Comment by ijiiijji1 | Hacker News Reader

Comment by ijiiijji1 | Hacker News Reader

Parent Full thread

The resolutions of 4K/5K/8K with completely lossless at high frame rates or at native sensor resolution in terms of storage for hundreds/thousands of hours of source material for marginal audiophile-like puritanism/placebo magical beliefs is just way too expensive for essentially no marginal return, because there are fundamental limits to psychovisual and psychoacoustic thresholds that no human can discern in a given playback venue (different for large IMAX, normal theater, home theater and small phablet; and different to a degree depending on an individual's would-be double-blinded or bio-electrically/fMRI measured thresholds.) The goal of pseudo-"lossless" compression is to be very conservative on what to throw out because transcoding to other formats can always (and must) optimize (throw away) data later... it really is marketing wank because there is some compression, but the idea is it's aimed to be minimal as to not matter in the final cut product.

Yes, for customers, I agree.

For VFX, no.

That extra resolution, and crucially colour dynamic range means that less time and money had to be spent cleaning up green screen masks.

It also means that there is more scope to do things like Day-for-night, change the colour temperature, and have full artistic freedom to change the look and feel of a movie without reshooting.

It's a bit more nuanced than that. Resolution is meaningless unless you know the size of the image and the distance to the viewer - so keeping a "lossless" copy makes sense for people professionally buying very expensive hardware. It's no different than keeping the original reels of a movie so you can scan them at higher and higher resolution as technology improves. In other word, the final product will always at best be limited by the format the source is kept in. That is not to say that even amateur videophiles have a use for it, they just want the best and mimic the professionals.

As for frame rate I'm pretty sure we know from experiments that the average human eye can see the difference in frame rate all the way up to 150-160 fps, and that the trained eye can detect images shown for less than 1/200 second.

So however marginal the gain is, it makes sense for professionals that take pride in having the absolutely best quality. It's the same people that will finetune encoder settings for the individual scenes in a movie just to get that tiny improvement.

Reply on news.ycombinator.com