Nikon D800 NEFs ==> SubIFD Tags corrupted ==> Rewrite possible?

[Phil Harvey]

Hi Gene,

Excellent.  Good job.  Thanks for the note.

- Phil

[joostkrebbers]

Phil,

Apparently the tool is not working for corrupted nefs from the Nikon D3; is it possible to make the tool work for the D3?

If so, highly appreciated.


Joost

[Phil Harvey]

Hi Joost,

How was the D3 image corrupted?  If it was corrupted by Nikon software, then send me a sample and I'll take a look.  My email is philharvey66 at gmail.com

- Phil

[Phil Harvey]

Hi Joost,

I got your image.  I has been edited by both Nikon Capture NX and PhotoMechanic.  The only thing I see obviously wrong are the X/YResolution tags in the SubIFD (the full-sized preview), but this shouldn't be a big problem.  Everything else looks OK, and I can open the NEF without a problem using Apple Preview and Nikon Capture NX-D.

What software are you using that doesn't display this image correctly?

- Phil

[joostkrebbers]

Hi Phil,

Thnx for your quick reply! I am using Capture 1 version 8.3 for mac and have a whole bunch of images previously edited with NX2 which don't display correctly in C1.


Rgrds. Joost 

[Phil Harvey]

Hi Joost,

I recommend complaining to Nikon.  They really need to work the bugs out of their software.

But as for what is causing C1 to have problems, I couldn't say for sure.  The way to figure this out would be to take an NEF, run it through NX2, then use ExifTool to compare the metadata (exiftool -a -G1 FILE).  Then, use ExifTool to change things one at a time or in groups until you figure out what was causing the problem.

- Phil

[joostkrebbers]

Hi Phil,

I would certainly complain if there would be a chance that Nikon would be listening but that's what I don't expect with a piece of aborted software. It is actually quite easy to determine what information has been added by NX2 but I have troubles to understand how I can delete that with Exiftool. I have found ExiftoolGui which is unfortunately too limited. Please can you explain me in a simple way how to delete the NX2 information(which is mainly in the makers notes)? It was by the way way my understanding that the utility "fix corrupted nef's" was more or less doing that for the more recent Nikon camera's?


Rgrds. Joost

[Phil Harvey]

Hi Joost,

As-is, ExifTool will not delete parts of the maker notes.  The reason for this is to avoid incompatibilities with manufacturer software.  If you really want to do this, you would need to create a user-defined tag for the MakerNotes tags you want to delete, with the Permanent flag set to 0 in each of your tag definitions.

The fix_corrupted_nef utility didn't mess with the maker notes.  It repaired damage done to the tags in SubIFD1, not the maker notes.

- Phil

[joostkrebbers]

Hi Phil,

I just checked the fix corrupted nef utility on a corrupted D800E nef and it apparently does not solve the problem with Capture 1 (I had the impression that it would); strange is that any other raw converter has no problem reading and properly displaying the corrupted nef. So, it seems Capture 1 specific. For now I will contact Phase One and find out whether they will be able and prepared to fix it.


Thanks and best regards, Joost

[spark]

Hi Joost.

I don't know if this is of any help, but in the past I experienced troubles between photomechanic and Nikon's Capture NX 2, which boiled down to one singular entry left by photmechanic, a tag which if I recall right is called or contains "endoffilepreferences".

One way of getting rid of this entry is via Photomechanic itself.
"Use the Delete Metadata command on the Tools menu to delete the 'Photo Mechanic end of file preferences'."

I read that other software, even some web services, had troubles with this tag too, so I reckon Aperture might have the same troubles

[spark]

Hi Phil,

it's been a while since I had a look at this thread and must say... WOW!
You put great effort into this project and potentially helped god knows how many people with your fix! (Have you actually any idea how many?)

Thanks for your efforts and eagerness to help, I'm happy to see that others could be helped too

[Phil Harvey]

Hi Spark,

Yes.  This problem has affected various other Nikon models and many Nikon owners.  I have no way to tell how many people have benefitted from fix_corrupted_nef, but the numbers were significant, and some people were able to rescue some images that were very important to them.

- Phil

[wengee]

Hi Phil
Another innocent caught out by using old Nikon Software (and of course I didn't read right through to page 59 of the manual where it coyly warns that 'using an earlier version may fail to transfer RAW files correctly' partway down the page. You'd think they might have it in big bold type right on the front page by now?) Does your software work with NEF files from the D5300? I tried downloading the software but it didn't seem to do anything. Generated a new folder, but nothing in it.
Did I use it wrongly (quite possible, I can use the computer OK, but know nothing of programming etc, sorry), if not, is there something I can do to get a version that will work?
I have over a thousand corrupted RAW files from my (at the time, brand new) D5300, from a 'once in a lifetime' trip around the UK, and visited a number of places specifically for the photo opportunities. I would very much like to be able to use them!
Thanks for any help.

[TorC]

Wengee, while I've never been caught because I've never had reason to touch Nikon's software with a barge pole, it looks like the 5300 is a new camera to the Nikon software corruption.

Check this post in the thread for details on what Phil probably needs to update the fix_corrupted_nef tool.  Duplicate that for all high-speed crops you need it for (and future users would appreciate if you made a complete set regardless).

Edit to add:  Not sure about the 5300, but some cameras appear to have obscure high speed crop sizes when used in movie mode for live view stills.

[Phil Harvey]

Send me a corrupted sample and I'll take a look to see if it can be fixed (philharvey66 at gmail.com).  If it can, then I'll need a full set of samples to add support for this version (explained in the thread that TorC linked).

- Phil