| Author |
Message |
    
John Tomaszewski
Rating: N/A
Votes: 0 (Vote!) | | Posted on Sunday, December 05, 2004 - 11:48 am: | 
|
Hi all,
I just built one of Henry Taber's Combination shutter/curtain speed testers and it works like a charm. My question is, what kind of precision should I expect, especially at the higher shutter speeds?
Here are my results for a Spotmatic I'm working on:
Setting: Seconds: 1st 2nd
1000 0.001 0.003 0.003
500 0.002 0.005 0.004
250 0.004 0.007 0.007
125 0.008 0.011 0.012
60 0.017 0.020 0.022
30 0.033 0.039 0.041
15 0.067 0.066 0.067
8 0.125 0.126 0.129
4 0.250 0.260 0.262
2 0.500 0.471 0.481
1 1.000 0.911 0.919
Any comments on what kind of precision and accuracy I should expect, both from the device, and from my camera, would be greatly appreciated. I'm completely new to all of this. Thanks.
John T. |
rick oleson
Rating: N/A
Votes: 0 (Vote!) | | Posted on Sunday, December 05, 2004 - 01:53 pm: |
|
i think +/- 1/3 stop is pretty good for an older camera, and probably as good as many of them were new. here are some numbers for that:
Marked__Nominal__+ limit__- limit
1000 __ .00098 __ .00078 __ .00123
500 ___ .00195 __ .00155 __ .00246
250 ___ .00391 __ .00310 __ .00492
125 ___ .00781 __ .00620 __ .00984
60 ___ .01563 __ .01240 __ .01969
30 ___ .03125 __ .02480 __ .03937
15 ___ .06250 __ .04961 __ .07875
8 ____ .12500 __ .09921 __ .15749
4 ____ .25000 __ .19843 __ .31498
2 ____ .50000 __ .39685 __ .62996
1 ___ 1.00000 __ .79370 __1.25992
Note that these nominal numbers are not exactly the same as yours. This is because they represent a perfect geometric progression, which deviates slightly from the "rounded off" figures on the shutter speed dial. Exact fractional speeds are:
1 2 4 8 16 32 64 128 256 512 1024 |
Henry
Rating: N/A
Votes: 0 (Vote!) | | Posted on Sunday, December 05, 2004 - 03:49 pm: |
|
Rick is absolutely right, going from a "perfect" 1 sec to an "imperfect" 1/1000. Going from a "perfect" 1/1000 to an "imperfect" 1 sec is also valid. When just testing a shutter either "top down" or "bottom up" will work fine. But when actually adjusting I usually first get the curtains traveling the EXACT SAME speed at approximately the factory rate. I don't get anal on trying to adjust exactly to factory curtain speed specs, but am always sure to error on the slow side. I do get anal on matching the curtain speeds! Next I set 1/250 dead on.
Why? The factory curtain speed is to maintain the proper sync speed for the flash. A little slow doesn't affect anything. Both curtains must travel at EXACTLY the same speed to eliminate all capping. And finally, perfect 1/250 because my most used speeds are 1/60 through 1/1000. On most mechanically timed cloth shuttered cameras there isn't adjustments for each shutter speed. 1/250 is in the middle of the speed range.
I assume that your "first" and "second" are shutter speeds at each photocell position. Both of these columns should be equal throughout the range if the curtains are travelling at the same speed. Yours aren't.
What were the curtain speeds?
Henry |
John Tomaszewski
Rating: N/A
Votes: 0 (Vote!) | | Posted on Sunday, December 05, 2004 - 08:03 pm: |
|
Henry and Rick,
Thanks for your help.
I think the small differences (except for 0.5s and 1s) in the measurements at the first and second photocells were due mostly to "eyeballing" where the signal started and stopped. Sometimes it was fairly ambiguous, especially at the trailing end of the signals. That's kind of what I was asking. It seemed to me that the last digit, measured by this method, was fairly imprecise. I did just notice that I only recorded at 44.1kHz but 48kHz was available. I should have at least done that.
The curtain times were 17ms +/- 1ms and again the measurements of the leading curtain were more precise and consistant than those of the trailing curtain, as the signals all started off with a spike but tapered off rather than ending abruptly.
If either of you have time, I could send the .wav file. It might tell you more than the numbers can about how well it's working. Or, if you could send me one of your .wav files from a well tuned camera that would tell me a lot as well.
Thanks again.
John T. |
rick oleson
Rating: N/A
Votes: 0 (Vote!) | | Posted on Monday, December 06, 2004 - 06:23 am: |
|
I don't use that kind of an instrument, I have one of the little digital handheld testers that used to be available (maybe still are?)... so I don't know what the .wav file should look like. |
Henry
Rating: N/A
Votes: 0 (Vote!) | | Posted on Monday, December 06, 2004 - 08:12 am: |
|
I don't use that tester anymore either. Although it worked very well while I did. I was fortunate enough to stumble upon a factory made unit with curtain speed attachment.
However, go ahead and send me the wave file. I'll download a Cooledit demo and take a look. And I'll check to see if I saved any of my wave files.
Henry |
John Tomaszewski
Rating: N/A
Votes: 0 (Vote!) | | Posted on Tuesday, January 25, 2005 - 08:22 pm: |
|
Henry and I corresponded a bit more on this topic by email so, with Henry's permission, I thought I'd post a summary of if it here for the archives:
John,
When you made the tester did you put the photocells exactly 18mm apart?
Henry
Henry,
Well, no. I don't have the means for that kind of precision. After I finished I measured them at 19.0mm apart and corrected for it in the curtain speed calculations.
John
Great, any distance will do as long as you know and compensate for it. I haven't downloaded Cooledit yet. Kind of under the weather lately...feel like crap.
Henry
John,
I loaded up Cooledit this morning...seems they were bought out by Adobe. But I located an old copy of Cooledit2000 on an archive CD. On most of your speeds the first curtain travels the 19mm in .007 sec while the second curtain does the distance in .009 sec. Clearly your Spotmatic is capping.
But your graphs look much different from mine. That is from the handful I found of when I built this tester back in 2000. Long time ago. This is, of course, a function of our different soundboards and photodiodes...mostly soundboards. But learning how to reach each of ours is the trick.
I'll attach a couple files so you can see mine. Notice how I got "shadows" of one diode to the other. I found this very useful. I don't even know what camera these files were of.
Pentax specs are for the curtain to traverse the entire 36mm of frame width in 14 +/- .3 ms. By your test, the first curtain is a bit fast while the second curtain is quite a bit slow. Btw, it is a standard practise to set the first curtain a bit faster than the second...but I don't.
Henry
Quote:N.B. I couldn't find an old copy of CoolEdit around but I had Turtle Beach's Audio Surgeon on hand and tried that. It worked but only reported time to milliseconds. I found Audacity for free and it has much higher resolution.
Henry,
Thanks for the info. and the wav files. I see what you mean by yours looking different than mine. I have a few other cameras that I'm trying to fix and a couple that I think work well so I need to test several more and get a better feel for what I should see.
Thanks also for the Pentax spec. I may have that number now in one of the repair manuals I've acquired but if I do I hadn't noticed it yet. I need to know what I'm shooting for. I was hoping it wasn't 10ms like I read for Nikons because I thought going from my present 17ms to 10ms is a big change, but 14ms seems like it's might be doable.
John T.
The Nikon spec might be for a vertically traveling, metal bladed shutter. Those usually sync at 125 instead of 60. The later Pentax's with Copal style shutter synced at 100, though it says just x on the mode dial.
And don't worry about what the graph looks like...looks is a function of the electronic components in the soundboard. Just learn to read your particular graphs by studying the graphs of good functioning shutters. This is why I didn't keep pics of my graphs on the article...too many questions concerning looks.
Henry
Henry,
Two last questions on this:
1) Given that the second curtain on my Spotmatic seems to be slow and I've already tightened both curtain tension screws 9 turns, do you think I should now consider cleaning the shutter mechanism, or can I just go ahead and keep tightening the springs until I get the curtains even and closer to 14ms?
2) Can I post a summary of our last few emails in the topic I started on the Repair Forum? I think it would be useful for others who are searching for info. in the future.
Thanks again.
John
John,
Nine turns? Yikes! I must not have made myself clear. Usually if curtains are slow they can be cleaned and properly lubed and their speed comes right back to spec. Nine turns, I'm surprised they didn't break or come loose.
Sure, you can post a summary.
Henry
That's it. I've checked several other cameras since then and gotten very good (apparently accurate) results. |
|
