On reduser someone said the k-10 is as accurate as the Hubble but faster.

I assume that means the k-10 is a more accurate device, needing less measurements to get the same results.
In a lab you always do a series of measurements to reach a more accurate result.

I have never used a CMS or a probe at that level. When I look at the results of Eizo's colornavigator software using my dtp94b to calibrate my EIZO cg241w, I get deltaE values in the order of 10^-1. If the human eye can not detect 3 deltaE or less, I wonder what probe one really needs. Wouldn't it be sufficient to stay one order of magnitude below human perception? I am curious.

Nope, both are about the same - both are filter based - and have similar accuracy.
You need the same number of measurements for calibration regardless of the probe - this defines the accuracy of the LUT - lress samples less accuracy, regardless of the probe.

In our evaluations the K-10 was slower than the Hubble when measuring dark colours.
But both are good probes, and work well with LightSpace

If you want ultimate accuracy go for a PR-655, or Jeti 1211.

But for very good general calibration the i1 Display Pro is fine.

And a deltaE value means nothing for true calibration as it is a 1 dimensional value.
You need to compare the colour space in true 3D space to see the variations.

Interesting!

Of course its more complicated in 3d, but how would you grasp the eye's minimal sensitivity in more detailed way? Are there studies? I just wonder. I hope not being too persistent.

The 3D display within LightSpace shows the variation between the target colour space and the display. If the 3D display is a perfect cube the display is 100% accurate...

Simples

my thoughts went more in the direction: 10 point cube would be enough, so considering the eye, a uneven distribution of such points, more density in the skin tones, less in the blues? So any 3D-LUT generation would be based on a color space determined from experiments generated from a number of test people. And each point you could describe an error by a coordinate vector and a radius, as an equivalent to the deltaEs in 1D.

@probes: so what makes a probe faster? I meant whats going on inside, which is send to the CMS for that one measured point. Or is the CMS addressing the probe on a physical level?

Sorry to say I really do not understand most of that last post?

A LUT has no idea in advance of what colour will be within a given image...

As for probes - it's down to how long it take a given probe to reach a stable reading for a given colour/brightness. When it has a stable reading it tells the CMS it's happy. The CMS then presents the next patch.

this stable reading is the result of a series of measurements I talked about.

I was referring also to the wavelength dependency of the human eye to detect differences in hue, saturation and brightness, and was wondering how to put it into numbers in 3d compared to 1d.

Every human eye is different...

well, i assume it does, the same function as Hubble, as mentioned in the site kleininstruments[dot]com/k10.html

Klein PC Program. Create calibration files and set measuring parameters. Then data log, graph and export. The PC program manages the data in an easy to use interface.

Whether it's true or how to do it is entirely another thing, i have no idea whatsoever. I saw there are calibration presets in their software, so adding should be possible. Klein Instrument site is such a joke that i have to confirmed through several sources & people that they sell a good product before i even started asking price point.

My view on Hubble vs Klein: if it's about the same accuracy, but K-10 is faster (a lot faster), then i'll take K-10 any day, after a year the price difference is pretty much paid off.
The only thing that i can't confirm until now is if K-10 can be used for projection profiling.

When the K-10 is connected to LightSpace, and you open the Probe Option menu you will see any available setting options.
As I don't have a K-10 here right now I can't test, but we have a number of customers using it, so all is good.

As for speed, the minimum reading speed for the K-10 and Hubble per profile patch is 3 seconds - identical. so no benefit with the K-10.

Hi all nice to join the forum...
I always have the same doubt with "destination" and "source" doing the lot.
Today i´m profiling a monitor with the 17cube6sec24fps del provided and conform on resolve, i have previously set up the monitor with the calibration patch Bright-contrast- set to120nits-and WB rec709.
So on now i have to make a lot for Rec709, and another to print with a film profile i have.
Could some one explain me again the "destination" "source" with this example?
Oh and what about the gamut check, peak luma/peak chroma, should i turn it on or set something there?

Regards
Thanks for help

Hopefully the user manual does explain the answers to the questions - See here.

In general the 'Source' is the space the images you are looking at 'expect' to be seen in, so for a project to be film-recorded this would be the film profile, and the 'destination. space is the space you are working in - so the measured display profile.

Peak Chroma and Peak Luma match the selected parameters when match profiles, so use Peak Luma is the main setting to use to match displays to a pre-set colour spaces, etc.

Hope this helps.

Just used closed loop mode yesterday. So fast! Thanks for that little feature Steve.
It took me a minute to figure out how to get the java app to connect. One must hit "enter" on the connected app in the network manager. Der.