Renaming pre-2000 files

[David Arnold]

Peter, after reading your discussion of file naming, I’ve decided to switch to your system(or possibly a slight variant of it). But first I need to deal with a couple of questions/problems:

I have 15,000+ photos shot before 1/1/00. Using YYMMDD for dating them, sorting by date will place the newer images (000101 and later) before the later ones (991231 and earlier). Other than simply “feeling” wrong, do see any problems arising from this sorting?

While most of the older photos are 35mm slides, some are 35mm, 2-1/4, and 4x5 negatives, and some 2-1/4 transparencies. Should I use a custom field in iView to indicate format, indicate it in the filename (which would add characters to the filename, which I don’t think I’ll have room for--see below), or handle it some other way?

My wife and I shoot together, often interchanging cameras, and copyrights are registered as copyright “David Arnold & Gail Rutman,” and our domain name is arnoldrutman.com, so the only thing that makes sense for Originator Name is 12 characters long. Add 6 for date, 4 for camera-generated unique ID, and we’re up to 25 characters. Then add up to 6 characters for derivative name (e.g., MstrBW) brings it to the 31 character limit.

Incidentally, one thing I’m nervous about is disclosing date shot, since it could kill some stock sales. Your thoughts?

David

[peterkrogh]

I'll address a few items.

As to the pre-200 work, I don't see a problem with the 99xxxx numbering, although I don't really feel the date string is as necessary for scans, since you will produce many fewer of these.  You might even be able to exclude the date string altogether and use a single sequence number.  Personally, I would have no idea the date most of my film images are shot.

I can't help you too much with the prefix string.  You should be able to make it work with your names, it will just take some attention to what your suffixes are.

As to disclosing the date shot, you may want to have a different naming system for images made available as stock, perhaps ArnoldRutman_00001.tif or something. 

As to the format annotation, either that could be in the name, or in the metadata.

Peter

[David Arnold]

Thanks, Peter, that helps. I'll give it some more thought and tried to get it pinned down by the end of the week so I can start implementing your system.

David

[JohnHHarrington]

David --

I am in the midst of going back and dealing with over 5,000 rolls of C-41. My system is the same as Peter's, with a few variations. The relative difference here is that I use YYYY instead of YY. So, my 1988 images organize nicely with my 2004 images.

All negatives were stored in hanging 3-ring binders. Frequent clients had their own binders, one-off or the infrequent client was stored chronologically. I've used my D2x as a copy camera to shoot NEFs of each page over a light table, which I then open in PS and use the INVERSE command to turn the negatives to positive, and then use the "set black point" in Curves (or levels) and click on the emulsion (black set to 10,10,10, not 0,0,0) and voila! a contact sheet in digital form!  During the process, I assign the "Job referrer" or "Job reference" field the name of the binder so that if I need scans from an individual neg, I can go back and get them. To give you an idea as to labor, it took 15 seconds per page to copy, and about 10 seconds per page to do the rename and inversing to a positive. So, that's about 25 seconds per roll.  Paying someone $15/hr, that works out to be about $0.25 per minute, or $0.125 per roll. Once they're in iView, I will go through the catalog and see which rolls I want to have frames individually scanned, but it's so much easier that way. For transparencies, each image is in a legacy database where I can do a keyword search, turn up the image id#, and then locate that page of images by category (images are stored by type: US news, international news, newsmakers, etc) and it takes me about 5-10 minutes from keyword search until I have the page of transparencies in hand. iView has a shot of the page which I can review prior to going to the original material to verify the contents. At some point I will transfer the legacy database contents into the metadata of the files, but currently there are over 60,000 transparencies in about 7,000 files (i.e. 20 or less per file), so that will take some time to make the transition.

Hope this helps.

John

[peterkrogh]

John,
Have you tried to do the inversion in Camera Raw? This will enable saving the file as a DNG, which saves space.  Additionally, you should be able to make the adjustments much faster, once you get the hang of it.
Peter

[JohnHHarrington]

Peter --

No, I've not tried it...I looked and couldn't find where to do it...any suggestions? Also, can I set the black point there too?

John

[G-Force]

Incidentally, one thing I’m nervous about is disclosing date shot, since it could kill some stock sales. Your thoughts?

David

Wouldn't the date shot also be disclosed in the EXIF data? How do you deal with this?

[peterkrogh]

John,
Flip the curve upside down in Camera Raw.

Depending on how you set the curve, the auto settings can work to set black.  I'll be working on this over the next several weeks. I will be making a set of RapidFixer buttons for B&W negs.
Peter

[JohnHHarrington]

Ok, so I've manually flipped the curves, but can't do a "set black point" in Camera Raw....suggestions?

John

[peterkrogh]

John,
You might try futzing with the toning controls on the first page of Camera Raw to see if you can get the Auto Brightness to work for you.

You may not need this, however.  If you load a lot of these images at once, you may generally be able to make one correction in Camera Raw and synchronize it to many pages at once.

Peter

[SeanD]

Then add up to 6 characters for derivative name (e.g., MstrBW) brings it to the 31 character limit.

David...I would suggest that you don't need 4 characters to denote a Master File. I just use "-M" to identify my masters. A color master would be "-MC" (the "C" for color) and a black and white master would be identified by "-MB" (the "B" for black and white). Whether this would work for you depends on how much you may be invested in a previously established naming scheme. I also will sometimes just have a single Master file and use a Layer Set to store the adjustment layers that create a BW effect. That way I only have one Master to deal with that contains different "versions" of how the image can look. Any derivatives from this Master can be labeled to indicate whether is is a color or BW derivative file.

--Sean

[David Arnold]

Wouldn't the date shot also be disclosed in the EXIF data? How do you deal with this?

My concern with respect to dates is for older images, primarily slildes, also some negs, so the only EXIF data displayed would be scan date, not capture. And anyhow, I'm not trying to prevent clients from locating date shot if it's important to them--I just don't want to hit them over the head with it.

David

[David Arnold]

John,
I had rejected YYYY because I was running out of room. But Sean's suggestion (limiting derivative name components to single characters) will give me enough space to work so I use 8-character (YYYYMMDD) date strings. I'll try it on a few hundred images and,  unless I run into an unforseen problem, that's what I'll go with.

Thanks guys.

David

[David Arnold]

A month ago on this thread I wrote:

I had rejected YYYY because I was running out of room. But Sean's suggestion (limiting derivative name components to single characters) will give me enough space to work so I use 8-character (YYYYMMDD) date strings. I'll try it on a few hundred images and,  unless I run into an unforseen problem, that's what I'll go with.

I may have found a way to get the best of both worlds (the proper secquencing of YYYYMMDD combined with the compactness of YYMMDD). Just as the computer sequences 1 before 2 and a before b (and numerals before letters), it sequences minus sign before underscore. So all files named:
    OriginatorName-YYMMDD_UniqueID_SequenceNumber
would show up before those named:
    OriginatorName-YYMMDD_UniqueID_SequenceNumber. ((.g. ArnoldRutman-050101_2728.dng)
would come before those named:
    OriginatorName_YYMMDD_UniqueID_SequenceNumber (e.g., ArnoldRutman_050101_2728.dng)

On the other hand, the subtle difference between - and _ could lead to errors or confusion. Anyone have any thoughts on this?

Incidentally, the full alphabetizing sequence is #, -, _, +, = (at least with Windows; is it the same with Mac?).

David

[DJ Webb]

Do 2 extra digits really matter that much? If you're worried about length why use your full name and not your initials or another shortening?