[T800]

Anamorphic means non square pixels. Why not just encode it at the aspect ratio it is supposed to be and allow the player to get it right. ie encode 16:9 at 1920x1080 and 2.35:1 at 1920x800 or x816. Would save masses of bitrate.

Why would it? It's just black bars.




From what I have read in the past black bars take up very little file size on a blu-ray 1080p image.

If you're re-encoding then it will be of better quality if you remove the black bars because they add noise to the image during the re-encoding process.

Blu-ray sticks to standard videos resolutions. 1920x800 isn't a standard video resolution guaranteed to be supported by all devices.

[nickr]

It's not the black bars, it's the sudden transition from black to [whatever colour the movie is] that takes a lot of bandwidth.

[T800]

What does that have to do with cropping black bars? That would happen in the image regardless.

[nickr]

No, you are not getting what I am saying and you are wrong, but I am off to bed and not going to explain again the problem with encoding a sudden transition in a video file.

[T800]

You say I'm wrong but your brief statement reveals the opposite to me.

[ertman]

Another question: Why don't studios encode a 2.35:1 (or any other aspect ratio with black bars) anamorphically? And then set a flag to have it play in the correct aspect ratio. Widescreen DVDs did this right? Wouldn't it be beneficial in the long run so when 4k goes mainstream, you'd have a greater picture (more resolution to be upscaled=better clarity?)

Personally, I reencode because it cuts the file size of my digitally filmed films in half or slightly more (32GB to 9-14ish GB)

Anamorphic means non square pixels. Why not just encode it at the aspect ratio it is supposed to be and allow the player to get it right. ie encode 16:9 at 1920x1080 and 2.35:1 at 1920x800 or x816. Would save masses of bitrate.

Also wouldn't anamorphic mastering of blurays only be useful to those with 21:9 screens? Or am I missing the some of the points of anamorphic resolution?

[magao]

What does that have to do with cropping black bars? That would happen in the image regardless.

If the black bars are there then you have to encode the transition from the image to the black bars. To get a sharp transition requires a lot of bandwidth. If you don't give it enough you will have a fuzzy edge.

If you crop the black bars then they are not there and there is no transition - so you don't have to waste any bandwidth encoding the transition.

FWIW I always crop the black bars when re-encoding, and I re-encode for space - at CRF 20 I usually end up with a file at least 50% smaller and often quite a bit smaller than that, with no perceptible difference in quality at normal viewing distances.

The only time I've seen an issue with cropping the bars was when I was testing out encoding using QuickSync using Handbrake. When the bars were cropped, QuickSync got the aspect ratio completely wrong - it stretched the 1920x800 -> 1920x1080. Maybe that bug has been fixed since I tried it, but the quality reduction by using QuickSync was way too great for me to use it.

[nickr]

You say I'm wrong but your brief statement reveals the opposite to me.

How about you replace your arrogance with research. Although this link refers to xvid, the theory is the same.
http://www.mplayerhq.hu/DOCS/HTML/en/men...mpeg4.html

[Prof Yaffle]

I think the simplest way to look at this is that the encoding process looks at a "field" of adjacent pixels and figures out what's changed between one frame and the next. The smaller these fields, and the more sensitive you want to be to changes, then the more processor-intensive becomes the encode and decode (playback) process. This is one of the reasons that H.264 HP can stress older processors.

Now, if one of these "fields" covers a group of pixels at the edge of the image, you've got half of it that's all black (because it's in the border) and half that's constantly changing (because it's in the picture). In the middle, you have a very distinct line - the edge itself. The point being made is that it's intensive to encode this accurately (and not get any "bleeding" into the black bit) versus encoding right up to the edge of the image and then allowing the playback software to "draw in" the black bars as required.

Vastly over-simplified, because I'm a vastly simple person - but that's the gist of it as I understand it all (and without an I-frame in sight).

(As an aside, is the conversation moot if the macroblock size coincides precisely with the border, i.e. if one block meets the next one exactly at the boundary between bar and image? Or is this just very unlikely unless forced, and then still likely to fall victim to variable-sizing somewhere in the process...).

[T800]

You say I'm wrong but your brief statement reveals the opposite to me.

How about you replace your arrogance with research.

No reading required.
You are correct I didn't understand what you were getting at. I do now and I had already said:

If you're re-encoding then it will be of better quality if you remove the black bars because they add noise to the image during the re-encoding process.

They add noise because of the transition. We were both on about the same thing.

Sounds like you didn't understand where I was coming from in the first place.

[pulsewave]

I can't imagine why anyone wouldn't want to remove the black bars, if you don't and then play the file on anything that is wider than 16:9 then you will have ugly black bars top and bottom, this is what happened on Youtube when people encoded 16:9 content to 4:3 with black bars, sure it was fine when watching on a 4:3 screen or in the 4:3 player but is an annoying pain when everything moved to 16:9.

One word of warning though, some movies most of the film is in 2.35:1 but then have some scenes at 16:9, on those movies you don't want any cropping.

[voochi]

The fact that bluray discs are encoded with black bars is massively retarded and is just one of many examples of how bluray is a a garbage media format/container/whatever.

LOL.

You know nothing about designing a stable, commercial format.

Restrictions are necessary. If you had a resolution free-for-all then ensuring compliancy would be a nightmare.

[Piers]

It's pretty simple, if a movie is filmed in 2.35:1 and you want to 'rip' it then it's obvious you crop the black bars to save on size. As it's been pointed out, the black bars and border between content and dead space require high rates of data to be used.

[Ned Scott]

The fact that bluray discs are encoded with black bars is massively retarded and is just one of many examples of how bluray is a a garbage media format/container/whatever.

LOL.

You know nothing about designing a stable, commercial format.

Restrictions are necessary. If you had a resolution free-for-all then ensuring compliancy would be a nightmare.

That must be why XBMC has had zero issues with multiple resolutions for almost 10 years. Or even DVD players that have DIVX playback support. Or even DVDs that used anamorphic encoding and no back bars at all.

Do I personally know how to design a stable, commercial format? No, but I also don't know how to direct a movie, but I can tell you when one sucks. I can also tell you what other experts have said about bluray, people who do know how to make a format. Including people who had a hand in designing bluray.

A lot of requests/requirements came down from execs who didn't know what they were talking about, or wanted to prioritize silly "wow" features over core functionality (like online connected content).

I'll admit, calling it garbage is a bit hyperbolical, but there really are a some negatives about the format.

[Bonzi]

Well I keep the black bars because I'm extremely lazy and I don't feel its worth my time. Matter a fact I just rip to mkv and store it. Hard drive space is so cheap...who cares if the movie is 15GB or 40GB..