Video Capturing Concepts: Interlacing Examples
Here are some examples of interlaced and non-interlaced video.
Simple deinterlacing methods lose image data and video quality. Because of the processing power required for an advanced high-quality deinterlace techniques, only simplistic methods can be used on-the-fly at capture time. There are four basic methods available, and none should be used if quality matters.
Drop-field deinterlacing is sometimes referred to as “odd” or “even” deinterlacing, and is a method that throws away 50% of the image to force a very low-quality and rough looking progressive image.
Notice the smooth, crisp lines in the Before image (left). This is the error-free image you’ll see on television, when the interlacing has been retained. The After image (right) is what happens when a field is dropped, leaving you with an incomplete picture. Notice the harsh stair-stepping, or aliasing, of the linear detail. It’s easiest to see in cartoon animation, but is present in live-action footage, too. In addition to the linear errors, there is a 50% reduction in image clarity — the image is much softer.
Blended deinterlacing is sometimes referred to as “double” or “merged” deinterlacing, and is a method that merges both two fields into a single frame. This deinterlacing method always causes significant blurring, ghosting or “mouse trails” to any moving content.
When properly interlaced, you will see Field 1 followed by Field 2, when viewed on a television. The Before image (left) shows interlacing lines, which is how the video is stored; this is only seen when viewed on a computer monitor. The After image (right) shows a blended deinterlace, which creates a low-quality ghosted image.
The raw interlaced image capture will look ugly on a computer monitor, but it will create a perfect image on a TV set. Interlaced video is two fields per frame, but a viewer will only see one field at a time.
Slightly more advanced than an odd/even drop frame. This method removes mice teeth, but still suffers from anti-aliasing lines (stair-stepping, zigzagging motions on linear objects and surface edges) and minor blurring.
Note the difference between the interlaced clip (left) and the adaptive deinterlaced version (right):
Compare again between the adaptive method (left) and the raw drop-field method (right):
Although better, the “adaptive” filter is still only a mild improvement over truly advanced methods available in hardware, and editors like Avisynth (using filters like Yadif, NNEDI, QTGMC, etc). The simple adaptive method will still ghost, create mild stair-step aliasing artifacts, and reduce some of the overall resolution.
Inverse Telecine (IVTC)
Attempts to quickly remove fake interlacing from video that was originally 24fps film. For the purpose of capturing, however, the IVTC is very rough and not at all suggested. Capture-time IVTC looks more like an adaptive deinterlace than a true high-quality IVTC from an advanced video editor, due to digital artifacts. It should also be noted that many broadcasted films have been edited (for special effects, or for censored content) — and in the interlaced domain — making IVTC a destructive process for these videos.
These methods all destroy interlaced footage.
Copyright Notice: All guides, articles and editorials found on digitalFAQ.com are copyright by The Digital FAQ and/or the respective authors. Articles may not be copied, borrowed, full-quoted or reproduced in any manner, online or in print, which includes blogs and forums, without the written email consent of Site Staff (which may or may not be given, for free or fee). Know that digitalFAQ.com staff does routinely monitor online plagiarism, and we do send takedown notices to site admins and/or web hosts (DMCA et al legal actions) as is necessary. If you would like for others to read articles found on The Digital FAQ, simply link to our content. (Note: Printouts for personal use is specifically allowed.)