Opinion: Page (1) of 1 - 09/24/01 Email this story to a friend. email article Print this page (Article printing at MyDmn.com).print page facebook

Garbage In, Garbage Out

Video Formats Suitable for DVD By Paulo de Andrade
Theres an old saying in the industry: Garbage in, garbage out. It couldnt be truer for DVD encoding. All DVD producers want their shinny discs to look great. After all, one of the advantages of the medium is to offer viewers virtually the same image quality that you and I can see in the studio. And DVD makes this possible for the first time ever.

Gone are the quality losses that weve had to live with for so many years when VHS and laserdisc were the only available consumer formats. But, in order to deliver this killer quality level, a lot of attention must be paid to the encoding (or transcoding) process. Several elements influence the encoding process, but the most important ones are the quality of the original material and the required media (DVD) space.

You should never expect a DVD created from a second-generation S-VHS edited master to look as good as a commercial DVD that started with professionally scanned 35mm film stored on uncompressed D1 or D5 digital videotape. MPEG-2 compression doesnt fix bad video. At best it replicates it. In reality, bad video can make the compressed movie look even worse and take up a lot more storage space. Thats because S-VHS (or any other consumer/prosumer analog format, for that matter) is very noisy. The colors are noisy because the frequency response of these formats is much lower than that of broadcast or digital formats. And the luminance part of the video is noisy for the same reasons. Plus lets not forget the resolution (or lack of) issue. Put these elements together and you end up with a recipe for sub-standard DVDs. After all, consumers do expect DVD video to look great.

Noisy video can be particularly bad because MPEG-2 compression is lossy as it discards redundant information in order do save storage space. It can be a very efficient CODEC when there is very little change within a shot or between frames. The problem with video noise is that its completely random and its pattern changes entirely from frame to frame. The MPEG-2 encoder doesnt know what is video information and what is noise and all it sees are a lot of changing pixels. Therefore, it will produce much larger files. And here is where the amount of available space plays a very important role. If the presentation is short (up to 1 hour long), you can use a very high bitrate and the encoder will still give you great looking video (as good as the original, anyway). But if your presentation is longer, you must throw away some information in order for the video to fit on a DVD. If you use fixed bit rate encoding, you will have to compromise the quality of the final product. And even if you use variable bitrate encoding, the noise causes so many changes between frames that theres not much that the encoder can do to create smaller files.

If you are creating a DVD from scratch, always use the best quality source you can obtain. Digital video formats are great for DVD because they are virtually noise free, have solid colors and high frequency response. Uncompressed formats such as D1 and D5 are the best, followed by compressed formats such as DigiBeta, D9 and DVCPro 50. Next come DVCam, DVCPro and DV, which in spite of being more compressed can produce outstanding results if the sources are of high quality. In fact, theres a huge difference in image quality coming from a broadcast-quality DV camera and a prosumer model. Most of the artifacts associated with DV are completely eliminated or become practically invisible. On the other hand, consumer DV cameras must be avoided because they produce video that looks very amateurish. As a minimum requirement, a DV camera must have 3 CCDs in order to produce decent colors.

All these digital formats use component video signals, just like DVD, which means that the encoded color signals will remain separate and the image will look better defined with no bleeding edges or chroma crawl. Beta SP or MII are analog formats that produce great results because of very low noise, stable colors and good resolution. They are also component formats, which is a big plus. Most other formats would not yield very good results for DVD production these days.

If you must produce a DVD based on existing footage, you cant always have the luxury of picking the ideal videotape format to work with. If you can obtain an original in the above formats, thats great. But you may be forced to work with a less desirable format. D2 and D3 are composite digital formats that produce very good quality. But because all the color and luma signals are mixed into a single one, you lose some of the advantages of DVDs component signals. Even thought a DVD can still output component, the signal will still be the original composite as it was encoded. A step down in quality, but still acceptable, is one inch videotape. This analog format offered very high quality at the time and was widely used as a master storage medium. Like D2 and D3, it is also a composite format.

If you are forced to work with U-matic, S-VHS, Hi-8, 8mm or VHS, you will end up with an inferior looking DVD, no matter how careful you are. These formats are very noisy and the colors dont look that great. They should only be utilized as a source for a DVD if theres absolutely no other option. If a client approaches you with one of these formats as source material, make sure that you explain ahead of time that the resulting DVD will not look that great in order to avoid future headaches.

Paulo de Andrade
senior producer

Page: 1

Related Keywords:DVD Creation, authoring, encoding, transcoding, Beta SP, MII, 3/4" U-Matic, D1, D2, S-VHS, VHS, Hi-8, 8mm, MPEG-2


Our Privacy Policy --- @ Copyright, 2015 Digital Media Online, All Rights Reserved