Encoding Video Materials for DVD

The DVD specification requires video movies to be in a DVD-compliant MPEG format. There are a number of methods for converting video from its current format to MPEG.

DVD Studio Pro includes an integrated MPEG encoder, which can quickly encode your QuickTime movies into DVD-compliant MPEG-2 files. See Using the Integrated MPEG Encoder for more information.

For analog sources, you need a hardware video capture card to digitize your video. Depending on the card’s output format, you may then need a transcoder to convert it to MPEG.

Elementary and Multiplexed MPEG Files

There are two common types of MPEG file structures: elementary and multiplexed. Elementary files contain only one stream of data, either the video or the audio. Multiplexed MPEG files, also known as system, transport, or program files, have the video and audio combined into a single file.

MPEG files that are imported directly into DVD Studio Pro must be elementary MPEG files. You can use an MPEG utility to convert multiplexed MPEG files into elementary streams. The integrated MPEG encoder converts QuickTime files into elementary MPEG streams.

When creating an HD project, you can import HDV and H.264 streams, both of which are multiplexed QuickTime streams.

Choosing Between MPEG-1 and MPEG-2

The DVD specification allows for the use of either MPEG-1 or MPEG-2 encoded video. In general, MPEG-1 provides for smaller file sizes but reduced picture quality (by scaling a half-screen image to full screen), while MPEG-2 provides for better quality with larger file sizes.

  • MPEG-1 is generally used for titles for which file size is critical and that will be played back on a computer (not a set-top DVD player).

  • MPEG-2 is usually used for DVD titles on DVD media. The integrated MPEG encoder encodes video into MPEG-2.

DVD Studio Pro works with both MPEG-1 and MPEG-2 video assets in your SD project tracks, as long as they are DVD-compliant. MPEG-1 video used as menu backgrounds is rendered to full D1 MPEG-2 when the project is built.

Important: HD projects cannot use MPEG-1 video.

Choosing a Bit Rate for SD Projects

Bit rate directly determines file size, and thus how many minutes of material you can fit onto a DVD disc. SD-based DVD players support combined bit rates of up to 10.08 megabits per second (Mbps), but this must include the audio and subtitles as well. The maximum video bit rate is 9.8 Mbps.

Note: Some DVD players have trouble playing video that uses sustained high bit rates. On such players, you may see dropped frames during playback.

Typical maximum bit rates used for video are around 8 Mbps. The maximum MPEG-1 bit rate is 1.856 Mbps, while 1.15 Mbps is most typical.

The actual video content determines how low a bit rate you can use and still get acceptable quality. Video with little movement works well at lower bit rates, while higher bit rates are best at handling video with a lot of movement (a quick camera pan, for example).

For more information on choosing a bit rate that ensures all the content in your project will fit on a DVD, see Making Sure Your Content Will Fit.

Choosing a Bit Rate for HD Projects

Choosing a bit rate for use in HD projects is similar to doing so for SD projects, but there are some differences.

The maximum overall bit rate for HD projects is 30.24 Mbps. While an HD project can use both SD and HD video assets, the maximum bit rate you can use is different for each:

  • HD assets: 29.4 Mbps

    Note: HD assets include the 480p and 576p resolutions.

  • SD assets: 15.0 Mbps

MPEG Encoding Methods

There are three common methods of encoding video: CBR (constant bit rate, also known as one-pass), one-pass VBR (variable bit rate), and two-pass VBR.

  • CBR (one-pass) method: You choose a bit rate and it is used for the entire video, regardless of its content. The advantage of this method is that you can reliably predict the final encoded file size. By adjusting the bit rate, you can ensure the video will fit on the DVD. The disadvantage is that the same bit rate gets applied to all of the video, regardless of the motion content. You must choose a high enough bit rate to provide acceptable results for the scenes with motion, even if they are only a small part of the movie.
  • One-pass VBR method: You choose a basic bit rate and a maximum bit rate. The encoder detects the amount of motion present in the video as it encodes and adjusts the bit rate appropriately—scenes with motion use higher bit rates (up to the maximum setting you choose) and scenes with little motion use the lower, basic bit rate. The disadvantage of this method is that the quality will not be as good as with the two-pass VBR method.
  • Two-pass VBR method: You choose a basic bit rate and a maximum bit rate, just as with the one-pass VBR method. The difference is that the encoder makes two passes through the video to encode it. The first pass examines the video to learn about its motion content. The second pass performs the encode, varying the bit rate to provide better results in scenes with motion (as compared to the one-pass VBR method). The disadvantage of this method is that it takes twice as long as the one-pass VBR method.

Important: With the VBR methods, the basic bit rate setting determines the encoded file size. The maximum bit rate setting does not affect the file size. Additionally, the VBR methods are most effective when the maximum bit rate is about double the basic bit rate, or at least 1 to 3 Mbps higher than the basic bit rate.

The integrated MPEG encoder allows you to select which of these methods you want to use.

Choosing a Video Resolution

The resolution determines the number of pixels processed within a frame. The more pixels used, the sharper the details in the image. Note that regardless of the resolution you choose, most DVD players play back at full screen. The following is a list of common resolutions used with MPEG-1 and MPEG-2 video with DVD Studio Pro.

Note: Menus must use full D1 resolution MPEG-2 video. If you assign any other resolution or an MPEG-1 asset, the video will be transcoded to full D1 MPEG-2 when you build the project.

SD Projects

SD projects can use the resolutions shown in the following tables.

NTSC

All frame rates use the interlaced scanning method (as indicated with an “i”). With the exception of 352 x 240, all resolutions are MPEG-2 only.

Resolution
Frame rate
Aspect ratio
Notes
352 x 240
29.97i
4:3
Also known as SIF format; MPEG-1 or MPEG-2
352 x 480
29.97i
4:3
Also known as 1/2 D1
704 x 480
29.97i
4:3, 16:9
Also known as Cropped D1; 16:9 is anamorphic
720 x 480
29.97i
4:3, 16:9
Also known as Full D1; 16:9 is anamorphic
PAL

All frame rates use the interlaced scanning method (as indicated with an “i”). With the exception of 352 x 288, all resolutions are MPEG-2 only.

Resolution
Frame rate
Aspect ratio
Notes
352 x 288
25i
4:3
Also known as SIF format; MPEG-1 or MPEG-2
352 x 576
25i
4:3
Also known as 1/2 D1
704 x 576
25i
4:3, 16:9
Also known as Cropped D1; 16:9 is anamorphic
720 x 576
25i
4:3, 16:9
Also known as Full D1; 16:9 is anamorphic

HD Projects

HD projects can use the resolutions shown in the following tables.

Note: MPEG-1 video is not supported.

NTSC

Frame rates with an “i” indicate the interlaced scanning method; those with a “p” indicate the progressive scanning method.

Resolution
Frame rate
Aspect ratio
Notes
352 x 240
29.97i
4:3
Also known as SIF format
352 x 480
29.97i
4:3
Also known as 1/2 D1
704 x 480
29.97i
4:3, 16:9
Also known as Cropped D1; 16:9 is anamorphic
720 x 480
29.97i
4:3, 16:9
Also known as Full D1; 16:9 is anamorphic
720 x 480
59.94p
16:9
HD only, also known as 480p; is anamorphic
1280 x 720
59.94p
16:9
HD only, also known as 720p
1440 x 1080
29.97i
16:9
HD only; 16:9 is anamorphic
1920 x 1080
29.97i
16:9
HD only, also known as 1080i
PAL

Frame rates with an “i” indicate the interlaced scanning method; those with a “p” indicate the progressive scanning method.

Resolution
Frame rate
Aspect ratio
Notes
352 x 288
25i
4:3
Also known as SIF format
352 x 576
25i
4:3
Also known as 1/2 D1
704 x 576
25i
4:3, 16:9
Also known as Cropped D1; 16:9 is anamorphic
720 x 576
25i
4:3, 16:9
Also known as Full D1; 16:9 is anamorphic
720 x 576
50p
16:9
HD only, also known as 576p; is anamorphic
1280 x 720
50p
16:9
HD only, also known as 720p
1440 x 1080
25i
16:9
HD only; 16:9 is anamorphic
1920 x 1080
25i
16:9
HD only, also known as 1080i

Picture Frame Rate

DVD Studio Pro NTSC projects support the NTSC frame rate (29.97 fps) and 23.976 fps assets that use NTSC video resolutions. PAL projects support the PAL frame rate (25 fps) and 23.976 fps assets that use PAL video resolutions. You can set a project’s video standard in the General tab in the Disc Inspector or in the Project pane of DVD Studio Pro Preferences.

Warning: You cannot mix NTSC and PAL MPEG assets in a project.

Working with GOP Settings

A major function of MPEG encoding is to identify redundancy not only within a frame, but also among a group of frames. A scene with no movement except a car driving by has a lot of redundancy over time. In this case, a single complete frame could provide the majority of the video; subsequent frames only need to provide for the areas that change. Each MPEG stream contains three types of frames that define the video. A Group of Pictures (GOP) setting defines the pattern of the three frame types used: Intra (I) frames, Bi-directional (B) frames, and Predicted (P) frames.

  • I-frames: Also known as reference or key frames, I-frames contain the complete image of the current frame, without reference to frames that precede or follow it. The I-frame is the only type of MPEG frame that can stand by itself, without requiring information from other frames in the GOP. Every GOP contains one I-frame, although it does not have to be the first frame of the GOP. I-frames are the least efficient MPEG frame type, requiring the most disc space. Markers on a DVD track can be placed only at I-frames.
  • P-frames: P-frames are encoded from a “predicted” picture based on the closest preceding I- or P-frame. P-frames are also known as reference frames, because neighboring B- and P-frames can reference them. P-frames typically require much less disc space than I-frames.
  • B-frames: B-frames are encoded based on an interpolation from I- and P-frames that come before and after them. B-frames are the most efficient MPEG frame type, requiring the least amount of disc space. While a GOP can begin with a B-frame, it cannot end with one.

The goal of encoding video to MPEG is to reduce the video file size as much as possible while maintaining good quality. The B- and P-frames are what make MPEG so good at compressing video into such small files. Because they only contain information about what has changed in the images, they use much less disc space than I-frames.

There are three aspects to choosing a GOP setting: the GOP pattern, the GOP length, and whether the GOP is “open” or “closed.”

Figure. Example of a closed GOP with an IBBP pattern.

GOP Pattern

The GOP pattern determines the ratio of P- to B-frames within a GOP. The most common patterns used for DVD are IBP and IBBP. All three frame types do not have to be used in a pattern; an IP pattern is also used in special circumstances.

IBP and IBBP GOP patterns, in conjunction with longer GOP lengths, produce the most efficient encodes. Smaller GOP patterns with shorter GOP lengths work better with video that has quick movements.

Some encoders can force I-frames to be added sporadically throughout a stream’s GOPs. These I-frames can be placed manually by someone watching for sudden video changes or automatically by an encoder detecting scene changes. You can place markers in Final Cut Pro to identify specific frames to have an I-frame, either for use as a chapter marker or to manually identify high motion areas. See Adding Markers to Your Video for more information.

The integrated MPEG encoder uses a GOP pattern of IBBP.

GOP Length

Longer GOP lengths produce more efficient encodes by reducing the number of I-frames but are less desirable during short-duration effects such as quick fades and quick camera pans.

  • SD-Based DVDs: The maximum GOP length for SD projects is 18 frames for NTSC or 15 frames for PAL. The minimum GOP length depends on the GOP pattern. For example, an IP pattern can have a length as short as two frames.

    The integrated MPEG encoder uses a GOP length of 15 frames for NTSC and 12 frames for PAL SD assets.

  • HD-Based DVDs: Because HD projects can use both interlaced and progressive video assets, they have different GOP length maximums.
    • NTSC interlaced assets: 18 frames per GOP
    • NTSC progressive assets: 36 frames per GOP
    • PAL interlaced assets: 15 frames per GOP
    • PAL progressive assets: 30 frames per GOP

    Note: 24p assets have their GOP structure based on their display rate, not the encoded rate. A 24p asset encoded to play in an NTSC project has a display rate of 29.97 interlaced frames (with an 18 frames per GOP maximum) or 59.94 progressive frames (with a 36 frames per GOP maximum).

    The integrated MPEG encoder uses a GOP length of 15 frames for NTSC and 12 frames for PAL HD assets.

Open or Closed GOPs

An open GOP allows the B-frames from one GOP to look at an I- or P-frame from an adjacent GOP. Open GOPs are more efficient, but cannot be used with multi-angle or mixed-angle titles. A closed GOP keeps all references within itself, standing alone with respect to other GOPs. DVD Studio Pro works with either type in single-angle titles and requires closed GOPs with multi-angle and mixed-angle titles.

It is important to know that the same pattern can provide different results when used with an open or closed GOP. For example, the integrated MPEG encoder uses a closed GOP type, and it starts its IBBP pattern with an I-frame. Other encoders using an open GOP may start with a B-frame. In this case, starting with a B-frame is a little more efficient; if you start with an I-frame, an extra P-frame must be added to the end (a GOP cannot end with a B-frame).

General Quality Issues

The quality of the MPEG encoding determines the quality of the title. It is well worth putting some effort into getting the best possible MPEG encoding environment. For example, start with the best quality source material you can. An uncompressed digital source is the best, followed by a DV source, and then an analog Component source. Of the other analog formats, S-Video (Y/C) is much preferable to Composite.

Noisy Video

Because a large portion of MPEG’s compression power depends on its ability to locate areas of the image that do not change between successive frames, “noisy” video can reduce the efficiency of MPEG encoding. While the actual video may be identical between two frames, if there are noise artifact differences between the two frames, they are detected as a change in video and require more bits to encode. A lot of noise causes poor encodes, because the encoding resources are tied up trying to faithfully reproduce the noise.

Transcoding Artifacts

Every time video is transcoded from one format to another—for example, from analog to digital, or DV to MPEG-2—artifacts are introduced into it. The visibility of these artifacts depends on the quality of the transcoding device and the content. While transcoding cannot be avoided, its effects can be minimized. Use the highest quality equipment and software you can, and choose products that work well together and support each other’s formats.