Steve - any thoughts on using a (single-link) ISMini in profiling quad-link displays (I'm think Sony X300 or Canon DP-V3010; but the Canon can be it's own patch-gen within LightSpace, so maybe not the Canon!) - is it reasonable to feed all four quads, use a full-screen patch in LightSpace and profile as normal? Any limitations using this when DCI-P3 or (a subset!) of Rec.2020 is what the monitor is set for?
Is it reasonable to derive a LUT this way in LightSpace which could then be used in a 6G/12G path using the new BM Teranex box that has a LUT internally?

At the moment the IS-mini can't do 4k... so I really am not sure if sending the dame signal into all 4 quad display inputs would work...

I would look at the BoxIO.
(We will have full intehration for the BoxIO, including Patch Generation, in the very near future)

See:http://www.lightillusion.com/boxio_lut_box.html

The Canon can be fully calibrated via LightSpace CMS, using it's internal 3D LUT, as well as internal patch generator.

Steve

Clearly I have to have a tinker! I shall profile my Canon using it's internal patch gen and then compare that to feeding it single-link from the ISmini and then looping all four quads with the o/p of the Fuji. Hopefully the results will be identical.

We'll be very interested in the results!

https://www.dropbox.com/sh/tmbjwftvlzmgxsf/AAAEyCKr7svDIkYzJgTFAptla?dl=0

So it all appeared to go OK - the folder (above) has the LightSpace PDFs for the same monitor - initially set up as factory default for Rec 709, 1886 Gamma (i.e. a vanilla TV display) and at default the 5% grey & 100% white were very close to 6504. Peak white set for 100Cd/m2.
Ran a 12-point profile twice (once using the Canon's internal patch gen, once with the Fuji ISMini) and then repeated with the monitor set at 4k res with DCI-P3 colour; same procedure, only difference was the ISMini was feeding all four quads of the SDi input.
I expected the two HD profiles to be exactly the same. But they aren't?! Neither are the two UHD/P3 ones.

What's the advice Steve; trust the ISMini or the Canon's patch gen? Nothing else changed in each case.

I would not know which result to trust - there are too many variables.
But, there are unexpected differences with the Black/White levels and Gamma between the different patch generators.
(The gamma difference is probably due to the Black/White differences)

This could be an error with the Canon internal patch generator, the video input path from the IS-mini, or the IS-mini itself.
You would have to test all variables with different sources to tell which is correct.

Steve

But the ISMini is driven USB from LightSpace and then SDi into the monitor; you'd think numbers are numbers?

No, numbers are not always numbers...
And the results you have tend to prove that.
It looks like there is an issue somewhere... and that could be anywhere within the chaing from the IS-mini into the Canon display.

Steve

Sorry Steve; I'm not trying to be dense! But it's an SDi cable from the ISMini into the Canon - everything else remains the same; I haven't recalled a different preset or anything like that.
Wouldn't you expect the pixel-values Lightspace sends to the ISMini to be the ones that travel down the cable into the BNC and (hopefully, although on the evidence maybe not?!) the values that Lightspace sends to the internal patch gen to match?

If all I have is the monitor, the ISMini, and Lightspace, what else could affect the pixel values?

If I have to investigate my ability to genarate patches what should I assume is at fault? Is it possible the ISMini is genarating patches of the wrong colour?

You need to assess if the display is responding the same to different inputs compared to its internal patch generator.
You then need to assess if different input sources result in different levels on the display.
Then you can define where this different is coming from - the difference you have already seen, so something is 'wrong' somewhere...

You cannot assume anything - you have to define exactly where the difference is being generated.
It could be anything, other than LightSpace, as that always just sends an RGB Triplet value that is always the same value.

Steve