What's your photography workflow like?

Hello all. My name is Mike & I’m super new here.
I’ve been a Lightroom user (v3 up to v5) and since I’ve since transferred my primary work-space from a Win7 laptop to a similarly aged MBP running OS X 10.6. :slightly_smiling:

Because of LR catalog issues I’ve started dabbling on the following software last year:

  • RawTherapee 4.2.270 : looks like the strongest contender to replace LR, however, I still have questions and certain roadblocks I’d like to get out of the way.
  • DarkTable 1.6.8 : longest learning curve for me, but found a way to use its SVG signature feature which has become the only feature so far I can’t let this tool go.
  • DXO Optics Pro 7 : found the auto-slight-HDR and auto lens distortion correction as great starting points upon ingest.
  • LightZone 4.1.2 : found this useful for one-off edits, leaning towards a not-so-friendly interface, ranks lowest for me.

Other supplementary tools I’m using:

  • XNview : for batch IPTC edits & batch renaming

So after several months of tinkering, I’d like to post some questions here hoping someone has had a work-around for the feature or maybe just some advice going the right direction. I have been wanting to use RawTherapee solely for my workflow, but as of today I find myself ingesting photos into DXO then run the batch on DT for the SVG signature and only use RT for more detailed processing.

I’ll post my questions separately as they come, thank you for your patience & I hope to learn from everyone willing to share.

Hello Mike,

I wrote some articles about photo watermarking using (free) software, how to do this by hand and how to batch process that. It’s written for a Linux machine but I guess you can use it as well on your Mac.
Have a look on my site if you like.

In the menu under Hybrid - Watermarking you can find some more ways to watermark your photos.

Hi @mgco! Welcome to the community!

I am in a similar OS state as well (Win7 at work and OS X at home - linux on various VM’s across the two).

This is one of the reasons I primarily work in RawTherapee. I like being able to work on images during lunch breaks at work and to be able to pick up and continuing processing at home without skipping a beat. Though I have been known to dabble in darktable from time to time. :wink:

My general workflow focuses on specific tasks that I’ll usually need to accomplish:

  • Quick contact-print view of all the images.
    Initial pass at exposure validating, focus, and composition.
    Usually done lately in native OS file manager or using imagemagick montage to create an actual contact sheet image.
  • Validation + work in raw processor.
    Verify focus and composition, sort by star ratings to winnow the results down to potential keepers, work on exposing the raw.
  • Export for further processing depending on workflow.
    • GIMP + G’MIC for various types of retouching
    • Hugin for panorama stitching or pixel-bending or creating lens correction parameters
    • Luminance HDR (usually happens earlier in the workflow lately with HDRMerge).
  • exiftool to maintain Exif data through the various pipelines
  • imagemagick for any batch-processing needs
  • Bash (batch renaming/moving/copying/etc…, processing needs)

In this case, why not just stay in darktable all the way through the workflow? :slight_smile:

Honestly, I’m not a fan of watermarks personally but if you need to add them for sure imagemagick is a super-fast and easy way to do it with a single line in your shell. Since this is usually the very last thing done to images it’s often something I’ll leave to imagemagick in batch as needed.

If I get a moment soon I can test a quick imagemagick script to apply a watermark and post it back here in case anyone wants it. (Pretty sure it’s a simple composite command, or a mogrify + composite overlay).

  1. My workflow goes like this:
  2. exiftool script runs on mounted SD card to import files into “raw” folder in the folder structure YYYY/MM/DD
  3. “raw” folder is a git-annex repo, all files are checked in to git-annex and (eventually) synced to other offline disks for redundancy
  4. preview images with geeqie
  5. one of the following:
  • if it is a multiple exposure shot (either bracketed or a pano) process with EV 0 & a neutral curve with ufraw, then process with hugin
  • if it is not a multi exposure shot, process with darktable or rawtherapee (I haven’t made up my mind either)
  1. pull into gimp for further retouching if needed
  2. gimp wavelet sharpen

My workflow works like this:

  1. Import into Filmulator (simultaneous import to main hdd and backup; Photos/yyyy/MM/yyyy-MM-dd
  2. Enqueue on import if I only used one camera, otherwise I import all the cameras and then manually enqueue them from the Organize tab.
  3. I go through the queue processing photos. Dequeue photos I don’t like, save photos I do.
  4. Use GIMP or RawTherapee (or even both) if needed (not often)

The end.

Just this will do it.

for i in *.jpg *.JPG; do echo $i; composite -gravity south-east my.watermark.png $i wm.$i; done

(Note: this is one line).
This says:

– take files ending on jpg and JPG in the current folder
– use the composite command that is part of the ImageMagick toolbox to overlay the file my.watermark.png on those jpg’s or JPG’s
– use the -gravity switch to say where the watermark must be placed; north, south, east or west and their intermediate directions/places: south-east, etc.
– write ‘wm.’ in front of the file name to indicate that it is watermarked.

Note that the original files will not be overwritten.

My main instrument is a (virtual) fridge. Basically I copy everything from the SD, splitting things between JPG and CR2 (assuming I also got them…), and go over the JPEG to check for technical errors (mostly focus or motion blur problems(*))(I use Gwenview for this). I’ll also often sort the JPEG into several directories. Then I run a couple of scripts to reconcile the JPEGs and their CR2s, and delete the CR2s of the JPEG I erased.

Then comes the fridge. I don’t look at the pictures for at least a couple of months. And when I do again, I’m much less proud of them. They could just as well have been shot by someone else. Not that many survive this. The fridge is my best critic.

(*) My next camera, with its better focus and somewhat higher ISOs will of course not change a thing.

1 Like

You must be joking! I work on photos when they are still “warm”, shortly after I took them. The mood, the colors, the sounds of the scene, everything is still in my head and I need that to recreate the moment I took the photo and to know why it was that I took the photo. In a perfect scenario I can bring back that ‘why’ part in the photo I show up to others.

I’m not. I believe the picture should speak for itself, if it’s good I’ll remember/figure this out.

1 Like

If a good, ruthless culling is required I often find time to be a good starting point at weeding out so-so from something possibly better.

I do understand the desire/ability to process in a way to capture the feeling you may have had at that moment. (It’s just not the way I personally tend to work).

  1. I use digiKam for DAM - I download photos from the SD card, one folder per shooting session e.g. /raw/2015-05-12 kittens/, review them, delete the bad ones and rename the remaining ones.
  2. Then I convert the raws to DNG using the latest version of Adobe DNG Converter.
  3. If needed, I merge bracketed shots using HDRMerge. I keep the original raws for the time being.
  4. I develop the raw files using RawTherapee.
  5. Final processing is done in other software - GIMP, Hugin, Autopano Giga, my own scripts, etc.
  6. I apply metadata to the raw files and the final processed images using digiKam. Sometimes I do this step before step 4.
  7. At this point, once I’ve processed the raws and I know that they had no defects, I delete the original raws and keep just the DNGs and the HDRMerged ones. These are what I backup.

I prefer that one program is best at one thing, than a swiss army knife made in china.
I try to be brutal when deleting bad ones. If a photo doesn’t excite me, if it doesn’t blow my skirt up, then what’s the point of wasting more time and space and money on it.
I store raws in /raws/ and final photos (developed and stitched) in /developed/, completely separate, as it makes file operations easier.

To put it mildly, I “have been known” to have a wee little 5 year gap between steps 1 and 2.

1 Like

My workflow is this:

  1. After photoshoot I review shots in DSLR and delete shots I dind;t like (I called this revision 1 & elimination)
  2. Then I copy the rest images to a dedicated folder stating the year, month, day and title of photoshoot if it is a unique project.
  3. Open RAW images in Darktable and view them in its Lightable mode and give them ratings or mark as rejected (revision 2 & elimination)
  4. Process only highest rated images and export them for retouching in GIMP
    (if I need to show images before retouching for a selection to whom I work on a project, I use converseen to scale them down so they would take less HDD space so I could send them fast to my colleague).
  5. I retouch only the most liked images in GIMP and GMIC (works like revision 3 & elimination)
  6. If i need to give away images to a model etc, I export them without water mark. For my own publishing I manually put my logo on each image as well as I manually scale down to HD resolution images that I put online. I do this manually because I want to be sure my images stay sharp the way I want them to be.
  7. After couple of months if I know I will not go back to the project, I delete raw files.
1 Like

Hey Morgan,

Why the conversion of the raw to DNG and not just work on the raw file directly?

Do you keep the raw and the DNG?

My workflow is pretty straightforward:

Import all files from the SD Card to the NAS, directory structure is:


So should a single day cover more than one major event, then 2 directories are created.

I then fire up RawTherapee and open up the directory in the browser tab and increase the thumbnails to the largest, normally I get 2 images in a row and quickly go through then and star, 1 star is generally what first lot get, noting that I can’t always see how well the focus is in the image.

Once done, I reduce the image thumbnails to the smallest and filter on 1 star and hit the Editor. I work through each image in turn. I use Auto White Balance on the camera, but many times images can take the previous settings, so I do get some time saving there with a copy paste between images. If I don’t have too many images to process I move through them all applying the changes I like, then once all done I click CTRL-A and CTRL-B to put them all on the queue and process.

That’s about it from that point of the processing as I don’t watermark and rarely do any additional post processing.

One thing I really want to do though is manage the images in a Digital Asset Management system. I’m using Windows 10 for the most part, plenty of linux servers in the house and a desire to move back to a Linux desktop, but at the moment, Digikam is a bit broken for Windows and I was using it for tagging and key wording.

One thing I have found over time though is, once you get a workflow or application under your control, it’s hard to switch between different ones. I tried using Darktable and Digikam’s processing tools, but found that RT’s workflow is much more to my liking.

I would be nice if there was a keywording/tagging capability, but I appreciate the Unix “do one thing well” philosophy it has.

1 Like

Use the Inspect tab for that. Activate that tab and place your mouse somewhere on a thumbnail (anyone), see screenshot.


Here is my typical workflow:

  1. I set the camera to Auto-WB and save RAW+Jpeg. The auto WB allows me to have always decent Jpegs for initial inspection, and since I always process the RAW files for the images I’m interested in it has no influence on the final result.

  2. I copy the images from the memory card to one of my hard drives, without changing file names and preserving the original folder structure in the camera’s memory card. This simplifies backups and mirrors across disks.

  3. I run a bash script that creates links to the original image files (RAW+Jpeg), the links being given an unique name based on a prefix + a growing numerical suffix. RAW and Jpegs are put in separate folders for easier searching.

  4. Before deleting the images from the memory card, I make sure that I have at least a copy of the images on a second hard drive. In fact, I’ve seen enough broken or stolen hard drives to understand than an HDD is far from being a permanent storage! Keep copies of your data on at least two hard drives in two different physical places (home & work, for example), and you have already taken good measures against 90% of the typical data loss causes.

  5. I open the Jpegs with picasa (shame on me!, I know… :disappointed:) and mark with a star all the pictures that are technically correct and artistically promising (or at least not totally boring). I also give labels to the images at this stage (family, landscape, macro, etc…) and also give a special label to those I’m really proud of.
    By the way, if you have a good picasa replacement that works fine on OSX to suggest, I’ll be more than happy!

  6. Sometimes I immediately jump on the good images and process the corresponding RAW files, sometimes it takes weeks or months until I start doing some editing… it depends on the mood and available time.
    I’m editing all images with my own PhotoFlow editor, and keep the files in a folder structure of this kind (where the Jpeg folders contain the in-camera Jpegs, the PP folders contain the RAW files + the photoflow files, and the Final folders the output TIFFs and web-scaled Jpegs):


|__ Jpeg
|__ PP
|__ Final
|__ Jpeg
|__ PP
|__ Final
|__ Dragonflies
    |__ Jpeg
    |__ PP
    |__ Final

I preview the images (raw+jpeg) directly from the card with Gwenview and copy only the best raws to the hard drives in a special folder of raws(with the name of the client, generally private or public enterprises). Then i process it in Rawtherapee and give the final touch (cloning of dirt) and optical corrections with Gimp 2.9.3.
The final images where all exported to 8 bit tiff and Adobe RGB profile. To the client, normally i send it full size Adobe RGB jpegs and reduced sRGB images converted with XNview. If neccesary, i batch rename the files with KRename. All the workflow 100% Linux on a old 2008 MacPro 3.1 with SSD (works spectacular in Manjaro JWM).

1 Like

I do a couple of different types of shooting, but the only “serious” work is an annual pilgrimage to the Cumbres and Toltec scenic railroad to take train pictures. This past fall, I attended a “photo freight” event, where they put together a train and do runbys for folks to shoot.

What I did last year:

  1. RAW+JPEG, exposing for highlights as most scenes were very contrasty.
  2. On my cheap tablet, copy the NEFs to a single directory, each day’s JPEGs to a separate directory
  3. Each evening:
  • scan the JPEGs for interesting images, and work them in Picasa. Typically, I pull up the shadows, maybe crop, save to a separate file.
  • from the worked images, select a subset for resizing and posting.
  1. Later, on my desktop computer, go through the worked JPEGs again to look for composition worth spending time with in Raw Therapee and GIMP. I have yet to do this… :slightly_smiling:

What I plan to do next year:

  1. Shoot RAW only.
  2. On cheap tablet, copy NEFs to a separate directory for each day. Scan the thumbnails for interesting compositions, open each in rawproc, add gamma, tone curve to suit, and save as a TIFF.
  3. Scan the TIFFs for the ones to post, open them in rawproc, resize and maybe crop or sharpen, save as JPEG.
  4. Later on the desktop computer, go through the TIFFs to cull the very best images for PP in RT and GIMP. rawproc saves the command chain in the TIFF EXIF, so I know what where to start.

That’s why I started writing rawproc; I had a vision of workflow, and I couldn’t quite put it together with current tools. Well, and the cheap tablet motivated me too, wanted something fast and to-the-point.

Also, I recently dug out my old F2 and lenses, and a film tank I bought years ago to develop a black-and-white roll of a few family shots for my wife’s school board campaign brochures. I’m going to shoot and develop a few film rolls of trains, and either scan or shoot the negatives for digital printing, try to relive some of the good ole days… :smiley: I’m going to put an invert operator in rawproc for that endeavor.

Thanks for the tip! :slight_smile: