Mamiya RB67 Lens Shutter tests

Alan Edward Klein · Nov 2, 2019

I decided to test my shutters to see how off they might be. They;re around 35 years old. The 90mm is an original lens and the 50, 180 and 360mm are "C" type bought new back then. I used my Samsung S7 Galaxy to record the sound. Then analyzed it using Audacity, a free audi app for my desktop. I watched this guy do it although he apparently used a dedicated microphone.
Over the years, especially during about a 15 year or longer period when I didn't use the camera, the lenses stayed cocked I believe. I've also had them uncocked for long periods as well. It doesn;t seem to have much of an effect as you can see from the results. The faster speeds were sometimes hard to see. The sound seem to get really condensed so it was hard to tell when the shutter opened and closed. The shutters seem pretty good at maintaining their settings and operation.

Please add any comments or testing you've done.

Alan Edward Klein · Nov 2, 2019

The first column in each lens groups is the shutter speed set on the lens. The second column is the length of the lens opened measured in Audacity. The 3rd column is the equivalent shutter speed fraction measure. It is calculated by dividing 1 by the time measure in column 2. Or = 1/time measured in the second column. The 4th column is the time of the shutter if it was equal to the shutter setting.

moto-uno · Nov 2, 2019

I enjoyed using this program , one of my most accurate shutters was on an old Welta Weltur ! Peter
The unit I used had a photo diode and with it you used the time between both peaks .

Alan Edward Klein · Nov 2, 2019

The peaks at faster speeds jus got blended in and there seem to be double peaks. It may be due to the cellphone microphone not being able to distinguish the sounds clearly at fast shutter speeds. Using a light probably is more accurate. But I figured I tried this method first. The shutters really seems pretty accurate. I was actually surprised considering how old they are.

Ariston · Nov 2, 2019

I tried doing this with an app and had no confidence in choosing where the shutter actuates and closes. I should try again.

AndyH · Nov 2, 2019

The shutter speed app provides a graph of the full opening, from beginning to end. Once you learn to read them, it's fairly simple. Second peak to third peak is the fully open phase.

Andy

Alan Edward Klein · Nov 2, 2019

Here are samples at 1/4, 1/30, 1/60 and 1/250 left to right, top to bottom. You can see it gets difficult to determined just where to manually highlight (that's the white fill) when you think the shutter is actually open. Easy at slow shutter. Very difficult at fast shutters. The numbers at the bottom define the length of the white selection. For example, it reads 00.279 in the 1/4 sec test upper left. But even at 1/30th on the upper right, it's hard to determine where to put it. The bottom two even are worse. It's probably due to the bounce and the sound being picked up by the poor microphone in the cell phone.

adelorenzo · Nov 3, 2019

I backed the Phocron shutter tester on Kickstarter. It's a really nice unit, if you have a lot of old cameras it's worth the price IMHO. I use it everytime I get a new camera or sell one.

https://www.phochronxa.com/

reddesert · Nov 3, 2019

A useful number to tabulate is (actual time / correct time). This gives you the ratio of exposure difference from ideal.

If you run the numbers through a calculator or program that does logarithms,
log( actual time / correct time) / log(2)
gives the over/under exposure in stops. So for example the 50mm at 1/125 had actual time 0.012 sec while correct is 0.008 sec. So it is overexposing by a factor of 1.5x, or 0.6 stops.

Alan Edward Klein · Nov 3, 2019

reddesert said:
A useful number to tabulate is (actual time / correct time). This gives you the ratio of exposure difference from ideal.

If you run the numbers through a calculator or program that does logarithms,
log( actual time / correct time) / log(2)
gives the over/under exposure in stops. So for example the 50mm at 1/125 had actual time 0.012 sec while correct is 0.008 sec. So it is overexposing by a factor of 1.5x, or 0.6 stops.

I was wondering if there was a formula. Could you show how that's done with a scientific calculator with Log button? I've got scientific calculator on my desktop computer (Windows 10) .

AndyH · Nov 3, 2019

Use the Shutter Speed app on your smart phone and it calculates the difference in f stops and lets you keep a record of how accurate each speed is on all of your cameras. Learning to read the sounds of your particular shutter takes some time, but once you see the pattern, it's easy to read the graphs with some degree of accuracy. If you're shooting a reflex, it's best to lock up the mirror at first, which will show you the pattern more clearly. Multiple trials also help.

Andy

Alan Edward Klein · Nov 3, 2019

Thanks Andy. This was my procedure.

The lenses were tested on their own disconnected from the camera body since the shutter is built into Mamiya lenses. So I didn't have to deal with the reflex mirror and its noises. The camera wasn't needed for the test. Only the 4 lenses.

The tests were done with the mirror up knob on the lenses set to Manual. I cocked the shutter manually on the lens for each shutter speed test. Then I manually released it back to the shooting position. I then set the shutter to the setting I was going to test. Then I turned on the recorder and placed the camera adjacent to the cell phone. Then I released the shutter with a cable release to avoid extraneous sounds and vibrations. I also left the aperture to its widest.

I then emailed the recordings to my desktop and loaded each one into the Audacity program for analysis. As you can see in the above 4 samples, the slowest shutter speeds were the easiest to mark their length. The faster shutter tests became more difficult due to the "hash" and overlapping noise . So it was hard to tell when the shutter opened and closed on fast shutters. I guessed a lot on those. These lenses have T, 1 sec, 1/2, 1/4, 1/8, 1/15, 1/30, 1/60, 1/125, 1/250 and 1/400. YOu can see the results on the hand written charts above.

Which app are you referring too?

Alan Edward Klein · Nov 3, 2019

All speeds were within 1/3 of a stop from 1 second to 1/60. Above that, a couple seemed to be more. But due to the difficulty in determining where open and close were, the results of these may be questioned. I used this chart to check. It's very impressive that these lenses hold their shutter speeds for 35 years especially since I left them cocked for years at a time. This may confort many who were concerned about this issue as was I. Note that these tests were for RB67 regular and "C" lenses, not the RL lenses or the RZ lenses for the RZ67 camera.

AndyH · Nov 3, 2019

Alan Edward Klein said:
Thanks Andy. This was my procedure.

The lenses were tested on their own disconnected from the camera body since the shutter is built into Mamiya lenses. So I didn't have to deal with the reflex mirror and its noises. The camera wasn't needed for the test. Only the 4 lenses.

The tests were done with the mirror up knob on the lenses set to Manual. I cocked the shutter manually on the lens for each shutter speed test. Then I manually released it back to the shooting position. I then set the shutter to the setting I was going to test. Then I turned on the recorder and placed the camera adjacent to the cell phone. Then I released the shutter with a cable release to avoid extraneous sounds and vibrations. I also left the aperture to its widest.

I then emailed the recordings to my desktop and loaded each one into the Audacity program for analysis. As you can see in the above 4 samples, the slowest shutter speeds were the easiest to mark their length. The faster shutter tests became more difficult due to the "hash" and overlapping noise . So it was hard to tell when the shutter opened and closed on fast shutters. I guessed a lot on those. These lenses have T, 1 sec, 1/2, 1/4, 1/8, 1/15, 1/30, 1/60, 1/125, 1/250 and 1/400. YOu can see the results on the hand written charts above.

Which app are you referring too?

Wow, that's a really comprehensive process. Well done!

The app is called Shutter Speed, and it's available on the Apple Store. I think it's also available for Android. Top item on these search results...

https://www.bing.com/search?q=shutt...s=n&sk=&cvid=7e3c56862a234a43d792b5fa202220bb

I use it to track speeds on all my analog cameras and lenses. You're right that the higher shutter speeds are more difficult to track, but you can get pretty good results with multiple trials looking for the same pattern. There is also a plug in light adaptor available for about fifty bucks, which will give much more accurate readings at higher speeds. I'm thinking of getting it. I like having the list of cameras right on my phone for field use - there are a couple of speeds on my Leicas that are off by as much as a full stop, but I can never remember which body and which speeds. It's really nice to have the lists at hand.

My Hasselblad lenses all appear to be in good shape (they're original C series) and no speed on my 50, 80, 160, and 250 is more than 1/3 stop off. My Rolleiflex is also pretty much spot on.

Andy

reddesert · Nov 3, 2019

Alan Edward Klein said:
I was wondering if there was a formula. Could you show how that's done with a scientific calculator with Log button? I've got scientific calculator on my desktop computer (Windows 10) .

I don't use the Windows calculator, but it should be the same way as most scientific calculators (excepting RPN calculators with a stack):

Enter 0.012, divide sign, 0.008, =
Now you see 1.5 which is the ratio of times.
Press the log button. Now it shows 0.176, which is the log10 of 1.5.
Press divide
Now enter 2, press the log button to take log10 of 2, you'll see 0.3.
Now press = (to complete dividing 0.176 by 0.3).
You should see 0.584...

It works this way because the log function has a higher precedence than divide, so the log button takes the log of whatever is showing immediately, while divide waits for you to enter the second number and press =.

Alan Edward Klein · Nov 3, 2019

reddesert said:
I don't use the Windows calculator, but it should be the same way as most scientific calculators (excepting RPN calculators with a stack):

Enter 0.012, divide sign, 0.008, =
Now you see 1.5 which is the ratio of times.
Press the log button. Now it shows 0.176, which is the log10 of 1.5.
Press divide
Now enter 2, press the log button to take log10 of 2, you'll see 0.3.
Now press = (to complete dividing 0.176 by 0.3).
You should see 0.584...

It works this way because the log function has a higher precedence than divide, so the log button takes the log of whatever is showing immediately, while divide waits for you to enter the second number and press =.

So that's 0.564 is that stop more or less than what it should be?

If reversed, I get -0.584. Is that stop more or less than what it should be?

Alex Varas · Nov 4, 2019

I use the Shutter Speed app with the sensor connected to the mini jacks phone, it does the job and I’m accurate in my exposures.

villagephotog · Nov 5, 2019

AndyH said:
There is also a plug in light adaptor available for about fifty bucks, which will give much more accurate readings at higher speeds.

I use the Shutter Speed app, too, and you're exactly right—the plug-in light sensor makes it much easier to get clean readings at shutter speeds faster than about 1/60. I think it cost me a little less than $50 (comes from Germany). I'm glad I got it, but I didn't learn anything crucial from it that I didn't know from the audio-only readings I had done before—at least on my lenses. They all use leaf shutters that max out at 1/500. I have no idea how much better it may or may not be with focal plane shutters or at speeds faster than 1/500.

Alex Varas · Nov 5, 2019

villagephotog said:
I use the Shutter Speed app, too, and you're exactly right—the plug-in light sensor makes it much easier to get clean readings at shutter speeds faster than about 1/60. I think it cost me a little less than $50 (comes from Germany). I'm glad I got it, but I didn't learn anything crucial from it that I didn't know from the audio-only readings I had done before—at least on my lenses. They all use leaf shutters that max out at 1/500. I have no idea how much better it may or may not be with focal plane shutters or at speeds faster than 1/500.

As far as I know 1/500 is the max reading. For me it was a change for better, the reading is clean and I can get the shutter speeds in the field for a better exposure.

Alan Edward Klein · Nov 5, 2019

The problem with focal plane shutters is that once you get past 1/60th or 1/125th, depending on direction of the shutter, the shutter is not open the full opening at any one time. A slit of opening traces across the opening because the open and close shutter (there are two), cannot move that quickly. That's why sync speeds for flash only work at lower speeds of 1/60th or 1/125th. Because it's only at lower speeds, that the film is exposed fully for a period of time so the flash can light the whole scene. If you use flash at faster shutter speeds, only a slit will be lit by the flash. So when you're examining shutter tests, it seems that's the point after which you can;t determine the actual shutter speed from the wave form.

I haven't tried it with a light, only a sound recording. But it seems that a light shone through the opening would present the same problem because there's a moving slit. Any thoughts on how to overcome this?

AndyH · Nov 6, 2019

Alan Edward Klein said:
The problem with focal plane shutters is that once you get past 1/60th or 1/125th, depending on direction of the shutter, the shutter is not open the full opening at any one time. A slit of opening traces across the opening because the open and close shutter (there are two), cannot move that quickly. That's why sync speeds for flash only work at lower speeds of 1/60th or 1/125th. Because it's only at lower speeds, that the film is exposed fully for a period of time so the flash can light the whole scene. If you use flash at faster shutter speeds, only a slit will be lit by the flash. So when you're examining shutter tests, it seems that's the point after which you can;t determine the actual shutter speed from the wave form.

I haven't tried it with a light, only a sound recording. But it seems that a light shone through the opening would present the same problem because there's a moving slit. Any thoughts on how to overcome this?

I think the app should be able to calculate this, so long as it's taking a reading of a wide enough section to capture the time element. After all, it's the duration time that light is shining at all, not the quantity of light in any way. The app doesn't need to know whether it's the first or last section of film being exposed, nor the width of the slit - it's just measuring how long any part of the curtain is opening and admitting light through the lens. No maths at all except for timing, I think...

Andy

Alan Edward Klein · Nov 6, 2019

Has anyone tried it with a light? What's been your experience?

Alex Varas · Nov 6, 2019

Alan Edward Klein said:
Has anyone tried it with a light? What's been your experience?

How I do my mettering is the phone with the app and the light-sensor at the back at the camera and another phone with the flashlight on, it works pretty well, I used for focal plane shutters as well such pentax 67 and reading were accurate (not 1000).
Is this answering your question?

Alan Edward Klein · Nov 6, 2019

alexvaras said:
How I do my mettering is the phone with the app and the light-sensor at the back at the camera and another phone with the flashlight on, it works pretty well, I used for focal plane shutters as well such pentax 67 and reading were accurate (not 1000).
Is this answering your question?

So you're able to tell when the shutter open and when the shutter close on focal plane cameras? To what shutter speed?

Alex Varas · Nov 6, 2019

I just did this for you
The phone I’m writing is the one I take a picture of the camera and takes the light sensor adapter, camera is 1/500 and this is my reading, pretty close to what this sensor can read.

I hope it helps,
Alex

Mamiya RB67 Lens Shutter tests

Privacy & Transparency

Privacy & Transparency