Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Capturing and Editing Video > Avisynth Usage

Reply
 
Thread Tools Search this Thread Display Modes
Old 21st November 2017, 20:41   #1  |  Link
jriker1
Registered User
 
Join Date: Dec 2003
Posts: 387
QTGMC double frame rate ok?

I have some interlaced video that I'm using AVISynth and QTGMC to make progressive. When I de-interlace with QTGMC it by default doubles my frame rate. So goes from 29.97 to 59.95. And larger file size of course at this initial conversion stage. Is it better to leave it at 59.95 or should I be setting it to throw out ever even frame or something in order to maintain the original frame rate? Any value one way or the other? Also does this effect the audio?

Thanks.

JR
jriker1 is offline   Reply With Quote
Old 21st November 2017, 21:13   #2  |  Link
johnmeyer
Registered User
 
Join Date: Feb 2002
Location: California
Posts: 2,090
If you have 29.97 interelaced, it has 59.94 "events" per second. This gives video its smooth feel. By contrast, 24p material has a "once-removed" feeling. Yes, I am comparing interlaced and progressive, but my point is that it is the number of unique events, whether fields or frames, that provides the smoothness that one associates with video, and not with film.

If you decimate the result of QTGMC back to 29.97 fps, so that you end up with 29.97 progressive, you are not far removed from 24p, and you will notice some judder on horizontal camera pans. This judder effect happens in your brain because you are on the threshold of the persistence of vision phenomena that creates the illusion of motion when viewing a series of still images.

If you keep it as 59.976 fps progressive, it will have all the fluidity of the original.

The file size depends entirely (100.00%) on the encoding bitrate and not on the number of frames per second. What's more, you can get pretty close to the spatial quality that you would get if you use the same encoding rate that you'd use for 29.97 fps. This is due to the very, very complex relationship between bitrate and the spatial offset between adjacent frames.

As a thought experiment, imagine encoding at 10,000 fps: the spatial difference between frames would be so small that in most adjacent frames, almost all pixels would be unchanged. You need virtually no bits to encode that difference. On the other end, if you go down to 12 fps, which is the speed used for some amateur movie film that I've transferred from the 1920s, the spatial gap between frames is gigantic, and you need a huge number of bits to encode those differences. So, with low fps material, you only need to encode, say, 12 frames per second, but you need a gigantic number of bits to faithfully recreate the difference between those frames, but when you encode high fps material, while you need to encode far more frames, you only need a handful of bits to represent the differences. Thus, the "equation" for equal quality is far from linear, and you actually may not need that many more bits, which is what creates a larger sized file, when encoding the higher fps material.

So, the relationship between bitrate and fps is far from linear, and as you double the framerate, you do not need to double the bitrate to get the identical quality (e.g., if you use a CQ encoder instead of a variable bitrate encoder with a target bitrate).

Last edited by johnmeyer; 21st November 2017 at 21:26. Reason: added two clarifying sentences in "thought experiment" paragraph
johnmeyer is offline   Reply With Quote
Old 21st November 2017, 22:12   #3  |  Link
lansing
Registered User
 
Join Date: Sep 2006
Posts: 1,050
If your video is 30 fps interlace, then its true intended frame rate is 60 fps, so bobbing it to 60fps is the correct way to do it. Throwing out the even frames means that you're throwing out half the information, which is not good.

If you're to encode it with x264, the difference in file size between 30fps and 60fps should only be about 20-30%.
lansing is offline   Reply With Quote
Old 23rd November 2017, 03:44   #4  |  Link
burfadel
Registered User
 
Join Date: Aug 2006
Posts: 2,235
If it is 29.97 it's probably telecined as 29.97 isn't standard. It's most likely 24 fps progressive, running at 23.976 fps, and telecined to 29.97 fps. You end up with 3 blended frames and 2 progressive frames. It's got to do with matching the power supply, and using film cameras that were 24 fps. Recording in native 29.97 means it's harder to sell overseas who are on 50 Hz power supplies, but more importantly would require special TV only cameras that would cost for the specialised purpose, and cost 25 percent more for the film due to the higher rate. Best thing to do would be to inverse telecine it and decimate the resulting still frame (fixed cycle of 5).
burfadel is offline   Reply With Quote
Old 23rd November 2017, 04:50   #5  |  Link
Katie Boundary
Registered User
 
Katie Boundary's Avatar
 
Join Date: Jan 2015
Posts: 579
The answer is that it depends on the frame rate of the original material. It could be 24 FPS telecined, or 30 FPS, or it could be 60 genuinely distinguishable images per second that each contributed only one field to the final stream, or any mix of the above and more. Only by looking at the resulting stream can you know what step to take next.
__________________
If I ask "How do I do X?" or "what happens if I do X?", and X is a very bad thing that no one would ever normally do, assume that I already know this, and that I have Katie reasons for asking anyway.
Katie Boundary is offline   Reply With Quote
Old 23rd November 2017, 12:53   #6  |  Link
Sharc
Registered User
 
Join Date: May 2006
Posts: 3,385
@jriker1
Why not upload a sample of few seconds duration of a motion/action scene? It's usually better for collecting advice than letting people speculate about your source being telecined film, true interlaced video, progressive encoded as interlaced, or whatever ....
Personally, I prefer to encode true interlaced video as interlaced rather than deinterlacing it. My second choice is bobbing (framerate doubling in order to preserve the time resolution) with QTGMC. Filesize can be controlled with the compression settings. Only you can tell what is good for your eyes

Last edited by Sharc; 23rd November 2017 at 13:15.
Sharc is offline   Reply With Quote
Old 23rd April 2018, 22:25   #7  |  Link
Logan9778
Registered User
 
Join Date: Mar 2017
Posts: 97
I LOVE QTGMC's double frame rate. Used it on some Doctor Who encodes. Went from choppy to smooth in the high motion scenes. To quote a little old lady on one American hot sauce ad - "I use that shit on everything!"
Logan9778 is offline   Reply With Quote
Old 24th April 2018, 14:05   #8  |  Link
Gser
Registered User
 
Join Date: Apr 2008
Posts: 387
60fps is is more relevant now then ever before since there is hardware support for it in pretty much in all h.265 decoders and in all resolutions. Back in the days of h.264 i would have to upscale to 720p or just not deinterlace 1080i material. I am so waiting for h.265 tv broadcasts. So tired of interlaced shit.
Gser is offline   Reply With Quote
Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 13:39.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.