Processing steps when processing raw pictures?

Hi there!

I recently purchased @RileyB’s excellent video lessons on open source photography (side note: excellent pack that covers everything you need to know to use Darktable and understand what it does), and I started processing my raw pictures instead of using the jpg files from my camera (an Olympus OM-D E-M5).

Two questions for the people out there who process their raw pictures:

  1. How long does it take you to process one photo?
  2. What is your usual process?

In my case, it still takes me 8-10 minutes per photo, which is pretty demotivating cause I usually have tens (if not hundreds!) of pictures from different sessions.

My usual process is something along these lines:

  1. lens correction
  2. exposure
  3. contrast/luminosity/saturation
  4. highlights/shadows
  5. levels
  6. crop/rotate
  7. denoise
  8. sharpen

Thanks in advance for your feedback!

3 Likes

I have a preset for the most basic things that saves me a minute or so. It includes denoising, sharpening and some saturation correction. It works for low iso values up to 400, for higher values I adapt the parameters manually, but I think one day I will do a high iso preset as well.

In general the time for one picture is somewhere between one minute and several hours. It depends on the image. Check the videos from Harry Durgin (https://m.youtube.com/user/harrydurgin) and your 10 minutes will feel ultrafast :wink:.

Furthermore, if you have several images in similar light, you could copy one edit to the remaining pictures, which I use a lot as well.

Edit: And I must admit that I have a backlog of several 100 if not 1000 unprocessed images :expressionless:.

I’m going to assume that I’m pretty slow, but it usually takes me at least 30 minutes, but usually more like an hour.

  1. lens correction
  2. exposure
  3. LAB curve adjustment
  4. Wavelets (contrast, edge sharpness)
  5. crop/rotate
  6. denoise (usually don’t need this as I shoot ISO 100 almost all the time)
  7. touch ups (clone tool, heal selection, maybe light masking and contrast adjustments, random GIMP adjustments)
  8. sharpen (g’mic richardson-lucy or octave sharpening)

Although I use darktable since a few years now, I find the video lessons from @RileyB really interesting. I guess they will save beginners a lot of time.

5 s to never really finished. Sometimes it’s enough to click a few presets or even copy over the development from a previous image (Ctrl+c & Ctrl+v) and finetune it. Sometimes I come back to an image after months, because I discovered someting new or I try new ideas.

A set of presets for modules like highpass, local contrast, profiled denoise, sharpen and some styles especially for denoising saves me a lot of time. So on everage maybe 2-3 minutes per final image including sorting, tagging, which is quite timeconsuming.

darktable has a shallow learning curve, because it has so much to discover.

I don’t have one, if you ask for specific darktable moduls.

As hobbyist I only shoot ~7-9000 RAWs per year and I try to keep the number of shots as low as possible. People around me take many shots of the same scene. The advantage of RAW for me really is that I can concentrate on framing, focus and exposure. So I don’t need 20 shots of the same scene, which saves time and much more :relaxed:.

I’m at the extreme of speed: I rarely spend more than 30 seconds on a photo. My editor Filmulator is practically a preset for what I like, with no other functionality to get in the way. Heck, I haven’t even implemented cropping because I don’t do it often enough.

It also helps that my lenses are decent enough that I don’t really need sharpening beyond the natural local contrast boost of Filmulation and they don’t distort so badly as to warrant correction.

1 Like

Like some (with the exception of the very speedy @CarVac!), I’m between about 5min to never done. :slight_smile: Importantly, I will usually have a series of images when I’m working that are all shot in very similar lighting conditions (either off-camera flash or at the same time of day/setting). This means that broad-strokes like exposure or overall color rendering can be adjusted on a representative image, and then easily copy/pasted onto any subsequent images (or batches of images).

This way the overall parameters can be pasted, and I can individually modify other parameters as desired. I find that from image-to-image I’m only usually adjusting a few small parameters once I get the first pass done (exposure tweaking mostly - maybe some color work if it’s a different complexion on a person).

The actual application of the adjustments is fixed in RT/dt, but for the most part I tend to:

  1. Exposure
  2. Highlight/Shadow compressions as needed
  3. Lightness/Contrast/Chromaticity adjustments to taste. (I find on my E-M5 the need to pull the lightness down a little bit).
  4. Noise Reduction (Impulse + chrominance noise - I often leave luminance noise alone).
  5. Color fiddling

Once I have something I like, I’ll bulk-apply as needed and usually only need to make minor (quick) adjustments to exposure/hightlight/shadows between shots.

I’ve been processing amsterdam.pef for six years now, I feel I’m really close to getting it right…

4 Likes

That’s great! I love hearing that people are learning a lot from my course :slight_smile:

If I am just editing RAW photos, generally from 30 seconds to 5 minutes. Of course, that is just an average. Some images will take longer.

I find that since darktable has so many different modules to play with, it is easy to spend a long time editing. Following a process helps a lot. And it looks like you already have one.

I will echo what others have said. When it comes to quick editing, presets and batch editing are you friends.

Like @patdavid already mentioned, I also find that I have a set of images with the same lighting conditions. I figure out how to edited one them (maybe try it a few different ways), then copy the edits to the other photos. Then I move on to the next set of images and repeat.

1 Like

Thank you everyone for your answers!

It looks like I’m not completely off-the-hook there :slight_smile:

I will try to bulk-apply changes from one photo to another, hopefully it will work well and save me some time.

1 Like

For me it’s:

  1. Tone curve 1 - custom / film style w/s curve.
  2. Tune curve 2 - parametric / shadow down, dark up, highlights down. Film style.
    After that, denoise, play with ciecam and color tab. I do sharpening in Gimp.

Most of these are saved in custom presets I created to get closer to Canon styles.

1 Like

I start with a basic raw conversion, just demosaic, white balance and colorspace defaults courtesy of libraw. Then:

  1. gamma 2.2 to stretch out the tones from the linear recording;
  2. black/white point to fill the histogram;
  3. tone curve for any other manipulations.
  4. rotate if straightening is needed

Save that as a TIFF as the basic image. From there, any of:

  • saturation, not much, but sometimes I overapply for effect.
  • crop, if the mood suits. This might produce a few versions, for different compositions.
  • grayscale. Been into black and white lately; I’ll sometimes use the channel mixer to darken skies or bring out a particular color.
  • for some websites, resize to 640xAspect, then apply the smallest sharpen to introduce acuity

These usually are saved as separate JPEG images, suitably named so I can keep track of them. My software stores the processing commands in the exif, so I can reconstitute any of the PP for a particular image.

If I need to do other correction or messing around, I’ll throw the TIFF or JPEG into GIMP.

I’m debating throwing in an input sharpening step, but it doesn’t seem to make enough difference to warrant the trouble. The output sharpening step for web resize makes a significant difference, however; reduction resizing loses resolution, and just the first sharpening step makes the reduced image look a lot crisper, for lack of a more technical term. Makes my D7000 look more like a D810… :smiley:

I may write a resize-sharpen script, but that’s about all the “automation” I’ll consider. I’m not into presets; I want to start as close to what the camera measured and go from there.

1 Like

Having a good preset really speeds up everything significantly. I have a standard darktable style I apply to every RAW image that includes:

  • Input color profile (I use Rec2020 because it produces much better results for A6000 than the standard or Adobe ones)
  • Base curve/tone curve (profiled for my camera)
  • Profiled denoise (heavy wavelet denoise on HSV color)
  • Profiled denoise (very low strength denoise on lumiance channel)
  • Automatic lens correction if it is detected in EXIF
  • Equalizer (for clarity boost with parametric mask to limit it in the mid-tone)
  • Color checker lut (to emulate fuji astia film)
  • Highpass sharpening

These gets me about 80% there. After that, I usually tweak (which will vary for each photo):

  • Exposure correction
  • White balance
  • Shadows and highlights
  • Color zones to selectively lighten/darken/(de)saturate each color
  • Tweaking denoise options for high ISO images
  • Vignetting, GND
  • Crop, rotate, keystone/perspective correction

Finally, it gets loaded into GIMP for healing, selective sharpening, dodging and burning, etc, if they are required.

Processing an image can take anywhere from 2 or 3 minutes to basically infinite. Most of the time problematic ones are those with mixed lighting, and sometimes my taste changes and I go back to re-process my older photos as well.

After processing a batch of photos, I also find it very useful to let the photos “settle” down for a while (if getting the photo out quickly isn’t the top priority). Go do something else for a few hours and come back to the image, you will usually discover flaws in your editing that you glanced over in the first time.

3 Likes

yes, same here, having presets. As an old Bibble user then switching to AfterShot - I know that’s not oss, sorry - I made up scripts which also write side car files (xmp) in which the processing commands for AS are written on basis of camera, lens, f stop, iso, time and measured light (if recorded). This way when AS opens the file it also read in the xmp and processes the there given commands. Thats almost the same like having presets, in the end it just saves one click or so :). As others said with this kind of presets one can get about 80% done.

I’m not very happy how AS is performing since Corel had taken over, so I’m now switching to something else.

I haven’t looked enough into RT or DT but I think this kind of “presets” should be possible there too.

Cheers
Martin

There are a lot of user contributed styles for darktable and rawtherapee has film simulation, and you can roll your own style in either.

Hi, sorry to be very late to that discussion, but could you provide more details on the darktable style? Like how you setup the different options… cause even though I ran into some of the options you describe, I have no idea how to set them up the way you mention!

Thanks :slight_smile:

https://dtstyle.net

Here you can find a lot of darktable preset Styles. You have to download and load them un style module.
And here you can download wonderful João Almeida’s film emulation for darktable too.

http://blog.joaoalmeidaphotography.com/en/tag/film-emulation/