A look at my digital photography post-processing workflow
by Mike Umscheid
. . . A little introduction . . .
My entrance into the digital SLR photography world was January 2005 when I purchased my first SLR camera – the Nikon D70. It didn’t take long before I realized that I had entered a whole new arena in regards to capturing storm (and other) images. Not only were the features of the SLR camera new to me, such as adjusting focal length and focus on the lens itself rather than on the camera (point & shoot) among other things, but shooting “in RAW” was something completely new to me. What exactly was “RAW” anyway, in terms of digital data, and how do I create photographs using these files? I had a lot of questions initially, not the least of which were terminology centered around post-processing RAW files. Terms like histogram, white/black point, tonal range, levels, curves, dynamic range, etc. were all foreign to me when I first got the camera. I did extensive reading on the Luminous Landscape website, and almost all of what I will share with you regarding my workflow and post-processing techniques were derived from the lessons and tutorials I found and read on that website, among other links from that website. Therefore, if you are first getting into DSLR photography, I strongly encourage bookmarking that site.
. . . The software . . .
· A little background/history:
I will state first off that I do not use Photoshop whatsoever. This is a personal choice, that was initially influenced by price, and the fact that I have been an avid user of JASC Paint Shop Pro (PSP) since 1997 and about version 5.0. Through all of these years, I have grown accustomed to how PSP works, and I am very fluent in the program. Until version 9.0 of PSP, it was primarily for the graphical artists, and not photographers. This began changing with the version of 9.0, with more advanced features of the program including digital camera noise reduction, and a crude RAW converter. A RAW converter is basically a program that turns the native digital data that the camera stores into a displayable image on the computer. Each camera manufacturer (Nikon, Canon, Minolta, etc.) have their own RAW file format (for instance, Nikon shooters deal with .NEF files, which are the Nikon RAW). Each individual DSLR camera also has their own unique RAW code, so it can get rather confusing which software to use to actually convert this data into an image.
· RawShooter Premium 2006 (RAW Converter):
You need software to interpret the RAW files and allow at least basic tonal adjustment tools to “tweak” the RAW. This has become my most important tool in processing of RAW images, in terms of making significant tonal shifts. For several months in 2005, I used PSP v9.0 RAW converter, which wasn’t too bad of a RAW converter. At least I didn’t think so. When I upgraded to PSPX (version 10.0) in September 2005, their RAW converter actually became worse. This may have been because of the fact they packaged an interesting little program called RawShooter Essentials by Pixmantec. I installed this program, and it blew me away! The rendered images far exceeded anything PSP ever rendered. This program really knew how to interpret the RAW data. This really opened my eyes. I was extremely thrilled with this program, even though it didn’t allow tonal adjustments using curves and levels, the most important tools for making tonal adjustments. It did, however, allow shadow/highlights contrast tweaks, color correction, and exposure compensation, which were good enough for me at the time. I could always adjust levels/curves in PSP. In January 2006, I upgraded to the $100 version of RawShooter Premium 2006 (RSP06). The major upgrade here was the ability to adjust levels and curves of the initial RAW image, before outputting to a TIFF or JPEG, along with more advanced color correction tools, cropping, and a straighten tool. All fantastic tools to apply to a RAW image before outputting to TIFF or JPEG.
RSP06 has a great workflow setup, in that you can easily organize/prioritize your work in different folders. All settings on a RAW image are stored in a file such that the RAW file is never really touched. You can always go back to the settings you had made at a previous time, and it will be applied to the RAW image.
Typical adjustments I make in RSP06, as necessary depending on the image:
Color temperature/tint (if necessary),
Levels/curves tonal shift (mainly levels),
Crop (if necessary),
Straighten (if necessary)
· Paint Shop Pro X (More advanced processing & prepare image for web/print):
Once I have the image processed the way I like it in RSP06, I will convert the image to a usable graphic file like TIFF or JPEG. In my workflow, I almost never convert to JPEG in RSP06, and almost always TIFF. A little sidebar here concerning color management before I move on, which is somewhat of importance: I have both my DSLR cameras setup for Adobe RGB (1998) color workspace, as well as RSP06 and PSPX. This was a recommendation passed down from my friend and professional photographer Jim Reed (www.jimreedphoto.com). Adobe RGB has a larger color gamut space than the standard sRGB color space. There is a bit of a problem here, though, as no monitor or printer exists that can reproduce all the colors within the Adobe RGB color gamut space, so an image file that was produced using the Adobe RGB space will have to be converted to sRGB color space for proper display on the web or more importantly when the image is ready to be printed at a photo lab. The whole color space issue has been confusing to me for awhile, but I understand enough about it now that I can at least talk about it .
After I convert the image, PSPX is brought up with my image (the screenshot shown includes annotation of my workspace, which I’ll elaborate on below):
I customize my buttons on top, as well as in my layers panel so that I can quickly access my “heavy hitting” items that I use repetedly in PSPX
Color working space: This quickly allows me to change the color working space to and from Adobe RGB/sRGB, in the menu it’s File — Color Management — Color Working Space….
Histogram: Quickly access the histogram for the image. View — Palettes — Histogram
Color balance: from Adjust — Color Balance…
Incr. to 16-bit: Quickly change image bit-depth to 16 bits per color channel
Decr. To 8-bit: Quickly change image bit-depth to 8 bits per color channel, for instance when I am ready to save the image in a final-form TIFF
DCNR: Paint Shop Pro X’s own digital noise reduction algorithm. It is very good. It’s so good that I will not go out and get Noise Ninja or Neat Image. The button I have is for the Adjust — One Step Noise Removal. I use this on a duplicate layer on about 90% of my images (When there is more noise to remove, or I am going for a “smooth” effect, I will use the Adjust — Digital Camera Noise Removal…
localContrast (script): A script I run to apply the local contrast technique, as described on various websites, including Luminous Landscape. This is basically the Unsharp Mask with Radius=70.00, Strength=10, Clipping=0.
saturationBoost (script): This is a script I run that applies saturation to portions of the image that are not already heavy in saturation. Areas with high saturation will see less additional saturation applied. Basically, it’s the HSL adjustment tool applied on a duplicate layer using a mask of the Saturation Channel, the saturation channel was adjusted using curves for the mask. More exact details on the processes this script does later in another article.
o Buttons to prepare image for the web:
Resize: pulls up the Image — Resize… dialogue box
USM LightnessChan (script): A script I run to sharpen the image. It uses the lightness sharpening technique, meaning pixels with higher lightness value will see more Unsharp Mask applied
versus darker pixels of the image.
©Mike Umscheid/www.underthemeso.com (script): A script I run to create the copyright stamp that you see on all my images.
Large watermark stamp (script): Another text script to apply a larger ©Mike Umscheid stamp with very high transparency (about 96-98% transparent)
. . . File Management . . .
Organizing your files, including your RAW files, TIFFs, and web-based JPEGs is important in a fluid workflow. There are a myriad of methodologies for organizing your files, but I’ll share my file naming structure/organization with you here. There’s nothing really involved here, and it is fairly “standard” I guess.
I organize my work “per shoot”, which is usually “per day”. My “root” folder for all my digital photography is NikonDSLR. I usually burn my images to an external hard drive for temporary storage, and this is from where I do all my post-processing. The next level in folders are in the format “yyyy_mm(mon)dd_description”. An example may be 2006_06(jun)02_day11. This file format allows me to order my folders chronologically, and I can quickly glance through the list to see what month the folder represents (jun), instead of just only the month number. I only append a description at the end of the folder if it is shoot other than storm chasing. I also append a reference day # if I’m on my annual storm chase trip. Within the shoot folder, are all my RAW (.NEF) files. Also in this directory are usually 3 or more sub-folders. These include .RWSettings, Converted_Pri#, webJPEG. Since I upload images to PhotoSIG photo critique site on occasion, I will also have a sub-folder called photoSIG, since these images are not uploaded to my personal www.underthemeso.com website. For image file names, I do not change my file names. In other words, the filename created by my Nikon DSLR cameras (i.e. _DSCN3234.NEF) is what I stick with. I keep this same filename throughout. This is because it is easy for me to reference back to the original NEF file.
.RWSettings: This is the directory created by RSP06 that stores individual files with the same name as the RAW files, that store every single setting (including levels/curves, color temp change, exposure compensation, straighten, crop, and more).
Converted_Pri1: Depending on the shoot, I will have 1 or more priority levels of images. Larger shoots, I will usually have 2 or even 3 priority levels. This is the folder that is created by RSP06 (based on my input) where all the converted TIFF images are stored. When converted, I have RSP06 append a “-nn” (i.e. _DSCN3234-01.TIF) to the filename, because for each RAW I may have two or more working TIFF images for blending techniques (in high-dynamic range situations). This TIFF image becomes my working file in PSPX. When I am done processing the image in PSPX, I resave the TIFF as an 8-bit per channel image using LZW compression (which is lossless compression). The final file name in this directory will be stripped of the “-nn” (i.e. _DSCN3234.TIF).
webJPEG: Once the TIFF is completed, I am ready to prepare the file for display on my website. For display in my gallery on Underthemeso.com, I resize to 800-850px on the largest side for horizontal orientation and ~725px for vertical orientation. After resize is when I apply sharpening and my copyright information/watermark.
After I process a shoot, I will then burn to a DVD for semi-permanent archival. I say semi-permanent, because my archival system is still a work in progress. I have developed a library system in Microsoft Excel to catalogue all my shoots. I will discuss this more in a future article.
. . . Summary of typical workflow . . .
To recap, here are the steps I follow to process a shoot
(this step-by-step is for one individual image):
1. Download RAW files from camera to shoot directory.
2. Prioritize important images in the shoot, and delete others. I usually do not delete a whole lot, unless the image is completely out of focus or has other technical flaws.
3. Make tonal adjustments/enhancements to the image using levels/curves/exposure compensation. Further tonal adjustments using highlights/shadows contrast may be necessary, depending on what how I want the image to look
4. Adjust color temperature/tint to my liking
5. Straighten and crop the image as necessary.
6. Create another adjustment tab for the image, if necessary (for very contrasty images, like highly backlit storm structure. This is part of my HDR [high-dynamic range] technique, which will be explained in a later article). NOTE: I do not use a third-party application for HDR. I use mask/layers with a combination of paintbrush/gradient fills on the mask.
7. Convert to a working TIFF file for PSPX
8. Apply local contrast technique (one of my scripts) to my liking. Sometimes on a layer
9. Apply saturation boost technique (one of my scripts) to my liking. Done on a layer
10. Address further saturation, if needed
11. Make a final color correction using a curves layer. Some images appear too “green” to me, so I usually make this adjustment. Again, I usually address color correction with a curves adjustment layer, although sometimes I’ll use PSPX’s Color Balance engine.
12. Noise removal using PSPX’s internal noise reduction engine. Done on a layer to control the effect.
13. Reduce to 8-bit color channel
14. Dust spec removal using clone brush (painting only works in 8-bit)
15. Save as final TIFF image
16. Move on to next image
17. After all images are completely processed to TIFF, begin web display preparation of each TIFF
Prep for the web (PSPX)
18. Set color working space to sRGB. Open image and acknowledge conversion to sRGB (from Adobe RGB) color space.
20. Run my sharpening script.
21. Add the copyright/watermark tags
22. Upload to gallery on Underthemeso.com!
23. Have a beer.
24. Rinse and repeat.