How can I create such a high quality video on YOUUBE?

[Stears555]

My HD TV records doesn't look so good after upload....
(watch in 1080p)

[noemi7]

do you need?

[jagabo]

Filter all the noise out of your video. Noise sucks up bitrate.

[netmask56]

Filter all the noise out of your video. Noise sucks up bitrate.

How would he do this assuming no prior knowledge of the appropriate techniques?

[jagabo]

Use strong temporal and spacial noise filters found in most video editors. And of course, shooting low noise video in the first place is a big help.

Also, whenever possible use completely static backgrounds or strong depth-of-field to get very blurry backgrounds.

[Mephesto]

What's preferable between a noisy and noiseless video at the same bitrate? The noisy video will have unintentional textures from the compression artifacts but the edges will be inferior even with Psy RDO enabled, and the noiseless video will have its oversmoothing of details exasperated by the compression and some distortion on the edges visible which are masked in the other video by the noise.

I prefer the latter but sometimes when I wasn't satisfied with some videos whose nosie couldn't be removed without oversmoothing, I encoded them with x264 with an intentionally lower bitrate than normal because I knew the compression would obliterate the noise anyway. The result to me looked better than encoding the denoised version at the same bitrate.

[bigass]

Sure, but would you ever intentionally add noise to a clean video to achieve some kind of encoding goal?

[pandy]

@ffmpeg -i %1 -vf "scale=iw*min(4096/iw\,2304/ih):ih*min(4096/iw\,2304/ih),pad=4096:2304:(4096-iw)/2:(2304-ih)/2" -c:v libx264 -preset slow -crf 18 -c:a libvorbis -q:a 5 output.mp4

[sanlyn]

Well, whether it's noiseless or noisy or oversmoothed or not, the techniques used here are for an A.D.D. crowd that doesn't see a scene long enough or focus on anything intently enough to see differences or problems, at least not until the vid's so bad that it completely falls apart. Do you think anyone would notice the chroma problems in the reprocessed/replay shots with the girl? I doubt it. With a more static video, such as two people talking where the camera stays focused on one shot for more than 5 seconds, it might matter. But since A.D.D. victims would start switching channels if anything lasted for 5 seconds, talk about subtle differences and refinements is moot. If you think I'm joking, you haven't been paying attention to what's happening on TV or in the video sample in post #1.

As for "not looking so good after upload", remember that you're looking at the results of UTube's El Crappo Destructo Re-Processor and La DooDooView Player. What do you expect? Have you just now noticed this about uTube?

[Selur]

Sure, but would you ever intentionally add noise to a clean video to achieve some kind of encoding goal?

if your source if 10bit and your target is 8bit banding could become a problem if you don't add some noise through dithering, so dithering would be a valid example of adding noise 'noise to a clean video to achieve some kind of encoding goal'