Lightroom and sRGB clipping

bohdank wrote in post #10526767
Are we talking about color gamut or just plain blowing out the reds ?

I use the aRGB histogram for reviewing the exposure of my photos. On my monitor it obviously is all within gamut because my monitor is effectively 100% aRGB coverage. However, I try on images that are particularly bright and colorful to reduce it to sRGB before converting for web purposes. It's quite difficult to do on some images.

I don't think switching to a sRGB histogram would help particularly. All that will result in as others have noted is me exposing a bit more "to the left" than I am currently. I can easily reduce exposure in post. However, that results in too dark an image as I mentioned before. Really, for some of these pictures its either selecting saturation adjustment or bust. I just wish I could do it in LR. Really its only my macro work that usually needs some adjustment.

I understand what you are saying. It seems everytime I take a picture of a flower, I struggle with the blown channels but it's blown not a gamut issue, imo, or can be both.

bohdank wrote in post #10529336
I understand what you are saying. It seems everytime I take a picture of a flower, I struggle with the blown channels but it's blown not a gamut issue, imo, or can be both.

It can be both. A red that is 250/0/0 in Adobe RGB is OOG for sRGB so when you convert it becomes 255/0/0. The pixel next to it that is 245/0/0 also is OOG for sRGB and also becomes 255/0/0 and there goes your detail. The colors that are outside sRGB are the highly saturated ones and these are often the ones where the primary channel has a high luminosity value which will blow when you try to pack it into the smaller space.

tzalman wrote in post #10529921
It can be both. A red that is 250/0/0 in Adobe RGB is OOG for sRGB so when you convert it becomes 255/0/0. The pixel next to it that is 245/0/0 also is OOG for sRGB and also becomes 255/0/0 and there goes your detail. The colors that are outside sRGB are the highly saturated ones and these are often the ones where the primary channel has a high luminosity value which will blow when you try to pack it into the smaller space.

Which is why I'm a big proponent that companies need to get off their rears and start accepting that newer display technology is among us and we should have OS's that are entirely color managed. Same thing for TVs. The new Sharp Aquos expands its color gamut by adding a yellow pixel. All this does is the same thing it does for sRGB images on a aRGB display without color management.

It seems to me that aRGB monitors are prevalent enough that pretty soon we should start expecting people to have a aRGB display. Of course there is still the printing issue. Not sure what the limitations are there but hopefully we start getting better there as well. Many print houses fall well short of even sRGB.

Yeh, what my TV did to the Holland uniforms was really ugly.

I would like to add two questions to the discusion:

1. Out of gamut clipping is not only a matter of looking for peaks on the right side of the histogram. If a colour channel after the profile conversion becomes negative (table profile conversions are just 3x3 matrix operations that can yield negative colour values), it will be clipped to 0. So out of gammut is not only a matter of not having clipped values in the right end of the histogram, but also on the left. So avoiding gamut clipping is not only a matter of setting a strong exposure adjustment down, there is more to it.

2. It is quite uncommon that a RAW file, after a totally neutral RAW development (i.e., no increased saturation was set, no contrast curves, no nothing) falls out of sRGB, and is very difficult to fall out of Adobe RGB (just try to develop your RAW files setting all parameters to 0 in ACR). Most out of gamut issues are created by the user who, in the image processing stage (and this includes RAW development settings), creates new colours (more saturated, more strange, more whatever) that may fall out of the typical standard profiles.

Regards

Sdiver2489 wrote in post #10528703
I don't think switching to a sRGB histogram would help particularly. All that will result in as others have noted is me exposing a bit more "to the left" than I am currently. I can easily reduce exposure in post. However, that results in too dark an image as I mentioned before. Really, for some of these pictures its either selecting saturation adjustment or bust.

It's not about exposure, it's about gamut.
If it were about exposure, then everything would be equally affected.
If you're going to jog a slider, the "saturation" slider would be a better bet then "exposure". The HSL panel would be even better.

Sdiver2489 wrote in post #10528703
I just wish I could do it in LR.

Yep.

Sdiver2489 wrote in post #10530029
Which is why I'm a big proponent that companies need to get off their rears and start accepting that newer display technology is among us

Among us, yes. But as long as the majority of the world still browses non-color managed and uncalibrated, who cares?
We photographers are a small minority on the web.

Sdiver2489 wrote in post #10530029
and we should have OS's that are entirely color managed.

Would be nice, and it seems we are slowly getting there.

Sdiver2489 wrote in post #10530029
Same thing for TVs. The new Sharp Aquos expands its color gamut by adding a yellow pixel. All this does is the same thing it does for sRGB images on a aRGB display without color management.

Not sure what you mean here: The Sharp tv supposedly has an expanded gamut (what I gather from the ad). So you'll be able to see more saturated yellows.
Yet colors might still suck, like on all TV's. Nothing new there?

Sdiver2489 wrote in post #10530029
It seems to me that aRGB monitors are prevalent enough that pretty soon we should start expecting people to have a aRGB display.

Heck no.
Most people still use a crappy TN display. If they are using an LCD that is.

Sdiver2489 wrote in post #10530029
Of course there is still the printing issue. Not sure what the limitations are there but hopefully we start getting better there as well. Many print houses fall well short of even sRGB.

Nature of the beast.
You should try having something printed on a (CMYK) printing press: even way less gamut.

You might find " Making Within the Stone" here interesting

tzalman wrote in post #10530044
Yeh, what my TV did to the Holland uniforms was really ugly.

Probably because those shirts are ugly

_GUI_ wrote in post #10535859
So out of gammut is not only a matter of not having clipped values in the right end of the histogram, but also on the left. So avoiding gamut clipping is not only a matter of setting a strong exposure adjustment down, there is more to it.

Obviously. Clipping to 255 is just as bad as clipping to 0

_GUI_ wrote in post #10535859
2. It is quite uncommon that a RAW file, after a totally neutral RAW development (i.e., no increased saturation was set, no contrast curves, no nothing) falls out of sRGB, and is very difficult to fall out of Adobe RGB

Might be. So?
I'm aiming for a nice image, not a perfect histogram.

_GUI_ wrote in post #10535859
(just try to develop your RAW files setting all parameters to 0 in ACR)

Default for ACR = 25 for contrast and 50 for brightness.
I know I can avoid clipping altogether in ACR. But I want the image to look like I remember it. Like it was if my memory serves me right.

Might be that some colors are out of sRGB gamut. If that is so, then so be it.
If you're shooting nature bright flowers, it's not uncommon to go beyond sRGB gamut. Even without extreme editing.

Interesting discussion

René Damkot wrote in post #10549671
Might be. So?
I'm aiming for a nice image, not a perfect histogram.

Of course, I agree. Just wanted to make clear that it is not usually the scenes in nature that produce out of gamut colours by themselves, but later processing to make them look pleasant to our eyes.

René Damkot wrote in post #10549671
Default for ACR = 25 for contrast and 50 for brightness.

What ACR designers consider default is irrelevant. 25 contrast and 50 brightness mean applying these curves (I reverse-engineered them) to the linear data, and that is post processing no matter if we call it default:

Bright:

René Damkot wrote in post #10549671
If you're shooting nature bright flowers, it's not uncommon to go beyond sRGB gamut. Even without extreme editing.

Correct, that is why I said it is quite uncommon. We must also bear in mind that the clipping thay may occur in the conversion from camera's space (let's call it that way) to for instance sRGB, is not unique. Every RAW developer will (and should, unless they share information) use different matrices. Some will be more accurate, some less. Some will tend to clip more, some less.

Regards