DaveC1 Loc: South East US

AzPicLady Loc: Behind the camera!

I'm not a whiz at PP work, so I try to do things in-camera. When shooting bright red flowers, I've learned to take the exposure down -1, sometimes less, sometimes more. It depends on how reflective the flower is. Then, if necessary, I can brighten it back up in post. Even then, sometimes the red prints as a blob with no detail.

DaveC1 Loc: South East US

I shoot in raw, expose for the green channel, and correct that in PP. Seems to work out with good results. Oh Nikon D300.

As always download to get a good view.

(Download)

selmslie Loc: Fernandina Beach, FL, USA

The exposure indicates that this was probably taken in full shade.

Green looks good but the red (pink) is at 255. Red channel may be blown but if you have the raw file you might recover that highlight.

In a normal scene the green channel is usually brighter. But in this case you should probably be exposing for the red channel rather than the green channel.

rmalarz Loc: Tempe, Arizona

John, As you are aware, and probably has been mentioned here, and other threads on UHH, digital cameras seem exceptionally sensitive to the colour red. I've been working on some processing techniques that helped slightly in this case. I was able to bring out some details in the rose petals. The largest obstacle is working with a .jpg image as the "original". What I've done here might have worked better starting with the RAW capture.

Another is color saturation. Although the illumination may have been suitable, the color red is saturated to almost 100%. Thus, leaving very little opportunity to bring out the subtle nuances of shades of red within each petal.

You might be able to help yourself by taking several photos of the same rose over a period of time, such as early morning, afternoon, early evening. That's just to see if the time of day and illumination may help. It would be far better to find a solution that works regardless of time of day just for the convenience of being able to take a photograph when you find a subject, as opposed to finding at a particular time.

Your posting this and your question has inspired me to do some research, and possible lab work to find an easy solution to "The Red Issue" with digital cameras. If I discover some magic elixir, I'll post it as a reply to this thread.
--Bob

(Download)

DaveC1 Loc: South East US

There's one spot in the red channel @255 in the area of x=7.6", y=4.2" (21.4% enlargement) but in looking at the enlarged image I don't see any loss of detail.

As far as the green channel exposure goes, here is an explanation of what I'm doing.

http://www.guillermoluijk.com/tutorial/uniwb/index_en.htm

So I really doubt I blew the red channel out in exposure.

DaveC1 Loc: South East US

Here's the original raw image converted to jpg without any pp.

(Download)

selmslie Loc: Fernandina Beach, FL, USA

No, the rest of the red is well below 255. But if you pixel-peep and find small areas that are uniformly at 255 in the red channel, either the red channel is blown out or you did not recover enough highlights.

I am familiar with Guillermo Lujik and his UniWB work. If you look in the last paragraph of the Introduction you will find, "A few days after publishing this article I knew that Iliah Borg, a regular participant in the Nikon DPreview forums, had already investigated the way to cancel the white balance on Nikon cameras." So Lujik did not originate the theory or even the name UniWB.

What's more interesting is that Iliah Borg is the developer of RawDigger. That's where you should go to study the question about which channel, if any, is blown out. The camera's histograms, even the separate red, green and blue versions, are too small and imprecise.

selmslie Loc: Fernandina Beach, FL, USA

If the original raw data converted the brightest part of the red to 254, it suggests that the red channel is not blown and that the 255 value we see in the original post was simply that the highlights were not completely recovered during PP.

You have the raw file and you can download RawDigger for a 1-month free trial and see for yourself. If there is no spike at the right end of the red histogram then you should be OK.

If this has peaked your curiosity about raw histograms, I bet you will happily fork over the $20 for a permanent copy of RawDigger.

Borg is probably no longer interested in UniWB.

DaveC1 Loc: South East US

Yeah, I already looked at the RawDigger website. I'll probably spend the $20 for what, at least for me, looks like a useful tool.

rmalarz Loc: Tempe, Arizona

John, not exactly an easy task. These aren't quite red, but more magenta. It can be done, but definitely a combination of exposure and processing, but more involved processing than I usually do. There are a lot of checks and balances along the way.

I'm quite glad you posted this topic, as it falls along the lines of some processing techniques with which I'm currently working.
--Bob

(Download)

DaveC1 Loc: South East US

Bob, in your estimation how close is this to the perceived daylight color of the original bougainvillea flowers (I assume they're bougainvilleas?)

rmalarz Loc: Tempe, Arizona

Pretty close. Probably not exact, but close.

Oh, and yes they are bougainvillea flowers.
--Bob

DaveC1 Loc: South East US

That strikes me as not being an easy hue/saturation to get to in the image of those blooms.

rmalarz Loc: Tempe, Arizona

Dave, it wasn't. It took a good deal of effort, probably more than it was worth, but it worked.
--Bob