Rendering Intents

[ColinR2016]

Hi,

I'm fairly new to printing and am struggling to understand the concept of rendering intents as there appears to be a lot of contradictory information around. I have read that the concept only applies to out of gamut areas of the image. What confuses me is that at the soft proofing stage I can see a small difference on-screen between Perceptual and Relative in areas which are in gamut according to my editing software Affinity. The rendering intent which does make a big difference is "Absolute Colorimetric" which appears to change the white point considerably.

Thanks

 

[Gesture]

This may help. Good question.

Choosing a Rendering Intent for Your Inkjet Print

There are four different Rendering Intents with which you may want to familiarize yourself with: Relative Colorimetric, Perceptual, Absolute Colorimetric, and Saturation. In thi

www.redrivercatalog.com

 

[Gryphon69]

Hi,

I'm fairly new to printing and am struggling to understand the concept of rendering intents as there appears to be a lot of contradictory information around. I have read that the concept only applies to out of gamut areas of the image. What confuses me is that at the soft proofing stage I can see a small difference on-screen between Perceptual and Relative in areas which are in gamut according to my editing software Affinity. The rendering intent which does make a big difference is "Absolute Colorimetric" which appears to change the white point considerably.

Thanks

Basically, RC compresses out of gamut colors into gamut without changing in-gamut colors, whilst Perceptual compresses out of gamut colors into gamut and also changes in-gamut colors so as to preserve the original color gradation as much as possible.

So if you have an image with a few colors that are just barely out of gamut, the difference between the two rendering intents will likely be minimal, whilst if you have an image with colors that are significantly out of gamut Perceptual will likely give you the more pleasing output as it preserves a smooth color gradient although it achieves that result by slightly adjusting (=changing) colors that would have been in-gamut in the original image.

Ultimately, unless you are doing some color-critical work, selecting a rendering intent boils down to which of the two (okay, technically four) gives you the most pleasing end result.

Hopefully other forum members will correct any inaccuracies in the above.

If I recall correctly, there is an excellent YT video by (I think) Rocco Ancora that gives an in-depth overview of color spaces and rendering intents. If I find it, I will post a link.

 

[Martin-C]

Hi,

I'm fairly new to printing and am struggling to understand the concept of rendering intents as there appears to be a lot of contradictory information around. I have read that the concept only applies to out of gamut areas of the image. What confuses me is that at the soft proofing stage I can see a small difference on-screen between Perceptual and Relative in areas which are in gamut according to my editing software Affinity. The rendering intent which does make a big difference is "Absolute Colorimetric" which appears to change the white point considerably.

Thanks

Basically, RC compresses out of gamut colors into gamut without changing in-gamut colors, whilst Perceptual compresses out of gamut colors into gamut and also changes in-gamut colors so as to preserve the original color gradation as much as possible.

That has to be the best explanation of rendering intent I've seen, thank you!

 

[Gryphon69]

Hi,

I'm fairly new to printing and am struggling to understand the concept of rendering intents as there appears to be a lot of contradictory information around. I have read that the concept only applies to out of gamut areas of the image. What confuses me is that at the soft proofing stage I can see a small difference on-screen between Perceptual and Relative in areas which are in gamut according to my editing software Affinity. The rendering intent which does make a big difference is "Absolute Colorimetric" which appears to change the white point considerably.

Thanks

Basically, RC compresses out of gamut colors into gamut without changing in-gamut colors, whilst Perceptual compresses out of gamut colors into gamut and also changes in-gamut colors so as to preserve the original color gradation as much as possible.

So if you have an image with a few colors that are just barely out of gamut, the difference between the two rendering intents will likely be minimal, whilst if you have an image with colors that are significantly out of gamut Perceptual will likely give you the more pleasing output as it preserves a smooth color gradient although it achieves that result by slightly adjusting (=changing) colors that would have been in-gamut in the original image.

Ultimately, unless you are doing some color-critical work, selecting a rendering intent boils down to which of the two (okay, technically four) gives you the most pleasing end result.

Hopefully other forum members will correct any inaccuracies in the above.

If I recall correctly, there is an excellent YT video by (I think) Rocco Ancora that gives an in-depth overview of color spaces and rendering intents. If I find it, I will post a link.

This is the video I was thinking of:

Rendering intents are discussed at minute 36 or so, but I recommend watching the whole thing.

If you (or others) are interested in an IMHO very good discussion about color management and color spaces in general, I would also recommend this other video:

 

[Gesture]

Thanks.

 

[Guillermo Shashte]

Hi,

I'm fairly new to printing and am struggling to understand the concept of rendering

Don't overcomplicate things, keep it simple for now. For Portrait and Nature photography (printing) use sRGB color space and perceptual rendering intent.

 

[NAwlins Contrarian]

I'm fairly new to printing and am struggling to understand the concept of rendering intents as there appears to be a lot of contradictory information around. I have read that the concept only applies to out of gamut areas of the image. What confuses me is that at the soft proofing stage I can see a small difference on-screen between Perceptual and Relative in areas which are in gamut according to my editing software Affinity. The rendering intent which does make a big difference is "Absolute Colorimetric" which appears to change the white point considerably.

Three points:

(1) For an explanation of rendering intents and many related concepts, go read what Andrew "Digital Dog" Rodney has written and posted at:

http://www.digitaldog.net/tips-and-tricks.html

especially Rendering Intents and ICC profiles and Soft Proofing explained.

(2) Not all ICC profiles contain the data to support all rendering intents. Just because the software lets you choose relative colorimetric or perceptual or one of the others does not mean that the profile contains the data to properly do that. In that case, the behavior is undefined: maybe the software uses another rendering intent, or maybe the software does nothing, or maybe the software tries to approximate what it thinks that rendering intent would do.

(3) The behavior of the perceptual rendering intent is not fully defined in ICC standards. In other words, depending on the software that built the ICC profile, perceptual may produce a different effect even if the printer and paper (and potentially even spectrophotometer) are the same.

 

[Gesture]

Great insights. If Keith "goes technical" in a book, this is something I'd like to see explained and demonstrated, if possible. Although I hope the book is more conceptual.

 

[NAwlins Contrarian]

Great insights. If Keith "goes technical" in a book, this is something I'd like to see explained and demonstrated, if possible. Although I hope the book is more conceptual.

Thanks. One of these days I'd like to take a sampling of ICC profiles and a test image or chart, do a bunch of conversions, and analyze just what the rendering intents (and black point compensation) do. You could start with a simple B&W step wedge, but saturated color behavior is probably more interesting.

 

[wilberforce_1]

Hi,

I'm fairly new to printing and am struggling to understand the concept of rendering

Don't overcomplicate things, keep it simple for now. For Portrait and Nature photography (printing) use sRGB color space and perceptual rendering intent.

Well, I think the opposite, use AdobeRGB and relative rendering intent as a default. SRGB is smaller space and perceptual often gives unexpected color shifts.

 

[ThrillaMozilla]

(2) Not all ICC profiles contain the data to support all rendering intents. Just because the software lets you choose relative colorimetric or perceptual or one of the others does not mean that the profile contains the data to properly do that. In that case, the behavior is undefined: maybe the software uses another rendering intent, or maybe the software does nothing, or maybe the software tries to approximate what it thinks that rendering intent would do.

Amen.

(3) The behavior of the perceptual rendering intent is not fully defined in ICC standards. In other words, depending on the software that built the ICC profile, perceptual may produce a different effect even if the printer and paper (and potentially even spectrophotometer) are the same.

Black point correction isn't fully specified either, and results can vary greatly depending on profile and paper.

 

[Rand 47]

Hi,

I'm fairly new to printing and am struggling to understand the concept of rendering

Don't overcomplicate things, keep it simple for now. For Portrait and Nature photography (printing) use sRGB color space and perceptual rendering intent.

Well, I think the opposite, use AdobeRGB and relative rendering intent as a default. SRGB is smaller space and perceptual often gives unexpected color shifts.

Well . . . both of these approaches are suboptimal.

Rand

 

[wilberforce_1]

Hi,

I'm fairly new to printing and am struggling to understand the concept of rendering

Don't overcomplicate things, keep it simple for now. For Portrait and Nature photography (printing) use sRGB color space and perceptual rendering intent.

Well, I think the opposite, use AdobeRGB and relative rendering intent as a default. SRGB is smaller space and perceptual often gives unexpected color shifts.

Well . . . both of these approaches are suboptimal.

Rand

Over an hour of video is a lot to watch. Any key takeaways?

 

[Rand 47]

Hi,

I'm fairly new to printing and am struggling to understand the concept of rendering

Don't overcomplicate things, keep it simple for now. For Portrait and Nature photography (printing) use sRGB color space and perceptual rendering intent.

Well, I think the opposite, use AdobeRGB and relative rendering intent as a default. SRGB is smaller space and perceptual often gives unexpected color shifts.

Well . . . both of these approaches are suboptimal.

Rand

Over an hour of video is a lot to watch. Any key takeaways?

Hummm … how do I respond w/o insulting you. It’s worth the education. There is no one better than Andrew Rodney re color management and printing.

But, here’s an example: Lightroom (my choice of editor and print utility) has by default “Melissa RGB” as it’s one-and-only default color space for all photos you ingest into Lightroom Classic (Lrc). You don’t even have a choice. Melissa RGB is a “flavor” of ProPhoto RGB. So, e.g. if you ingest a raw into LrC and then send it to Photoshop (PS) for some editing, and PS’s working color space is “smaller than” ProPhotoRGB, you risk throwing away color information (that you CAN print, by the way) in your raw, that is beyond the gamut of AdobeRGB - and WAY beyond the gamut of sRGB. Why do that? So, by having your working colorspace LARGE you preserve all the available data that your expensive camera captured.

Where printing is concerned, many modern papers/inksets have gamuts larger than Adobe RGB. And you can print them. Usually it provides better tonal transitions in prints, rather than “startling amount of more color.” I do a demo of this when I do color management seminars. Students can see the difference. This, of course, assumes that these colors were “out in front of” your camera when you took the photo.

In the last video I posted, Andrew even provides an image where you can test this out for yourself on your own printer.

If “good enough” is good enough for your intended end product, then perhaps this is not worth pursuing for someone. But when in pursuit of the best fine art print possible, this is extremely valuable information.

Rand

 

[Rand 47]

Hi,

I'm fairly new to printing and am struggling to understand the concept of rendering

Don't overcomplicate things, keep it simple for now. For Portrait and Nature photography (printing) use sRGB color space and perceptual rendering intent.

Well, I think the opposite, use AdobeRGB and relative rendering intent as a default. SRGB is smaller space and perceptual often gives unexpected color shifts.

Well . . . both of these approaches are suboptimal.

Rand

Over an hour of video is a lot to watch. Any key takeaways?

Hummm … how do I respond w/o insulting you. It’s worth the education. There is no one better than Andrew Rodney re color management and printing.

But, here’s an example: Lightroom (my choice of editor and print utility) has by default “Melissa RGB” as it’s one-and-only default color space for all photos you ingest into Lightroom Classic (Lrc). You don’t even have a choice. Melissa RGB is a “flavor” of ProPhoto RGB. So, e.g. if you ingest a raw into LrC and then send it to Photoshop (PS) for some editing, and PS’s working color space is “smaller than” ProPhotoRGB, you risk throwing away color information (that you CAN print, by the way) in your raw, that is beyond the gamut of AdobeRGB - and WAY beyond the gamut of sRGB. Why do that? So, by having your working colorspace LARGE you preserve all the available data that your expensive camera captured.

Where printing is concerned, many modern papers/inksets have gamuts larger than Adobe RGB. And you can print them. Usually it provides better tonal transitions in prints, rather than “startling amount of more color.” I do a demo of this when I do color management seminars. Students can see the difference. This, of course, assumes that these colors were “out in front of” your camera when you took the photo.

In the last video I posted, Andrew even provides an image where you can test this out for yourself on your own printer.

If “good enough” is good enough for your intended end product, then perhaps this is not worth pursuing for someone. But when in pursuit of the best fine art print possible, this is extremely valuable information.

Rand

OK, I’m replying to myself here, but I thought it might be helpful to see what I’m talking about above. Look at the image below:



The wireframe is the gamut representation is Adobe RGB 1998 in ColorThink Pro.

The solid color gamut plot is the color gamut of Ilford Gold Fibre Gloss paper, when printing to an Epson P5000 printer.

Even from this single-angle view you can see that the P5000 printer w/ Ilford Gold Fibre Gloss paper can print “a lot of colors” (deeper greens / blues & and on the other end lighter tones of yellow, yellow-orange, etc. than Adobe RGB 1998 can contain.

If you can conceive of a colorful “fall color” image that has these colors / tones “out there” for the camera to capture in the raw data, and you bring that file into a working color space of Adobe RGB 1998… those colors that your expensive camera / sensor captured are … GONE. And you could have printed them! Why would you do this? As I mention above, most of the time this “appears in print” as smoother tonal gradations within color ranges. But it is visible. Andrew Rodney taught me this.

Rand

 

[wilberforce_1]

Appreciate the replies!

I have more enthusiasm for watching the videos now, but have not done so yet.

I agree on using the largest color space available, which is what I do as much as reasonably feasible. However, at some stages sometimes the best available is AdobeRGB. I think we can agree that using sRGB is not desirable, except possibly for final export for the web, as almost always larger is available.

In LR the only real choices on rendering intent are Perceptual and Relative. I have found the Perceptual sometimes gives unexpected color shifts (perhaps the fault of the ICC profile?), especially in blue skies. I suppose one should try both for each image, but this usually means printing everything twice, as I have found the "soft-proofing" in LR or PS to not reliably represent the actual printed colors for Perceptual vs Relative, at least for some ICC profiles, for some reason. However, for images with lots of highly saturated colors obviously out of gamut, Perceptual would make more sense.

 

[Rand 47]

Appreciate the replies!

I have more enthusiasm for watching the videos now, but have not done so yet.

I agree on using the largest color space available, which is what I do as much as reasonably feasible. However, at some stages sometimes the best available is AdobeRGB. I think we can agree that using sRGB is not desirable, except possibly for final export for the web, as almost always larger is available.

In LR the only real choices on rendering intent are Perceptual and Relative. I have found the Perceptual sometimes gives unexpected color shifts (perhaps the fault of the ICC profile?), especially in blue skies. I suppose one should try both for each image, but this usually means printing everything twice, as I have found the "soft-proofing" in LR or PS to not reliably represent the actual printed colors for Perceptual vs Relative, at least for some ICC profiles, for some reason. However, for images with lots of highly saturated colors obviously out of gamut, Perceptual would make more sense.

You’re welcome re the replies. I can’t think of a time when Adobe RGB is better as a working color space. Of course when we’re ready to export for various end uses, we’ll convert to the appropriate color space. E.g. sRGB for stuff going to sRGB devices like phones and tablets.

I find LrC’s soft proofing to be very good. Having said there is never a perfect match from screen to print. But with a properly calibrated wide gamut monitor, and consistent viewing conditions (in my case 4700k Solux lamps or my GTI Lightbooth), I find that the screen to print match is VERY good indeed. I only very rarely see unexpected or different colors, and that’s more likely due to my being sloppy with soft proofing.

I think you’ll find the time you spend watching Andrew’s videos well worth it. Perhaps watch the last one first. It’s the main tutorial on the benefit of wide gamut color spaces. The first two were rebuttals to some of the internet FOD about sRGB.

Rand

 

[Guillermo Shashte]

I think we can agree that using sRGB is not desirable, except possibly for final export for the web, as almost always larger is available.

I disagree. The OP said. "I'm fairly new to printing and am struggling to understand the concept of rendering intents." Using a large working color space maybe fine if you know how to use it. sRGB is a safe working space for newbies because it will allow to work without getting out of gamut colors. Moreover, it will give them a baseline that will help evaluate prints when they decide to try a larger working space.

In LR the only real choices on rendering intent are Perceptual and Relative. I have found the Perceptual sometimes gives unexpected color shifts (perhaps the fault of the ICC profile?), .

That could be a problem when you are using a working space that is larger than your printer profile. Very unlikely you will have this problem if you are using sRGB as your working space.

For users who have no idea what " ColorThink Pro" is, just keep it simple and use sRGB and perceptual rendering intent. At least until you don't need to ask questions.

That's all I want to say about this and will not be reading any other replies to this thread.

--
GS

 

[Rand 47]

I think we can agree that using sRGB is not desirable, except possibly for final export for the web, as almost always larger is available.

I disagree. The OP said. "I'm fairly new to printing and am struggling to understand the concept of rendering intents." Using a large working color space maybe fine if you know how to use it. sRGB is a safe working space for newbies because it will allow to work without getting out of gamut colors. Moreover, it will give them a baseline that will help evaluate prints when they decide to try a larger working space.

In LR the only real choices on rendering intent are Perceptual and Relative. I have found the Perceptual sometimes gives unexpected color shifts (perhaps the fault of the ICC profile?), .

That could be a problem when you are using a working space that is larger than your printer profile. Very unlikely you will have this problem if you are using sRGB as your working space.

For users who have no idea what " ColorThink Pro" is, just keep it simple and use sRGB and perceptual rendering intent. At least until you don't need to ask questions.

That's all I want to say about this and will not be reading any other replies to this thread.

Well, I guess if the OP is using a relatively low end camera and shooting only jpgs, your advice is OK. Other than that, using sRGB as a working color space is like buying a 3 thousand dollar stereo set and playing it through two dollar speakers.

It’s not that complicated. Just watch Andrew Rodney’s video on large gamut working spaces and be done with it. And, if the OP happens to be working exclusively in Lightroom Classic, there’s no choice anyway, it IS ProPhoto RGB with zero ability to have sRGB as working space. And one might even ask, “Why did Adobe make Lightroom Classic that way?”

Rand