Yes, those 3 plus one more if I do not abide by CIE rounding rules, use original observer data and calculate the D50 spectrum following late Dr. Schanda suggestions (things I normally do).
Will we ever get Ps to do error diffusion with blue noise when rendering to display?
Sure, but it should be obvious that there are significant differences in the data and not just display for my adjusted Granger rainbows. This is readily confirmed when I take the data from both into Lab and compare the values there.
Now I STILL know.
Those are reflectance coefficients, there is no unnatural light in the data. I doubt that they are average unless you have macro in mind. Most likely, they measured small spots. About the variations - colors outside the target space are usually mapped inside with some compression and necessary distortion but that does not mean that small detail is lost and that the variations go far away from the average.
I have my own examples of flowers where a wider color space creates a better photo. I do not doubt that sometimes, wider spaces are useful. I am just skeptical that the data presented serve as a convincing evidence of that.
Yes, it can happen, I have seen it.
To repeat (something you enjoy doing) - it's 'enough' for a good 95% of the samples here.Yes, it’s not enough (sRGB), colors clipped! Ditto with Adobe RGB (1998).The samples Iliah put forward, to illustrate it being not 'enough'.95% of just 218 samples, mostly measurements of vegetation.Or rather... yes.No!So, sRGB might be enough after all?I did a quick and dirty analysis of the 219 natural colors by hand. It appears that about 10 of them fall outside sRGB gamut, which are shown in the plot below. Each dot shows one natural color. Cyan = In Gamut. Magenta = Out of Gamut.
Natural Color Set. SRGB space. Vector length for each primary = Respective Unit Stimulus.
The plot shows that these particular natural colors are concentrated cone like volume.
On my screen I can rotate the diagram and see stuff a bit better. Difficult to find a view that shows every thing from a good vantage point in a still.
According to Joofa's analysis, 'enough' for a good 95% of the samples here.
I'll speak as and for whoever I choose.
Your president - a president you deserve.
Most is probably understating it.
There's nothing 'foolish' in the matter.
Only a small minority of the world's photography/photographers work from RAW - and noboby is stopping them from choosing alternatives to sRGB.
Why have you posted a link to my DPR 'gallery' ?
You're very fond of tiresome, clichéd, cut & paste literary quotes - your having no original thought of your own.
If you have the same whitepoints and the first RGB space has all three primary CIExy chromaticites encompassing those of the second (in 2D space), then the first XYZ (3D) space will not have any colors not included in the second (3D) space.Well you can have the same white point and as long as primaries of the 2 color 'spaces' are pointing in different directions, you can still have one color 'space' composed of more colors than the other.When discussing absolutes (XYZ), It goes without saying that any two RGB colorspaces with differing "D" whitepoints, regardless of where their CCT, will each have colors the other won't.Well, when some colors that are in gamut in standardized Adobe RGB but are out of gamut in standardized ProPhoto RGB in linear color space terms, then they are just not there. One can throw in various flavors of rendering intents. But, that is just to maintain certain relationships between colors in one way or another.Only an issue in CIE Abs. Colorimetry as a result of D65 to D50 conversion. However, the ICC, defines Abs. Col. differently and clearly states that Abs. Col. conversion from D65 <> D50 should use adapted white points. This is how Adobe's color engine does things as well. Microsoft's ICM deviates from the ICC standards and preserves white points during Abs. Col. conversions. The Microsoft engine also has some other significant issues introducing fairly large errors during simple conversions from Adobe RGB to sRGB as I've itemized over on the printer forum.Without changing anything, especially normalization, it appears that 4 colors are out of gamut in ARGB.
BTW, here is a comparison of human visual spectral locus, Adobe RGB, and ProPhoto RGB below.
Adobe RGB vs. Pro Photo RGB
ProPhoto RGB is a bit bigger and seems to encompass more of human visual space. Except near the blue areas there are colors that are in Adobe RGB but outside of Pro Photo RGB gamut. As shown in the link below.
When is Pro Photo RGB bigger than Adobe RGB.
--
Dj Joofa
http://www.djjoofa.com
Of course. That's just linear algebra.
Yes, of course. I agree.
Yes, and so? I don't disagree with your statement or reasoning where and why Adobe RGB can exceed ProPhoto RGB in the "blue." I've simply pointed out that all* RGB colorspaces, where the white points D* differ, will also each have colors the other will not in the CIE sense of XYZ intersection.
Neither measurements conditions nor D50 are photographically natural light, they are flat light field, intensity & spectra. XYZ samples are for D50. To calculate natural color in natural conditions one needs to apply naturally varying light field to coefficients, or to use different instrumentation.
Average happens at instrument's aperture or in a light sphere.
https://www.dpreview.com/forums/post/59584445To repeat (something you enjoy doing) - it's 'enough' for a good 95% of the samples here.Yes, it’s not enough (sRGB), colors clipped! Ditto with Adobe RGB (1998).The samples Iliah put forward, to illustrate it being not 'enough'.95% of just 218 samples, mostly measurements of vegetation.Or rather... yes.No!So, sRGB might be enough after all?I did a quick and dirty analysis of the 219 natural colors by hand. It appears that about 10 of them fall outside sRGB gamut, which are shown in the plot below. Each dot shows one natural color. Cyan = In Gamut. Magenta = Out of Gamut.
Natural Color Set. SRGB space. Vector length for each primary = Respective Unit Stimulus.
The plot shows that these particular natural colors are concentrated cone like volume.
On my screen I can rotate the diagram and see stuff a bit better. Difficult to find a view that shows every thing from a good vantage point in a still.
According to Joofa's analysis, 'enough' for a good 95% of the samples here.
Then your writings can and should be dismissed and ignored.
Yes, you got that right; I never stated who’s president he was, only that he’s the US president and his statements sound as foolish as yours when you speak for everyone.
Assumptions.
Assumptions without fact.
We have a difference of opinion then.
Prove that’s accurate.
To illustrate good enough.
That’s incorrect too. You’re on a roll.
This does not change the fact that those are reflection coefficients. If you want to say that they are very unreliably measured, this does not make the case for sRGB enough or not based on that data more convincing.
Average over what area? You cut the rest of the paragraph.
Got it.
Man, you are seriously confused. Get back to your Word and Excel.
I fully understand both.https://www.dpreview.com/forums/post/59584445To repeat (something you enjoy doing) - it's 'enough' for a good 95% of the samples here.Yes, it’s not enough (sRGB), colors clipped! Ditto with Adobe RGB (1998).The samples Iliah put forward, to illustrate it being not 'enough'.95% of just 218 samples, mostly measurements of vegetation.Or rather... yes.No!So, sRGB might be enough after all?I did a quick and dirty analysis of the 219 natural colors by hand. It appears that about 10 of them fall outside sRGB gamut, which are shown in the plot below. Each dot shows one natural color. Cyan = In Gamut. Magenta = Out of Gamut.
Natural Color Set. SRGB space. Vector length for each primary = Respective Unit Stimulus.
The plot shows that these particular natural colors are concentrated cone like volume.
On my screen I can rotate the diagram and see stuff a bit better. Difficult to find a view that shows every thing from a good vantage point in a still.
According to Joofa's analysis, 'enough' for a good 95% of the samples here.
You misunderstand the analysis and the results.
I understand it, and accept it insofar as less than 100% not being 'enough' for some.
You don't have either that capability, or that discipline.
A president that the likes of you deserve.
You can't see it - but there are considerable similarities between Andrew Rodney (alias 'digidog') and Trump behaviour.
Prove it isn't.
In 'old money' - many an image started with a 'negative', but only a minority printed/enlarged for themselves from their own neg's (and FWIW even fewer using colour).
Sure - keep believing it.
Thanks but - whilst it's my/DPR web space - actually not my photos (with just a couple of exceptions).
Actually, it is correct - and you, are not on a roll.
