Different result spectral and matrix correction

[chris-to-fer]

Hey,

I use an i1 pro to correct an iDisplay pro. In the interactive adjustment I can see already the difference for white between spectral and matricx correction and when I do a verification report with sim. profile as display profile just with the different hardware settings from interactive adjustment with the spectral correction my display is a bit too far off in blue (DE2-3) and with the matrix it’s even more off in blue (DE5), measured with or i1 pro or i1Display pro. With i1Display pro with no correction red is off (DE 3-4) but not as much as blue before.

So is one of my probes that bad? Is it normal to see that difference between spectral and matrix correction? If I chose spectral it locks to refresh, but it’s a a-Si TFT Active Matrix LCD?

Thanks

Chris

Calibrite Display Pro HL on Amazon  
Disclosure: As an Amazon Associate I earn from qualifying purchases.

[Vincent]

A matrix correction is used to match the measuremenst from one device to another. It does not need that corrected value (same or very close to reference measured value) is accurate, just that it matches reference value closely.
Measurements from an xrite spectro can be not very accurate in 10nm mode if measured spectral distribution has very narrow spikes.

Spectral correction uses colorimeter firmware data with its sensivity curves to correct its mismach with standard observer while measuring a sample spectral power distribution (CCSS)

So they are doing different things.

For not very demanding spectral power distribution like sRGB LEDs, or GB-LEDs or QLEDs or TV WOLEDS, an i1pro 3nm readings and i1displaypro readings with CCSS 3nm CCSS (or some of the Xrite reference 1nm samples) should be close to each other.
For some new backlights even 3nm readings can lack of resolution, like blue laser projectors or some new WLED PFS backlights.

[chris-to-fer]

Thanks. I understand the difference now. I have no idea about what nm number is relevant for me, how can i find that out or understand that better? Nonetheless I wonder what i can do now to find out what is better? Just colorimeter without correction with spectral correction or with matrix to spectro? I actually only want to do some hardware correction over interactive menu. The monitor has RBG gain and bias controlls. Just want to get those right for Rec 709 d65. Shall I only use i1pro?

[Vincent]

Use 3nm by default, plot spectral power distribution, look for potential “averaged” narrow spikes. There are lot of examples in forum, like those P3 multimedia displays WLED PFS and their 1nm spectral corrections bundled with displayCAL. “(i)” button next to correction combo box will plot them. Choose of these 1nm bundled ones for WLED PFS and see hot they look like, or WG CCFL for old widegamuts like PA271W.

3nm readings & colorimeter corrected values should be close to each other. I would choose spectral and share it with community.

[Author]

Posts