DVD to XviD Conversion

[DJRumpy]

Use whatever the documentation for the MPEG2Source plugin your using indicates. Normally it would still say MPEG2Source("dvd2avi.d2v").

[TheRainmaker]

I have a very good idea but i don't know it works correctly or not!

You know that your guide is everythings about make a .AVS file
So as you know if you use GK for rips its makes automaticly .AVS file
And i tried to open .AVS file by VDM and its correctly opening...
For exmp GK .AVS file:

DirectShowSource("E:\Rips\VIDEO_TS\VTS_01_1.VOB",fps=25000,audio=true)
#deinterlace
crop( 4, 34, -2, -32)

BilinearResize(720,384) # Bilinear (Soft)
#denoise

ConvertToYV12()

So i can edit that codes as like resize script and also i can add HSV filter too by VDM...

But i don't know what about quality and size of output video file?
And Why DirectShowSource is not MPEG2Source there... Are they same or similar inputs...?

So what is do you think about this idea?

[DJRumpy]

Using a script creator isn't a bad thing. It's just up to you. You can certainly use DirectShowSource. The only difference is that Mpeg2Soure is geared towards MPEG sources while DirectShowSource relies on the underlying OS codecs to deliver the video content to the script. The end result is the same.

The quality is completely subjective. If you find it acceptable, then use it as is. There are some options you can change in AutoGK which can affect quality. Read on of the guides to find the options and their affect on your output.

[TheRainmaker]

Thanks for your answer...
I really don't know when i should add filters?
Before save 1st pass job or 2nd pass job?

[DJRumpy]

The filters should be there for the entire process, not for an individual pass.

[TheRainmaker]

I mean as you know i just wanna add HSV filter but i don't know script for that filter and i can only add in VDM if i don't know script as you said before..
And i dont know when i will add HSV filter?

[DJRumpy]

The filter would be in place for everything you run through virtual dub mod.

I think in your case, I would run your .AVS through VDM initially, and save your output as a Huffman .AVI file. You can drop that directly into your MPEG encoder.

Have you googled for an HSV filter for AVISynth?

[TheRainmaker]

GK created that codes below for the movie and i changed some codes as in your guide and its opening correctly in VDM and i previewed by WMPlayer..And i adjusted only XviD Compression settings as like in your guide...
So that codes are right?

#  PLUGINS
LoadPlugin("C:\PROGRA~1\GORDIA~1\DGMPGDec\DGDecode.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\decomb.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\KernelDeInt.dll")
LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\UnDot.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\dgbob.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\Convolution3d.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\FluxSmooth.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\TomsMoComp.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\VSFilter.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\SimpleResize.dll")

#  SOURCE
mpeg2source("E:\VIDEO\VIDEO_TS\video.d2v")

#  CROPPING
crop(8,70,708,436)

#  RESIZING
LanczosResize(720,384)

So sorry but, i really don't know; whats "googled" means?

[DJRumpy]

www.google.com is where you 'Google' things. Just type in what your looking for.

If the AVS plays in Windows Media Player, then it's right.

[TheRainmaker]

www.google.com is where you 'Google' things. Just type in what your looking for.

If the AVS plays in Windows Media Player, then it's right.

Yes i googled HSV filter of course but results are unfortunatelly its not good...
I didn't find any source for HSV filter...
Its strange normally i am using HSV filter in VDM but i dont know where is the .dll or source...

[DJRumpy]

There are commands built into AVISynth that will do the same. Try the TWEAK command. See the AVISynth documentation to find out how it works, or post your questions under it's own topic in the main forums. None of these questions are related to this guide.

[TheRainmaker]

I will try sometings to do that...
Because i just know i can only add filter with your guide...

And normally i asked twice but it was with GK/AGK...
If i am not success i will try to ask again in main forum...


Thanks for your time and helps...

[TheRainmaker]

When you said Tweak Commands i googled them and i found them and its amazing cause i tried find many times them but i dont know what i need to make with that pages;
http://www.avisynth.org.ru/docs/english/corefilters/tweak.htm

http://neuron2.net/tweak/tweak.html


my adjusts:
Hue: +5.0
Sat: +150.0
Value: +0.0

How can i write a code for them to my avisynth (AVS) file?

If you help about this i really will be happy...

[manono]

Hi-

Tweak is now built into AviSynth. Nothing has to be loaded before you can use it. Just add this line near the bottom of the script:

Tweak(Hue=5,Sat=1.5)

There's no value setting in Tweak, and Saturation doesn't go to 150. So be sure to check out and tweak the settings in VDub(Mod) before encoding.

[TheRainmaker]

I made everythings correct in your guide and i suprised about output video file!
The code was opened by VDM:

#  PLUGINS
LoadPlugin("C:\PROGRA~1\GORDIA~1\DGMPGDec\DGDecode.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\decomb.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\KernelDeInt.dll")
LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\UnDot.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\dgbob.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\Convolution3d.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\FluxSmooth.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\TomsMoComp.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\VSFilter.dll")
#LoadPlugin("C:\PROGRA~1\GORDIA~1\AviSynthPlugins\SimpleResize.dll")
# LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\Tweak.dll")
#  SOURCE
mpeg2source("E:\VIDEOS\VIDEO_TS\video.d2v")

#  CROPPING
crop(8,70,708,436)

#  RESIZING
LanczosResize(720,384)

#  TWEAK
Tweak(hue=8,sat=1.2,cont=1.1)

Video Size(without audio and 2 passes in VDM): 134 MB !

What is that means?

[TheRainmaker]

File Name ..........: video.avi
Total Size (Bytes) .: 141.344.768 Bytes
Total Size (KB) ....: 138.032 KB
Total Size (MB) ....: 134,80 MB
Total Streams ......: 1 Stream(s)
File Size Correct ..: Yes

Video Size (Bytes) .: 141.336.596 Bytes
Video Size (KB) ....: 138.024 KB
Video Size (MB) ....: 134,79 MB
Video Length .......: 01:28:07
Video Codec Code ...: XVID
Video Codec Name ...: XviD MPEG-4 codec
Video Bitrate ......: 213 KB/s
Resolution .........: 720 x 384
Aspect Ratio .......: 1.87:1
Total Frames .......: 132165 Frames
Framerate ..........: 25 FPS
Color Depth ........: 24 Bits
Quality Factor .....: 0,030 B/px
Packed Bitstream ...: Yes

Audio Size (Bytes) .: n.a.
Audio Size (KB) ....: n.a.
Audio Size (MB) ....: n.a.
Audio Length .......: n.a.
Audio Codec Code ...: n.a.
Audio Codec Name ...: n.a.
Audio Bitrate ......: n.a.
Channels ...........: n.a.
Sampling Rate ......: n.a.

[manono]

What is that means?

Video Bitrate ......: 213 KB/s
Resolution .........: 720 x 384

Well, among other things, it means your bitrate is way too low for your resolution, and your video looks like garbage.

[Toastie]

Some questions on XviD settings:

Max Consecutive VBOPs: Most how to guides which state that this should be set to 2, but one I found insisted that it be set to 1 for compatablity with some standalone players like the DVP-642. I don't have a 642, but does that really make a difference if someone wanted to play an XviD file I created on a 642? I know about Qpel, Global Motion Compensation, andPacked Bitstream, but does the Max Consecutive VBOPs matter too?

Adaptive Quantization: some guides say to check it, others don't. Any comments on what this is doing and why you would or wouldn't use it?

Chroma Motion: Guides say to check Chroma Motion. However I read somewhere that checking this will make the encode really a CBR encode, so its not worth setting this if you are doing the 2-pass method. True or total nonsence? (I use the 2 pass method)

Also - "Chroma" sounds like something to do with color - what if the video being encoded is in in black and white?

Turbo: One person suggested checking - I'd think that this might make XviD encode faster, but at lower quality? I haven't norticed a diference on my machine, but if anyone can comment on their experiences....


I'd sure like to hear what you guys have to say on these settings, and if maybe they should be different for the 1st pass, as compared to the 2nd pass.


Thanks!

[TheRainmaker]

What is that means?

Video Bitrate ......: 213 KB/s
Resolution .........: 720 x 384

Well, among other things, it means your bitrate is way too low for your resolution, and your video looks like garbage.

But i pointed target size bitrate 1768 so i got output video file 213 kbps...

[manono]

Hi-

But i pointed target size bitrate 1768 so i got output video file 213 kbps...

Did you do the whole thing through GKnot, or only have it make the .avs for you? If you used GKnot, do you have the log for the encode that you could post?

[TheRainmaker]

Hi-

But i pointed target size bitrate 1768 so i got output video file 213 kbps...

Did you do the whole thing through GKnot, or only have it make the .avs for you? If you used GKnot, do you have the log for the encode that you could post?

Not I used .avs file with VDM...

I dont wanna use GKnot again because i am not touching extra settings manuelly...
For example color depth:
All the time when i use GKnot i saw output file's color deph:12 Bit...
But in VDM its 24 Bit pointed already...
If you know and tell me how can i adjust color depth in GKnot i will return GKnot again...

[manono]

Hi-

All the time when i use GKnot i saw output file's color deph:12 Bit...
But in VDM its 24 Bit pointed already...

So you're using Full Processing in VDubMod? You know that defeats one of the reasons to use AviSynth as a frameserver, don't you? The much faster encoding speed. Anyway, all XviD/DivX encoded video is 12 bit. After using VDubMod's Full Processing Mode, it erroneously shows it to be 24 bit. It's really 12 bit.

You're unnecessarily converting to RGB by using VDubMod in Full Processing Mode, slowing the encoding and lowering the quality, because of the unnecessary color conversions. By using Fast Recompress, you keep it YV12 (if from a DVD) all the way.

http://forum.doom9.org/showthread.php?t=109281
http://forum.doom9.org/showthread.php?t=88691

[DJRumpy]

Some questions on XviD settings:

Max Consecutive VBOPs: Most how to guides which state that this should be set to 2, but one I found insisted that it be set to 1 for compatablity with some standalone players like the DVP-642. I don't have a 642, but does that really make a difference if someone wanted to play an XviD file I created on a 642? I know about Qpel, Global Motion Compensation, andPacked Bitstream, but does the Max Consecutive VBOPs matter too?

You should be fine with it set to 2. Unless the DivX player your using complains, you can get better quality with 2. It allows for better compression.

Adaptive Quantization: some guides say to check it, others don't. Any comments on what this is doing and why you would or wouldn't use it?

This setting compresses dark areas of video more than most. It saves bitrate as a result, although some people can notice this. This is a personal preference as to whether or not you notice the compression in those dark areas.

Chroma Motion: Guides say to check Chroma Motion. However I read somewhere that checking this will make the encode really a CBR encode, so its not worth setting this if you are doing the 2-pass method. True or total nonsence? (I use the 2 pass method).

This does not make your encode CBR. It simply uses color information in the video to also detect motion. It is a good thing.

Also - "Chroma" sounds like something to do with color - what if the video being encoded is in in black and white?

You can skip Chroma Motion if the video is true black and white, as it only has luma information.

Turbo: One person suggested checking - I'd think that this might make XviD encode faster, but at lower quality? I haven't norticed a diference on my machine, but if anyone can comment on their experiences....

If you don't notice a difference, then use it. Quality is a subjective thing. This setting takes a few shortcuts to speed things up, with a loss in quality as a result. If you don't notice that loss, then use it. If you don't mind the slower encode speed, then leave it unchecked.

I'd sure like to hear what you guys have to say on these settings, and if maybe they should be different for the 1st pass, as compared to the 2nd pass.

I personally always do both passes at the same time, although I know many people will tweak the settings based on the compressability of the file after the first pass. They examine the stats file and tweak from what they find there. Search the forums for more info on that.

[meldavid]

Great guide Rumpy! I am new at making Xvids -- can you tell me if I am doing something wrong please:

I am converting a PAL DVD of Contact (Jodie Foster) to Xvid AVI. DGIndex reports the moofie as 25 fps Interlaced.

So do I need to use VirtualDub's Smart Deinterlace filter?

Or do I just tick Interlaced Encoding in the Profile tab of the Xvid configuration window.

Thanks.

[DJRumpy]

Actually no. PAL movies always seem to be reported in DGIndex as interlaced when most times they are not (this affects PAL video only). Chances are it's progressive rather than interlaced. For PAL video you can always double check your source with VirtualDub but I would suggest you treat your source as progressive.

[manono]

PAL movies always seem to be reported in DGIndex as interlaced when most times they are not (this affects PAL video only).

Your advice was correct, but the reasoning wasn't. DGIndex reports how a video was encoded and not what the source is like. Therefore, what it reports is accurate. PAL movies on DVD are usually encoded as interlaced even though the sources are almost always progressive. And if the source is progressive, it can be treated as such when converting to AVI. Therefore no deinterlacing is needed. But to be sure, the poster should check all his movies for interlacing. If he doesn't see any, it can be treated as progressive. And if he does see interlacing, that's no guarantee of it being interlaced. It could just have phase-shifted fields and a simple field matcher (Telecide by itself, or TFM by itself) can realign them again.

And it doesn't always apply only to PAL DVDs. Any progressive 29.97fps NTSC source is almost always encoded as interlaced for DVD as well. But so far there isn't very much of that.

[DJRumpy]

I don't recall giving a reason

[meldavid]

I always seem to get confused with interlacing.

If I capture a Foxtel cable TV broadcast using my video camera onto DV tape -- I use this method because I already have this equipment and saves me buying a capture card -- this captured DV footage is interlaced video, right?
I always used Smart Deinterlace when converting this DV footage to Xvid but I don't deinterlace if I am converting to MPEG-2 to make a DVD (I learnt how from one of your other guides! ).

I just assumed that all PAL DVDs are interlaced material because everyone had interlaced TVs when DVDs were released to market and it's only the later DVD players that output a progressive signal.

How exactly do I tell if the video needs to be deinterlaced? Is it when I see pin-striping, like on some sports footage? I've seen some other guides that tell you how to check for interlacing by advancing the frame on action sequences, but no matter how hard I look I'm still not sure what exactly to look out for.

Or do I assume that if it's a high budget cinema release movie, like Contact, that it was shot on 24 fps film and therefore the "original source" was progressive, even though the target medium (DVD) is interlaced?

As you can see I'm still quite confused. Lol.

[DJRumpy]

DV video video always used to be interlaced, although I don't know if that's still the case given the advent of progressive capable video camera's these days. I never use that format and it's been years since I even looked at it.

To tell if a video is interlaced you can create your .D2V project file in DGIndex, while leaving the Force Film option off. That D2V file can be imported into VirtualDub or you can use AVISynth to format it into something that just about any editor can use.

From there it's just a matter of examining each frame to see if it exhibits a combing affect (like pulling a rake through sand). That is the "pin-striping" as you call it. It will be very obvious on fast action scenes. Just step through the video one frame at a time and look at the results.

If in doubt, you should always check, but most PAL video's that are converted from NTSC FILM sources (like Contact) will be progressive,

[djagt3]

I have been working on a GUI for mencoder and have included support for many de-interlacing filters which can be very effective with the preview mode to recognise which filter will do the best job at dealing with DV sources or interlaced videos, I am working on adding this program to the tools selection on this site and will do a tutorial very soon but it realy is easy to use and a tutorial is not essential to make it work, check it out its free and it has many settings available for Xvid encoding. Movie Library ++ download it here.