Mixed Aspect Ratio Mini-Guide

[Mister Hatt]

AR calc is totally faster and more sugoi and everyone should use it, write their own alternative, or get their hands on MAGICAL apps that are even better and do exceedingly hax things.

[Cannonaire]

I have no hate for the AR calc; it is my preferred method. Except that it doesn't work if you are mixing 4:3 and anamorphic footage in the same video.

[Mister Hatt]

It totally does if you resort to dirty hacks like ordered chapters with different AR's. IIRC most windows players hate this though because they don't reset the decoder on each segment like mplayer-uau does. But multiple AR's for anything never really works properly so it's understandable that the AR calc doesn't know. You need to pick either 4:3 or 16:9 or whatever else and rescale or pad to it. THEN you can calc it.

[mirkosp]

It totally does if you resort to dirty hacks like ordered chapters with different AR's. IIRC most windows players hate this though because they don't reset the decoder on each segment like mplayer-uau does.

I recall Fluff saying that he didn't require it in his encode test for that and, IIRC, because even mkvmerge itself didn't support it properly. Might just be recalling wrong, though.

[Mister Hatt]

Why would it make a difference to mkvmerge? You're muxing each file separately. I think you are recalling it wrong. Fluff's test was also ridiculously hard for what it was meant to cover. Apparently writing a test that messy-chan lazied out of and never finished makes you good enough to skip it tho, he's so smart.

[Cannonaire]

I wouldn't want to split an AMV into tens or hundreds of tiny chunks and have the AR and possibly the window size changing every few frames, or at best every few seconds.

Though it might be a fun experiment just to see how the media players will respond to such a thing.

[mirkosp]

Though it might be a fun experiment just to see how the media players will respond to such a thing.

I wanted to especially use this idea to make an AMV around this concept. I had to drop the idea during the "test" phase because the player was utterly failing. Just make a flash animation or a bat that launches jpg images or something along those lines. Or box everything in a bigger video, although it won't have the same impact.