Gary Merson's bob and weave tests

mdanderson -- Thu, 07/06/2006 - 23:03

Does anyone know which issue of TPV contains the follow up to Mr. Merson's excellent article in the Dec. issue of TPV in which he conducted bob and weave tests on HDTV deinterlacing? He stated he would be conducting more tests on new sets as they came out. Thanks.

Matt

scottwilkinson -- Fri, 07/07/2006 - 00:20

Gary and TPV parted ways before he could follow up on that article, which was just about the time I came onboard as video editor. The information he presented was interesting, but incomplete and somewhat misleading. It’s important to understand that bobbing and weaving are the simplest forms of deinterlacing, and each is better suited for a different situation. Weaving is better for static (non-moving) images or pixels (Gary was using a static test pattern in his tests), whereas bobbing is better for fast-moving images or pixels. The problem is that most video images include a combination of stationary and moving objects. If the deinterlacer incorrectly bobs stationary pixels, the result is a half-resolution progressive image as Gary discussed. If it incorrectly weaves moving objects, the result is “combing” artifacts. Robust, per-pixel, motion-adaptive deinterlacing makes the right decision for every pixel in the image, preserving detail and preventing artifacts. Advanced deinterlacers also employ diagonal filters to reconstruct some of the lost detail in moving areas that need to be bobbed. As Gary asserted, weaving is better than bobbing in that the resulting progressive frames contain detail from both fields, but only when accompanied by intelligent motion-adaptive decisions.

Scott Wilkinson
Video Editor
The Perfect Vision

mdanderson -- Fri, 07/07/2006 - 19:42

Mr.Wilkinson,

Thanks so much for your quick reply and for the valuable info you gave on Gary Merson's article. I am trying to make a decision on the new Sony SXRD A2000 sets which are scheduled to come out in the next few weeks. The ability to pass the deinterlacing test is important to me and I will have to wait until the new sets come out and are reviewed before I can make an intelligent decision. Like you said, the article is not a be all and end all test but in does provide useful info.

Thanks again for your contribution and for all the great articles in TPV.

Matt Anderson

scottwilkinson -- Fri, 07/07/2006 - 21:04

Thanks for your kind words. If you don't want to wait for a review, I suggest getting a copy of the HQV Benchmark from Silicon Optix; go to www.hqv.com to order. This is a great test DVD to evaluate the quality of any video processor. If you take the DVD to a store that has the Sony available to look at, be sure to play it from a DVD player set to interlaced output to evaluate the quality of the display's processor. The disc does not provide tests for bobbing and weaving specifically, but it does have many tests that let you clearly see how well the processor does deinterlacing, noise reduction, diagonal filtering, etc. The DVD also includes good explanations of all the tests and what to look for. I suggest running through it before you go to the store so you are familiar with it.

Scott Wilkinson
Video Editor
The Perfect Vision

mdanderson -- Sun, 07/09/2006 - 12:52

Scott,
I forgot about the HQV Benchmark test disc. I have heard good things about this dvd and I appreciate you suggesting it. Thanks again.

Matt