The Virtual Dub thread

[Brian Luce]

So there is some consensus that VD is the best way to make a SD DVD from an HD file in conjunction with Vegas and DVDA.

What does VD do? A lot of stuff, but from what I know, us DVD makers want our HD clips resized to SD for SD DVD burning. So we use it to resize.

First, VD is free, second, there's not a lot of info on it, though there is a youtube tutorial:
YouTube - Virtualdub Tutorial - Free Video Editor For Windows

It's a simple light little software.
Here are the basics (I think) of importing and resizing a clip:
Open VD>video>filters>add>resize>size options>select 720x405
-Aspect Ratio: Same as source
Filter Mode: Lanczos 3
-Do Not letterbox
-Codec Friendly sizing: I have no idea about this one, though I like the way "Friendly" sounds.
-hit OK
-Save as

[Gareth Watkins]

Hi there
I've used Virtual Dub for a number of project and it works well... you do need to treat the images once again in your NLE before authoring your DVD which takes extra time..

More recently I've been saving out my editied clip in HDV and using MpegStreamclip to make my final file for the DVD. It seems to do a pretty good job of downscaling and saves out to Mpeg2 with loads of parameters. I use it too for MP4 files to upload to Vimeo and Youtube.

That said I find the recent encoders with Adobe CS4 far better than the previous version for rendering out an SD Mpeg2 file from an HDV timeline for DVD.

Regards
Gareth

[Thomas Moore]

Can you list your workflow for those of us who have not tried this but may wish to try?

[Kenneth Anderson]

http://www.dvinfo.net/forum/what-hap...d-quality.html

[Brian Luce]

Originally Posted by Kenneth Anderson

http://www.dvinfo.net/forum/what-hap...d-quality.html

If you don't want to go through 12 pages of thread, here are the highlights:

Most of the below is from Perrone Ford.

Virtual Dub Test: I took 2 frames of a RED 4k file and laid it onto a timeline. I conformed that timeline to 1024x512 to preserve the 2:1 layout of the original. I took a frameshot of that 1024x512 file. I then went back to a 4k timeline, rendered that 4k RED file to a 4k uncompressed AVI file. Pulled that into virtualdub, did a lanczos resize to 720x360, and rendered back out uncompressed. Put that onto a 720x360 timeline, and took another frameshot.

As for the VDub workflow it's simple.

1. Import interlaced high quality .avi file
2. Select "Resize"
3. Input desired size (720x480)
4. Select Lanczos as the rescaler
5. Choose compression type for output
6. Save new .avi file.

Done.

Q: Perrone, with creating the initial .avi, what sort of hard drive space are we talking about here - what .avi settings are you using?
I use the Lagarith codec (works in 32 and 64 bit modes and is lossless). For audio I am using PCM 16 or 24bit 48KHz.

A: It's about 460 Mbps so works out to roughly 3GB/min. Which is about 1/3 what uncompressed 10bit AVI would be or just less than half what uncompressed 8-bit uncompressed would be.
The 720x405 works fine for SD. It's a bit tricky for those with 16:9 TVs. 720x480 with the widescreen flag on should work fine for making SD discs and it's what I tend to do these days.

I avoid non-square pixels like the plague. They complicate matters more than necessary. Once you get to square pixels everything becomes perfectly simple.

DVDA is DVD Architect. If you open that application, choose New > Single Movie. It will let you select a video and audio file. It will not create a menu or separate chapters. It just makes a simple DVD that auto-plays when inserted into a DVD player.

Virtualdub ONLY understand .avi files (and a couple specialized files with some hacks) so you will need to create an .avi master to bring into virtualdub. This should be done with a lossless .avi codec like Lagarith, HuffYUV, or uncompressed. Cineform is also a legal type to bring into Virtualdub, but it is not lossless.

7d workflow:
1. Convert 1080 .MOV into 720 .AVI (take the lossy hit)
2. Edit video, add titles, cross fades, credits, whatever - finish content of movie.
3. Render in Vegas Studio to ? (.AVI again?)
4. Bring finished rendered .AVI into Virtualdub and resize (what size?) with Lanczos scaler.
5. Within Virtualdub, export a Lagarith or HuffYUV compressed AVI
6. Bring that into DVDA and select single movie and it will automatically convert it to a MP2 for DVD burn.

When I bring the material into DVD Architect, I get asked what kind of DVD I want to create. I choose NTSC 60i.

Q: So my 5D Mk2 shoots 1080/30fps so should i produce an AVI in Vegas at that? (ie no resizing) I had been rendering in Vegas to 720.

I did find Lanczos scaler.
Where is the HuffYUV or Lagarith lossless in Vegas and in VirtualDub. I've spent hours looking for those suckas!!

A: In Vegas choose Render As and pick a template. Then click on Custom/Video/Video Format. They are in the drop down list.

In Virtual Dub they are under the Video/Compression/Select Video Compression list

1. Pull .MOV (native 1920x1080p/30fps) into timeline
2. Edit video, add titles, cross fades, credits, whatever - finish content of movie.
3. Render in Vegas Studio to 1920x1080p/30fps with Lagarith pixel ar 1:1 (if i play resultant AVI in Windows Media Player it'd like learning how to drive a stick shift - figured that was my poor computer cound'nt process the full HD but the MP2 will be smooth)
4. Pull the full sized AVI into VirtualDub
5. In VirtualDub set the filter - add resize with Lanczos scaler to 720x405 (Match original AR)
6. In VirtualDub choose compression type for output to a Lagarith or HuffYUV AVI (Think i may have skipped this step in VD though i did this already in Vegas when i rendered....do i need to do it twice?)
7. Save .avi file
7. Bring that into DVDA and select single movie and it will automatically convert it to a MP2 for DVD burn.

Q: Found the MPEG-2 Main Concepts type but could not customize the video to change the output size to same VD produced which was 720x405 - so all i could do is bring the product of VD into DVDA and burning to DVD now.....
A: This is correct. DVDA or Mainconcept in Vegas will both create a 720x480 project. You want that, so don't worry.

When you create DVDs, the video portion and the audio portion are compressed separately. So you compress the video in Vegas to Mpeg2. Then you compress the audio separately with the Dolby Digitial AC-3 Studio Template. Name this the same thing as you named the video file, and when you are in DVDA and drop the video file on the timeline, the audio will come with it automatically.

Q: What I am wondering is if we shouldn't all just start shooting the dreaded interlaced (1920X1080i60) if we are not shooting 24P?
A: Good gracious why? If you plan to deliver for broadcast (over the air or on optical) then yes, shoot 24p or 60i. Otherwise, shoot what you want.
Q:We are all rendering back to interlaced anyway?
A: No, we are not all rendering back to interlaced.

Q: I am starting to wonder what the purpose of 30P is since there is no native 30P playback.
A: I don't typically shoot 30p, but if I have significant motion, and I am delivering for non-broadcast, then 30p becomes viable. I typically shoot 24p or 60p though.

Q: Still don't have an answer to what to do with progressive footage shot at 1080/30p from an EX1.
A: What do you mean? What you "do with it" is totally dependent on how you will deliver. Web, DVD, BluRay... all demand something different.

Q: 1. Render out an intermediate avi lossless file at 1920 x 1080 using which template for the Lagarith or Huffy UV?
A: Make a template. 1080p, PCM 16bit/48khz audio (or whatever your camera shot) Huffyuv or Lagarith codec. Done. Simple.


Q: 2. Resize in VD at 720 x 405?? or change to 480? Use a lossless codec. Interlaced?? Where is Progressive in VD? What do I do with progressive footage in VD?
A: 720x405. Interlaced? Why? You shot progressive. Let it stay progressive. VD is expecting progressive and likes it. Leave it alone.

Q: 3. Load resized avi into vegas and render to an Mpeg 2 for DVDA as what? Interlaced, Progressive?
A: NTSC 60i Widescreen. It's a pre-built template. Use it.

Q: 4. Field order problems?
A: None if you keep everything progressive until you build the mpeg2 for the DVD. Don't over complicate things.

Set your template in DVDA to NTSC widescreen 720x480 and you're golden.

COMMENT: I just did this workflow on my editing machine as a test. Rendered a one second 1080/30p with Lagarith from Vegas, all the way through DVDA. Works fine.

[Perrone Ford]

Man, that's an ugly read!

Especially for something that really is so simple.

[Brian Luce]

Originally Posted by Perrone Ford

Man, that's an ugly read!

Especially for something that really is so simple.

Okay, I cleaned it up. It's more of a Q&A now.
Thanks for all the info.

In theory, HDV should yield a better SD DVD than something that originates in mini DV but a lot of people don't get good results and swear that you're better off originating in SD. Hope this puts that belief to bed.

[Sigmund Reboquio]

DVDA is DVD Architect. If you open that application, choose New > Single Movie. It will let you select a video and audio file. It will not create a menu or separate chapters. It just makes a simple DVD that auto-plays when inserted into a DVD player.


>>>Would this work if you want menus and submenus in your DVD? I believe most people here
have menus/chapters on their DVDs.

[Perrone Ford]

Originally Posted by Sigmund Reboquio

DVDA is DVD Architect. If you open that application, choose New > Single Movie. It will let you select a video and audio file. It will not create a menu or separate chapters. It just makes a simple DVD that auto-plays when inserted into a DVD player.


>>>Would this work if you want menus and submenus in your DVD? I believe most people here
have menus/chapters on their DVDs.

No, in that case, you'd just do a normal new project and add the video/audio assets the normal way.

[Dale Guthormsen]

Perrone, and others,


One of the great aspects of using Hd for your base file was that it gives you so much room for cropping and such, then rendering it to SD.

Is the render to SD in Vegas not as good as in virtual dub?

If you are going to do cropping and all that what would the procedure be, work it out in editor, render it as a m2t file, then put that in virtual dub?



Dale Guthormsen

[Perrone Ford]

Originally Posted by Dale Guthormsen

Is the render to SD in Vegas not as good as in virtual dub?

It's not the render. It's the rescale. It's poor in Vegas and most other NLEs as well.

Originally Posted by Dale Guthormsen

If you are going to do cropping and all that what would the procedure be, work it out in editor, render it as a m2t file, then put that in virtual dub?

Crop to what you want, render to a lossless .AVI file, then do the rescale in VirtualDub or other software. M2T is not only not understood (natively) by VirtualDub, it's quite lossy as well.

[Randall Leong]

Originally Posted by Dale Guthormsen

One of the great aspects of using Hd for your base file was that it gives you so much room for cropping and such, then rendering it to SD.

Is the render to SD in Vegas not as good as in virtual dub?

If you are going to do cropping and all that what would the procedure be, work it out in editor, render it as a m2t file, then put that in virtual dub?

The biggest problem with resizing from HD to SD in Vegas or any other NLE is that they simply do a horrible job of resizing interlaced footage. If the original footage is interlaced, the fields must be pulled apart and properly deinterlaced before the resize; otherwise, there will be visible (and often severe) artifacts and/or blurring in the SD downconversion. Unfortunately, Vegas and most other "prosumer" NLEs by themselves offer only two deinterlacing methods: Blend and Interpolate. The former degrades temporal resolution (and also degrades overall spatial resolution in scenes containing motion); the latter, effective vertical spatial resolution (and often creates stuttering in scenes with a lot of motion). (If you have inexpensive consumer editing software, you're often permanently stuck in the Blend deinterlace mode with no way to override that at all whatsoever simply because such software does not support progressive footage at frame rates higher than 30 fps at all.) And once deinterlaced, the NLEs cannot properly re-interlace the footage - which means that you have lost not only spatial resolution, but also temporal resolution. At worst the end result will either be the equivalent of a 360x240 (instead of the normal 720x480) image after the downsizing in the NLE (that's right, this resizing will strip away up to 75 percent of the newly downsized resolution!) or an image with severe artifacts.

As for cropping, the only way to do so without loss in quality (that I know of) would be to render as an uncompressed or lossless AVI. If you render to an m2t file and then put it back into VirtualDub via AVISynth and the DGAVC or DGMPEG video plugin plus an audio plugin (since VirtualDub does not natively support lossy compressed video or compressed audio), the video will be lossy compressed even further when you re-render the resulting AVI file back to a lossy format.