KALK4L wrote: ↑13 Apr 2023, 23:22
I have two questions that I'd appreciate an answer for:
- Do OLED monitors have per pixel illumination in SDR mode?
If no, then how can they achieve those deep blacks and high contrast in SDR mode?
- Why OLED monitors' max brightness is always higher in SDR than in HDR mode when both are at 100% APL (Displaying full white screen)? For example the Asus OLED PG27AQDM has a max brightness of ~250 nits in SDR mode and a max brightness of ~200 nits in HDR mode when both displaying the same full white content. I understand why the same monitor would be brighter in HDR mode when at less than 100% APL since the monitor can concentrate more of it's power illuminating a smaller area, but it's confusing to me why HDR is worse than SDR when both are illuminating the same amount of pixels at 100% APL
1. Yes, OLED always do per-pixel illumination in all modes, this applies to all OLED's I've ever seen, including those OLEDs on phones and televisions by all brands. Did anybody strangely tell you otherwise?
2. I can't say for sure, as I don't work with those firmware engineers.
For the 100% APL thing -- there are several algorithms at play with completely different coding paths. Now computer programming the code for the firmwares by their software developers.... Programming OLED SDR and OLED HDR is sometimes like programming firmwares for two completely separate computer monitors, and they sometimes have very different behaviors. There's more commonalities if you program to an LCD, but for the OLEDs, due to APL and due to HDR, you pretty much have much more separate code for SDR and for HDR respectively. Could be a bug where they didn't port a tweak from one of the modes (SDR) to the other mode (HDR). But it may not be a bug. It's giants amount of separate code paths, in the scaler. Sometime code changes for one isn't completely ported over to the other. There's a large number of very analog-like floating point mathematics involved with auto-APL HDR, especially with video signals containing floating point HDR pixels. It's not as easy 24-bit RGB, and then exponentially difficult to program the firmwares when APL is involved. Now you have the white subpixel, which can sometimes be brightened further by also illuminating R,G,B at the same time. There might be a bug factor involved. There might be a power budget factor involved. The power budget available when a picture suddenly switches from extremely bright mixed HDR, suddenly to a 100% APL white image, then suddenly bright mixed HDR in the next frame (e.g. scene change effect, or explosion flash effect) -- may (in theory) create power transients bigger to the power supply (or internal VRMs) than when you're already capping the pixel brightnesses in SDR mode. An early prototype of an OLED monitor I got, kept shutting down when the picture suddenly changed when the whole screen was very bright, so there's kind of a possible power-surge effect between frame changes in certain weird conditions. I'm only speculating based on observation on that lineitem. But it could very well be behaviors of different coding paths as SDR and HDR firmwares is like two completely separate scaler firmwares, in a manner of speaking. In summary, integer-based SDR pixels and floating-point-based HDR pixels can create very different quirks, possibly intentional (aka power transients from refreshing HDR gamut during very fast APL changes), or possibly unintentional (aka a bug!). Either way, many possible reasons -- and I am not sure if bug or feature caused by the separate coding paths.
I'm going to edit the title of the post to better describe its content.