←back to thread

20 points kristianp | 2 comments | | HN request time: 0.4s | source
1. davidhyde ◴[] No.44392308[source]
> “ While testing was conducted at 60 FPS, the response times even fall short of this low bar, with 16.67 ms being the slowest response time required for the pixels to refresh between frames”

Well it’s actually 16.66 ms (1000 / 60). You can’t round up. A render loop that takes just a fraction more than the absolute minimum above would neck down to 30fps.

replies(1): >>44393859 #
2. addaon ◴[] No.44393859[source]
> Well it’s actually 16.66 ms (1000 / 60). You can’t round up. A render loop that takes just a fraction more than the absolute minimum above would neck down to 30fps.

This is about response time, not render loop time. Response time is the analog time required to transition a physical subpixel from one brightness level to another, usually measuring from when the transition starts. A 16.67 ms response time does indeed mean that with a 16.66(6) ms frame period it's still 0.003(3) ms away from completing the transition... but that just means that it's an immeasurable fraction away from the commanded brightness at that time, not that there will be rippling effects to the next frame or anything.