Are DC time stamps off?

The best solution for finding and removing duplicate files.
HiTechHiTouch
Posts: 15
Joined: Sun Jul 29, 2012 4:33 pm

Are DC time stamps off?

Post by HiTechHiTouch »

The base point is that the create and modified dates in the DC Duplicate Files listing seem to be off.

(Bare with me please, I've tried to document everything with screen shots. The /img/ tag on this bbs seems not to work. It could be that the bbs server can't access the screen shots, but I can see them in google with an alternate id, so I'm pretty sure the are public. Plus, it looks like the BBS has an attachment limit of 3.)

What got me started was noticing in Beyond Compare, then in Windows Explorer, that the timestamps in one file tree were off by 1 hour from the timestamps in the other tree being compared. This hour difference is a known "feature" because Windows does not correct for daylight saving time.

My question was, of two identical files, one created by copying the other, which of the two timestamps (differing by and hour) is the correct one? Fortunately with most photos, there is date information inside the photo that can be used to decide between the two windows file system values.

I was also running DC, and chose an example file from the Duplicate Files tab. I quickly saw that we had different dates in DC than in Explorer! Further research shows the DC dates do not match any other dates, including those in EXIF and IPTC data!

OK, here's the audit trail. This runs Duplicate Cleaner 3.0.6 to locate F:\Photos\2009\DSC00590.jpg and its duplicate F:\Photos-Do_NOT_Delete\2009\DSC00590.jpg.

Image https://docs.google.com/file/d/0B0tbUZR ... 1hQTQ/edit
dc-search.png
Image https://docs.google.com/file/d/0B0tbUZR ... VOOW8/edit
dc-scan loc.png
Image https://docs.google.com/file/d/0B0tbUZR ... lzOEE/edit
dc-result.png
Image https://docs.google.com/file/d/0B0tbUZR ... hoYWs/edit

Now here's the Windows property page for the Photos\2009\ version
Image https://docs.google.com/file/d/0B0tbUZR ... diTXc/edit

The DC created time of day is "3:22:04 PM". The windows time of day is "3:29:13 PM".
The DC modified time of day is "2:06:26 PM". The windows time of day is "2:13:26 PM".
Neither created nor modified times of day are close.

Now lets look at the file internals. First is XnView
Image https://docs.google.com/file/d/0B0tbUZR ... dPanM/edit

Image https://docs.google.com/file/d/0B0tbUZR ... 81RjA/edit

Next is IfranView

Image https://docs.google.com/file/d/0B0tbUZR ... NGMEE/edit

Now I use IfranView to add some text to the IPTC caption field, showing timestamps seen by IfranView. The template is:

Image https://docs.google.com/file/d/0B0tbUZR ... VvOFE/edit

Resulting in
Image https://docs.google.com/file/d/0B0tbUZR ... QwYUk/edit
Last edited by HiTechHiTouch on Thu Aug 02, 2012 5:28 am, edited 5 times in total.
HiTechHiTouch
Posts: 15
Joined: Sun Jul 29, 2012 4:33 pm

Re: Are DC time stamps off?

Post by HiTechHiTouch »

OK, here's the audit trail. This runs Duplicate Cleaner 3.0.6 to locate Photos\2009\DSC00590.jpg and its duplicate F:\Photos-Do_NOT_Delete\2009\DSC00590.jpg.

Image https://docs.google.com/file/d/0B0tbUZR ... 1hQTQ/edit

Image https://docs.google.com/file/d/0B0tbUZR ... VOOW8/edit

Image https://docs.google.com/file/d/0B0tbUZR ... lzOEE/edit

Image https://docs.google.com/file/d/0B0tbUZR ... hoYWs/edit
dc-dup files.png
Now here's the Windows property page for the Photos\2009\ version
Image
dc-prop.png
dc-prop.png (19.41 KiB) Viewed 4729 times
The DC created time of day is "3:22:04 PM". The windows time of day is "3:29:13 PM".
The DC modified time of day is "2:06:26 PM". The windows time of day is "2:13:26 PM".
Neither created nor modified times of day are close.
HiTechHiTouch
Posts: 15
Joined: Sun Jul 29, 2012 4:33 pm

Re: Are DC time stamps off?

Post by HiTechHiTouch »

Last edited by HiTechHiTouch on Thu Aug 02, 2012 5:44 am, edited 1 time in total.
HiTechHiTouch
Posts: 15
Joined: Sun Jul 29, 2012 4:33 pm

Re: Are DC time stamps off?

Post by HiTechHiTouch »

Next is IfranView
Image https://docs.google.com/file/d/0B0tbUZR ... NGMEE/edit
dc-ifran-exif.png
Now I use IfranView to add some text to the IPTC caption field, showing timestamps seen by IfranView. The template is: exif date=$E306 org=$E36867|file date=$T sys date=$U|IPTC rel= $I30 $I35 create=$I55 $I60.

Resulting in
Image https://docs.google.com/file/d/0B0tbUZR ... QwYUk/edit
Attachments
dc-ifran-iptc.png
User avatar
therube
Posts: 634
Joined: Tue Jun 28, 2011 4:38 pm

Re: Are DC time stamps off?

Post by therube »

I'm thinking that he is trying to "normalize" date/time in order to be able to ignore 1 or 2 hours, & also +- a couple (two) seconds differences.

Only there's a bug in the algorithm ;-).

Also a time of >11:PM <11:59 PM, "normalized", will also cause the date to be one day different.


In addition to daylight savings time, there can also be differences in time due to FAT/NTFS issues.


Or possibly related to what I said here, viewtopic.php?p=3587#p3587, "it must be using short date, long time".


Bug in date/time change of file

Time stamp changes when copying files
User avatar
DigitalVolcano
Site Admin
Posts: 1864
Joined: Thu Jun 09, 2011 10:04 am

Re: Are DC time stamps off?

Post by DigitalVolcano »

Thanks for the reports - the times are a few minutes off in some cases. Nothing to do with daylight savings, or any normalisation. It seems to be an issue with the .NET library behaving oddly. I'm experimenting with reverting to Windows API methods (as in duplicate cleaner 1.0!) as they seem to return the correct times...
User avatar
DigitalVolcano
Site Admin
Posts: 1864
Joined: Thu Jun 09, 2011 10:04 am

Re: Are DC time stamps off?

Post by DigitalVolcano »

Try 3.08 - this should be fixed now.
User avatar
therube
Posts: 634
Joined: Tue Jun 28, 2011 4:38 pm

Re: Are DC time stamps off?

Post by therube »

It seems that this fix was not ported to 3.0.1 Free.
(BTW, the same issue exists in DC 2.1.)

---

In my current situation ...

+Ignore Content
+Same Size
+Same Modified Date

would be a highly efficient search, but Free is not returning the expected results.
I'm pretty sure I was way off on that comment.
What I think happened was that in my Search Paths, I had a child tree excluded, while at the same time I had it's parent included. Didn't realize & was expecting the entire tree starting from the parent to be scanned, but because of the exclusion, much was not scanned.

I'll verify this when I get some more free time.
Post Reply