HP spectre x360 15-bl112dx LED type?

[Dennis]

Hi there!

Could anyone advise me on how to find out what LED technology is used? Specifically for my HP spectre x360 15-bl112dx, model late 2017/early 2018?

I read that most laptops/notebooks use White LED?

Thanks in advance!

Dennis

[Dennis]

And an additional question I hope someone can help me with:

I tend to prefer post-processing photos on a higher cd/m2, than the recommended value of 120. Is it ok to calibrate on a value quite a bit higher than 120 cd/m2? Let’s say, 300ish? I use a Spyder X. When I perform a pre-measurement in DisplayCAL, and adjust the brightness, I don’t notice a shift in RGB values. So I suppose calibrating on a brighter screen than 120 cd/m2, would not negatively influence a created profile?

Thanks for all the work!

Dennis

SpyderX Pro on Amazon  
Disclosure: As an Amazon Associate I earn from qualifying purchases.

[Vincent]

Measure native 255 green and 255 red, get their color coordinates (you can copy them here). There are several ways to do it, from a DisplayCAL default vendor profile verification to MS paint and ArgyllCMS spotread or something like that.
Even you may try to get them from display EDID data, or vendor display ICM without a measurement.

If they are far from sRGB green & red… then you do not have a “common” WLED display. When we use “WLED” name usually it is for blue led with yellow phosphor like in WLED IPS/TN/VA displays with a colorspace close to sRGB/Rec709 or smaller.

Since you own a SpyderX if you get a more saturated 255green/red coordinates than sRGB, use “Widegamut LED”, “W-LED PFS” correction or whatever name close to those Datacolor uses for this kind of measurement mode. In that situation do not use “Standard LED”, which means “sRGB-like WLED”.

[Dennis]

Thanks so much for your reply Vincent! Regarding your suggestion to measure the 255 green and red colour coordinates, how would I do this in DisplayCAL? I suppose I go to the “Verification” tab/menu, and what option do I select there?

I do notice a huge green colour cast when I previously calibrated the laptop with the ‘White LED’ and ‘Generic’ options, compared to my more accurately calibrated desktop (compared to for instance an Iphone, and my memory of how my processed photos looked in the past). So I’ll try to calibrate with the ‘LCD Phosphor WLED RGB LED’ option this evening. I’ll come back to it.

Thanks again! 🙂

Dennis

[Dennis]

Ok, back. I can’t seem to get a reasonable calibration performed on any LED technology setting, with the Spyder X. They all result in a too greenish colour cast, and one is too yellowish.  With the Spyder 4, I was able to have a succesful calibration result, matching other calibrated screens… Any suggestions are welcome. Thanks!

• This reply was modified 4 years, 9 months ago by Dennis.

[Vincent]

Ok, back. I can’t seem to get a reasonable calibration performed on any LED technology setting, with the Spyder X. They all result in a too greenish colour cast, and one is too yellowish.

There is a strong design limitation for SpyderX: AFAIK it does not store colorimeter spectral sensivities in its firmware, or if by chance they exist (and it looks like they do not exist) they are not readable by ArgyllCMS.
Hence “upgradeability” of SpyderX to some display types CANNOT be don in a generic way with an spectral power distribution sample, like those CCSS you see in DisplayCAL bundles for xrite i1d3 colorimeter, or CCSS in displaycal colorimeer database. You are limited to SpyderX display modes, a closed not upgradeable list.

The only chance to correct a SpyderX is like other colorimeters: to buy/rent a good spectrophotometer and build the typical 3×3 correction matrix, which is not portable between devices, so you must make a sutom matrix for YOUR unit and YOUR display.
Also keep in mind that some new backlights like those WLED PFS (newer widegamut LEDs) require very good spectral resolution or you won’t see some things, so if you end using a Xrite spectrophotometer, use High res mode @3nm.

Your other alternative is choosing whitepoint in a visual way (like in DisplayCAL visual whitepoint editor), and trust that RGB primaries coordinates are measured in a sufficient accurate way for profiling.

[Florian Höch]

Could anyone advise me on how to find out what LED technology is used? Specifically for my HP spectre x360 15-bl112dx, model late 2017/early 2018?

These laptops seem to use a range of different panels, the early 2017 model white LED (no PFS) according to a review. The newer 2019 model is available in an AMOLED variant (this should be easy to determine though, because with AMOLED you get perfect zero black level and quasi-infinite contrast).

When in doubt, just look at the gamut coverage and volume after doing a quick profile without calibration. Anything with around 70% AdobeRGB coverage or lower is likely white LED, anything with around 90% P3 likely PFS Phosphor.

There is a strong design limitation for SpyderX: AFAIK it does not store colorimeter spectral sensivities in its firmware, or if by chance they exist (and it looks like they do not exist) they are not readable by ArgyllCMS.
Hence “upgradeability” of SpyderX to some display types CANNOT be don in a generic way

Well, but a generic (in terms of not accounting for the spectral power distribution of the exact specific display in question, just like “generic” CCSS) correction matrix for the specific SpyderX can be created relatively easily even if no spectrometer is available, it just requires taking readings with the colorimeter and an already existing spectral power distribution of the display type in question (e.g. a CCSS file). The i1D3 is only cooking with water here, too, it just skips the “take readings with colorimeter” step and creates a 3×3 matrix from spectral data (CCSS) and internal spectral sensitivity on-the-fly.

@whentostoptime, bring up the colorimeter correction creation dialog (click the “plus” icon next to the correction dropdown) and do a measurement with the SpyderX. Attach the resulting TI3 file.

[Vincent]

 

There is a strong design limitation for SpyderX: AFAIK it does not store colorimeter spectral sensivities in its firmware, or if by chance they exist (and it looks like they do not exist) they are not readable by ArgyllCMS.
Hence “upgradeability” of SpyderX to some display types CANNOT be don in a generic way

Well, but a generic (in terms of not accounting for the spectral power distribution of the exact specific display in question, just like “generic” CCSS) correction matrix for the specific SpyderX can be created relatively easily even if no spectrometer is available, it just requires taking readings with the colorimeter and an already existing spectral power distribution of the display type in question (e.g. a CCSS file). The i1D3 is only cooking with water here, too, it just skips the “take readings with colorimeter” step and creates a 3×3 matrix from spectral data (CCSS) and internal spectral sensitivity on-the-fly.

i1d3 RGB to XYZ matrix computation based on CCSS in ArgyllCMS uses for reference observer and for colorimeter observer the same SPD sample (AFAIK, maybe I miss something).
So it’s like using the same “virtual display” with two observers like you would do with an spectrophotometer and a colorimeter measuring the same display in the same comfiguration.
Your suggested approach would need a  whitepoint adjustment (W = k1*R+k2*G+k3*B if we assume non WOLED displays) if you want to use different spectral samples as references (CCSS and actual SPD that arrives at colorimeter sensor)
Otherwise it would be like measuring different displays, for example a community CCCS could be greenish 6700K and user display could be close to D50.
Although is possible to rewrite a CCSS with a desired WP (W’=R+G+B, R’=k1*R, G’=k2*G, B’=k3*B ), that WP coordinates you choose as target (the ones measured by colorimeter)  would be wrong since you are using an unreliable source (uncorrected Spyder, uncorreted i1d3… etc) for that transformation.

So I see no point doing that correction for a SpyderX, i1d2, Spyder3… and all those devices without spectral sensivities stored & readeable in firmware.
You’ll need an spectrophotometer to take actual SPD of that display so you can calculate a transformation from uncorrected XYZ of the same SPD (if we ignore spectrophotometer limitations) but with “not so perfect observer” to correct XYZ with your desired observer.

Maybe I miss something.

• This reply was modified 4 years, 9 months ago by Vincent.
• This reply was modified 4 years, 9 months ago by Vincent.

[Florian Höch]

You’re right, whitepoint equalization (i.e. adapt both sets of XYZ to a common whitepoint) would be needed in this case before creating a correction matrix (and also probably ignoring white when creating the correction). If and how much (or not) the resulting correction then improves colorimeter accuracy would be a point of interest.

[Dennis]

Hi guys! Thanks for all the help so far. Most of the discussion goes beyond my knowledge in this area, but I do appreciate you guys thinking with/for me.

Florian, hereby I attached the TI3 file.

Also, the screen is probably white LED. With quick profiling (for white LED) I got the following percentages:

Gamut coverage:
87,5% sRGB
63,8% Adobe RGB
70,4% DCI P3

Gamut volume:
99,4% sRGB
68,5% Adobe RGB
70,4% DCI P3

Attachments:

You must be logged in to view attached files.

[Author]

Posts