WLED instrument mode for Spyder 4/5 is off

[p.dada]

Hello Florian,

I just took 11k measurements for a madVR 3D LUT, only to realise that HCFR wouldn’t agree with DisplayCAL’s readings. There’s something wrong with WLED instrument mode for Spyder 4/5 (other modes are fine). HCFR and OEM software agree for all display types.  I confirmed this on a second PC, as well.

Please check the attachment.

Attachments:

You must be logged in to view attached files.

[Florian Höch]

Hi,

How did you get the xyY readings? DisplayCAL typically only provides these during interactive display adjustment, when there is no videoLUT active. You can use Argyll’s spotread to check measurement consistency, i.e.

spotread -v -e -ye -x

The difference in the attachment above is around ~2 dE 2k for the WLED column in both cases. Was the instrument in the same spot for all measurements?

[p.dada]

Yes, the instrument was in the same spot for all measurements.

I just used the “Report on calibrated/uncalibrated device” function under the tools menu and took note of the (rather cool) white point coordinates of a WLED panel. I repeated the measurement for all available instrument modes. Then I measured the 100% white in HCFR for every display type. Finally, I used the CheckCAL function in the OEM software without any curves loaded in the video LUT and took note of the reported white point in the results page.

I repeated the above procedure on a different PC with a CCFL panel (again using all available display type corrections) and the results were consistent.

Because:
1) OEM software and HCFR agree and
2) HCFR uses ArgyllCMS
I concluded that there is an error in DisplayCAL’s interpretation of the readings. But only for the WLED mode; other modes seem fine.

0.008 deviation in y axis is enough to give ΔΕ >4 in HCFR…

• This reply was modified 8 years, 1 month ago by p.dada.
• This reply was modified 8 years, 1 month ago by p.dada.

[p.dada]

I found the problem. You are supposed to use the switch “-y 2” with spotread, not “-y e”.

-y n Spyder4: Generic Non-Refresh Display [Default,CB1]
r|c Spyder4: Generic Refresh Display [CB2]
f Spyder4: LCD, CCFL Backlight
1 Spyder4: Wide Gamut LCD, CCFL Backlight
2 Spyder4: LCD, White LED Backlight
B Spyder4: Wide Gamut LCD, RGB LED Backlight
x Spyder4: LCD, CCFL Backlight (Laptop ?)
3 Spyder4: CRT (Hitachi CM2112MET, Diamond View 1772ie)
l Spyder4: LCD CCFL IPS (CCFL AC EIZO HP with CORRECTION)
L Spyder4: LCD CCFL Wide Gamut IPS (WG CCFL NEC241 271)
4 Spyder4: LCD RG Phosphor (Addition of Dell Ultrasharp U2413 to the RG-Phosphor Family EDR (created originall
b Spyder4: LCD RGB LED IPS (RGBLED HP SOYO)
e Spyder4: LCD White LED IPS (WLED AC LG Samsung)
o Spyder4: LED OLED (EDR for OLED created with primaries R, G, B, White)
m Spyder4: Plasma (EDR for Plasma created with primaries R, G, B and White)
p Spyder4: Projector (Marantz HP Panasonic Projectors Hybrid EDR)

Should be a quick fix.

(Btw I’m the same guy that helped you realise that Windows calibration loading from the vcgt tag sucked big time and DispcalGUI was doing a much smoother job. I re-registered with a new nick.)

[Florian Höch]

Seems like something is wrong with your system. Spotread output on mine (which matches expectations from http://www.argyllcms.com/doc/instruments.html#spyd4 which is the same for Spyder4/5)

spotread -?
Measure spot values, Version 1.8.3
...
-y n Spyder5: Generic Non-Refresh Display [Default,CB1]
r Spyder5: Generic Refresh Display [CB2]
f Spyder5: LCD, CCFL Backlight
L Spyder5: Wide Gamut LCD, CCFL Backlight
e Spyder5: LCD, White LED Backlight
B Spyder5: Wide Gamut LCD, RGB LED Backlight
x Spyder5: LCD, CCFL Backlight (Laptop ?)
c Spyder5: CRT (Hitachi CM2112MET, Diamond View 1772ie)
l Spyder5: LCD CCFL IPS (CCFL AC EIZO HP with CORRECTION)
1 Spyder5: LCD CCFL Wide Gamut IPS (WG CCFL NEC241 271)
2 Spyder5: LCD RG Phosphor (Addition of Dell Ultrasharp U2413 to the RG-Phosphor Family EDR (created originall
3 Spyder5: LCD RGB LED IPS (RGBLED HP SOYO)
4 Spyder5: LCD White LED IPS (WLED AC LG Samsung)
5 Spyder5: LED OLED (EDR for OLED created with primaries R, G, B, White)
6 Spyder5: Plasma (EDR for Plasma created with primaries R, G, B and White)
7 Spyder5: Projector (Marantz HP Panasonic Projectors Hybrid EDR)

The letters n, r, f, L, e, B, x are the default measurement modes for the Spyder4/5 (they are exactly the same for both instruments) as per http://www.argyllcms.com/doc/instruments.html#spyd4

When you have the X-Rite *.ccss files also installed, you get the list above (c, l and 1-7 are coming from CCSS in my case, but note that the mapping of letters/numbers to CCSS files for these is not guaranteed because it depends on the files installed!). You should probably nuke all *.ccss files in %APPDATA%\ArgyllCMS and %APPDATA%\color (the latter are installed by HCFR) and re-import them.

• This reply was modified 8 years, 1 month ago by Florian Höch.

[p.dada]

Veeery interesting.

Not sure what I did here, but I deleted spyd4cal.bin from %appdata%\ArgyllCMS and every CCSS from %appdata%\color and now I am getting consistent readings across all software and the correct spotread output.

 -y n|l Spyder4: Generic Non-Refresh Display [Default,CB1]
r|c Spyder4: Generic Refresh Display [CB2]
f Spyder4: LCD, CCFL Backlight
L Spyder4: Wide Gamut LCD, CCFL Backlight
e Spyder4: LCD, White LED Backlight
B Spyder4: Wide Gamut LCD, RGB LED Backlight
x Spyder4: LCD, CCFL Backlight (Laptop ?)

But I lost the rest modes (plasma, projector etc). I’m not sure if I should have those installed in the first place. They must have come from HCFR.

The interesting part is that I managed to screw this on both systems. Btw after I reimported the corrections, neither spyd4cal.bin nor the CCSS files in color reappeared. I chose system-wide installation. Any clue where they must be? Nvm, I reimported them, chosing “user-only” mode and spyd4cal.bin was recreated.

• This reply was modified 8 years, 1 month ago by p.dada.

[Author]

Posts