Film to digital workflow

[kshapero]

I would like to hear about the workflow, people do to go from film to digital.

My steps are:

1. Develop negs only
2. Scan negs with Vuescan 24 bit tiff
3. Save as tiff
4. Photoshop Elements 5.0 for minor adjustments
5. Save as a JPEG 8 file
6. Upload keepers to my website.

 

[charjohncarter]

Nothing wrong with that! If you have to do a bunch more, then it is at the camera end not the PS end.

 

[GeneW]

Very similar to my workflow. I simplify a bit though. I can't see any improvement in Vuescan between scanning B&W negs in col and scanning them in Vuescan's B&W mode. I also scan straight to 8-bit jpg in modest resolution. The idea, for me, is to whip them up as fast as possible, adjust as needed in Photoshop, and post the better ones to the web. If something is really good, I'll rescan at high res to 16-bit TIFF. Not that many of these, so I save scanning time by lower res and 8-bit.

Gene

 

[kshapero]

Very similar to my workflow. I simplify a bit though. I can't see any improvement in Vuescan between scanning B&W negs in col and scanning them in Vuescan's B&W mode. I also scan straight to 8-bit jpg in modest resolution. The idea, for me, is to whip them up as fast as possible, adjust as needed in Photoshop, and post the better ones to the web. If something is really good, I'll rescan at high res to 16-bit TIFF. Not that many of these, so I save scanning time by lower res and 8-bit.

Gene

that makes sense. BTW I get inconsistent color profiles when I scan with Vuescan.

 

[DavidH]

1. Dev negs
2. Scan at high resolution and large size - using Silverfast Ai
3. Pshop - clone out dust and scratches if any, basic tidy up using levels and curves
4. Save as a master file - this is archived (into the same structure as shots from digital slr - those are processed in CaptureOne - most jobs have a mix of film and digital originals)
5. Process further as needed - depending on size and style requirements for a client, for news/magazine, for printing, for the web, using additional tools such as noise ninja, for a particular effect etc...

I use TIF file format throughout.

 

[wilt]

1. Develop films. Rodinal, Tri-X, Efke KB 100.

2. Scan in KM Scan Dual IV. I use VueScan to drive the scaner. Batch mode, 16 bit grayscale. Saved as tiffs. Scan each and every picture on the films.

3. Editing, deciding which pictures to process further: first browse through the thumbnails in Photoshop, then pictures that I want to do something about are moved to a "pipeline" folder. Straight out of Vuescan, the pictures are grey, so in Photoshop I do curves and levels, dust.

4. Some of the pictures appear on my photoblog http://www.5063.com/, those are resized and saved for web.

At first I edited in the scan phase, but now I rather scan the lot. It takes much disk drive space, but I can return to the thumbnails and find new ones I didn't care for at first.

Recently, I got hold of a flatbed scanner with transparency adapter, an Epson Perfection (duh) 3200. Optics are of course not as good as with the film scanner. Only just begun to learn how to scan my MF slides and negs. It is a slower process.

 

[mike_j]

I use colour neg film and get it developed and put on CD by a minilab, they also supply an index print. The CD resolution is only 1895x1272 but for web use and small prints this is OK.

This system saves a lot of boring routine scanning work. The few shots that that I want printed carefully I rescan on an Epson 4990 flat bed at high resolution. All cataloguing and indexing is done in Lightroom. I also use Lightroom for colour adjustment, sharpening, B&W conversion and printing. Actual retouching work is done in Photoshop.

 

[kshapero]

I use colour neg film and get it developed and put on CD by a minilab, they also supply an index print. The CD resolution is only 1895x1272 but for web use and small prints this is OK.

This system saves a lot of boring routine scanning work. The few shots that that I want printed carefully I rescan on an Epson 4990 flat bed at high resolution. All cataloguing and indexing is done in Lightroom. I also use Lightroom for colour adjustment, sharpening, B&W conversion and printing. Actual retouching work is done in Photoshop.

Nice, makes sense, too. I have a prolab that will scan at 3200 x 2300 onto a CD but they charge $10 a roll. Ouch!!

 

[Tuolumne]

Anyone tried Mpix? (www.mpix.com). I mail them my film; they develop and scan it. I buy a CD of all the scans (JPG file size about 1.2 MB) and get the original developed film back. (In the mean time you get to preview the scans on their Web site where you can also order prints from the scans.) Load CD into computer and you have decent size files for the Web or small prints. If I like one alot, rescan negative at higher res. Mpix is cheap, fast and reliable. Unfortuynately, they only handle c-41 films.

/T

 

[nightfly]

1) Develop film or if it's color pick up from lab un-cut.
2) Cut and place in Clear File Archival Elite Sleeves.
3) Number it and pop it down on my Epson 4990 with film area thingy (flexy piece of crap)
4) Scan whole sheet to Photoshop.
5) Look over scan in Photoshop as I would a proof sheet.
6) Create new folder for roll and save proof.
7) Scan only the negs I think I want using Epson Scan Twain driver in Photoshop
8) Create adjustment layers in Photoshop for levels, curves and if color, color balance.
9) Tweak and save as PSD (master)
10) Convert to 8 bit, collapse layers, save as tif (for printing)
11) Resize, sharpen and save as JPG (web)
12) Go into Iview Media Pro and add ratings, tags and selects to catalog sets.
13) Sync annotations back to originals (this writes the tags to the files)
14) If they are going to Flickr, select the jpeg versions and send them to Flickr uploader from I View (this adds the tags I just created to Flickr)
15) Tweak order in Flickr.

 

[Tuolumne]

1) Develop film or if it's color pick up from lab un-cut.
2) Cut and place in Clear File Archival Elite Sleeves.
3) Number it and pop it down on my Epson 4990 with film area thingy (flexy piece of crap)
4) Scan whole sheet to Photoshop.
5) Look over scan in Photoshop as I would a proof sheet.
6) Create new folder for roll and save proof.
7) Scan only the negs I think I want using Epson Scan Twain driver in Photoshop
8) Create adjustment layers in Photoshop for levels, curves and if color, color balance.
9) Tweak and save as PSD (master)
10) Convert to 8 bit, collapse layers, save as tif (for printing)
11) Resize, sharpen and save as JPG (web)
12) Go into Iview Media Pro and add ratings, tags and selects to catalog sets.
13) Sync annotations back to originals (this writes the tags to the files)
14) If they are going to Flickr, select the jpeg versions and send them to Flickr uploader from I View (this adds the tags I just created to Flickr)
15) Tweak order in Flickr.

Wow. I'm exhausted just reading this. That's alot of work! How much time does it take you to do all of that for a roll of film?

/T

 

[nightfly]

I know writing it out was exhausting.

The initial scan of the whole sleeve takes maybe 3-5 minutes.

Then each high dpi scan of individual frames is another few minutes but I can select from two strips of five in my negative carrier and batch scan. Still it's about 3 minutes per frame while it scans them in. This is with the scanner connected to Firewire 400 and the external drive connected via Firewire 800 and a 7200 rpm external drive.

The photoshop work is another maybe 5 minutes per frame. I have some actions going in photoshop but should add some more since some things are repetitive.

Then the iView stuff is another 5-10 minutes per roll.

Probably an hour to scan say 10-15 frames off a 36 roll. I don't feel any need to scan every frame. I've always got the scan of the proof sheet and the negative. On a positive note, it makes you edit a bit more since it's such a hassle which I think is a really good thing and where most photographers need some help.

I'm not very good with iView yet but I have it set up for folder watching and auto importing the new scans. The biggest pain for me is the syncing in iView is a bit retarded. I feel like it should just automatically do this. I also don't have the management of 3 versions of each file very well under control in iView. I wish it bound files of the same name but different suffixes together somehow and automatically made any annotations to all files.

I just added the iView portion recently but I find I'm doing way more with my files now that I can actually view them quickly and easily and find things again. Some go to my blog, some to flickr etc.

I'm also slowly moving over my older scans into this system and this just involves some batch re-naming, tagging and syncing annotations which iView makes pretty easy.

 

[shenkerian]

Do any of you incorporate JPEG2000 into your workflow? I'm considering scanning to TIFF and then converting to JPEG2000 for smaller files. I'd be curious if anybody else does this.

 

[haagen_dazs]

for those who use VueScan , do you guys always use the film profiles that is available? are the colours accurate and reliable?

 

[kshapero]

for those who use VueScan , do you guys always use the film profiles that is available? are the colours accurate and reliable?

Not in my experience. very inconsistent. is it because I have a cheap scanner?

 

[GeneW]

shenkerian, JPEG2000 is one of those 'good ideas' that never caught on widely. Only a few products support it. TIFFs are large but they're universally supported.

Mark, to tell the truth, I rarely use VueScan for colour work. I switch over to my Minolta software for that because it lets me use Digital ICE (on a Minolta 5400). When I do use VueScan for colour, I use the closest colour profile I can find. When I pull the scan into Photoshop, a click on 'AutoColor' usually removes odd colour casts from the scan.

Gene

 

[sienarot]

After developing/cutting/sleeving:

1. Scan only the frames that look like keepers (ie: good composition, exposure, etc...) as 24bit TIFF, 2400 dpi (larger image makes for easier dust clean up for me).
- B&W I use the Epson branded software
- Slide I usually use Vuescan, but I've been tempted to try Silverfast AI again
- Negatives I haven't shot a lot of, but will likely use Silverfast AI for it.

2. In photoshop, fix tilting, crop, clean up dust/scratches and save as 24bit TIFF.

3. (Optional) If more significant editting from the original scan is done, I save it as a different file, again 24bit TIFF.

4. Resize, covert to sRGB, add border/watermark if necessary, save as JPG for web images.

5. Upload.

I've learned from the past that it can be almost impossible to go back and reproduce a picture you've editted, even if it is trivial stuff sometimes. That's why I make sure I always have a full sized 24bit TIFF of the final version I can resize later if I need to go back and make changes to the JPG version. Normally photos I post are no larger than 600 pixels on the longest side, not including the border. When I need to go back for a larger version, a ready copy is waiting for me in TIFF format so it only takes a couple seconds to convert and upload a new one.



My file naming convention is as follows:

YYYYMMDD{abcd...}_XX.[format]

ie: 20060312_03.jpg

where,
- YYYYMMMDD is the date in which the roll was developed, even if the first and last frame are months apart
- {abcd....} is for days when I've had more than one roll developed that day (this is optional)
- XX is the frame number on that particular roll. Makes for easy reference in case I have to go back to the original negative
- [format] is tif or jpg

Images are sorted by folder according to the date the roll was developed.

 

[stumar]

My steps,

1 Develop Negs, Ilford HP5+ in DDX.

2. Scan with Epson V100 to 3200Dpi Jpeg, in BW negative mode on Epson software.

3. Import to Lightroom, boost contrast then possibly add tint and other effects such as vignett.

4. Export to Photoshop to crop to 300dpi and print to desired size.


I would like to be more technical with photoshop in areas like dodgeing and burning etc.. allthogh lightroom i think is superb, simplicity itself with good results!

Stu.