1
\$\begingroup\$

I now own LG 4K OLED TV (HDMI 2.0a) and LG 32UD99 4K monitor (DisplayPort 1.2) connected to nVidia GeForce 1070 card. The monitor has true 10 bit LCD panel and the video card also supports 10 bits output (enabled in GeForce control panel).

I want to get the maximum out of that setup to display photos to my friends on the TV as well as on the monitor. So I guess I need to go beyond sRGB?

My source images are Canon RAWs (5D Mark II) and 16 bits film scans from Nikon Scan 9000 (it embeds sRGB profile). I do all of the processing in Capture 1. Both the TV and the monitor support HDR10 but I'm well aware that HDR only matters for video. However, I'd like to enjoy 1:20000 static contrast ratio of my OLED somehow. I'm OK if the images processed for my TV and monitor won't be displayed correctly on other systems.

What would be the optimal workflow for that?

Update

My calibrator (i1Display Pro) is on its way, though I don't understand how it would fit into that imagined workflow. As for wide gamut, I don't think it's applicable for TVs, but my guess is both OLED and the monitos's true 10-bits panels guarantee much better colors than sRGB.

\$\endgroup\$
2
  • 2
    \$\begingroup\$ You don’t mention calibration. Unless you calibrate the monitors all bets are off. \$\endgroup\$ Commented May 6, 2018 at 18:17
  • 1
    \$\begingroup\$ Are both displays wide gamut (as in >100% sRGB)? Because high bit depth != wide gamut. \$\endgroup\$
    – flolilo
    Commented May 6, 2018 at 18:51

1 Answer 1

1
\$\begingroup\$

Those displays seem to cover (most of) the DCI-P3 colour space, which is wider than the sRGB colour space. So at the very least, your editing should be done in a colour space that covers all of DCI-P3, and preferably at 16 bits/channel.

Also, the final image should be in a colour space that is at least as large as DCI-P3, and that colour space should be in the metadata (most programs assume sRGB if no colour space information is present).

And the final image should have at least 10 bits/channel to get full profit from a 10-bit display. So that means you shouldn't export to JPG ... But do test to make sure you do see a difference between 8-bit jpeg and say 16-bit png, and test it on a range of images with different colours and different saturations.

Calibration and profiling should have no impact on your editing workflow, which is ideally device-independent*.

Normally, calibration and profiling are done once every so often. Iirc, for LCD-type displays once per month is recommended, but I've also seen intervals of 2 or 3 months. But that depends also on how much time you want to spend on it and how precise you need/want your colours to be (some professionals need very accurate colour, I'd say most amateurs don't, but may of course want high accuracy).

Calibration consists of using the display's settings to get as close as possible to the wanted output without any display profile active. This minimises the corrections to be applied by that profile, and ensures you have the optimal colour resolution. It also sets the black and white points for the display. The profiling gives you a device profile that will be applied on display of the image (it should not be baked in the images). Ideally, it corrects for the remaining deviations in the display, so that the image as shown corresponds exactly to what the rgb values say it should be (given a defined colour space).
For a computer display, it should be enough to set the device profile in the operating system (it's even possible to use different profiles for each screen in a multiscreen setup); for a TV it might be less simple (unless the TV is seen as a display by a computer). In that case, you have to rely on the calibration only.

Oled and 10 bits/channel in themselves don't give you any larger colour space, by the way, that depends also on the rest of the display's hardware.


* In practice, you'll have to take into account whether the image is for screen display (additive colour, or light emission) or print (usually CMYK, and subtractive colour or light absorption).

Standard jpeg only handles 8 bit/channel, newer versions like jpeg2000 or jpeg XR can handle larger bit depths, but are less common at the moment.

\$\endgroup\$
4
  • \$\begingroup\$ In C1, JPEG 2000 & JPEG XR have options for 16 bit. PNG, however, does not. \$\endgroup\$
    – flolilo
    Commented May 7, 2018 at 10:58
  • \$\begingroup\$ @flolilolilo I don't get the "In C1" part of your comment, but I've been saving 16 bit/channel pngs for several years now. I edited my answer to include the other jpeg formats. \$\endgroup\$
    – remco
    Commented May 7, 2018 at 11:51
  • \$\begingroup\$ OP stated I do all of the processing in Capture 1. - C1 (Capture One) only can process PNG in 8 bit. It's just meant to be an extra information, not a "you're wrong, remco!"-type of statement. ;-) \$\endgroup\$
    – flolilo
    Commented May 7, 2018 at 13:01
  • \$\begingroup\$ OK, I see. I never used Capture One, so I wasn't aware of that quirk. Thanks for the information :) \$\endgroup\$
    – remco
    Commented May 9, 2018 at 5:57

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.