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

The sad thing about Popcorn Time is the way it shows how little people actually care about AV quality. One of the biggest reasons why I personally resort to illegal options (especially with video material) is because of the higher quality they offer compared to legal alternatives, but this aspect is not present in Popcorn Time at all since it sources all the video and audio from YIFY.

YIFY is basically a bunch of morons producing nothing but total garbage who would be better off encoding things in SD with the low bitrates they use. Sadly, since their fork is likely going to be considered the "main" one for Popcorn Time, decent AV quality will probably never be a thing with it. (I found a couple issues on various GitHub forks about this subject and ran into this image[1] which demonstrates the problem quite well.)

EDIT: Well, I guess there's some hope for the future[2]. Not holding my breath, though.

[1] https://f.cloud.github.com/assets/1736009/2426834/15aa358e-a...

[2] https://github.com/Yify/popcorn-app/issues/51




> The sad thing about Popcorn Time is the way it shows how little people actually care about AV quality.

What's sad about the fact that different consumers value different things?


In the various discussions about Popcorn Time here and elsewhere, the technical quality of the content has been the subject of discussion multiple times, along with mentions of illegal alternatives offering higher quality than legal ones. By sourcing from YIFY, we're talking about the opposite, though - legal options like Netflix will most definitely offer higher video/audio quality than Popcorn Time.

It's certainly true that different consumers value different things, though. Which is why it's sad that with Popcorn Time my only choices are shit quality or nothing.


>who would be better off encoding things in SD with the low bitrates they use

DCT quantization always beats down-scaling at the same bit rate.


Look at the image I linked and say that again.

And it's not like this is particularly hard to test on your own. If you encode the same video at low enough bitrate X in 480p/720p/1080p, the higher resolution versions can have more definition at times but the 480p version will generally end up having the highest overall quality and is usually the most consistent in its quality as well. I've done this kind of comparing a couple times myself and the results are exactly that.


Comparing still frames from VBR video streams is spurious on multiple levels.


Only if you have a really crummy upscaler on playback; with a good upscaler, DCT starts losing at moderate bitrates and starts to get terrible at low bitrates.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

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

Search: