In this training video we look at the different methods of deinterlacing footage like the field combination, field extension, and motion detection methods to help you make your footage look its best.
One of the most unfortunate inventions of the 20th century was interlacing. It may have worked well for broadcast but splitting video into two fields has wreaked havoc on editors the world over ever since. That's why we'll be showing you the different methods of deinterlacing footage like the field combination method, field extension method, and motion detection method so that you can know which method will be the most efficient at making your footage look its best.
Before we start going over how to deinterlace footage, it's good to know what interlacing is and how it came to be. Interlacing was invented back in the 1930's in order to make video frame rates appear to be twice as fast as they were without increasing transmission or file size in broadcast. By recording in this method, fast action in videos like sports games could look smooth despite the limitation in frame rate. To do this, each signal was split into two fields: an odd field and an even field. Each field was recorded or displayed every 1/60th of a second. Unfortunately this means that when working with this kind of footage in an editing system, it can be difficult to do accurate frame by frame edits since each frame of is really two fields of video taken at separate times. In addition, this also means that playing interlaced footage on progressive screens like LCDs and Plasmas can end up looking pretty bad since the interlaced lines will be visible. This is why so many editors end up choosing to deinterlace their footage and why we'll be looking at the different ways it can be done.
The first method of deinterlacing is called the field combination method. This is the most straight-forward way to get rid of those pesky interlaced lines. Basically field combination takes both fields of video and overlays them into one whole frame. Here's what our video looks like with interlacing and here's how it looks after using the field combination method. Doing this makes frame by frame editing easier and keeps the vertical resolution of your picture but also results in a low quality movement called ghosting or combing. Combing is when something in motion in your video appears to have a low opacity double of itself near the object in motion. This appears as a result of showing two fields which are captured at slightly separate times together in one frame of your video. Due to this fact, the field combination method is the least preferred amongst editors since temporal resolution is so low in the image. However, this method is the fastest way to deinterlace video and is the only way some editing programs have to deinterlace footage so you may still have use for it after all.
Another common way to deinterlace footage is by using the field extension method. Field extension is done by copying either your odd or even field and pasting it over your missing field of video. Here's what our video looked like before deinterlacing and here's what it looks like after using this method. As you can quickly see, this method reduces the vertical resolution by half but keeps a smooth motion since our frame rate is now the same as the field rate in our video. Unfortunately, this method is often used by editing programs to deinterlace on the fly due to the simplicity of rendering the effect. However, much like field combination method it's not a preferred way of deinterlacing due to the perceivable drop of quality in the image.
The last and best method of deinterlacing footage is to use motion detection. This advanced method works by having your software reconstruct the missing field of video in your image by looking at the frame that comes before and after the current field and building a new one based off of the position of the objects in motion. Here's how our footage looked with interlacing and here's what it looks like with motion detection deinterlacing applied. As you can see, this method looks the best since it's actually building new fields of video instead of throwing them away or combining them. The only downside of this method is the sheer amount of time it can take to process a piece of deinterlaced footage. Some encoding programs such as Compressor and Sorenson Squeeze can literally take hours to render a 5 minute piece of video using this method. Not great news for someone who's on a tight deadline. This method can also produce some odd motion in your video when the software has trouble predicting motion in your series of frames as you can see in this video here. Nonetheless, this method is often worth the time it takes to render your footage due to how much better it can make your interlaced footage look. Keep in mind, it's always best to deinterlace footage before you bring it into your editing software as any scene change in your final video will tend to confuse your encoding software which can lead to some interesting results.
Most editors will be relieved to hear that even if you're stuck with interlaced footage, it doesn't mean your footage can't look its best. Instead, by using one of the deinterlacing methods we've shown you can help your footage look great no matter what screen it's on.
