Hacker News new | past | comments | ask | show | jobs | submit login

> Modern OLED displays are superior in every way and CRT aesthetics can be replicated in software, so a more practical route would be probably to build some "pass-through" device that adds shadow mask, color bleed, and what-have-you.

OLEDs are still behind on motion clarity, but getting close. We finally have 480 Hz OLEDs, and seem to be on track to the 1000Hz needed to match CRTs.

The Retrotink 4k also exists as a standalone box to emulate CRTs and is really great. The main problem being it's HDMI 2.0 output, so you need to choose between 4k60 output with better resolution to emulate CRT masks/scan lines, or 1440p120 for better motion clarity.

Something 4k500 or 4k1000 is likely needed to really replace CRTs completely.

Really hoping by the time 1000 Hz displays are common we do end up with some pass-through box that can fully emulate everything. Emulating full rolling CRT gun scan out should be possible at that refresh rate, which would be amazing.






1000Hz is enough to match CRT quality on a sample-and-hold display, but only when you're displaying 1000fps content. A great many games are limited to 60fps, which means you'll need to either interpolate motion, which adds latency and artifacts, or insert black frames (or better, black lines for a rolling scan, which avoids the latency penalty), which reduces brightness. Adding 16 black frames between every image frame is probably going to reduce brightness to unacceptable levels.

How many nits of brightness did high end CRTs reach?

The brightest CRTs were those used in CRT projectors. These had the advantage of using three separate monochrome tubes, which meant the whole screen could be coated in phosphor without any gaps, and they were often liquid cooled.

Direct-view color CRTs topped out at about 300 nits, which is IMO plenty for non-HDR content.


Why stop there? We can simulate the phosphor activation by the electron beam quite accurately with 5 million FPS or so.

And the difference between 480 and 1000 Hz is perceptible?

For smooth and fast motion, yes. Although I don't have such fast displays for testing, you can simulate the effect of sample-and-hold blur by applying linear motion blur in a linear color space. A static image (e.g. the sample-and-hold frame) with moving eyeballs (as in smooth pursuit eye tracking) looks identical to a moving image with static eyeballs, and the linear motion blur effect gives a good approximation of that moving image.



Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: