Portland, Multnomah County, Oregon

My Raw Workflow – in detail

As promised earlier, here is my RAW format workflow.  Not because I think it’s great or the best, but because it is what I use, and sharing it might help someone else and might even help me, through ideas readers have gained from their own experiences and are willing to share via comment.  I’ve seen other workflows and they are nice but usually not detailed enough to help me actually make improvements to my own workflow.  Although I’ve attempted to explain things in detail here, there may STILL be parts of this workflow that need further explanation.  If that is the case and you want more information, please ask your question in a comment below and I’ll do my best to answer.  Grab a beverage, sit back, and try to stay awake!

I import my RAW photos from my compact flash cards into Lightroom 2 (LR).  LR requires me to import my images, which also automatically catalogs them.  I wouldn’t do keywording or tagging of photos at this point since most of the shots will probably be deleted once the process that follows is completed.  Do keywording after all deletions have been made.

I have LR set to automatically apply a baseline set of adjustments to every RAW photo imported. These adjustments do not change the original RAW file, they are saved in small sub-files (sidecars) that the program executes against the original RAW file.

My auto Pre-sets for imports in Lightroom are as follows:

Exposure 0

Blacks 5

Brightness 50

Contrast 25

Clarity 31 (this is not sharpness)

Vibrance 31

Saturation 2

Sharpening 15

Radius 1

Detail 25

Color NR 25

Luminance NR 0

For some photos, these base adjustments are all I need in Lightroom (but there is always more to do in Photoshop for the photos I pick to process).  But many times I need to +/- exposure, +/- fill light, +/- blacks, etc., in LR, depending on the image.

I have a histogram displayed in LR and use it to get the exposure as correct as possible.  This is the beauty of RAW.  You have quite a bit of latitude for exposure and other adjustments but, always try to get the right exposure in the camera as SOME noise will be introduced when you change exposure, fill light, etc.  When too much adjustment is needed and it adds too much noise, then the photo may not be salvageable.  I also have a 2nd chance at adjustments in Photoshop but more on that later.

Here are the steps I take once the images are in Lightroom:

1. I quickly look at each photo and delete the ones that are obviously not going to cut the mustard!  Way OOF, bad light, bad pose, obstacles in front of bird, etc.  I look at each one at a 100% crop–1:1.  A good sharp photo should show the bird’s eye and surrounding feathers still pretty sharp even at a 1:1 crop.  You should be able to see individual feathers.  In LR, a click of my mouse jumps to a quick 1:1 crop.  This is MY benchmark–you may have your own.  But delete the photos that you wouldn’t possibly want to show anyone else.

2. An exception might be to keep photos of birds you have not ever shot before, even if they aren’t the best–just for documentary purposes.  When you get a better image of that species, then you can delete the old one.    

3. After I’ve completed step 2, at least half of my shots are now deleted.  But now I go thru the ones that are left and try to choose the ones I want to take time to process.  These are shots that have the best pose and are the sharpest with the best lighting, etc.  You might have multiple shots of the same bird in almost the same pose.  You need to compare these and select your best– most in-focus shot(s), best pose, etc., out of the bunch that you want to process (it can get very tedious).  I flag these as a “pick”.  So out of 500 shots, I might end up with, say, 30 shots of initial picks.  These are good enough to show to someone else and ones that are worth taking the time to process further.

4. I then look at each of my picks and crop them to where I want them.  If they look too pixilated after a crop, I’ll either crop lighter or just throw out the photo.  LR allows some spot removal, too.  This is a type of clone tool where you can remove some simple spots from the photo if they are distracting.  An example is glare in the eyes when I use flash. I’ll clone over the glary spots usually with black or whatever color the bird’s eye is. LR’s spot removal tool is not nearly as powerful as the clone tool in Photoshop though—it really is meant to do spots only.  LR also has a graduated filter and an adjustment brush (both of which are quite handy).  I can make parts of the bird lighter if it has shaded areas but there is a limit to the corrections you can make before it starts looking weird or adds too much noise.  The graduated filter tool is very helpful in shots where an adjustment is needed to a larger area of the image.

5. Note that my original auto adjustments (listed at the top) include a low amount of sharpening.  This is because I do the major part of my sharpening in Photoshop (PS).  I’ve read that LR’s sharpening is not up to par with PS, so I do it this way. (This may not be the case though with Lightroom 3, which I don’t have as of Nov. 2010).  Now, with Bibble, PSP, DPP, and other RAW editors, you will have to experiment to figure out which one gives you the best result (sharpening in the RAW editor or in PS).  Another reason I sharpen in PS is because some pros say that sharpening should be the last adjustment you make to your photo.  I suppose this is debatable like everything else!  I also make sharpening and noise reduction adjustments in PS because I can do it selectively, where Lightroom’s masking tool is not precise enough for me.

6.  Now that I’ve done everything in LR that I can, I select all the adjusted ‘picked’ photos I want to export to a subfolder of where the original RAW photo is located on my drive, then run LR’s export function.  I made a preset in LR so I don’t have to select a bunch of attributes each time I export a batch.  My export preset name is “TIF photo for PS”.  I export as the largest TIF possible in 16 bits/channel and 300 pixels/inch.

7. I export as TIF’s because TIF is not a lossy format and I can adjust it as many times as I need to and save it repeatedly if need be without worrying about image degradation.  I can also save layers (if needed) if I save the image in the TIF format.  Many people probably export their RAWS as JPG.  It’s an individual choice but much of my workflow may not apply if that is the case.  I then bring the TIF photos into PS and select the bird and any other branches, leaves, etc that should be sharp, and save that selection (many times this is easier by selecting the NON-SHARP stuff first (like a clear blue sky) and saving that–then invert the selection to get the bird selected).  I then run the photo thru neat image, de-noising the background and sharpening the bird ‘some’.  I then might do some contrast (curves) adjustments to the photo if I think it gives it a better look.  I then run an Unsharp Mask (USM) filter adjustment on the bird only and sharpen it to whatever level makes it look the best.

Details of Sharpening and Noise Reduction in Photoshop:  Lately, I have started to use the USM filter in steps.  I’ll set the Amount slider to about 80%, the radius to a range of .7 to .9, and Threshold to 1.  This would be for a photo that I think is pretty sharp already (not real soft that I am hoping to resurrect!).  With the part of the image that I want sharp selected (like the bird, perch, nearby leaves & branches, etc.) I run the USM filter 2 or 3 times on the image, actually over-sharpening it a little.  I then use either Photoshop’s noise reduction filter or Neat Image and use it to “back off” on the over sharpening I just did.  The NR strength using the PS filter is usually in the 3 to 6 range; it depends on what setting gives you the best trade-off between good detail and low noise.  Another option to use is, before you apply sharpening, go to Image/Mode/Lab Color and then select the Channels Tab in Photoshop.  Select the Lightness channel, then apply the sharpening and NR steps above to the Lightness channel only.  When done, go back to Image/Mode/RGB Color to return to the normal setting and see how it looks.  Some say this process introduces less noise into the image.  For me, the jury is still out.

In Photoshop is also where I have been known to clone out distracting branches and undesirable elements.  Some photographers believe this is cheating but I think most are okay with it as long as the natural essence of the scene is not severely modified.  Learning to use your clone tool and healing tools is real important.  I also use the dodge, burn, sharpen, and blur tools quite a bit.  Also, in Photoshop CS5, they have a new feature called “Content-Aware fill.”  It does an amazing job of filling in areas of a photo based on the surrounding pixels.  This tool can be found as a choice on the Spot Healing Brush tool screen.  Have you ever wished a stray branch wasn’t crossing in front of the bird’s body?  This tool just may be able to remove it and fill in behind it well enough so that with a little bit of clean up, you’d never know the branch was there.

8. Once the photo is looking its best, I ‘save as’ and add “modinPS” to the name so that I now have the TIF original exported file plus I have identified the modified TIF file with a name that tells me it has been modified.  I then change the TIF image from 16 bits/channel to 8 bits/channel.  I also change the size of the photo to about 900 pixels on the longest side for web upload purposes.  I then convert the color space in the TIF image from Adobe RGB (1998) to sRGB (which is recommended for web viewing).  Then, I finally ‘save as’ again but this time I choose to save as a JPG file at the quality in PS so that the photo ends up at about 150-250 MB in file size.  Now, I have the 2 TIF images and one JPG image of the same photo.  I now close out that image still open in PS, WITHOUT SAVING because I don’t want my modified TIF image to be overwritten with a smaller size, different color profile and an 8 bit/channel picture.  I just made those 3 changes for the sake of the JPG ONLY.  I then upload the photo to Flickr, etc., depending upon how good it is and who I want to see it.  I realize that the final format of your image (printed, web, etc.) dictates to what extent the image is sharpened.  I don’t do much printing of my shots (at least currently), so the above steps prepare the images for web posting.  If I were to print shots, I would probably load the modified TIF version of the image and apply the extra sharpening needed for printing and then save that in JPG with a new name identifying it as the print version with extra sharpening.  I would leave the modified TIF the way it is and not write over it.

This may not be the best way to process RAW photos but it is the way the process has evolved for me over the years.  For every photographer, there is a different RAW workflow.  So I hope if you discover a trick or setting that really works well for you, that you please share it with me and others here on THE BLOG.  One thing I’m not doing presently is using layers in Photoshop in my routine workflow—I have tried it and was not satisfied with the results I got.  This may well be due to the methods I used.  This is one area I am open to any tips on from others to help me make layers work for me in the selecting, sharpening and NR phase of my process.

By the way, I have my camera set to take photos in color space Adobe RGB(1998), not sRGB.  Adobe RGB has a wider color gamut and I think a lot of pros use it.  If you set your camera to this color space, your images will be numbered in this format: “_MG_”  You need to set up your RAW editor and Photoshop so that they know what color work space you want it to use to display your pictures.  Make sure your camera, your RAW editor, and Photoshop are all set to the same color space regardless of which space you choose to use.  Color management is another issue that is complicated and I know just enough to be dangerous. 

Oh, and one other thing you can spend more money on–— a screen calibrator.   Have you calibrated your monitor?  If not, you could be making the wrong adjustments to your photos so they don’t look right when printed or displayed on an adjusted monitor.  I use Eye-one Match 3 software and colorimeter to do this.  There is also a Spyder brand that is pretty good.  I paid about $200 for the software and colorimeter.  This is expensive but it is important if you want to ensure your brightness and colors are correct. 

Thanks for taking the time to read through this.  I hope it gave you something to compare to your own workflow and maybe even some new ideas to try on your own photos.  See you next time on THE BLOG!

This entry was posted on Tuesday, November 16th, 2010 at 1:52 am and is filed under The Blog. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.

4 Responses to “My Raw Workflow – in detail”

  1. Dennis,

    Thanks for sharing your “Raw Workflow”. It’s well done and clearly took a lot of work. I appreciate it.

    I have some comments and questions:

    I also work entirely in Raw. All my images are taken in “Adobe Raw”. My cameras allow the selection of Pentax Raw or Adobe Raw. I have selected Adobe Raw because I feel that it is more generic and will survive the years as a more universal format. Additionally, I find it easy to work with.

    I have “Lightroom 2” and have attempted to use it repeatedly. However, since all of my images are final processed in “Photoshop CS2”, I have found it much more convenient to use “Breeze Browser” for previewing and downloading raw images. Although it is designed to work with Canon, it works great with Pentax and Adobe. “Breeze Browser” uses very few computer resources and I can have Photoshop running at the same time without bogging down. Whereas, whenever I used Photoshop and Lightroom at the same time, the computer virtually locked up with much Photoshop processing.

    Question: I see you save in TIF format. I save everything as a PSD. Does TIF offer some advantage? Otherwise, to me, it is just an extra step. For every image I end up with a Raw, PSD and JPG file.

    You mention cropping and pixilation as a concern in Photoshop. I do all my cropping and enlargement with the Photoshop plug-in, “Genuine Fractals”. This is great piece of software that virtually eliminates pixilation. I have gotten some very nice, large prints from rather small image files using this software.

    I, like you, do my sharpening as the very last step. However, I use Nik Sharpener plug-in as my primary tool. I find it to be much superior to the Adobe tools. I use it on a layer, usually with a layer mask. I am going to try your suggestion of sharpening in “Lab Color”. I have heard this before, but never tried it.

    Every single thing I do in Photoshop, I do as a separate layer. Using layers allows me to always know what was done in processing and correct or adjust later.

    If I have any extraordinary noise, instead of “Neat Image” I use “Noise Ninja”.

    I also understand that Adobe RGB has a wider color range and is better for printing. However, since almost all my images are shown either on the Internet or in digital slide shows, I have my cameras set for sRGB. As I see it, having to convert is just another unnecessary step in the process. I have printed from both files types and my Epson printer does well with both.

    I use a Spyder monitor calibrator.

    As I have reported with my “Bull Moose” image on Flickr, it looks like my computer hard-drive is dead. I am currently using Adobe CS2 but anticipate upgrading to CS5 or beyond. My current computer is a dual-drive gaming computer. Clearly, I have to replace it. As you are running CS5, I would appreciate any recommendations for a replacement PC with necessary functionality. Budget: $2000 max. without monitor.


    Wily Fawlet
    http://www.flickr.com/photos/[email protected]/

    • Dennis says:

      Hi George, thanks for kicking off the raw workflow discussion. Post processing is such an important part of creating good images.

      Yeah, I agree with you that Adobe RGB is the color space to go with rather than Pentax RGB for the reasons you state.

      You import your raw photos into Breeze Browser(BB) (which I know nothing about!) and not LR2. I assume BB doesn’t let you do a lot of raw editing?? If that is the case, you must be doing basically all your editing in Photoshop (PS). I see no problem with this setup since photoshop can do everything LR does (and more obviously). So your RAW images are left as they came out of the camera, then you send the converted image to PS and save the edited images as PSD and JPG files. I have used PSD files in the past and do not notice any difference (from a working standpoint) between PSD and TIF. Both save layers, channels, and take up about the same space on your hard drive. One might believe that TIF is more universal than PSD but if you save a layered tif image, even programs that say they open tif images can have trouble opening specialized layered tif images saved in PS. So using PSD is probably no more risky than using tif.

      Regarding your computer lockup (and your question regarding computers at the end of your comment), that is a bummer that you have trouble running both programs simultaneously. I had similar trouble with my previous pc. It would take LR2 30 seconds to load an image sometimes–and it drove me crazy. Also, LR had to be re-started about every 20 minutes due to my limited 2GB of memory. I ended up getting an Intel i7-920 2.67GHz processor, Win 7 64-bit, and 12 GB of DDR3 memory back in February of this year. Now, both PS and LR2 run fine together. When I upgraded from the 10MP 40D to the 18MP 7D, I did notice a slight delay in LR’s ability to load the bigger images, but they still load within a few seconds. [The 7D raw image can be over 30MB.] Also, do some research on graphics cards–be sure you get a good enough one that can handle CS5’s features. Check Adobe’s site to determine which models work best for your needs.

      You mention that I had a concern with cropping and pixilation in Photoshop. In item #4 of my workflow, I was speaking of cropping images and adjusting exposure, etc., in Lightroom. This of course, is a universal problem no matter what program you crop and adjust in. I have heard that Genuine Fractals is a good program for enlarging images. At this point, I haven’t had much need to enlarge but I have done some of that in PS when I send a shot into a photo contest. At some point in the future, Fractals could be a great addition to my tools.

      Regarding Nik Sharpener, I haven’t used it. That is certainly something to look into. Anything that can make sharpening better is something to research. Can you be more specific as to how it is superior to PS sharpening tools such as USM? Thanks George!

      Doing everything in layers in PS—this is really a smart way to go! I would do it too if I could make layers work for me when I do my NR and sharpening. I obviously need more training at that because it’s obvious it can be done. Generally speaking, I have no problem with layers but the times I did try sharpening with them, I ended up with ugly artifacts around the mask edge, even when I had feathering turned all the way off. I quickly reverted to my old way of sharpening by inverting selections! Maybe you could give me (and readers) a step by step lesson of the way you use layers for the sharpening phase. I would really prefer using layers if I could make it look okay!

      I have not used Noise Ninja but I know folks who do…I think both programs do a comparable job of reducing noise–it’s just a matter of choice here (IMO).

      Regarding the use of Adobe RGB vs. sRGB for printing–we’re in agreement. The color differences are probably undetectable to the naked eye. Which one a person uses is really just a personal preference. I don’t mind the extra step as I am so used to it. It’s tough to teach on old dog (me) new tricks!

      Again, it’s personal preference on the calibrator one chooses to use. Both Spyder and Eye One do a comparable job in my opinion.

      Sorry about your dead hard drive George! Refer to my recommendations above for power in a PC enough to run almost any programs simultaneously. I got the setup I described for less than $1500 through Costco so this (or even something better) is well within your budget –with enough left over to upgrade to CS5! 🙂 I really try to buy my PC’s through Costco due to their 90 day return policy. Prior to getting my current HP machine there, I tried one of their ZT Systems PC’s. The ZT had a big bang for the buck but the computer refused to boot most of the time and I took it back and got a refund. That’s when I ordered the HP which I’m quite satisfied with. I used Costco’s HP customization page to get the features I wanted on it. They also have great support services. (I don’t have stock in Costco–I just like their policy on returns and support!). 🙂

      Also, since we’re talking computers I thought I might mention if you are getting a new PC and upgrading to Win 7and have any old programs that may not run correctly in Win 7, consider upgrading the OS to Win 7 Professional (from Home Premium) and get the ability to download Windows XP Professional Virtual Machine for no extra cost. This has worked amazingly well for me. Without this, I would not have been able to load my Microtek scanner software and would probably have had to replace the scanner (a good $500). Microtek did not have a Win 7 driver that would work correctly. I’ve got several other old programs loaded in Win XP that also will not work in Win 7. However, in order to use Windows XP Professional Virtual Machine, you must have a processor that supports hardware virtualization, and it requires an extra GB of memory. This link explains more about it: http://www.microsoft.com/windows/virtual-pc/support/faq.aspx

  2. Im obliged for the blog post.Thanks Again. Great.

Leave a Reply