Vidding 101

[thuviaptarth]

I should probably post this to vidder, but I am too embarrassed, because I feel like I should have figured it out already fter reading the A&E Guide five zillion times. I am hoping one of you will take pity on me and answer here anyway.

The problem is (I mean, the problem apart from not being able to tell where the beat is, and Premiere crashing on me like a really bad pun, and ...), it feels like I'm doing a lot of processing on the files I rip from DVDs and I can't shake the feeling that I really shouldn't have to do so much to recent footage to get it viddable. If I do, that's cool -- I'm doing it, I've done it. I'm just terrified I'm doing extra and unnecessary processing and degrading visual quality because I don't know what I'm doing and am too freaked out to be able to tell anymore whether one picture is better than another.

So, I'm ripping the Supernatural DVDs (I know you're shocked), and this is what I'm doing:

1. I rip VOB files from the DVDs.
   
2. I run the VOBs through DGIndex (usually I create one index for each DVD) to get D2V indices.
   
3. I create AVS files for the indices which look roughly like this:
   
   

   MPEG2Source("C:/Thuvia Ptarth/spn2.1.d2v",cpu=4,upconv=true)
   Telecide(order=1, post=4,show=false) 
   Decimate(mode=1, threshold=1.0)
   ConvertToRGB32()
   

   
   
   --except for when, just to make me extra crazy, the DVDs switch from Top Field First to Bottom Field First for a disc or two in Season Two, completely destroying the false sense of confidence I gained from anticipating the Pilot and first disc might be in a different frame rate than the other discs, but. Anyway.
   

So, um. Is that right? Should I be doing something different? Should I be doing less? Should I be doing more? Should I be going for a particular frame rate or just converting the odd files to whatever frame rate most of the files are in?

Help?

Comments

[obsessive24.livejournal.com]

Should I be doing less?
See, I don't know about this "should" business because it all looks really complicated to me and I'm not sure if each of these lines serve a really useful function?

Could well just be I am lazy, but all I usually do is

MPEG2Source("file location", false)

And that's about it (false being kill audio). If I use different framerate sources, I add the AssumeFPS line.

I export using the "no fields" option in Premiere, neither top nor bottom, and on the occasion that the interlacing looks really off, I use the TomsMoComp(1,5,1) script and that usually smoothes things over.

So this is obviously the lazy and "you didn't actually fix things properly" way of doing it, and I'm sure I'm breaking a lot of rules and Ian will be all gritting his teeth and shaking his head if he reads this, but hey, the end product looks all right from my end and that's pretty much all I care about.

Er... not sure if that was helpful. Just letting you know how I go about it. :)

[shati.livejournal.com]

I do all three steps, but my AVS file is just MPEG2Source("C:/shati/omganepisode.d2v"). I think Telecide and Decimate in combination will reduce your frame rate, which will reduce video quality if you don't need them, but it sounds like you do (they're for interlaced footage, iirc). I am surprised that SPN is interlaced, though. Can you see interlacing or does DGIndex just tell you it's interlaced?

[heresluck]

FWIW, that's pretty much what I do. What the scripts look like varies from source to source, for me. So, for example, the scripts for my latest vid look like this:

mpeg2source("F:\Gilmore Girls\scripts\GG1x01-11.d2v",cpu=4)
Telecide(order=1,guide=0)
Decimate(cycle=5,mode=2)
ConvertToRGB32()
AssumeFPS(24)

For me, at least, the last line is crucial because Premiere Pro 1.5 doesn't like 23.97whatever footage.

On the other hand, the Due South footage needed a lot more tweaking, at least for "Out Here," so I ended up with:

mpeg2source("D:\due South\scripts and clips\S3 eps 1-4.d2v")
Telecide(order=1,guide=1)
Decimate(cycle=5,mode=2)
#ColorYUV(gain_u=5)
#ColorYUV(gain_v=-5)
#ColorYUV(cont_y=20)
#Tweak(sat=1.2)
#Tweak(cont=1.2)
#msharpen(threshold=8,strength=80,highq=true)
ConvertToRGB32()
AssumeFPS(24)

I checked the look of the filtered source in VDub (verdict: garbage in garbage out, but at least it's higher-contrast and color-corrected garbage), commented out the filters for vidding purposes, turned the filters back on once the vid was basically done, and did a final check on color and contrast on the filtered version before final export.

[doki.livejournal.com]

I don't have much experience with TFF/BFF shifting source, but TIVTC (http://bengal.missouri.edu/~kes25c/) is a different IVTC process that might help. (scroll down to find it -- there's a bunch of good stuff on that page)

You call TFM and TDecimate instead of Telecide and Decimate, and TFM has a d2v parameter that will "analyze to see if there are any illegal field order changes and optionally set the order parameter using the field order of the d2v file." (see docs for more info)

Also note there's an updated version of the A&E guide here (http://www.animemusicvideos.org/guides/avtechbeta/). That's where I learned about TIVTC. This guide's been "beta" for a while, and I've been hoping Zarxrax will finish it.

Edit: I just noticed the sources mentioned. (Supernatural, Gilmore Girls, Due South) Are these TV shows really shot in 24fps and telecined? I thought they'd be shot in 30fps for TV. Are 2 out of every 5 frames interlaced as seen in the left column here (http://www.animemusicvideos.org/guides/avtechbeta/video2_2.htm), or is every frame interlaced?