Vidding 101: Getting started, Common terms, FAQs

[Elvira]

I've found the opposite to be true, in all editors that I've tried. I recently did a test in Vegas, where I inadvertently saved the file as 25 fps, when the source was 29.97, and the flow of the video was "jerky." (I didn't know that I'd made the mistake until I looked at the video and couldn't figure out why it was jerky. Then I saw the frame rate I'd used and I went )

It's not always detectable when you have fast editing, but with a longer scene or clip, you can often see it. But, if you think a video can look okay by downconverting it to 25 fps (when the source includes 29.97) then go ahead and do it. All my testing and samples have been pretty consistent in this, however, which is why I'm going to continue to recommend exporting with the higher frame rate, rather than the lower.

[rhoboat]

Maybe it also depends on project settings, not just render settings, because I've frankly had the exact opposite problem on a vid I did last year until I completely changed the project as well as render settings. I'm a firm believer that adding something that isn't there (i.e. rendering 29.97 fps with 25 fps sources) isn't something you want to do.

Edit: I knew I'd read something about this before, back when I still used Ulead VS 10. Not exactly vidding-specific, but it's something that stuck out in my mind.

From the Ulead boards:

PAL video runs at a frame rate of 25 frames per second whereas NTSC runs at 29.97 frames per second. This particular difference is the major cause of trouble in the conversion process. Any conversion of PAL to NTSC is somehow going to have to create 4.97 frames per second in order to maintain sync with the audio track or use up the available frames faster resulting in a shorter video file and somehow decimate the audio track to maintain sync. There are two problems with the "creation" of these extra frames. First is the existence of movement from one frame to the next, simply repeating a frame will cause "jumpiness" and the motion will lose fluidity. Secondly, if the software is going to use the existing frames it can only use "half frames" because the original footage will be interlaced.

[Elvira]

Well! Again I've learned a lot of new things, thanks to you, rhoboat!

First off, much to my amazement, I discover that Vegas Pro has a feature for inverse telecine, which is (from my understanding) the desired way to convert a video's frame rate down. Here's AMV.org's tutorial on Inverse Telecine, and here's Ishtori.net's tutorial on performing inverse telecine with Cinema Tools. (I've done inverse telecine in Cinema Tools a few times, but there are other programs for the Mac as well.)

I found this tutorial on using inverse telecine within Vegas. Which is pretty exciting! I tried it out, and sure enough, the "2-3 pulldown" in Vegas worked (the motion was smooth in my samples). However, I saw a consistent loss of quality (sharpness). I tried several times, with different codecs, settings, etc, and still got this loss of quality. Rather than show you lengthy video clips, the difference can be summed up as this:

The picture on the left is from an NTSC clip exported at 29.97 fps, the picture on the right is the same clip, exported using Vegas's inverse telecine so the clip was changed to 24 fps. (I used the exact same settings as much as I could, exported out as uncompressed, and converted over to H.264 MP4 using identical settings. )

Perhaps some Vegas Guru can show us what settings to use to get it looking better, but all I know is that I don't like the quality hit. Until I figure out how to get it sharper, it's just not happening.

So anyway, rhoboat, yes, it looks like Vegas Pro (but not Movie Studio) does have a way to convert down to 24 fps (or 23.976) without the jerky quality, which is great. But I still don't like the way it looks.

Now onto mixing PAL and NTSC footage, here are some visual demonstrations, so everyone knows what we're talking about: (video quality in these samples is so-so. I made them quickly! )

NTSC clip, saved as NTSC, edited in NTSC timeline. (Streaming MOV file, 11 MB)

The same NTSC clip, rendered down to PAL in PAL timeline. It's not glaringly obvious in all shots, but there's a definite "jerky" quality in the movement. That's because the clip was imported straight into the editor without preprocessing (inverse telecine). At least that's always the result I get, in the different video editors I've worked with.

PAL and NTSC clip, project rendered as NTSC. The Doctor Who clip is PAL, but I converted it to NTSC in MPEG Streamclip and then added it to an NTSC timeline with an NTSC Robin Hood clip.

PAL and NTSC clip, project rendered as PAL. The same two clips. The Doctor Who clip still looks fine, but the Robin Hood clip (converted down to PAL in MPEG Streamclip) looks "jerky."

PAL and NTSC clip, rendered as PAL, w/ NTSC clip "converted" down to 25 fps in Avidemux. I think I see a teensy bit of improvement in the "jerkiness" (not quite so bad) in the PAL clip, but it's still kind of "meh."

I plead ignorance to how the best way to go about using inverse telecine on your footage—I know how to do it in Cinema Tools, and I recall that there's an Avisynth script to do it in VirtualDubMod. Perhaps zetaminor (our Avisynth guru) can shed some more light on this.

So far, based on my research, the best way to convert a clip down to a lower frame rate is to inverse telecine (usually in Cinema Tools, AVISynth of some other app). Or, if they know the magic combination in Vegas Pro and are able to get inverse telecine in their fan videos without losing detail, then that is fabulous. If there is another way to do it while maintaining quality, I haven't heard much about it (but await to be enlightened ). The focus seems to be mostly on inverse telecine.

There are tutorials on ishtori.net and on amv.org for reverse (inverse?) telecine, if a vidder is interested in that. (And they certainly are encouraged to do some research on it! ) But I don't see how rendering away frames (i.e. not using inverse telecine) looks better when done within your video editor. But, whatever. If some vidders are able to get it looking good, then they should go for it!

[Elvira]

Also . . . I took a look at the tutorial on Ulead's forums. This jumped out at me:

A conversion either way requires either frame padding or frame decimation which will result in less fluid motion especially in high motion or panning scenes.

The "less fluid motion" is the "jerky" quality, which is what you get unless you inverse telecine. Adding frames (frame padding) doesn't create the "jerky." The tutorial also says that this conversion can be used either way (NTSC to PAL, or PAL to NTSC). But since you lose fluid motion when converting down, then it seems that editing in NTSC would look better. It sure does in all the tests I've done.

So far I haven't had trouble mixing NTSC and PAL clips in the same timeline, because I'm often not using the source audio (so there are no possible sync issues). (Also, Final Cut Express claims to support mixed frame rates. ) But it sounds like converting up to NTSC (if some clips are PAL, and some are NTSC) would be the more desirable way to do it in some video editors. (That, or inverse telecine.)

[zetaminor]

Think I saw my name in there somewhere?

I've always been under the impression (though I'm not definite on this at all) that Inverse Telecine is for use on Film Telecined Sources only.

Telecine is a way of taking a TV or Movie that has been shot in Film and transferring it to DVD or Video. It uses a type of field scanning very similar to Interlacing, but they're not the same.

While using Inverse Telecine on Film Sources will reduce your frame rate, you should never purposely change your frame rate for non-Film sources using it. There are ways to change frame rate that will not damage your source much, but you will always get some trouble.

You'll get less trouble increasing your frame rate than reducing it, from what I read, but I've never tried it. I've never needed to go through it, though I may need to soon If I ever want to make a BSG vid. Those source DVDs are all Film.

When you use IVTC on a Film source, your frame rate will be reduced, usually to 23 (ish) FPS. Most editors can't edit in 23 FPS, in which case you'll have to increase your FPS manually, which is where Decimation comes in.

Anyway, that's what I've understood from what I've read, though I've not tried to put any of this into practice as of yet. Too much else going on...

But, when I do get a few projects finished and am looking to start ripping BSG or House MD clips, I'll do an AVISynth Film and IVTC Tutorial. ;D

If either of you have more informationa and have tried more things out recently, let me know. I'd love to know if any of this is confirmed or not, before I go and make my own mistakes.

[Elvira]

Jul 15, 2008 8:59:40 GMT -5 zetaminor said:

Think I saw my name in there somewhere?

So happy to have you here! I was eager to know your opinion on this.

I've always been under the impression (though I'm not definite on this at all) that Inverse Telecine is for use on Film Telecined Sources only.

This has been a source of confusion for me too.

Now, Vegas's built-in inverse telecine was able to smoothly convert down my TV footage, but it lost quality. But it did it. Whether it can look better, I don't know. All I know is that unless I find out some easy way to make it look better with TV footage (PAL or NTSC), then it's not worth it, and perhaps never meant to be?

While using Inverse Telecine on Film Sources will reduce your frame rate, you should never purposely change your frame rate for non-Film sources using it. There are ways to change frame rate that will not damage your source much, but you will always get some trouble.

I've inversed telecine a few times, but decided that for me, it was easier to stick with the same frame rate as the source. During my last adventure with AVISynth, I did follow those video tutorials which did (I believe) instruct on inverse telecine, and it seemed to work, but I got caught up with the other parts of AVISynth and didn't explore further.

You'll get less trouble increasing your frame rate than reducing it, from what I read, but I've never tried it.

Yes, this has been my experience too. For the sake of simplicity (especially when trying to write a tutorial for all the different software we use here), it's been easier to just say, "export/edit in the higher frame rate, because you won't get the 'jerky' look." So far this seems to be working out well.

But, when I do get a few projects finished and am looking to start ripping BSG or House MD clips, I'll do an AVISynth Film and IVTC Tutorial. ;D

I was hesitant to ask . . . but YAY! Thank you!

[zetaminor]

Jul 15, 2008 9:15:26 GMT -5 Elvira said:

So happy to have you here!

LOL, procrastination.... Sing along with me now. Can't seem to get motivated today... ;D

Now, Vegas's built-in inverse telecine was able to smoothly convert down my TV footage, but it lost quality. But it did it. Whether it can look better, I don't know. All I know is that unless I find out some easy way to make it look better with TV footage (PAL or NTSC), then it's not worth it, and perhaps never meant to be?

But, TV sources are almost never Film. They're Interlaced fields.

TVTC is for Film sources only, don't use it to deinterlace pure interlacing you you will get that loss of detail and clarity.

The only way I know of to find out if your source is Film or Interlaced is to use DGIndex. There may be others, but I've not used them...

Best to use Deinterlacing if you're not sure.

I've inversed telecine a few times, but decided that for me, it was easier to stick with the same frame rate as the source. During my last adventure with AVISynth, I did follow those video tutorials which did (I believe) instruct on inverse telecine, and it seemed to work, but I got caught up with the other parts of AVISynth and didn't explore further.

Yeah, the tutorials do touch on it. There's more involved descriptions on the script page tutorials, if you can wade through it all.

That, and the fact that all the sources I've used so far are not Film, is why I've not gotten in to it yet.

Yes, this has been my experience too. For the sake of simplicity (especially when trying to write a tutorial for all the different software we use here), it's been easier to just say, "export/edit in the higher frame rate, because you won't get the 'jerky' look." So far this seems to be working out well.

True enough! Simple is good, for now.

I was hesitant to ask . . . but YAY! Thank you!

if I ever finish the one I'm still working on...

[Elvira]

Jul 15, 2008 9:31:11 GMT -5 zetaminor said:

But, TV sources are almost never Film. They're Interlaced fields.

THANK YOU!!

I've always been confused about this. In my gut I felt this might be the case, but I didn't know the technobabble term for it!

THANK YOU. Well! That explains a lot.

The only way I know of to find out if your source is Film or Interlaced is to use DGIndex.

This is interesting. DGIndex said that some of my Robin Hood DVDs were progressive, but that others were interlaced. None have interlacing (at least that I can see, or that warrants deinterlacing).

MacTheRipper is very good about doing "force film," so usually I know right off if a DVD is progressive. And then I edit in FCP, as PAL (or NTSC) with a 24 fps (or 23.976) frame rate. Works great. The rest of the time, I either edit in PAL, NTSC, or in the case of mixed frame rates, NTSC.

That, and the fact that all the sources I've used so far are not Film, is why I've not gotten in to it yet.

Now I'm curious about that one season of Robin Hood. I believe that DGIndex said it was progressive. But DGIndex also said it was PAL? Oh, never mind. I'm editing it as PAL! (Edited: I think it was Vegas that said it was progressive ... getting confused now. )

True enough! Simple is good, for now.

if I ever finish the one I'm still working on...

We know you'll get there!

[zetaminor]

Now I'm curious about that one season of Robin Hood. I believe that DGIndex said it was progressive. But DGIndex also said it was PAL? Oh, never mind. I'm editing it as PAL!

And confusion Reigns!

It can be Progressive PAL, Progressive NTSC, Interlaced PAL, Interlaced NTSC, or FILM.

Progressive means the DVD was coded Progressive and tells the DVD player to interlace on its way to the TV as it's being played.
So, you shouldn't need to Deinterlace at all.

I should put up the DGIndex Tutorial soon.... Was waiting until I could get my example Clips sorted and post it at the same time as the AVISynth one, since they're connected, but...

Can you get a screencap of the DGIndex Info screen for that DVD? I'll get a look and tell you what I think I see.

[Elvira]

I'll get that screenshot later on. (I'm on the Mac side now. ) I can't recall now if it was Vegas or DGIndex that said it was progressive. I just remember being confused.

So far, the PAL Robin Hood DVDs (neither Season One or Season Two) have needed deinterlacing. At least, they all look fine as they are, no deinterlacing needed when converting them with MPEG Streamclip or AMVapp, and they look gorgeous after they've been edited. I was just confused as to why one set of DVDs was listed as progressive, and the other interlaced, when neither seemed to require deinterlacing! Well, when you see the screenshot of DGIndex, perhaps you can sort it out.

[val]

Ooops sorry Elvia I wrote '740' by mistake ! I meant '720' .
Thanks for the explanation ! So basically I should always be watching out for frame rate and NOT frame size and what goes in must come out right !?

Thanks for your inputs too Rhoboat and Zetaminor !

This is alot of info though and I think I`ll need some time to read and digest it ;D

Okay so just tell me one thing...what are the best SV project settings that I should set when I`m using
A. WMV clips with 720x480 25 fps clips (PAL settings I know but what are the other things that I should keep in mind?)
B.Avi Divx clips

[Elvira]

Jul 15, 2008 22:56:00 GMT -5 val said:

So basically I should always be watching out for frame rate and NOT frame size and what goes in must come out right !?

Frame rate is really important, because as we've discussed, if you take away frames from the original, you get that "jerky" look.

Frame size is important, because it may indicate aspect ratio. But it doesn't always. You need to know the real aspect ratio of your source footage (you can't always tell by the frame size of your footage because it could be using "unsquare pixels"—which is explained in this 101 tutorial). Then you export out of your video editor (in this case, Vegas) with the proper aspect ratio.

Okay so just tell me one thing...what are the best SV project settings that I should set when I`m using
A. WMV clips with 720x480 25 fps clips (PAL settings I know but what are the other things that I should keep in mind?)

Convert the clips to Lagarith AVI, DV AVI, or HuffYUV AVI by following this tutorial, or this tutorial, or this tutorial, or this tutorial. Then set up your project by following this tutorial (select the project settings with the right frame rate and aspect ratio).

B.Avi Divx clips

Convert the clips to Lagarith AVI, DV AVI, or HuffYUV AVI by following this tutorial, or this tutorial, or this tutorial, or this tutorial. Then set up your project by following this tutorial (select the project settings with the right frame rate and aspect ratio).

(I know, I know, I'm being mean! )

This tutorial explains why you should convert to a better format first.

I am not saying you can't edit in XviD, DivX, WMV or whatever anyway, but I can't give you advice on it. If I give you advice and something goes wrong (and it eventually will, because that's how it usually is when editing in XviD, DivX, WMV, etc) then you'll ask me to troubleshoot! (And that would be the natural thing to do, if I'd advised you in the first place. )

Working in DV AVI, Lagarith AVI, HuffYUV AVI or one of the other editable codecs require less troubleshooting, (and look so much better), so that's what we're trying to stick with here.

[val]

Wow ! Okay BIG THANKYOU for all the info Elvira !!!!!!!
I have so much to learn ! Thanks !!!!!

[randhrshipper1]

Hey! Another question. I checked my ripper program and it says that the AVI has an "Internal MPEG4 codec" with a Video Bitrate of 1200 kbps. After having read other tutorials here (which are full of great info!), that seems to be the format I have access to that I want to be working in. Am I right about this?

[Elvira]

Jul 17, 2008 21:03:21 GMT -5 randhrshipper1 said:

Hey! Another question. I checked my ripper program and it says that the AVI has an "Internal MPEG4 codec" with a Video Bitrate of 1200 kbps. After having read other tutorials here (which are full of great info!), that seems to be the format I have access to that I want to be working in. Am I right about this?

No, a video bitrate of 1200 is very low for editing. And MPEG4 is not, I don't believe, typically an editing codec. (XviD, DivX, and 3ivX are MPEG4.)

We have tutorials here which recommend using the free DVD Decrypter, which will rip your DVD's VOB files (digital video files) and place a copy of these VOB files on your computer's hard drive. It is strongly recommended that you use this kind of program (a DVD ripper/extractor, that only copies the VOB files and nothing more). Then you can open the ripped DVD (.vob) files in a program like VirtualDubMod, MPEG Streamclip, or Avidemux, and let them convert them to an editable codec (like DV AVI, Lagarith AVI, or HuffYUV AVI).

If you are short on disk space, you can conserve it by just trimming down to only the scenes you need to AVI, instead of converting the whole DVD over. DV AVI (and Lagarith, and especially HuffYUV) take up more disk space than DivX or other kinds of codecs used with AVI, but that's because they don't compress the video a lot (or at all) and therefore they lose very little (or no) detail. When you edit in DV, Lagarith, etc, your finished video is going to look a lot sharper and better (and, Vegas is probably going to behave more smoothly, since it likes DV AVI and Lagarith, but may or may not tolerate DivX, MPEG4, etc).