1 second is more like .80 seconds, CLA?

[kipkeston]

Greetings,

My M6 1 second is more like .8 according to my stopwatch and the 1/2 second is more like .38 on average. Is this within normal expectations or should I get a CLA? I think 1/4th is 1/4th and the rest are fine from what I can tell. I don't use 1/2 and 1 hand held really so it's not a major issue, but I was planning on upgrading the viewfinder in the near future anyway. Just looking for thoughts. Thanks.

 

[pvdhaar]

I'd think it very hard to check shutter accuracy with a handheld stopwatch. You're working in a timeframe where the time it takes your brain to process what you see/hear and control your fingers plays a major role..

Why not test this with some slide film, and check exposure consistency over a couple of aperture-shutter combinations? If you can't spot the difference in the slides, there's no real world problem at all..

 

[kipkeston]

Indeed, I think I should try a few shots with slide. I will compare them tomorrow and see what happens. But honestly, if we're talking about delay between the time I see the shutter close and the time I hit stop on the stop watch, I'm still coming up short on 1 second, about .2 seconds short. The way I did it wasn't really to wait for the shutter to close, but to sync myself to the shutter, meaning I hit stop on the watch at the moment I felt the shutter would close. 1 second is a long time if you're listening and you can tell if you hit stop the same time the shutter closes.

 

[kipkeston]

Well, if I wanted to get fancy, I suppose I could load up logic or some sound recorder, record the shutters and look at the wave after. Maybe I'll do that right now.

 

[Mackinaw]

0.8 seconds is perfectly normal and would fall within the acceptable "tolerance range" for any mechanically-controlled shutter as is used in the M6. I wouldn't worry about it.

Jim B.

 

[payasam]

You probably don't need to worry until the slow speeds begin to stick.

 

[JohnL]

It would probably be possible to write a simple program to display a trace on a computer screen that moves at an appropriate speed. Take a shot and the accurate shutter speed would be related to the length of the trace. Getting the exposure right might need some experimenting. A bright trace on a dark screen would probably work best.

 

[kipkeston]

That is simple, I'd like one. Ok, I'll just keep shooting then. Thanks.

 

[waileong]

1 s is simple to test. Just use a watch with a second hand.

Expose film and see if the second hand covers 1/60 of the circle.

Ditto for 1/2 s, although it may be harder to judge 1/120 of a circle.

 

[rogue_designer]

don't forget that the amount of noise the shutter makes, doesn't necessarily correlate exactly to the amount of time the film is exposed.

A photocell tester is best, if you're that worried about it.

 

[Assaf]

Another method

Another method

If you have some experience with simple sound editting, you can try the following:

1) record the shutter "click" with you computer microphone

2) open the file with a sound editor (many free ones in the net). Look on the sound wave - you should be able to detect the openning click (a first peak) and the cloasing click (the second one). Then the sound editor should tell you the elaped time.

This method should work for much higher speeds as well

 

[Tom A]

The speed of a mechanical shutter is rather "flexible". The tolerance level is around +/- 15 to 20%, particularly at the extreme ends. Your 0.8 sec versus indicated 1 sec falls within these levels (particularly condsidering your measuring method).
It is rare that you have a consistent speed with mechanicals. I used to check my M's against a shutter speed tester and even out of the box M6/MP's would vary. The 1/1000 was mostly in the 1/750 to 1/800 and slow speeds could vary even more.
You can have them calibrated and corrected, but they rarely remain at that set speed. The variations are not bid enough to significantly affect exposure in most cases.
The infamous Contax/Leica "war" in the 30's with Contax proudly announcing the top speed of 1/1250 (vertical running shutter) versus the Leicas 1/1000. I suspect that the Contax was closer to 1/1000 than the Leica, but none of them ever got close to indicated topspeed!
If you are shooting at 1/2 or 1 sec, you are most likely using a tripod or support the camera somehow. Just bracket the expoure with a +/- set (bl/w or color neg, bracket 1 stop each way, color transparancy film, bracket 1/2 stop each way).

 

[Ororaro]

20% equals 1/5th of a stop, which is nothing.

 

[tetrisattack]

I wonder why someone doesn't make a shutter speed tester at an economical price anymore. The technology would be dead simple these days. I have a shutter speed tester Calumet sold back in the early 1980's that works great. Don't remember what it cost then, but it's electronics are pretty simple inside.

I bet a lot of us gear heads would buy one.

Buy a photo cell from radio shack. Solder it to a cord, put a standard headphone jack on the other end.

Plug it into your computer's sound card. (!)

Use a program like Audacity -- free -- to record the voltage change. Zooming in on the waveform, you should be able to see for how long the light is hitting the photo cell. And it should be accurate to 1/44100th of a second, to boot. 🙂

 

[tetrisattack]

If you have some experience with simple sound editting, you can try the following:

1) record the shutter "click" with you computer microphone

2) open the file with a sound editor (many free ones in the net). Look on the sound wave - you should be able to detect the openning click (a first peak) and the cloasing click (the second one). Then the sound editor should tell you the elaped time.

This method should work for much higher speeds as well

I've tried this; I had a really hard time differentiating between the 1st and 2nd curtain open/closing. You'd want to start your measurement when the first curtain begins to open and end it when you hear the second curtain begin to move, but in practice, I found it hard to get a good audio representation of this. I tried a couple different mikes from different points but had no solid results past the flash sync speed.

 

[Assaf]

20% equals 1/5th of a stop, which is nothing.

-20% is -.32 stop, almost third of a stop ( log2(.8) )

Indeed, not a big issue if you are shooting negative

tetrisattack, can you please send me the audio file? I'm curious about that. Must admit that I never tried the method I proposed 🙂

 

[Ororaro]

-20% is -.32 stop, almost third of a stop ( log2(.8) )

Indeed, not a big issue if you are shooting negative

tetrisattack, can you please send me the audio file? I'm curious about that. Must admit that I never tried the method I proposed 🙂

You're probably right. Nonetheless, 0.32 of a stop combined to an approximate speed film and approximate developing all come up to accurate results 😀

 

[John Rountree]

As Tom A said, a variation of +/_ 15-20% is considered within the normal tolerance for a mechanical shutter. Not to make you paranoid, but have you considered how accurate the iris is in your lens? Does it really allow half as much or twice as much light at each stop? Depending on your film developing technique (or processor) you might not be achieveing the actual ISO rated sensitivity. If you are really concerned about improving exposure you should probably calibrate your shutter-lens-film combination. That will give you better and longer lasting results than simply knowing the variation in your shutter speed.

 

[Roger Hicks]

Google ZTS. A very good tester, and a light meter!

I'd not trust stopwatch timing for speeds. Nor would I worry about a 20% error.

Cheers,

R.

 

[kipkeston]

I can't get my sound recorder to work. But thanks for the piece of mind. I will get a CLA next spring with upgraded viewfinder so it can wait til then. Thanks again.