PDA

View Full Version : Determining The True Number of Clicks on 1D Series Cameras REDUX


dfindr
7th of June 2007 (Thu), 17:45
Sorry for the long post but this has been bugging me for a long time and I would like to get a definitive answer.

Below are readings of click counts on two seperate 1D Mark 1 cameras.

1D-A Unknown (0093)3,9 : 18,0,31766,0,0,0,65535,65535,1
1D-B Unknown (0093)3,9 : 18,0,3131,0,0,0,4,0,1

We need to clear something up for me and probably for many others here. There are various Exif reader programs out there including CANCOUNT, EXIFREAD, and EXIFTOOL which claim to accurately count clicks in 1D series cameras.

CANCOUNT only give you the current number of clicks in the current cycle. If the clicks are past 65535, than the counter starts over at 1 and CANCOUNT would read that as the number of clicks after 65535, and doesn't count the first 65535 clicks giving you a way way to low reading. For example if a camera is checked with CANCOUNT and the click count states 5000, with CANCOUNT you don't know if the camera has 5000 clicks or 70535 clicks.

Now that is ok if you know for certain you have less than 65535 clicks on the camera you are checking. EXIFTOOL works the same way. So how do you know how many times the camera has passed 65535 clicks using these two programs? From what I can tell you can't.

Now EXIFREAD gives you the (0093) line as illustrated above. What I need to know is what the other numbers mean. I have seen it posted on this forum as well as others that the last number in the string, and in the examples above the number "1", means the camera has turned over one time after reaching 65535 clicks. Is this true or not?

I have also read on this forum as well as other that the number after the "18" in the examples above "0", is where the turn over counter would read "1" or "2" or "3" depending on how many times the camera has passed 65535 following a reset. Is this true or not?

The problem is obvious, if the last number indicates the turn over has occurred, than the cameras noted above have 65535 + 31766 and 65535 + 3131 clicks respectively.

If the turnover counter is the "0" immediately after the 18 than these 1D's have 31766 and 3131 clicks respectively. BIG DIFFERENCE when
evaluating used 1D's.

Also does any one know what the other numbers mean? I know the (0093) is a line number, I figure the 3,9 and 18 are exclusive identifiers to the 1D as I have never seen them change when evaluating many used 1D Mark 1's. Each model 1D is different I believe at least it is for the 1D Mark 2. So can someone please fill in the blanks?

Unknown (0093) 3,9 : 18, A, B, C, D, E, F, G, H..

A - is the subject of the above question
B - is the click count
C - ??
D - ??
E - ??
F - ??
G - ??
H - is the subject of the above question as well.

Thanks!

dfindr
7th of June 2007 (Thu), 22:25
Anyone? Please!

sports_photog_inmd
7th of June 2007 (Thu), 22:31
http://www.foxbat.me.uk/myDownloads.php

This was posted on here several months ago. Works on all 1 Series cameras.

Hermeto
8th of June 2007 (Fri), 00:10
1D-A Unknown (0093)3,9 : 18,0,31766,0,0,0,65535,65535,1
1D-B Unknown (0093)3,9 : 18,0,3131,0,0,0,4,0,1


I am not sure what youre asking..
In the example above, Id say that camera has 31,766 + 65,535 + 65,535 = 162,836 actuations..

I could be wrong, `tho..

foxbat
8th of June 2007 (Fri), 04:58
Anyone? Please!In pseudocode the total 32 bit count is e[2] + e[1]<<16 where e is a 16 bit array of unsigned integers in exif#93. Applied to your examples you are looking at counts of 31766 and 3131 respectively. Use my utility (linked above) if you're a non-programmer.

pcollyer
8th of June 2007 (Fri), 09:51
Anyone? Please!

Found this on another site, I think posted by Tim Grey...

In PS CS2 using a RAW file:

File
File Info
Advanced
Click on the + line: "http://ns/adobe.com/exif/1.0/aux/" (this line is not available for jpg). [Edited]
There will be a line that reads: ImageNumber: xxxx

It works.

dfindr
9th of June 2007 (Sat), 15:00
Come on guys. There has to me a mega geek out there that lives for this stuff. We need to get this figured out so we can check old 1D bodies and get an accurate click count.

Foxbat -- neat utility of you know 100% certain your camera does not have more than 65535 clicks on it. So for folks that buy 1D bodies new, great utility. However, if you are evaluating a 1D mark 1 body, it could very well have over 65535 clicks on it, or for that matter 2 or 3 times that many. There are folks out there selling these things when the cycle counter has turned over 1 or 2 times at 65535 clicks claiming the 1D has a whole lot less clicks than what is the true count.

The Adobe CS2 suggestion is good however the same thing applies as it cannot tell you if the cycle counter has turned over at 65535. Same problem as above.

Regading HERMETO's opinion, the second camera string in the OP cannot be true per Hermeto's opinion as I know that particular camera does not have just 3131 clicks on it which leads me to believe the last number and not the 2nd number after the colon is the cycle counter (How many times the shutter has passed 65535 clicks).

Now based on a post in the FM forum, one of the responders stated that he had his shutter replaced and the second number after the colon was a "1", with 994 in the third position. Now this makes sense perhaps in that canon when replacing the shutter, put the "1" in the second position.

Any canon camera repairs techs out there?

segasaturn
9th of June 2007 (Sat), 19:36
I bought a used 1D. The seller said it had about 78000 clicks, but the reader says only 71,000. However, in the exif, it does have 65,535 TWO TIMES!!! So now I'm concerned that it actually has closer to 135,000 clicks rather than 78,000. 135,000 is a lot closer to the 150,000 MTF than 78,000 is.

cosworth
9th of June 2007 (Sat), 19:45
Found this on another site, I think posted by Tim Grey...

In PS CS2 using a RAW file:

File
File Info
Advanced
Click on the + line: "http://ns/adobe.com/exif/1.0/aux/" (this line is not available for jpg). [Edited]
There will be a line that reads: ImageNumber: xxxx

It works.

This results in incorrect numbers.

dfindr
9th of June 2007 (Sat), 22:26
I bought a used 1D. The seller said it had about 78000 clicks, but the reader says only 71,000. However, in the exif, it does have 65,535 TWO TIMES!!! So now I'm concerned that it actually has closer to 135,000 clicks rather than 78,000. 135,000 is a lot closer to the 150,000 MTF than 78,000 is.

Segasatun "I feel your pain" I can't believe that there isn't someone out there that can clear all this up for us. The numbers mean something and we need to figure it out. Someone please help!

garypasz
9th of June 2007 (Sat), 22:40
Try this macro counter for 1D models only, can not remember who first provided this but quite a few people on here recommend it.

http://www.sonnycao.com/files/1dcount.zip

dfindr
9th of June 2007 (Sat), 23:21
Try this macro counter for 1D models only, can not remember who first provided this but quite a few people on here recommend it.

http://www.sonnycao.com/files/1dcount.zip

I have tried it. It only tells you how many clicks up to 65535. At that point the cycle counter turns over and the count starts again at 1. That means it is accurate only if you know for a fact your camera has not exceeded 65535 clicks.

malla1962
10th of June 2007 (Sun), 02:04
Have you tried exiftool?its free and gives a lot of info inc shutter count.
http://www.sno.phy.queensu.ca/~phil/exiftool/

cosworth
10th of June 2007 (Sun), 02:15
1dcount works perfectly on my Vista PC and has accurately determined the frame count on my old 1Ds mk.I to 86K, matching Cancount perfectly.

dfindr
10th of June 2007 (Sun), 13:43
1dcount works perfectly on my Vista PC and has accurately determined the frame count on my old 1Ds mk.I to 86K, matching Cancount perfectly.

Cosworth:

That is interesting, because I used 1dCount on an 1D that I know has lots of shutter clicks, it tells me that it has only 3158 clicks, and I know that is not accurate. When you ran your camera through 1DCount did it actually give you a number more than 65535?

Anyone else have this same experience with 1DCount in actually getting a click number in excess of 65535?

Interesting, more research is required!

cosworth
10th of June 2007 (Sun), 13:45
Do you own a 1 series body with over 65k clicks? It's easy to see the errors.

cdifoto
10th of June 2007 (Sun), 15:55
Do you own a 1 series body with over 65k clicks? It's easy to see the errors.

He will if he stops kicking the tires on mine. ;)

dfindr
10th of June 2007 (Sun), 19:58
Ok. this is getting a little more interesting. I just did a check on a jpg sent to me by another forum member, he told me the 1D had 71385 clicks on it. Sure enough, I ran it through 1DCount and it has 71385 clicks on it.

I ran it through ExifReader and got the following:

Unknown (0093)3,9 : 18,1,5849,0,0,0,65535, 65535, 0

65535 + 5849 = 71384 --- off by 1 click.

It appears that the number after the 18 is indeed a cycle counter which advances 1 after 65535 clicks.

I also ran it through CanCount and got 71385.

I also ran it through ExifTool and got 71385.

So much for the theory that these shutter count programs do not report clicks accurately past 65535.

I need more data and research.

If anyone is of the mind to email me small jpg's directly and unprocessed from your 1D cameras, and tell me what you think the shutter count is. I would like to test out this theory and will post all results.

Thanks!

MKII
11th of June 2007 (Mon), 03:10
i have a mark 2n...what will happen when i shot the 200,000 shot...? ? will the cam stop functioning...? ?

foxbat
11th of June 2007 (Mon), 03:32
Ok. this is getting a little more interesting. I just did a check on a jpg sent to me by another forum member, he told me the 1D had 71385 clicks on it. Sure enough, I ran it through 1DCount and it has 71385 clicks on it.

I ran it through ExifReader and got the following:

Unknown (0093)3,9 : 18,1,5849,0,0,0,65535, 65535, 0

65535 + 5849 = 71384 --- off by 1 click.

It appears that the number after the 18 is indeed a cycle counter which advances 1 after 65535 clicks.

I also ran it through CanCount and got 71385.

I also ran it through ExifTool and got 71385.

So much for the theory that these shutter count programs do not report clicks accurately past 65535.

I need more data and research.

If anyone is of the mind to email me small jpg's directly and unprocessed from your 1D cameras, and tell me what you think the shutter count is. I would like to test out this theory and will post all results.

Thanks!
You clearly don't have a programming background and you are trying to understand a computer data structure so I'll try to keep this as non-technical as possible. This is not guesswork, theory or hearsay. These are the facts:

The first number (a constant 18 ) is the byte count for the entire size of the field. There are 9 2-byte words in the field including the size itself. 9*2 = 18. This value enables exif readers to know how much data to expect to read.

The second and third number are the 32-bit count split over 2 16-bit words, high word first. When the low word reaches 65535 and you click again it rolls back to 0 and the high word increments to 1. Therefore the formula is (65536 * high) + low. It's basic 1st year computer science and we call it 'big-endian' or sometimes 'motorola' encoding.

The rest of the field (all those zeros and 65535 values) are not related to the count and can be ignored. We don't know what they mean but we do know that they don't change as you click the shutter.

Note. I've read that if you are in posession of two 1D cameras of the same model it is possible to export the personal settings data from one of them and import it into the second one, overwriting the shutter count (and other owner registration data) thereby making it appear to counter readers that it has a lower count than it really does. Bear this in mind if you are dealing with a camera sold by a dealer who you do not trust.

schmoelzel
11th of June 2007 (Mon), 07:56
I bought a used 1D a couple of years ago and was told that it had about 78000 actuations......with a shutter that is rated for at least 200000, I honestly never checked the actual # and have never been disappointed with this camera. I would say that if the camera works and all it's essential tools are functioning, nevermind about the shutter count! A new shutter will run about $300 so it wouldn't be the end of the world!!

dfindr
11th of June 2007 (Mon), 12:58
Foxbat wrote: "The rest of the field (all those zeros and 65535 values) are not related to the count and can be ignored. We don't know what they mean but we do know that they don't change as you click the shutter"

Foxbat, thanks for some clarity. And you are right I know nothing about programming so thanks. One question however, some 1D's have the 65535 value where others have "0". Do you know why that is? Also a very interesting point about being able to import/export settings from camera to camera.

number six
11th of June 2007 (Mon), 13:10
65535 + 5849 = 71384 --- off by 1 click.

Foxbat nailed the explanation, I just wanted to mention that you're *not* off by one click. We always start counting with zero, don't we?;)

-js