Do Pentax DSLRs (K5) smear of poorly render reds?

I have started a new thread here to discuss the technical reasons for saturated red in flower photos.
http://forums.dpreview.com/forums/read.asp?forum=1036&message=38349987

paulkienitz said:

You saturated the red channel. Pentax's display tells you when you clipped the overall color to white, but not when you have clipped just one primary color, or two. Just lower the exposure a bit. It may also help to use the highlight extension feature (which I wish my old k10d had).

Click to expand...

No, you can actually see the individual RGB color channels on the camera's LCD at the time of capture. Go to Image Review and press the Info button until you get the histogram. Then press the down/flash button to see the individual color channel histograms. In this way, you can see if you have clipped one of the colors. Press the down/flash button again to return to the main histogram.

Rob

paulkienitz said:

You saturated the red channel.

Click to expand...

Yup, I always underexpose a little for red flowers, or pink ones!

They need some post-processing after to revive the colors, but that way, detail is maintained.

paulkienitz said:

Pentax's display tells you when you clipped the overall color to white, but not when you have clipped just one primary color, or two. Just lower the exposure a bit. It may also help to use the highlight extension feature (which I wish my old k10d had).

Click to expand...

Nah, just press the info button while reviewing, and you get all 4 histograms!

Here is a discussion about the K5 and red and to some extent Pentax and red.

http://forum.getdpi.com/forum/showthread.php?t=21931

Personally I'd rather have much weaker magenta & blue presence in the photos I take than in the samples I have seen from the latest Pentax models. Real sky is mostly just cyan, very rarely deep blue..
--
Ask not what your camera can do for you, but what you can do for your camera
http://www.fluidr.com/photos/gs-photos

I think it was the OP who commented that the colours - especially red- seemed to get more vibrant at higher ISOs. In this weeks AP mag, in the sensor comparison (K5/d5100/A55) they comment that as the ISO gets higher Pentax appear to enhance the colour more to make it vibrant, and suggest it is possibly due to the way they use noise reduction.

As a side issue, I'm no expert, but I vaguly thought red was the hardest colour for the Bayer sensor to cope with. I read an article recently but cant find it now to post it.

woody
--

After 10 years of picture taking - you'd have thought I'd have learnt to remove the lens cap...

ukwoody said:

I think it was the OP who commented that the colours - especially red- seemed to get more vibrant at higher ISOs. In this weeks AP mag, in the sensor comparison (K5/d5100/A55) they comment that as the ISO gets higher Pentax appear to enhance the colour more to make it vibrant, and suggest it is possibly due to the way they use noise reduction.

Click to expand...

Woody, as you surmise the more saturated red colour with higher ISO, if it exists, is simply a matter of colour processing in the conversion from sensor raw data to the output JPEG by the camera, an may include the Noise Reduction (NR) processing. However, the NR effect should be able to be tested in that Pentax DSLR's offer a JPEG Noise Reduction Off setting that should eliminate this effect if that is what is doing it.

ukwoody said:

As a side issue, I'm no expert, but I vaguly thought red was the hardest colour for the Bayer sensor to cope with. I read an article recently but cant find it now to post it.

Click to expand...

Actually, both red and blue are the hardest to render with high resolution as there are half as many of these colour photosite detectors in the Bayer Colour Filter Array (CFA) as compared to green; however that does not affect the saturation or clipping as seems to have become the subject of this thread. It does affect the perception of smearing, especially when NR is applied since the resolution may be lower in the first place.

Channel clipping is both a matter of conversion of colour space in the case where the raw sensor data was not clipped and, where the raw data especially for the red and blue channels are clipped, due to most cameras exposure metering systems detecting only luminance signals (which is close to equivalent to the green) and not therefore detecting strong reds and blues. The red channel in such cases is often not actually clipped in the raw data for typical daylight illumination but appear to be clipped after application of white balancing factors to make the whites appear as they should.

In rarer cases it may also come about that the raw sensor channel data is actually clipped in very strong off daylight illumination as in fire lit scenes, such as would also tend to make the reds stronger. This is due to the red channel actually then being of a larger amplitude than the green channel in those cases.

Regards, GordonBGood