So, been a while since I've done any of this, recently just bought LTE.
I am firstly verifying a Postium 17" N170G monitor and I just want to check that my general observations so far are on the right track.

I am using Resolve on a remove machine to do patches, networked to Colourspace. The machine has a DeckLink SDI out to the display. I believe everything is set up correctly. The monitor itself is set to 709, D65, 2.4 Gamma and 112 brightness.
I am using an older model i1D3 and I also have another one on its way to me to check with as well.

I have Colourspace set up with delays, and what I believe are good settings for this hardware based on the page on the website.

I am doing grey and primaries at the moment and I am seeing results like the attached. I think the RGB balance is telling me that the white point is incorrect. But I assume setting 709 as the profile has a D65 target whitepoint? So either the monitor is out of spec or the probe is (hopefully the second probe will help confirm that).

Looking at the EOTF it seems pretty good if I flip between 709 profile and 1886, the white target gamma appears different. Should they not be the same?
Setting EOTF to absolute error shows me the blue, which I think is confirming the RGB Balance.
So is this a white point issue on my side or are the profiles set for a different whitepoint. I couldn't find anything in the guides so far about that?

Kindest
Paul

It is really better to post the profile, not screen grabs.
But the white point is well off, assuming the target colour space is Rec709.
The Absolute EOTF graphs is showing the luma error vs. the target min/max values, so is not necessarily the same as the colour temp error the TGB Balance is showing.
There will be a difference between Rec709 as target and BT1886, as the latter used the measured black point to alter the near black EOTF.
(See the info on BT1886 on the website, and why we do not recommend it.)

Steve

Hi Steve, not quite at the full profiling stage, just been working through and checking things.

The BT1886 vs 709 - the 1886 Gamma appears to be closer to 2.2 than 2.4. These are the built-in colourspaces and I've not modified them. I can see that both 709 and 1886 are set to 2.4 and sRGB is 2.2 but when flipping between them on the EOTF graph the 1886 is closer to the 2.2 of sRGB (apart from the toe) than the 2.4 of 709.

I can't find the info you mean on the website - which guide is that in?

I've been running profile tests with two different 1D3 probes but are there any tools that let me see the difference between two profiles? They are both in the ball park but do have slightly different bias in the colours. Just trying to work out how I can determine which is better (no easy task with nothing to go on). I do have a Sekonic colour light meter. I wonder if somehow I can use that to verify - if that is accurate enough!

Kindest
Paul

Yes, if the display has a lifted black BT1886 will indeed look closer to an EOTF of 2.2.
As said, this is explained in the website - see the 'Calibration Issues' page for example, or just search this Forum.

And sRGB display EOTF is a pure 2.2 EOTF, no toe as that is just Encoding - again, try a search of these Forums.

And no, there is no way to know which i1D3 is more accurate without a high-end spectro to compare to.
See the example of the issues on the i1D3 page of the website.
(And certainly not the Sekonic.)

Steve

Clearly I'm missing something. I read the guide and the graphs there look as I expect - a small difference between 709 and 1886 but with the lifted shadows.

But taking an initial profile of this monitor in it's most native mode and then flipping the colourspace (Target Gamma / EOTF option) between 709, 1886 and sRGB I am seeing 1886 and sRGB being a much closer match in terms of the white EOTF curve than the 709. Yet if I look at the profile definitions, 709 and 1886 are set to 2.4 and sRGB to 2.2.

(The monitor is set to 2.4 but according to these results seems to track 2.2 closer)

I have obviously got something wrong. But just flipping between these colourspaces shouldn't look like this should it? I should be seeing a closer match between 709 and 1886 than sRGB, like the guide you mention.

Yes, as said, with a display with lifted black that is exactly what you get with BT1886.
I guess your black is higher than 0.01 nits, as in the example on the website.
And that is why we do not recommend it!

Steve

Steve

So that white line when I swap the target EOFT - what does the white line mean? I took it to mean the 'correct EOTF curve for the profile that was selected' which is irrespective of whatever other profiled data/curve is shown (the coloued curves)?

As per the User Guides, the white line is the Correct Target EOTF for the selected target colour space.
The RGB lines are the actual measured EOTF values for the display.

Steve

Steve
So then why do I see a different curve here than the guide shows in the BT1886 section (https://lightillusion.com/calibration_issues.html)

That's my question. On your curves in the guide above the 709 and 1886 are quite similar apart from the shadows and sRGB is 2.2 so quite different. In the screens above when I do that the 1886 and sRGB are similar but the 709 is 2.4

This is my confusion. I am seeing a different EOTF than the guides show me, so wondering what I have done that is different. If I look at the definitions then 1886 and 709 are 2.4, sRGB is 2.2 and that is not what I am seeing here.

Sorry, just trying to understand what's going on and thinking I'm going mad. So ignore the profile in the 3 screens above and just look at the white curves and flip between them.

Because the black level of your display is different, as I did say.
As per the guide, BT1886 alters the target EOTF (specifically the lower range, reducing the change as the brightness increases) depending on the black level relative to the white level.

If you have a display with a ZERO black level, the target EOTF will be pure 2.4.
As the display black level increases the BT1886 target EOTF changes - gets lifted.

And that again is why we do not recommend it.

Steve

Just for the hell of it, here is an example of a display with a 0.1 nit black level compared to targets Rec709, BT1886, and sRGB.
Hopefully you can see why we do not recommend BT1886.
(The display was accurately calibrated to BT1886!)

Steve

Rec709

BT1886

sRGB

Steve

I hope a picture can replace 1000 of my rambling words and thanks for replying...

In that guide you wrote about a black level of 0.1 and the curve you show then has a raised toe but the mids are still a 2.4 gamma.

See the graphic. At the top are the graphs from the guide for 709, 1886 and sRGB. What I see is that 709 and 1886 are both 2.4 curves but 1886 has shadows raised as per the discussion. sRGB is 2.2 Gamma and quite different. On my plots underneath, which are from a profile of my monitor which has min of 0.17 and max of 164, I see 709 perfect as I expect and sRGB obviously 2.2. But this time, unlike the version from the guide, the 1886 curve looks more like a 2.2 gamma and so with a 0.17 nits is that level of luma raising expected? It looks too much to me, this would raise the midtones as well. But what I think you're saying is yes. The 1886 plot is relative to the min but the degree of brightness across the whole image is really striking.

Obviously like you say, ignore 1886 and that's fine. I am just dead curious if the level of black raising can be this much. I'd assume it was just shadows only.

Hope this makes sense!

See my post above.
The User Guide for BT1886 uses a black value of 0.01.
The post above uses a black value of 0.1 nits.
If you display is 0.17 nits for black it will be even worse, and to be fair, will be a display that is basically not fit for purpose...
And all this is EXACTLY how BT1886 works (or doesn't depending on your view... raising near black, slowly reducing the amount as brightness increases.)
(This discussion also shows why posting your actual profile makes things so much easier.)

Steve

Ah, this makes sense now. 0.17 really does make that much difference - I will be able to go back and check profiling and see if I've made a mistake - this is exactly why I need sanity checking. I just didn't think it would be *that* far out.

The monitor is a Postium LCD - should be reasonably good.

Thanks! (And yes in future will post profile)
Paul