1.866.840.3062

Video has come a long, long ways in the last few years in terms of web friendly video. High definition video cameras can now record video as progressive footage, meaning there is no need to de-interlace the footage prior to viewing it on a computer monitor. By nature, high definition (or HD) footage is also convenient for computer monitors since it uses square pixels, thereby eliminating headaches associated with pixel aspect ratios. By having similar characteristics to that of a computer monitor, HD video is ideal for the web.

But what if you’re not shooting in HD? What if you’re shooting DV footage? How do you make DV footage look good on the web? Let’s talk about DV footage first.  Native DV records at 720×480 non-square pixels. DV footage is also interlaced meaning that each “frame” of video is actually made up of two fields (each containing half of the information) that are combined together to create an image. Computer monitors on the other hand, use square pixels and display video progressively (or non-interlaced.) So as you can see, DV footage, in nature, clashes with the web.

How to get DV footage to look good on the web

Now let’s talk about how we get our DV footage to look good on computers and (more importantly) the web. There are two major adjustments that need to be made to the video. Remember that no matter how little the alteration is, you’re going to lose some quality. The amount you lose depends immensely on the methods you use when altering your video.

First, we need to de-interlace the video.

As I said before, computer monitors display video progressively. What this means is that each “frame” of video is one image, unlike interlaced video that is broken up into two separate fields. So what we have to do is make the video progressive or “de-interlace” the video. This will eliminate those annoying “saw teeth” or jaggy edges you see on footage that wasn’t de-interlaced.

Most NLE (or non-linear editors) contain built in de-interlacing methods or filters. In most cases, the software will either blend the fields together or interpolate (eliminate half the lines and use the average of the two lines to fill in the missing data.) There are also a large variety of third party plug-ins and filters available for specifically de-interlacing video footage. 

For example, Virtualdub, a free video utility, contains several advanced de-interlacing filters that can provide very nice results. For myself, lately I’ve been using a plug-in for Adobe After Effects called Topaz Enhance to do my de-interlacing. It’s made by Topaz Labs and is by far the best I’ve used. A free 30 day trial is available. 

Second, we need to change the pixel aspect ratio.

Since DV footage natively records in non-square pixels, we need to change it to match the square pixel ratio that computer monitors use. Without doing this, the footage will look stretched and disproportional. There are several ways this can be achieved.

For myself, I use Adobe After Effects. I like to import the footage and change the interpretation to “square pixels.” Then I’ll move the footage into a new composition and render it out at 640×480. This tends to work very well and is quite easy to do, especially if I’m already de-interlacing the footage.

Many NLEs will also contain presets for doing this. For example, Sony Vegas contains a specific video size preset for “640×480 square pixels.” [Note: 640x480 is the resolution at which DV footage will look correct when using square pixels.]

Again, remember that when applying these changes you’re going to lose some quality. Therefore, be sure to experiment a bit and figure out which method works best for you.

So now you’ve got yourself some footage that was once native DV and is now 640×480, using square pixels and de-interlaced. What are you waiting for? Upload that!