A rose by any other name...oversaturated in 60D!

Drop both versions into Photoshop and mask as necessary.

Why are you using ISO 400 in the day for your shots?

Take Care,
Cheers, Patrick

I don't think it makes any difference whether I used ISO100 or 400. I have never seen any...

Well it does make a difference, that is why your reds are over saturated and blown out. Any time you over expose a red shot it will look like that.

Take Care,
Cheers, Patrick

The original shot isn't generally overexposed and ISO has nothing to do with it anyway.

In common with virtually everyone else, I use the full range of ISOs at any time of day or year, depending of course on what aperture and shutter speed I need.

Your correlation of higher (but only two stops higher) than minimum ISO with overexposure baffles me.

I had the same issue last week shooting the roses in my wife's garden. I ended up taking a series of shots, lowering the exposure compensation by 1/3 for each shot. I had no blown reds at -1 2/3 with "Picture style" set to "Faithful."

IMAGE LINK: http://www.flickr.com …scubadude1960/8​696376854/

ScubaDude wrote in post #15909922
I had the same issue last week shooting the roses in my wife's garden. I ended up taking a series of shots, lowering the exposure compensation by 1/3 for each shot. I had no blown reds at -1 2/3 with "Picture style" set to "Faithful."

IMAGE LINK: http://www.flickr.com …scubadude1960/8​696376854/
5469 by ScubaDude1960, on Flickr

Thank you. This can probably take care for shooting in jpg. How was the color/brightness for the non-red part of the photo? Did underexposure by 1 1/3 affect it drasticaly?

I was mainly looking for any remedies for RAW shooting, I will have to try underexposing this much (or whatever it takes looking at the histogram).

scubadude's shot seems to be in diffuse light, not direct sun, which would make the light cooler.

cooler light = more cyan

since cyan is the opposite of red, the reds would be pulled back to a more normal level.

i'm not saying that underexposing won't help, but imo, since you are shooting RAW, you're better off with a red channel adjustment in post.

also, depending on metering modes, and the spot being metered, the amount of EC would differ.

cheers!

MakisM1 wrote in post #15910139
Thank you. This can probably take care for shooting in jpg. How was the color/brightness for the non-red part of the photo? Did underexposure by 1 1/3 affect it drasticaly?

I was mainly looking for any remedies for RAW shooting, I will have to try underexposing this much (or whatever it takes looking at the histogram).

It was shot in RAW and converted to TIFF in DPP (which does apply the Picture Style to the image). As you can see, noise is not an issue; a little n/r was applied in post. Also, I reduced some of the red saturation (in ACR's "HSL/Grayscale" tab) and boosted overall saturation on the main tab to balance the colors.

The costume and lighting designers for our dance company often use very saturated reds. I have learned to plan on dropping the reds in post. It's been that way on all 3 Canons I've used. I thought it was just because they were far from the cream of the crop. Now it seems like it may be a Canon thing.

ScubaDude wrote in post #15911117
It was shot in RAW and converted to TIFF in DPP (which does apply the Picture Style to the image). As you can see, noise is not an issue; a little n/r was applied in post. Also, I reduced some of the red saturation (in ACR's "HSL/Grayscale" tab) and boosted overall saturation on the main tab to balance the colors.

hes gone wrote in post #15910369
=he's gone;15910369]scubadud​e's shot seems to be in diffuse light, not direct sun, which would make the light cooler.

cooler light = more cyan

since cyan is the opposite of red, the reds would be pulled back to a more normal level.

i'm not saying that underexposing won't help, but imo, since you are shooting RAW, you're better off with a red channel adjustment in post.

also, depending on metering modes, and the spot being metered, the amount of EC would differ.

cheers!

graydragon1 wrote in post #15912843
The costume and lighting designers for our dance company often use very saturated reds. I have learned to plan on dropping the reds in post. It's been that way on all 3 Canons I've used. I thought it was just because they were far from the cream of the crop. Now it seems like it may be a Canon thing.

He's gone has it right and I have no problem correcting it to my exact taste/memory in post. My question was whether I could change something in my shooting to have a better photo coming out of RAW.

I believe the answer is spot meter the reds and check your histogram for the reds. I would then have less way to go in PP to get back to the photo as it was in reality.

I don't use DPP or anything else that would change the original file. All the PP is in a sidecar file and I have various choices for output files, so no problem there.

Thank you guys for your interest and your valuable advise.

I do not know if it will run under wine but downloading and looking at your RAW files in RAW Digger can be very helpful. RAW Digger will actually provide you with a full linear RGBG RAW histogram. I know that for both of my cameras, and I beleive it is actually the general situation is that the red channel sensels are actually less sensitive than the blue or green, so in the linear RAW space you are least likely to saturate the red channel. The issue comes when you apply the WB as for daylight at around 5200 - 5500 K the channel multipliers are around 1 for the green channel, 1.5 for the blue and 2 for the red, this is when the red channel is likely to hit it's limit, and also why it is normally possible to pull the information back out by desaturating the channel a bit.

Personally I would rather keep the exposure levels up so that I do not have to boost all of the channels in post, with the added noise that will be introduced and deal with reducing the red channel where I may have to.

Alan

Alan, Darktable has the same facility, and I can confirm that using the linear RGB profile for input and that the multipliers you are describing are roughly correct. The resultant image corresponds quite closely to my first desaturated photo. However, the extended color matrix profile seems to be the most faithful, (and is the default) so I will stick with it. In my mind, all that remains to be seen is what the results will be if I spot-meter for the reds, and whether I have to push the red histogram (in-camera).

In PP, I think that we are all aware how to deal with this issue.

Quick hijack: Glad you brought up Darktable. Hadn't heard of it, but now I'll be following it. And glad also that Alan mentioned Raw Digger and the multipliers.

If shooting by histogram with an eye for RAW (getting the data), you might want to change the default in-camera processing to neutral or faithful (or whatever it is). At least in my 5Dc and other cameras around that time, the histogram is generated (roughly) from the interpreted RAW file (the thumbnail, basically), not the RAW data itself. It interprets according to the jpg processing setting, so if the camera is set for vibrant colors, it will alter the reported histogram. However, the histogram (again, for the cameras around my 5Dc's time at least) is generated off only a low quality image. There might be many individual pixels (fine detail) with blown channels that may not show up in the in-camera histogram.

Alan's comment has me wondering if I should also be selecting a specific WB for best evaluating the RAW file in camera. And whether or not there is a role for colored filters here.

I've played around with a variation of this problem as well, though not very seriously. Blown color channels bother me, too (tonal flat spots), so I can appreciate the concern.

I think most (all?) Canon cameras suffers from the histogram being generated out of a jpeg conversion after having taken all the image settings into account.

So the histogram can seriously lie about clipping of the actual raw data.

This means that even raw shooters should make sure they have as neutral settings as possible in their camera even when not intending to capture to jpeg.