Ad
Ad
Ad
Pages: [1] 2 »   Bottom of Page
Print
Author Topic: What Does This Histogram Mean?  (Read 5316 times)
JohnKoerner
Guest
« on: August 20, 2011, 06:02:47 AM »
ReplyReply




Does this mean I have no shadow clipping, no highlight clipping, but DO have color clipping that is outside the Adobe RGB color space?

Forgive the simplicity of the question, but I am trying to better understand what the histogram is telling me.

Would this be an instance where, say, a "print" rendered in the ProPhoto color space would give me a substantially more "vivid" output, color-wise, than what I am seeing on my Adobe RGB screen?

Thanks for any help,

Jack

.
Logged
Peter_DL
Sr. Member
****
Offline Offline

Posts: 413


« Reply #1 on: August 20, 2011, 07:30:15 AM »
ReplyReply


[/center]

Does this mean I have no shadow clipping, no highlight clipping, but DO have color clipping that is outside the Adobe RGB color space?

With Lightroom’s histogram referring to this sRGB-TRCed ProPhoto-gamut Melissa space,
it seems to me hard to predict if saturation clipping would occur upon conversion to smaller Adobe RGB.

But then I’m not a Lightroom user, and the question may fit to the current discussion in the LR forum on its Clipping Indicators.

Peter

--
Logged
pegelli
Sr. Member
****
Offline Offline

Posts: 580



WWW
« Reply #2 on: August 20, 2011, 07:36:55 AM »
ReplyReply

Jack. Having nothing flow left or right from the histogram means that you have no channels clipping and nothing maxed out on saturation in the lightroom colour space. Beware though that the lightoom colour space is quite a bit bigger than sRGB and AdobeRGB, so if you export to a jpg or tiff you could experience clipping or oversaturation there, allthough on this picture I think that risk is low.
Having the middle of the histogram above the maximum has no other meaning than that you have a lot of pixels with those channel intensities. It does not indicate any saturation and/or clipping.
Logged

pieter, aka pegelli
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1277


WWW
« Reply #3 on: August 20, 2011, 07:40:38 AM »
ReplyReply

Does this mean I have no shadow clipping, no highlight clipping, but DO have color clipping that is outside the Adobe RGB color space?

Fun photo and a common question! While the left and right histogram "walls" represent black and white clipping, the "ceiling" does not represent clipping. The height of a histogram scales dynamically to be more visually informative and sometimes scales past this ceiling. This often happens in situations like this where your tonality is isolated to a fairly narrow range.

In short, your histogram is telling you that you've got lots of midtones, no black or white clipping, and, as always, it's not telling you anything about gamut or gamut clipping.

For now, you'll have to switch to Photoshop with a ProphotoRGB rendered version to analyze gamut clipping for your print spaces. Enjoy.
Logged

JohnKoerner
Guest
« Reply #4 on: August 20, 2011, 08:54:02 AM »
ReplyReply

Thank you for taking the time, fellas, I appreciate that.

Peter, thanks for the correction: I forgot, Lightroom deals in the Melissa Color Space, which is even larger than Adobe RGB;

Pieter, thank you for the information. I always export to .tiff in the ProPhoto Color Space from Lightroom Photoshop, so I hope there will be no clipping or oversaturation.

Scott, thank you for the info too (and glad you like the photo Grin ). Does the Photoshop histogram tell more a complete story about gamut than does Lightroom? Or is there some other means by which I should analyze the gamut in PS?

Jack


.
Logged
Guillermo Luijk
Sr. Member
****
Offline Offline

Posts: 1273



WWW
« Reply #5 on: August 20, 2011, 09:00:31 AM »
ReplyReply

What you see there is not any kind of clipping, is just that the Y-axis scale in which LR decided to show you the histogram doesn't reach its maximum, so it shows you a truncated version of the complete histogram. But it's just a matter of plotting, nothing related to the content of your image gamut.

The following image:




has the following histograms (left Photoshop, right complete histogram):




As can be seen, PS truncates the Y axis of the graph in order to make the plot more representative.

To analyze gamut clipping in the histogram, just look at histogram ends: in my picture, the Glencoe landscape in the Highlands was so intensely green that sRGB didn't manage to encode it, clipping the blue channel in 0.

For photographers who never heard about histograms, I think it's a good idea to look at the statistical concept of histogram, much wider that its particular application to photography.

Regards
« Last Edit: August 20, 2011, 09:04:56 AM by Guillermo Luijk » Logged

Scott Martin
Sr. Member
****
Offline Offline

Posts: 1277


WWW
« Reply #6 on: August 20, 2011, 09:24:06 AM »
ReplyReply

Scott, thank you for the info too (and glad you like the photo Grin ). Does the Photoshop histogram tell more a complete story about gamut than does Lightroom? Or is there some other means by which I should analyze the gamut in PS?

Histograms aren't helpful for gamut or gamut clipping info. Photoshop's soft proofing capabilities are designed for help analyze gamut clipping.  View>ProofSetup>Custom... and View>GamutWarning are the two tools worth getting to know if you're not familiar with them already. 
Logged

Guillermo Luijk
Sr. Member
****
Offline Offline

Posts: 1273



WWW
« Reply #7 on: August 20, 2011, 09:46:51 AM »
ReplyReply

Histograms aren't helpful for gamut or gamut clipping info.

Totally incorrect (just look at my example). Photoshop's soft proofing capabilities are only useful to find out if clipping occurs at converting to an output colour profile. For any other situation (increasing saturation for instance), the histogram is the best tool.
« Last Edit: August 20, 2011, 09:52:33 AM by Guillermo Luijk » Logged

bjanes
Sr. Member
****
Online Online

Posts: 2714



« Reply #8 on: August 20, 2011, 10:02:18 AM »
ReplyReply

With Lightroom’s histogram referring to this sRGB-TRCed ProPhoto-gamut Melissa space,
it seems to me hard to predict if saturation clipping would occur upon conversion to smaller Adobe RGB.

But then I’m not a Lightroom user, and the question may fit to the current discussion in the LR forum on its Clipping Indicators.

Peter

Quite true, but the histogram indicates that the scene is of rather low contrast, and setting of better black and white points would drastically improve the image. With ACR one can see gamut clipping in the chosen color space, but with LR one would have to export to ProPhotoRGB and use Photoshop's soft proofing tools.

Regards,

Bill
Logged
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1277


WWW
« Reply #9 on: August 20, 2011, 10:10:36 AM »
ReplyReply

Totally incorrect (just look at my example). Photoshop's soft proofing capabilities are only useful to find out if clipping occurs at converting to an output colour profile.

Your example shows 'white point clipping', or we could even call it 'dynamic range clipping' which is quite different from color saturation clipping or gamut analysis. I'll rephrase my previous statement as "Histograms aren't helpful for determining color saturation clipping."

Photoshop's soft proofing is useful for perhaps more than you think, including comparison to a working space like the original poster specifically asked about. If he wants to see if his images has out of gamut colors relative to AdobeRGB he can find this out with those tools.
« Last Edit: August 20, 2011, 11:58:55 AM by Onsight » Logged

digitaldog
Sr. Member
****
Offline Offline

Posts: 8036



WWW
« Reply #10 on: August 20, 2011, 10:18:05 AM »
ReplyReply

Totally incorrect (just look at my example). Photoshop's soft proofing capabilities are only useful to find out if clipping occurs at converting to an output colour profile. For any other situation (increasing saturation for instance), the histogram is the best tool.

I’d agree with you too. The colors seen in the histogram are useful to see the gamut clipping of one, two (saturation) or more channels (white/black all three). This tool is more useful in ACR because you can see the effect of gamut clipping as you change the RGB encoding options (toggle from sRGB to ProPhoto RGB as an example). Unfortunately, in LR, you get to see clipping based on ProPhoto primaries and if you export in anything but ProPhoto, the resulting histogram is different and you probably clipped saturation “blindly”.

To answer the OP’s question about all three channels clipping, I would suggest he alter the various rendering controls (Exposure, Blacks) to push the histogram out and see the results of the image. IOW, having a fixed histogram appearance can often make the image appearance ugly and awful! Edit images to appear as you desire on a calibrated and profiled display, not to produce a certain appearance of a histogram. Also using the clipping indicators (Alt/Option key) as you drag various sliders (again, Exposure, Blacks, Recovery) can be useful to see not only what is and isn’t clipping but where you might want to set the end of the tone scale and then viewing the image. Do you have and do you want a highlight that’s close to clipping (a specular?). Do you want to block up shadow detail as part of an artistic expression of the image? Look at the work of Greg Gorman and see how his style has no regard for shadow detail by design. There is no rule that says because you have shadow detail, you need to render and express it in your image. Clip it to death if you like that rendering.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
digitaldog
Sr. Member
****
Offline Offline

Posts: 8036



WWW
« Reply #11 on: August 20, 2011, 10:23:01 AM »
ReplyReply

In terms of what the histogram colors tell you about channel clipping in these Adobe raw products:


Red=Red
Green=Green
Blue=Blue
Yellow=Red+Green
Magenta=Red+Blue
Cyan=Green+Blue
White=RGB highlights (255-100%)
Black=RGB shadows (0)

The easy way to remember the dual channel clipping is if you see a color (Yellow), ask yourself, what’s the complement? Blue. So no Blue clipping (Red+Green channel clipping saturation).
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1277


WWW
« Reply #12 on: August 20, 2011, 11:56:48 AM »
ReplyReply

The colors seen in the histogram are useful to see the gamut clipping of one, two (saturation) or more channels (white/black all three).

I think you and Guillermo are taking this conversation out of context for it to be constructive to the OP. I understand what everyone is saying but let's look at the OP's post. He has no black or white clipping in his histogram - the left and right histogram walls aren't touched. Sure, in other images touching these walls indicates clipping and in channels that the histogram indicates - blah blah blah - that's another conversation.

The OP is asking about the portion of the histogram that scrapes the ceiling of the histogram. If in this context, what the both of you are saying could easily be taken the wrong way. I think we need to be fair and clear to the OP that touching the ceiling doesn't indicate that he's exceeding the limits of AdobeRGB. That's his original question after all. If he wants to see if any of the colors in his image exceed AdobeRGB (or any other color space) then PS's soft proofing tools are a better tool for this. And more importantly, that he shouldn't be alarmed that his histogram is touching it's ceiling.

Logged

digitaldog
Sr. Member
****
Offline Offline

Posts: 8036



WWW
« Reply #13 on: August 20, 2011, 12:18:12 PM »
ReplyReply

I think you and Guillermo are taking this conversation out of context for it to be constructive to the OP.

Actually no, I addressed the OP’s questions about histograms san’s any clipping.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
Guillermo Luijk
Sr. Member
****
Offline Offline

Posts: 1273



WWW
« Reply #14 on: August 20, 2011, 12:43:00 PM »
ReplyReply

I think you and Guillermo are taking this conversation out of context for it to be constructive to the OP

Logged

bjanes
Sr. Member
****
Online Online

Posts: 2714



« Reply #15 on: August 20, 2011, 12:54:31 PM »
ReplyReply

Your example shows 'white point clipping', or we could even call it 'dynamic range clipping' which is quite different from color saturation clipping or gamut analysis. I'll rephrase my previous statement as "Histograms aren't helpful for determining color saturation clipping."

Photoshop's soft proofing is useful for perhaps more than you think, including comparison to a working space like the original poster specifically asked about. If he wants to see if his images has out of gamut colors relative to AdobeRGB he can find this out with those tools.

Wrong again. Acutally, the ACR histogram is quite helpful in detecting saturation clipping. This is demonstrated in the two histogram previews shown below. In the first histogram with sRGB the red channel shows severe clipping. This is totally removed by using ProPhotoRGB.

Regards,

Bill
Logged
JohnKoerner
Guest
« Reply #16 on: August 20, 2011, 01:27:01 PM »
ReplyReply

What you see there is not any kind of clipping, is just that the Y-axis scale in which LR decided to show you the histogram doesn't reach its maximum, so it shows you a truncated version of the complete histogram. But it's just a matter of plotting, nothing related to the content of your image gamut.
The following image:
xxx
has the following histograms (left Photoshop, right complete histogram):
xxx
As can be seen, PS truncates the Y axis of the graph in order to make the plot more representative.
To analyze gamut clipping in the histogram, just look at histogram ends: in my picture, the Glencoe landscape in the Highlands was so intensely green that sRGB didn't manage to encode it, clipping the blue channel in 0.
For photographers who never heard about histograms, I think it's a good idea to look at the statistical concept of histogram, much wider that its particular application to photography.
Regards

Okay, thank you for explaining. What you described "a truncated version of the complete histogram" I was interpreting to mean "can't represent all of the colors" when was looking at the histogram.

But, in other words, I am not seeing color clipping at all here, what I am seeing is the limitatons of that histogram(?).




______________________________
______________________________




Histograms aren't helpful for gamut or gamut clipping info. Photoshop's soft proofing capabilities are designed for help analyze gamut clipping.  View>ProofSetup>Custom... and View>GamutWarning are the two tools worth getting to know if you're not familiar with them already.  

Okay, thank you.




______________________________
______________________________




Quite true, but the histogram indicates that the scene is of rather low contrast, and setting of better black and white points would drastically improve the image. With ACR one can see gamut clipping in the chosen color space, but with LR one would have to export to ProPhotoRGB and use Photoshop's soft proofing tools.
Regards,
Bill

Thank you. I know how to set the white point, but how do I set the black point?




______________________________
______________________________




I’d agree with you too. The colors seen in the histogram are useful to see the gamut clipping of one, two (saturation) or more channels (white/black all three). This tool is more useful in ACR because you can see the effect of gamut clipping as you change the RGB encoding options (toggle from sRGB to ProPhoto RGB as an example). Unfortunately, in LR, you get to see clipping based on ProPhoto primaries and if you export in anything but ProPhoto, the resulting histogram is different and you probably clipped saturation “blindly”.

Thank you for that. I always use ProPhoto.





To answer the OP’s question about all three channels clipping, I would suggest he alter the various rendering controls (Exposure, Blacks) to push the histogram out and see the results of the image. IOW, having a fixed histogram appearance can often make the image appearance ugly and awful! Edit images to appear as you desire on a calibrated and profiled display, not to produce a certain appearance of a histogram. Also using the clipping indicators (Alt/Option key) as you drag various sliders (again, Exposure, Blacks, Recovery) can be useful to see not only what is and isn’t clipping but where you might want to set the end of the tone scale and then viewing the image. Do you have and do you want a highlight that’s close to clipping (a specular?). Do you want to block up shadow detail as part of an artistic expression of the image? Look at the work of Greg Gorman and see how his style has no regard for shadow detail by design. There is no rule that says because you have shadow detail, you need to render and express it in your image. Clip it to death if you like that rendering.

You see, this is how I always do edit: to "my eye" not any histogram or need to conform. However, I just wanted to try to understand the histogram more, to have it help me. In other words (as Michael likes to say), "The difference between science and art." Recognizing this difference, I will always side with what looks good to me, but I just want to try to understand the science more.

I have many images that I edit, where my histogram is all off, but I like the way it looks. And yet, I have other images where the histogram is "perfect" but the image lacks punch to my eyes. And yet, I am sure in many other ways my being able to fully understand the histogram can only help me.

So, good point, and thanks for your time.

Jack


.
« Last Edit: August 20, 2011, 01:33:19 PM by John Koerner » Logged
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1277


WWW
« Reply #17 on: August 20, 2011, 01:55:44 PM »
ReplyReply

Acutally, the ACR histogram is quite helpful in detecting saturation clipping. This is demonstrated in the two histogram previews shown below. In the first histogram with sRGB the red channel shows severe clipping. This is totally removed by using ProPhotoRGB.

OK, yes, good example. LR's histograms, as the OP asked about, won't show this. I'll rephrase my previous statement as "LR's Histograms aren't helpful for determining color saturation clipping, and ceiling scraping isn't a concern."
Logged

N Walker
Sr. Member
****
Offline Offline

Posts: 291


WWW
« Reply #18 on: August 20, 2011, 05:53:41 PM »
ReplyReply


Thank you. I know how to set the white point, but how do I set the black point?


Jack,

You set the black point to your liking - this might involve clipping shadows to produce a higher contrast image.
Logged

MarkM
Full Member
***
Offline Offline

Posts: 230



WWW
« Reply #19 on: August 20, 2011, 06:10:44 PM »
ReplyReply

OK, yes, good example. LR's histograms, as the OP asked about, won't show this. I'll rephrase my previous statement as "LR's Histograms aren't helpful for determining color saturation clipping, and ceiling scraping isn't a concern."

Lightroom's histogram will show saturation clipping, but only those colors that exceed ProPhoto's gamut. The histogram in Lightroom seems to represent the ProPhoto space regardless of the color space of the image you're looking at.
Logged

Pages: [1] 2 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad