Stuart Bowie was kind enough to send me his RAW file of the White-Faced Duck he posted yesterday. The whites in that image were blown, and a good discussion ensued about how to properly expose black and white birds. I asked Stuart to send me the image so that I could do some RAW processing of my own to try to recover the whites. What you see above is the end result of that Lightroom processing with a little CS4 thrown in the mix. I was able to recover the whites entirely using Lightroom; the only thing I did in PS was some NR, sharpening, and a little eye and BG work. No linear burns to pull anything out of the whites.
I advised yesterday that I overexpose the bird by 1/3 to 2/3 of a stop to get more details out of the blacks and then pull the whites back in Lightroom. The problem with Stuart's image was that the whites were overexposed by close to 2 stops. Nonetheless I was able to get the whites looking good using a combination of the exposure slider, the highlight tone curve, and the exposure brush.
Nice work on recovering the whites! I can clearly see detail there now.
There has been a little shift in the color balance of the image, and I prefer the slightly warmer colors in the original, esp. the background and the rusty/browns on the bird, and the original eye.
I think this has been a very helpful thread, and will bookmark it for future reference. Thanks for the thoughtful rework.
Very interesting to see the different versions. I've never seen this species and am not sure what the colors should look like but agree that there is a color shift.
With the exception of the whites the colours in Stuarts original are what the bird looks like in real life.
The reposts may sort out the whites but do nothing for retaining the true colours of the rest of the bird.
Doug, thank you so much for your time on reposting and fixing the blown whites on my original post. You have done a great job on the whites on the face. With regards to the neck, I feel my colour is correct, and apart from my previous post of one in flight, and Mark Dumbleton ( page 2 ), you will notice if you take out the reddish cast, you will get to the right colour. I realise, you not knowing the species, its difficult to know which way to go. Thanks again Doug, and Fabs, and each of you,for your input. Thats why this is such a great site.
Here's one last crack at it. As Stuart said, I didn't know what these birds looked like when I processed the image. This was more of an exercise in extreme (2 stops) highlight recovery. I hope this is more faithful to what the bird looks like in real life.
Great works and comments by all above and certainly nothing offensive.
One important note for Doug. Tehcnically speaking (a rarity for me), the whites on the original could not have "been blown by two stops." Had they been truly overexposed, my belief is that they are not re-covereable as you cannot recover detail that is not there. What you did and did well, was recover detail that was present in the RAW file.
BIRDS AS ART Blog: great info and lessons, lots of images with our legendary BAA educational Captions; we will not sell you junk. 30+ years of long lens experience/e-mail with gear questions.
BIRDS AS ART Online Store: we will not sell you junk. 35 years of long lens experience. Please e-mail with gear questions.
Check out the new SONY e-Guide and videos that I did with Patrick Sparkman here. Ten percent discount for BPN members,
Thanks Artie. Perhaps I should clarify. I said that the original was overexposed by about 2 stops; that was based on the histogram displayed in Lightroom and the adjustments required to tame the whites. As you know (but maybe some forum members don't), the histogram that we see on our cameras is derived from the JPG thumbnail, not the RAW file. The RAW file has many more brightness levels on the right side of the histogram than does a JPG. It's all those extra brightness levels that allow us to recover detail from seemingly blown whites. That's why I asked Stuart to send me the RAW file.
Not to bust your _ _ _ _ s, but here is what you stated (referring to the RAW file):
"The problem with Stuart's image was that the whites were overexposed by close to 2 stops."
re:
Thanks Artie. Perhaps I should clarify.
YAW, but you cannot clarify an statement that is incorrect :)
I said that the original was overexposed by about 2 stops;
Again, that statement as I explained above is incorrect.
... that was based on the histogram displayed in Lightroom and the adjustments required to tame the whites. As you know (but maybe some forum members don't), the histogram that we see on our cameras is derived from the JPG thumbnail, not the RAW file.
If you read my post above carefully I did state exactly that.
The RAW file has many more brightness levels on the right side of the histogram than does a JPG. It's all those extra brightness levels that allow us to recover detail from seemingly blown whites. That's why I asked Stuart to send me the RAW file.
As above, your repost showed folks what we taught them in the post in the How White the Whites post in ER. And while it is rare that I know much about anything technical (and this has not a been a high level technical discussion) I am good with language and do not let folks slide with inaccurate statememnts.
Please do not take it personally as I appreciate the great work that you do here.
BIRDS AS ART Blog: great info and lessons, lots of images with our legendary BAA educational Captions; we will not sell you junk. 30+ years of long lens experience/e-mail with gear questions.
BIRDS AS ART Online Store: we will not sell you junk. 35 years of long lens experience. Please e-mail with gear questions.
Check out the new SONY e-Guide and videos that I did with Patrick Sparkman here. Ten percent discount for BPN members,
It got me thinking about what "blown" whites mean. Any given area of white tone in an image could have some pixels with luminosity = 255, i.e., pure white. First question- are the whites blown in this case and should we be always striving for exposures that produce all white tone pixels will luminosity < 255? If these blown pixels are distributed in very small patches about the area, with darker pixels in between, I think most people would not consider the area blown because the darker pixels would provide what we call "detail". If the pure white pixels are clumped to a larger degree, or there were more of them, you might consider the area blown because whole regions of the image would have luminosity = 255.
Incidentally, I did an experiment to see what the Recover slider does in Adobe Camera Raw. I took the same image through ACR, one unrecovered, one recovered by 13% and created layers with each. I then selected a 400 pixel oblong area of the brightest part of the image (the white throat of a harshly-lit Tree Swallow) and copied the same area from the unrecovered and recovered layers to two other layers, then had a look at each with the Histogram in Ps CS3. Both unrecovered and recovered white areas had 80 pixels with luminosity = 255, i.e., pure white, and signified by a vertical line at the far right of the Histogram. So the Recovery slider in ACR does nothing to pure white pixels. Then there was a gap in the histogram with the rest of the white-tone pixels in each distributed with a pointy peak. The difference between the two was simply in the size of the gap in the histogram. The recovered area had a larger gap with the white-toned pixels pushed more to the left than in the unrecovered area.
Not sure if the Lightroom tool that Doug mentioned behaves in the same way.
In some cases--heck in many cases if folks do not know what they are doing with a histogram, truly blown overexposed whites (at 255 by definition) cannot be saved. In the image posted here that appeared over-exposed, they most obviously could be saved. So it is not the behaviour of Lightroom or any other recovery efforts that matter, it is a matter of whether the detail was there to begin with....
As for me, the only thing I want at 255 in my frontlit images are specular highlights.
BIRDS AS ART Blog: great info and lessons, lots of images with our legendary BAA educational Captions; we will not sell you junk. 30+ years of long lens experience/e-mail with gear questions.
BIRDS AS ART Online Store: we will not sell you junk. 35 years of long lens experience. Please e-mail with gear questions.
Check out the new SONY e-Guide and videos that I did with Patrick Sparkman here. Ten percent discount for BPN members,
Agree Artie. I just thought it was interesting to note that the Recovery algorithm in ACR does not touch luminosity = 255 whites so it's not doing anything magical to those, only the <255 surrounding whites. The developers could have decided to actually subtract some luminosity from 255 whites and offset them like they do for <255 whites in Recovery, but they didn't.