graphical interface in Audacity.)

I'll be the first to admit that integrated light readings could be useful. However, they'd more useful, I'd think, to determine your real f/stop ratings once you know you have a well-timed shutter.

Film was traditionally used for this purpose. Probably still most useful for determining the system integrity (i.e. light leaks).