logo Sign In

24p- Why, again?

Author
Time
Why do we IVTC if our DVD players are just going to insert pulldown so that our TVs can display the 29.97 interlaced image? Does the 24p help with software DVD playback on a PC (no deinterlacing required?), or maybe over an LCD projector? Perhaps to preserve it in its original form for the future? Or is it just to save space when compressing to MPEG-2? All are fine reasons. I'm just asking because I'm getting some artifacts with Decomb on my Waterworld edit, and I wonder if it would have been better to just leave the footage as interlaced 29.97 fps.
Author
Time
Originally posted by: Zion
Mostly it's to save space and allow you to encode at a higher bit rate.

That's a good enough reason, considering compression is going to be 25% better.

Originally posted by: Zion
What kind of artifacts are you getting? Macro blocks or something similar? That shouldn't be caused by any IVTC process but perhaps somewhere else in your workflow.


Here are some pics from all of my different TV caps:
http://i70.photobucket.com/albums/i82/thegreathunger/Deinterlace-TelemundoSplash.jpg
http://i70.photobucket.com/albums/i82/thegreathunger/Deinterlace-USASplash.jpg
http://i70.photobucket.com/albums/i82/thegreathunger/Deinterlace-Sci-FiSplash.jpg
http://i70.photobucket.com/albums/i82/thegreathunger/Deinterlace-BravoSplash.jpg

It happens in areas of motion, so I picked a shot with a lot of motion. It is evident, though extremely subtle, in almost all shots. It is most evident on the Telemundo cap, and least on the USA. I thought it might be the station compressing the broadcast as MPEG2, but the artifacts are horizontal lines, which leads me towards IVTC as the culprit. The lines are non-existent on the source (29.97) captures, except, of course, for the two interlaced frames. The artifacts on my IVTC'd version are present on all frames, though barely noticable on any but the frames derived from interlaced frames. I used Telecide and Decimate in AviSynth, and no matter what "post" setting I use in Telecide, the result is the same.

There's not much else in my workflow that I can think would be causing this. The caps are DV, I have to convert the colorspace, and then decomb. That's all so far.

Here's another example:
http://i70.photobucket.com/albums/i82/thegreathunger/Deinterlace-Telemundo.jpg

Author
Time
Have you tried manual IVTC; i.e. use DoubleWeave().Pulldown(x,y)?

Guidelines for post content and general behaviour: read announcement here

Max. allowable image sizes in signatures: reminder here

Author
Time
Originally posted by: Moth3r
Have you tried manual IVTC; i.e. use DoubleWeave().Pulldown(x,y)?


Originally I did, but Decomb was more highly recommended and I figured it could automatically keep up with the changes in the 3:2 pattern that arose each time an extra scene was cut into the original film. Thanks for the recommendation, I'll try manual IVTC again.
Author
Time
I tried Doubleweave again, and strangely, both Doubleweave and Decomb yield IDENTICAL results. Both are of lesser quality than the source (29.97) frames. Still, when I stare enough at the source frames, I can see the artifacts there as well. See for yourself:

SOURCE
http://i70.photobucket.com/albums/i82/thegreathunger/Deinterlace-TelemundoSplash2-Source.jpg

IVTC
http://i70.photobucket.com/albums/i82/thegreathunger/Deinterlace-TelemundoSplash2-IVTC.jpg
Author
Time
Originally posted by: mcfly89
... I thought it might be the station compressing the broadcast as MPEG2...
Looking at the source image, I think you could be right. Is it just on shots with high motion?

Guidelines for post content and general behaviour: read announcement here

Max. allowable image sizes in signatures: reminder here

Author
Time
Originally posted by: Moth3r
Originally posted by: mcfly89
... I thought it might be the station compressing the broadcast as MPEG2...
Looking at the source image, I think you could be right. Is it just on shots with high motion?


I may be obsessing a bit too much over "restoring" quality lost in the broadcast. The horizontal distortion is present in even still shots (though barely visible), but I just realized it's present in noisy areas of the shots, which is "motion" as far as an MPEG encoder is concerned. I'm ready to chock this one up to MPEG and close the book.

I was obsessing all day over a slightly different issue, leaving the broadcast noise alone or accepting blocky artifacts from the FFT3D denoiser, and when I finally previewed the two side by side on my TV, the difference was barely visible. I think I just need to accept that these are going to be great on a TV and crap on a PC monitor.
Author
Time
Could these be chroma errors? The artifacts in your screenshots (although very hard to see on my screen) look similar to the chroma errors you get when using a DVD player that has the chroma bug. It may be down to the original mpeg compression producing the chroma artifacts.

Does your software have any chroma filters that you can use? If so, it may be able to cleanup the chroma channel (although maybe at the expense of a slight loss in chroma detail).
Author
Time
I was obsessing all day over a slightly different issue, leaving the broadcast noise alone or accepting blocky artifacts from the FFT3D denoiser


Just a quick aside, as don't want to derail the thread, have you thought about combining the your multiple captures to reduce noise/improve picture quality? It's the technique Citizen used to create his very clean Star Wars laserdisc trasfers.

Some info from the doom9 forum: Look for the section '7.2.12 Using multiple captures to reduce noise'

There's also the Average and TooT plugins that can be used to get the best out of multple captures


Author
Time
24p- Why, again?

I asked myself this the entire sixth season once Jack Bauer's brother and dad were behind it all.

Oh, wait, what are we talking about again?
I am fluent in over six million forms of procrastination.
Author
Time
Does your software have any chroma filters that you can use? If so, it may be able to cleanup the chroma channel (although maybe at the expense of a slight loss in chroma detail).

Someone recommended Reinterpolate411, though that had no effect. Is that the type of filter you mean? I'm running all my files through Avisynth, so I can use any Avisynth filters.

Just a quick aside, as don't want to derail the thread, have you thought about combining the your multiple captures to reduce noise/improve picture quality? It's the technique Citizen used to create his very clean Star Wars laserdisc trasfers.

Some info from the doom9 forum: Look for the section '7.2.12 Using multiple captures to reduce noise'

There's also the Average and TooT plugins that can be used to get the best out of multple captures


All of my captures are from different broadcasts, which means I wouldn't be able to apply this filter because they have different framing, edits, etc. Am I correct in assuming that? Even if that is the case, a couple other people captured the same broadcast as me. Perhaps their versions, combined with mine, could cut the noise?
Author
Time
which means I wouldn't be able to apply this filter because they have different framing, edits, etc. Am I correct in assuming that?


Well, it would be an absolute bitch to line up framing, synchrionize frame cadences, and the like...but if you're a glutton for punishment....

<span class=“Italics”>MeBeJedi: Sadly, I believe the prequels are beyond repair.
<span class=“Bold”>JediRandy: They’re certainly beyond any repair you’re capable of making.</span></span>

<span class=“Italics”>MeBeJedi: You aren’t one of us.
<span class=“Bold”>Go-Mer-Tonic: I can’t say I find that very disappointing.</span></span>

<span class=“Italics”>JediRandy: I won’t suck as much as a fan edit.</span>

Author
Time
Not being a glutton for punishment, I waited for an identical airing on the Sci-Fi channel to roll around to test the average filters. Unfortunately, "Average" and "Toot" are both very hard to find (broken links). I tried a number of scripts from "7.2.12 Using multiple captures to reduce noise," and the only one I could get to work was "Overlay" (the first one).

It works nicely to reduce noise, but I guess I was hoping for something more. Here's my question: these filters all blend the two captures together; why not identify the similarities between the two (the actual image) and subtract the difference (the noise)? Is there a filter like that? Or am I imagining the impossible?
Author
Time
This link still works:Toot. I don't know if it's the latest version.

If I'm not mistaken, Toot looks for the similarities - where two of the three caps agree on a pixel.


Sounds like what you're describing would look at two and denoise on the differences? That might be an idea.

One day I found... 10 years had got behind me. Next day was worse.

 

Download  shows from Cable DVR (Updated! Yes, it needs a rewrite, but it's worth slogging through, anyway).