Bluebottle fly shot and rendering intent.

[xpatUSA]

I found a dead Sweat Bee on my bench and was struck by the diffracted colors of its body. Took a raw shot with the GH1 and the Leica macro lens - the shot itself could be improved but those beautiful saturated colors came out well in RawTherapee:



Almost too well, I thought. So I opened the above sRGB JPEG image in a program called ColorThink where the gamut of an image can be compared with that of a color space. Sure enough, most of the blues were hard up against the boundary of the sRGB gamut, in other words those nice vibrant blues in the above image were gamut-clipped (not the same thing as blown highlights, BTW).

So I redeveloped the raw file, this time setting the working space to ProPhoto and the output to sRGB but with the perceptual rendering intent. In ColorThink, it was clear that most of the blues were now inside the sRGB boundary and a little more spread out. Here the two images are screen-captured side-by-side:



The differences are very subtle but I hope you can see that the right hand image has more gradation of color in it and is just a tiny bit less harsh. Perhaps most easily seen in the violets in its hairy back-end.

[DanK]

Interesting comparison. Thanks for posting it.

I'm pretty sure that this is a sweat bee, not a fly. The eyes are bee-like, and it doesn't have the paddle-shaped antennae you generally find on flies. However, I certainly could be wrong.

[xpatUSA]

Thanks Dan,

I'm not from here (Gulf Coast) so "bluebottle fly" was from my English experience!

(blush),

[davidedric]

Thanks, Ted, very interesting to see the effect of colour spaces and rendering intent used in practice.

(Confession - I have bought a copy of ColorThink. Pure self indulgence - but I love experimenting, and it is beautiful software. Having had to learn to think in 3+ dimensions when I was researching in crystallography, I find the plots easy to appreciate)

Dave

[xpatUSA]

(Confession - I have bought a copy of ColorThink. Pure self indulgence - but I love experimenting, and it is beautiful software. Having had to learn to think in 3+ dimensions when I was researching in crystallography, I find the plots easy to appreciate)
Dave

I bought the cheaper version and it does have one flaw from a purist point of view. I use the Lab 3D view mostly and colors that should show as points actually get spread out radially (a lot). The test is to take a full frame shot of the MacBeth card, and you would expect each patch to show as a point. Or, even more precise and less time-wasting, analyze Lindbloom's or Koren's simulations thereof. Lindbloom's on mine shows up 6800 colors, even though the patches are filled, not photographed.

There's a thread on LL about it (link forgotten), recommending down-sampling an image to a certain size before opening it in ColorThink, but what a waste of time and it didn't work for my version anyway :-(

Doesn't bother me much, BTW, just adds fun to interpreting the result :-)

Hopefully your version is better.

[Richard Lundberg]

I wish you had output to aRGB. My monitor isn't quite wide gamut, but it's considerably bigger than sRGB in blue and red. Any chance you could post that?

[DanK]

BTW, LR allows you to switch back and forth between perceptual and relative rendering, and if you turn on soft proofing, it will indicate out-of-gamut areas. Of course, it only shows that it is out of gamut; it doesn't show the color space of the image.

[xpatUSA]

I wish you had output to aRGB. My monitor isn't quite wide gamut, but it's considerably bigger than sRGB in blue and red. Any chance you could post that?

Sorry, I don't use Adobe RGB (1998) at all.

You appear to have a most unusual monitor! If it is "considerably bigger than sRGB in blue and red", then it is also considerably bigger than aRGB in blue and red - because the blue and red primaries are identical in either space:



I would interested to know the primaries for your monitor. Is there a diagram for it like the above?

Cheers,

[Richard Lundberg]

I'll try to post the colorsync.

Still assembling the composite, but have discovered that you are absolutely right. I had compared the dell out of the box to the cal profile in error.

[Richard Lundberg]

I am unable to upload any images for some reason, the "Click Here" thing doesn't do anything
If I could, the profiles would show that the Dell profile is smaller than sRGB, the calibrated profile is identical to sRGB, aRGB is shown for reference, and my Epson six color printer is quite a bit larger than aRGB. Which brings up the question, how do you take advantage of that when starting from RAW Pro Photo.

[Richard Lundberg]

So here it is, finally. Upper left Dell out of the box. Upper right Dell calibrated identical to sRGB, lower left aRGB for reference, lower right Epson six color printer. Profiles are in Lab space.

[xpatUSA]

So here it is, finally. Upper left Dell out of the box. Upper right Dell calibrated identical to sRGB, lower left aRGB for reference, lower right Epson six color printer. Profiles are in Lab space.

Thanks Richard. Interesting that there is a difference between aRGB and sRGB (slightly more saturation) in the monitor reds/purples in L*a*b* space yet no difference in xyY. I checked in ColorThink and it is indeed so, no difference in Luv space either.

In ColorThink L*a*b* 3D, the aRGB reds/purples go outward and upward (L*) a bit compared to sRGB.

Which brings up the question, how do you take advantage of that when starting from RAW Pro Photo.

I don't do any printing but will try to answer in principle (forum printing cognoscenti will be along to help shortly, I hope). Assuming that your captured image contains colors that are out of gamut in aRGB, but within gamut for your printer, you would use ProPhoto or Adobe Wide Gamut as your working space and for intermediate saves. 16-bit working is usually recommended. When done, you would print in the printer color space, using relative colorimetric rendering intent because your captured image colors were within the printer gamut.

To post that same image would need a conversion to sRGB with either perceptual or relative colorimetric dependent on your taste.

Cheers,

[Richard Lundberg]

I am trying to figure exactly how to get from Pro Photo to print right now. It's not a problem to stay in PP 16 bit until ready to print ( actually for me it'a problem to get to Pro Photo in the first place with my RAF images, but that's another discussion). The space above for my six color printer would be nice to use. The problem is that when you use Photoshop manages colors, there's huge list of profiles to choose from, staring with the Epson model combined with several paper types, followed by device independent profiles Pro Photo, a RGB, sRGB and then many many others. It's not clear to me that I should use the Epson model with matte profile or the Pro Photo profile to get all the deep blues and cyans. I have used both and gotten good results, but I don't have a spectrophotometer. I can't find any explanation on the web why the profile choices include device independent profiles, but I have Schewe's book The Digital Print, maybe I'll find it there. I haven't had time to open it because I'm fighting the battle of getting my RAF files to open properly. The apps that use the lens correction won't save as a Pro Photo 16bit tiff. I'll probably have to use ACR and do the lens manually until I can calibrate using PTLens. It's a nightmare.
I hope someone with vast or half vast experience printing will chime in.

[Macmahon]

I am trying to figure exactly how to get from Pro Photo to print right now. ... The problem is that when you use Photoshop manages colors, there's huge list of profiles to choose from, staring with the Epson model combined with several paper types ...

That's where you need to be looking. If, for example you're using Epson Premium Semi Gloss, choose the Epson profile for that paper, then make sure you select the same paper in the print driver media setting. Photoshop will then manage the colour data conversion form ProPhoto to the print profile while preserving all the colours that are in the print gamut. I don't know what printer/inkset you use, but I'm guessing if it's a mainstream Epson photo printer it will manage the cyans and blues beautifully.

followed by device independent profiles Pro Photo, a RGB, sRGB and then many many others. ... I can't find any explanation on the web why the profile choices include device independent profiles...

Photoshop's list of profiles (every colour profile on the computer!), including the device independent ones is very confusing and not very helpful. You might assign or convert to one of the device independent profiles if you were exporting the image to be used in some other setting - e.g. if some publisher wants a TIFF version converted to sRGB, for some reason. But the Epson+paper profile is the right choice for your printing.

I have Schewe's book The Digital Print ...

Great resource. Read together with his earlier one "The Digital Negative" you should find the answers to every question!

Another very useful resource is this one: a free download from Native Digital's site. The tutorials on CiC are good too. For a simple-minded and possibly less complete approach you could try my website.

Cheers

Tim

[Richard Lundberg]

It's an Epson Artisan 725 six ink, which uses something like Dura Bright. I get excellent matches to my monitor using the Epson 720 with Premium Presentation Matte profile and very importantly the additional media set to matte, but of course these all within sRGB. I have used the Pro Photo profile and it didn't make any difference, I assume because the image wasn't outside sRGB.
I do have The Digital Negative as well.
I'm hoping to find a Pro Photo test image file somewhere on the web, so that I don't have to create one using invisible primaries and guesswork. Thanks for the references.
I have learned over the past few weeks that when an app "supports " your camera it a) doesn't, b) opens your raw but doesn't see the lens,or c) opens the raw and sees the lens but won't save as a Pro Photo 16 bit tiff. So I'll have to do it the hard way in ACR and Photoshop and take out perspective with DxO Perspective, a very user-friendly little app.

[flashback]

Other than the distinction between a fly and a sweat bee I don't understand a single thing otherwise posted. But that doesn't mean I can't learn (something).

[Richard Lundberg]

Here is the Pro Photo test image I linked in another thread:
http://www.outbackphoto.com/printing...048/essay.html
The first image on that page will no longer load but it's not the Pro Photo. The second image is the one you want.

[xpatUSA]

Other than the distinction between a fly and a sweat bee I don't understand a single thing otherwise posted. But that doesn't mean I can't learn (something).

It's a long hard road, Jack, with diminishing returns along the way. The point of the OP was just to demonstrate something that we're told often but not often shown. We're told that there are colors in nature that are so saturated (like intense or brilliant) that, even if the camera can capture them, it's not possible to show those exact colors on-screen or in-print even though our eyes saw them in the original scene. However, some software "working spaces" (meaning color spaces) can capture and manipulate many of those saturated colors. But what you see on the screen is not what's in the working space because there's translation software in the computer to translate (but not convert, i.e. change) any image to "look right" on your screen or print. This is one possible source of the standard whine "looks right on the screen but the print looks like cr@p".

There are several ways to go from a source image in a software working space (in RAM, not yet saved") which are explained here: https://www.cambridgeincolour.com/tu...conversion.htm

There is also the subject of "ICC color profiles" waiting to confuse the living mojon out of you

I perceive that you do have an interest, so good luck!

Here is the Pro Photo test image I linked in another thread:
http://www.outbackphoto.com/printing...048/essay.html
The first image on that page will no longer load but it's not the Pro Photo. The second image is the one you want.

That link is bad - gives error 404 'not found'

[Richard Lundberg]

When I copy/paste from your post it is bad. I'll check it out and repost. I did dl the image and use it with good results.

[Richard Lundberg]

Outback has become outback photo.net. There is a link there to their old blog of 2012. This is the page http://www.outbackphoto.com/CONTENT_...ing/index.html
You will see Printer Evaluation Target and how to use Printer Evaluation Target.
I am not sure why I was able to reach the target page using a link in another page, but can't reach it directly using copy/paste.