Digital Sensors and Red Autumn Leaves...

Okay, so I have long noticed that the red in new england autumn landscapes doesn't seem as vibrant in the digital sensor as it does on film. I know that Velvia did it's own thing, but it always seems washed out in digital.

Earlier this week I was reading about how different sensors are more sensitive to different colors and why the 40D's RGB histogram is useful compared to the regular histogram.

So my technical questions...
(1) what is the 'regular histogram' measuring the tones of (grey I assume)

(2)Am I crazy or does red seem to blow out in digital faster than other colors

(3) Before it's too late, how do I best expose for autumn scenes without cranking the saturation slider?!?!?

Thanks,
Jim

Yes, I have to be very careful not to blow out the red channel on New England foliage with my D2x. However, it has an RGB histogram that shows me a blown red channel. I assume the regular histogram is showing luminosity (gray) so that if any channel is not blown it won't register any problem.

As for how to expose, since you don't have an RGB histogram all I can think is to bracket and take a shot a bit underexposed.

Also, I found my reds were a lot truer after I added calibration for my D2x to the ACR calibration tab. If you use any version of ACR read this.

Hello Jim,

I haven't read up on the current state of digital sensors recently, however in the past when I did, it was obvious there are many design elements and materials involved such that one could not really make many general statements about such sensors. What an engineer implemented in one design may not apply to another. And there are both hardware and software aspects of such designs involved. The technical issues around spectral response and sensitivity to materials are rather bewildering for anyone without specific advanced color theory education. It is not only the sensor responses but also the color materials used as filters.

One can look at consumer technical reviewes of the various digital cameras and see how reviewers regularly describe the various tested cameras as all somewhat different in how they display color. Currently I own a Nikon Coolpix 7900 compact digital camera and it's color response is quite different than the two previous such cameras I owned. I have to set the contrast control to the lowest minus setting to get close to natural color and even then need to manually reduce the red response later in Photoshop about -15% and shift cyan-blues about 5 degrees toward blue. Of course the marketers of such cameras know most customers prefer vivid color and especially red so this is what they are giving the public. Not something I personally appreciate. The other issue is unless one buys an X-Rite colorimeter or spectrophometer, the pricy Eye-One commercial software, the big Gretag-McBeth color chart for digital cameras, and figure out how to calibrate one's camera like commercial product photographers do, obtaining accurate color is crude, at best limited to adjusting white balance and eye-balling the results say on an monitor while tweaking the many controls on todays D-SLRs. Camera manufacturers thus far have had no pressure to provided accurate color so don't. That's one of the main advantages to using EPN100, Provia, or Astia film if properly exposed on a not too contrasty subject. What you get already is quite accurate. If someone gave me one of the new D-SLRs for free, I would likely buy a cheap point and shoot 35mm and fill it with one of those films. Then whenever I took a serious image with the D-SLR would take a reference pic with the P&S. ...David

Hi Jim
I used spot metering on my 30D and under exposed by 1/3 stop & still I felt the exposure could be better tuned to get the exact Red.
Have you used Sekonic light meter to get the exact exposure? If not, may be it's worth a try.

I've noticed on my 20D (yeah, a different camera) that the red channel blows out extrememly easily. If there is a lot of red *and* I remember, I'll underexpose by 1/3 or sometimes even 2/3's of a stop and deal with the other two channels in Photoshop.

I don't think a light meter really is the answer here. From what I can see in the histograms and from basic exposures, the sensor in my camera is more sensitive to reds. This is something a light meter simply wouldn't pick up on and you would have to manually adjust for anyway.

We don't get the New England fall colors where I live, so it isn't a problem I've ever noticed.

Canon 1D cameras have picture styles that can be selected at the users option. Like different types of film, they give a different look, sometimes quite subtle, to the picture. Like Velvia compared to Astia compared to Kodachrome. If you shoot in RAW, these can be changed after the fact.

The 1D Mark III also allows one to choose between a 'brightness' histogram and an RGB histogram.

When digital cameras first started becoming popular, I wondered why there was no discussion regarding the look or color or saturation among the sensors of digital brands of cameras the way there was about different brands or types of film.

Jim:

I run into this problem here in Moab all year 'round because the sandstone at sunrise/sunset is so vibrantly red. I haven't found a real useful solution to the issue, either. Often the only way I am able to control it is through a blended exposure.

I'll be interested to hear experiences and thoughts from others on the forum.

Just wondering which color space you are using. I seem to find better reds with adobe RGB than with sRGB.

Carson,

I've used both color spaces, although my camera is set to Adobe RGB. For me, it is an issue with both, when enough red is in the scene.

Have you tried setting your WB manually, Jim?