POV-Ray : Newsgroups : povray.unix : library finding request : Re: library finding request Server Time
6 Oct 2024 10:37:43 EDT (-0400)
  Re: library finding request  
From: Le Forgeron
Date: 16 Aug 2002 08:34:09
Message: <3D5CF14D.1020409@free.fr>
ABX wrote:

> On Fri, 16 Aug 2002 11:39:49 +0200, Le Forgeron <jgr### [at] freefr> wrote:
> 
>>Interesting concept: extracting a frame of a movie,
>>
> 
> I'm happy that first response is not completly agains ... :-)
> 
> 
>>but I wonder if it would not be better to have the timeline as the time 
>>elapsed since the beginning of the movie.
>>
> 
> There is nothing difficoult to use Interpolate macro or range function to move
> cursor from [0..1] to real time I think.


It's not difficult, BUT then, you have to exactly know the whole length
of the animation and put that knowledge in the script.
Why put the burden on the user when the code could do it.

> I think idea of 0..1 is simliar to
> putting images in <0,0>..<1,1> region even if images has some resolution,
> placing, dimensions already written in it format.


I'm not sure that the similiraty could hold for time as it can for 
spatial dimension. It's nice to be able to give the position in a 
picture by using percentage. Even if the resolution of the picture 
changes, the same element would still be there.
For movies, I'm afraid the resolution will not changed, but instead, 
most of the time, the duration of the movie will be extended (see all 
the DVD with Director'/Cameraman'/Script'/Producer'/Janitor's cuts, it's 
always the same movie with some additional scenes. I would find it 
easier to have a timeline which state in plain the time, with a comment 
which gives the expected scene such as:

  timeline 0:13:30 // Luke Skywalker enters his room with R2D2
(Ok, format of timeline is hard for the parser, but that's another story)

This way, whenever instead of SW-EP4v1(1978), someone want to use 
SW-EP4v2(198x) or even SW-EP4v3(199y) (until we get SW-EP4v4(20xx),
all he has to do is, using any player, to find that same scene, probably 
  around the original 0:13:30 .

With interpolate, he would have to guess the length of MY version in 
order to guess the time of the original, then perform also a precise 
measure of the full movie length (I have some DVD which have the bad 
authoring Idea of having absolutly NO time code information, which means 
that you cannot jump back twenty minutes after the start [because you 
previously look these twenty minutes two day before], the only way is 
fast-forward... BTW, the player cannot tell you it's twenty minutes 
away, you have to use the wall-clock!).

Side note: beware about the numbering scheme for frames inside a 
seconds, the usage in video software is to number each frame 
continiously from the first up (tagged 0) to the rate (tagged rate -1),
Thus there is 'rate' frames between 1:12:34 and 1:12:35, and they are
1:12:34.0, 1:12:34.1, 1:12:34.2, and so on.
The strange thing is that on a xx fps movie, the next after 1:12:34.xx-1 
is 1:12:35.0 (and not 1:12:34.xx)
Also, there is some fun with the NTSC setting (29,97 is real fun, best 
to considere it at 30 anyway, just droping a frame sometime to keep the 
sound in sync). Also pull-down (24 fps in 30 fps) is an interesting 
perversion.

> I think it should be easy to
> get specified moment of movie by writing in sdl:
> 
>   #macro Time(Hours,Minutes,Seconds,S100)
>     (((Hours*60+Minutes)*60+Seconds)*100+S100)
>   #end
> 


So far so good, but why limit the resolution to 1/100th of second.
#macro Time(Hours, Minutes, Seconds, Fraction)
   (((Hours*60+Minutes)*60+Seconds+Fraction)
#end
With Fraction between 0 and 1.


>   image_map{
>     mpeg "movie.mpg"
>     timeline Time(2,30,46,23)/Time(3,20,21,10)
>   }
> 


That's the part I do not like, having to know and code the total length 
of the movie.


> Making timeline dependant on clock and using motion_blur (I hope it will be
> patched to 3.5) would made movie playing very smooth.
> 
> 
>>the beginning of the files might not be at timecode 0.
>>
> 
> IIRC GIF format can have written position of image. Is it considered when gif
> is loaded as image_map?
> 


Gif support has been removed from 3.5, so that's irrelevant!
But I believe that only the first picture only was decoded...
I do not know, you will have to get the code of 3.1

Was partial image and animation part of the 87a Gif format ? or did they 
come later with the next Gif format format ?


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.