DV to DVD guide

The following guide aims to help you through the first steps in creating a DVD from a DV source.

You'll need the following software for this guide

your favorite DV editing program (see below for a list of programs)
AviSynth
BeSweet & GUI
CinemaCraft Encoder (CCE), TMPGEnc, Canopus Procoder or MainConcept MPEG Encoder
your favorite DVD authoring program

0. Transfer video from Camcorder to your PC

Before you can start editing your video, you first have to transfer it to your PC. Most DV camcorders have digital outputs that allow for a quick and lossless transfer.


1. Cutting, Editing, Transitions, etc.

The first step in DVD creation is the most creative part: Editing your video. This can be done using a variety of basic DV editing programs like Ulead VideoStudio, MGI VideoWave, Pinnacle Studio, or using a professional software, e.g. Adobe Premiere, Ulead MediaStudio Pro, or SonicFoundry Vegas Video. The large variety of editing programs and the different editing options they offer is beyond the scope of this guide. So capture your DV file from your camcorder using the editing software of your choice, and edit it until the result pleases you (obviously I can't tell you where to cut and which effects to use, as everybody has a different opinion on this subject).

Once the editing step is done, save the DV file as either type-1 or as a type-2 AVI. Save the audio separately to a WAV file (your DV editing program should offer this option - below is an example how this could look).



2. Encoding the audio

You can use either MP2 or AC3 (Dolby Digital 2.0) audio for a PAL DVD. MP2 is not allowed on NTSC DVDs, but certain players still support it. Make sure your DVD authoring software supports AC3; if not, you have to use MP2. My preferred authoring program is DVDlab, which supports AC3. Use BeSweet to convert your WAV file to the desired target format. There is a BeSweet GUI available, which you can use for MP2 encoding; for AC3 I prefer AC3Machine, which is a GUI for BeSweet as well, but focused on AC3 encoding.

2.1 MP2 encoding

Start up BeLight, drag and drop your WAV file onto BeLight, select the MP2 DVD preset and start encoding. For more information on MP2 encoding with BeSweet, check out this guide.

2.2 AC3 encoding

Start up BeLight, drag and drop your WAV file onto BeLight, select the AC3 192 kbit/s 2 channel preset and start encoding. For more information on AC3 encoding with BeSweet, check out this guide.

3. Encoding the video

DVDs use MPEG-2 as video format. There are several good MPEG-2 encoders available: CinemaCraft Encoder (CCE), TMPGEnc, Canopus Procoder, MainConcept MPEG Encoder, and a few more. In this guide I'll focus on CCE 2.66, TMPGEnc 2.5, Canopus Procoder 1.5, and MainConcept MPEG Encoder 1.4.2.

If you aim for a specific size, you need to calculate the video bitrate first. I suggest to convert all file sizes to Megabytes, so that you don't get confused. We start with the size available. Let's say you want to fill a whole DVD-R, so the size available is 4.38 GB or 4485 MB. Subtract the size of your audio file (in MB), e.g. 167 MB, so the space available for the video is 4485 MB - 167 MB = 4319 MB. Calculate the duration of the video in seconds, e.g. if the video is 1h 44min 10sec, the total length in secs is 6250. So the video bitrate should not exceed 4319 * 1024 * 8 / 6250 kbps = 5660 kbps. Because there's some muxing overhead, and you probably want to add menus and chapters, you should get lower than that, so in the example above I'd use an average bitrate of 5500 kbps.

3.1 CCE

3.1.1 Create an AviSynth script

The way you write your AviSynth script depends on the source and on your personal taste. Here are two examples to start with; of course you may change filter settings, use different filters, etc.:

Progressive DV

code:
AviSource("C:\MyDVFiles\DV_type2_file.avi") Convolution3D(1, 6, 10, 6, 8, 2.8, 0)
or
code:
DirectShowSource("C:\MyDVFiles\DV_type1_file.avi", 25) Convolution3D(1, 6, 10, 6, 8, 2.8, 0)
(the 25 above is the frame rate for PAL; for NTSC you'd use 29.97.)

Interlaced DV
code:
AviSource("C:\MyDVFiles\DV_type2_file.avi") SeparateFields() odd=SelectOdd.Convolution3D(1, 6, 10, 6, 8, 2.8, 0) evn=SelectEven.Convolution3D(1, 6, 10, 6, 8, 2.8, 0) Interleave(evn,odd) Weave() DoubleWeave.SelectOdd()

This way each field will be filtered separately, thus the interlaced structure will be preserved. Replace AviSource with DirectShowSource like above if your DV AVI is type-1.

NTSC issues

As NTSC DV uses the 4:1:1 colorspace, there is an issue with the conversion to the 4:2:0 colorspace used in DVD MPEG-2. To address this issue use trbarry's filter ReInterpolate411().

To get improved color conversion, simply add a
ReInterpolate411()
command below the AviSource or DirectShowSource command, respectively.

3.1.2 EclCCE

EclCCE is a nice tool provided by RB, which extends CCE's capabilities, e.g. it enables the CCE trial version of saving ECL (project settings) files. But the addition I like best is the integration of a shutdown checkbox, so that you can shutdown your computer after the encoding has finished.

3.1.3 CCE settings

Start EclCCE (or CCE directly). Open the AVS file you created in step 3.1.1 by right-clicking the CCE window and selecting Add. The AVS appears in the CCE window. Double-click the new entry to change the encoding settings:

You can use MPEG-2 (ES, One-pass VBR) if the video you want to encode is not very long, you don't need to meet a given size exactly, and you want to save time during encoding. In this case select a Q factor of 60 or less - I often use 40.

Usually you'll want to have a predictable size with the best quality possible. You'll choose MPEG-2 (ES, Multipass VBR), set the calculated bitrate as your average bitrate, and set minimum and maximum bitrate to 2000 and 8500 respectively. Use at least 2 passes; 3 passes is even better, but more is useless.

Set the correct aspect ratio, 4:3 for most DV sources, 16:9 if you recorded in 16:9 mode. Don't process the audio in CCE (we will use BeSweet for better quality).

In the Video settings select DVD video compliant , and uncheck the Upper field first box (field order correction is performed in AviSynth script).

In the menu you get when clicking the Quality button, deselect Low pass filter, Effect restricted vertical filter, and Dithered quantization. I leave the flat part priority at the default of 16. Set Intra block DC precision to 9 or 10 Bits. For interlaced sources set Block scanning order to Alternate. For progressive sources select Zigzag and check Progressive frame flag.


 

For more information on the various settings in CCE, check out the CCE guide. If you're using CCE Basic, follow the CCE Basic guide instead, but keep the DV specific settings in mind (block scanning order, progressive / interlaced and the field order)

3.2 TMPGEnc

It is really easy to use TMPGEnc for DV to DVD conversions. Start the encoder, cancel the wizard (if applicable), and open your DV type-1 or type-2 file via the Browse button next to the Video source field. Your AVI file will automatically be entered in the Audio source field, too, and an Output file name will be suggested.

We are going to create elementary streams as input for a DVD authoring program. In the Stream type box select ES Video+Audio, if you want to create an MP2 file in TMPGEnc (skipping the BeSweet step above). Otherwise, select ES Video only (MP2 or AC3 created by BeSweet). Click the Load button, and open a template, either DVD (PAL).mcf or DVD (NTSC).mcf.

Click on the Setting button to make further adjustments. Because you loaded a template, many options are greyed out. If you'd like to change these for some reason, load the template unlock.mcf, which you will find in the subdirectory Extra. In this guide we won't need that.

Let's have a look at the Video tab first. You may leave the Rate control mode settings to their respective defaults (Constant Quality, quality factor 65, min. bitrate 2000 kbps, max. bitrate 8000 kbps). If you need to fit your movie to a certain size, select 2-pass VBR(VBR), and click the Setting button next to the rate control mode. Adjust the Average bitrate to your needs.

The Encode mode should be set according to your source material. If it's interlaced, select Interlace; if it's progressive, select Non-interlace.

Leave the DC component precision at 9 bits, or set it to 10 bits.

The Motion search precision has a large impact on the final quality, but on the processing time as well. Motion estimate search (fast) is a good compromise between quality and speed. If you want the best quality possible, select High quality (slow) or even Highest quality (very slow). But be warned: your encoding session can take many hours!

Move over to the Advanced tab. Set Video source type according to your source (should be the same as Encode mode on the Video tab).

The Field order is for interlaced sources only. DV should be Bottom field first (field B), unless you have a strange source...

Under Source aspect ratio select one of the 4:3 or 16:9 settings, depending on your source again. E.g. I use 16:9 625 line (PAL) for my 16:9 recordings, and 4:3 625 line (PAL) for my 4:3 recordings.

Also select Full screen (keep aspect ratio) as the Video arrange Method.

You may activate filters here, too. I'll leave this topic to your experiments.

As we won't make changes to the GOP structure or the quantize matrices, skip to the Audio tab (which is not present if you selected stream type ES (Video only)). You may change the audio bitrate here, if you want TMPGEnc to encode your audio.

Confirm the tabbed dialog by clicking the OK button, then click on Start in the upper left and wait until TMPGEnc is finished.

For more information on TMPG, check out this guide.

3.3 Canopus ProCoder

Start ProCoder. In the Source tab, add the DV file you want to encode. Once the file is opened, you'll see the Source Parameters on the right, containing information about the video and the audio. The only thing you may want to change here is the Aspect Ratio setting, which defaults to 4 x 3, but has to be changed into 16 x 9 for 16:9 sources.
Switch to the Target tab, and click the Add button. The Add Target dialog box opens, and you continue on its VCD/DVD tab. Select the appropriate template, either MPEG2 - DVD PAL, or MPEG2 - DVD NTSC. Confirm by clicking OK.

The Target Parameters contain all the settings you can change before starting the conversion. First you may want to change the Base Name and the Path, which specify filenames and location of your output files (video and audio). Under Stream Type, make sure MPEG2 Video + Wave file is selected.

Interlacing will usually stay on Lower/Odd field first, unless your source is progressive. In that case select Non-Interlaced. Set Video Aspect Ratio to 4:3or 16:9, depending on whether you shot your DV in 4:3 or 16:9 mode. And there's the Pixel Aspect Ratio setting, which has to be set to one of the D1 entries (D1/DV NTSC, D1/DV NTSC 16:9, D1/DV PAL, or D1/DV PAL 16:9), depending on your video format and whether you shot in 4:3 or in 16:9 mode.

Under Speed/Quality you may select Highest Quality. If you don't care for processing time, Mastering Quality may give slightly better results.

CBR/VBR Type should be set to either VBR 1-pass or VBR 2-pass. The latter needs twice the encoding time, but you'll get closer to the specified bitrate, and thus your filesize will be predictable. For the same (average) bitrate VBR 2-pass reaches better quality. There are two video bitrates to be set: Video Bitrate(kbps) is the average bitrate you calculated at the beginning of step 3, whereas Max. bitrate(kbps) specifies the limit for the encoder. Set that one to 8000.

There are a lot more options available when you click the Advanced button, and you can add video and audio filters. As a beginner you can safely leave everything to the defaults here, but of course feel invited to experiment! E.g. if you recorded at 24 fps, you'll need to add a Video Pulldown filter.

Now you may want to save your profile ;-) You can create so-called droplets, too, but you have to find out yourself how to do that :-)

Finally switch to the Convert tab, click the Convert button and take a cup of coffee. You should end up with an m2v video file (MPEG-2 elementary stream) and a WAV audio file.

Special thanks to Dali Lama for the inspiration!

3.4 MainConcept MPEG Encoder

Start the MainConcept MPEG Encoder (MPE in brief). Open your DV video file by clicking the Open button behind the Video source line. Among MPE's supported file types are DV type-1 / type-2 and AviSynth scripts (AVS files). As soon as you open the video file, MPE will show the same file in the Audio source line, and it will suggest an output file.

You may verify the file properties in the Video input info and Audio input info boxes. In the Output format box, select DVD as MPEG type, and set the Video mode to your needs, either PAL or NTSC. If you want MPEG-1 Layer II audio, you can let MPE create the audio together with the video. Because you'll need to author the DVD later on, you better select Elem. Video and Audio as Stream type, not Program (Video+Audio). If you want to encode the audio separately, select Elementary Video here, and the output file's extension will automatically change to .mpv.

Now click on the Details button for the detailed settings. Here you can change the video size, which you should keep at 720x576 for PAL or 720x480 for NTSC. You can crop and scale the video, if desired, but I guess this is more interesting for analogue sources. An important setting is the 2-pass encoding checkbox, which you may want to check for best quality at lower bitrates. I doubt that you'll see an effect except for the doubled processing time, if you go for bitrates above 6000 kbps, but in general 2-pass encoding will produce better quality at a given bitrate.

Set the Frame rate to 25 for PAL or 29.97 for NTSC. Most camcorders don't have a progressive scan option, so most likely you have to select Bottom Field First as Field encoding. If your DV video is progressive, you should select Progressive Frame here, of course.

The Search method and Search range sliders allow you to modify the motion estimation search and thoroughness. The more you move the sliders to the right, the more precise will motion estimation be (and thus the better quality), but this also increases the encoding time. You can live with the defaults, but you may as well increase the values for better quality MPEG-2.

If you encode audio, the settings should be correct automatically (48000 Hz, Stereo). You should not enable file splitting in the Muxer box.



Let's dig a little deeper into MPE's dialog structure by clicking on Advanced. Here you find a true El Dorado for tweakers, but amazingly you'll find some important basic settings here, too.

On the Video Settings tab, select the proper Aspect ratio, i.e. the aspect ratio you set in your camcorder when filming. You may select either 4:3 Display or 16:9 Display, the latter for anamorphic video. If your camcorder supports letterboxed 16:9 only, you probably have to select the 4:3 option. Don't touch the GOP structure settings, and you'll get DVD compliant MPEG-2 video.

You can set Variable bitrate or Constant bitrate as the Bitrate type, no matter whether you chose 2-pass encoding or not. The variable bitrate option requires a maximum, an average, and a minimum bitrate. I suggest to leave max. and min. bitrate at the defaults of 8000 kbps and 2500 kbps, respectively. You may set a higher max. bitrate, but that may lead to a few problems with authoring and playback later on, depending on the software and standalone player (and even the DVD-R media) you use. The most important setting is the Average (kbps), because this one is responsible for the file size you'll get, so set it to the value you calculated. The constant bitrate setting is easy: simply set your bitrate, there's only one. Leave the Rate control mode at its default, which is Mode 1.

The Advanced Video Settings tab has some more settings to play with: You may switch between Noise sensitivity and Noise reduction by clicking the NR button, and you can play with the value, which may improve noisy low light shots. I won't cover all the settings here, but there's one which is useful specifically for DV encoding: You may want to check Input video is RGB16-235, if your DV codec is known to reduce the color range, and you feed the encoder with RGB video. You may verify this using the histogram in VirtualDub, but remember that this setting applies only if your source is RGB, e.g. converted in AviSynth or via another frameserver, so normally you don't have to worry about this setting.

Confirm all settings with their OK buttons, and click Convert in the main dialog. Now all you need is patience until the encoding is done. MainConcept's MPEG Encoder is pretty fast, almost as fast as CCE, so the coffee break won't be too long.

4. Author the DVD

There are several DVD authoring programs available on the market. Make sure you choose one which can import your video and audio files without re-encoding E.g. DVDlab is a good choice, because it is low-priced, and you can import the video, MP2 audio, and even AC3 audio without re-encoding

The process of creating the menus and the DVD structure depends very much on the authoring program used, so I won't focus on this topic. Low-budget programs often use a "single-VTS" approach, i.e. all your video and audio files are muxed into a single video title set. As a consequence you have to make sure that you use the same kind of audio in all your clips (if you plan to put more than one on a DVD). For an extensive list of DVD authoring guides, click here.


Any comments are very welcome. bb

This document was last updated on January 15, 2006