Once Again, Micro$oft Lies?
Dear Gary: Shocking! Absolutely shocking! You've been running articles about Microsoft’s VC-1 compression. I had my doubts about their claims that they were better than MPEG 4. It turns out that once again M$ has claimed their product is better than the competition and isn’t according to tests done by SMPTE. M$ has also made exaggerated claims about the licensing. Turns out they don’t have all of the patents to the intellectual property they are using. How predictable. How does this company stay in business? You think they would learn their lesson at some point! Here’s a link to the article: http://www.eetimes.com/showArticle.jhtml?articleID=50500181 I emailed you a while back that Widescreen Review should do some testing to see if M$'s claims were true but SMPTE has done that work for you. I look forward to your continued coverage of these important issues.
Kraig Bailey
Contributing Editor Joe Kane Comments:
On October 18, Junko Yoshida of the EE Times, as you noted in your letter, posted an article, “M’soft Coup Starts Media Codec Fight” (http://www.eetimes.com/showArticle.jhtml?articleID=50500181), which uses many unnamed sources to criticize the progress being made with VC-1 within SMPTE. Among other charges, the EE Times indicates that the SMPTE process is broken and that Microsoft has somehow bullied its way into being considered. It would appear as if the issue you brought up in your letter to Widescreen Review about Windows Media Video is not going to go away. This reminds me of the discussions on the Internet forums and the Presidential election this year. There is a reply to the original EE Times article that appeared a week later. Perhaps studying this and subsequent replies in EE Times, as well as attending one of our educational demonstrations regarding Windows Media Video, will help to clarify what is actually happening. This week (December 5) we are in the final stages of encoding materials from Digital Video Essentials in Windows Media Video (WMV). We’re finding that it is good enough to challenge any playback system we’ve been able to obtain. Of course I wonder about that part of the controversy, knowing that not all digital is created equal. We continue to be reminded of the role played by the playback system, independent of anything in WMV, as we swap video output cards and see a large difference in image quality. We could condemn the quality of the entire DVD format based on the lack of quality of a few DVD players. So how good was this demo? From what we read in the article I’d question everything about what they did including their motivations. But then I have my own motivations, something about wanting to see the best possible picture on the screen. It’s obvious to anyone following the politics of the next generation of codec systems that the competitive stakes are high, with the media, such as Widescreen Review, playing an important role in how this is portrayed to the public. Letters like Kraig’s and the articles in EE Times put out controversial assertions based on who knows what in the reality of their evaluation. My pictures and my demonstrations tell a completely different story, one that has surprised every one who has seen them, including people from Microsoft. I care enough about my demonstrations to let the video speak for itself. If it’s good, the picture will look great. You know my pictures look good and Windows Media Video, along with Samsung, have made that possible. We don’t know how much of a role politics or playback system capability is playing in the original EE Times article, but the reporting is certainly less than ideal. As I said, this reminds me of things that go on in elections. I don't know how far into the article I should go but from my experience as Chair of the SMPTE Working Group on Professional and Studio Monitors, the Society’s role is to create workable standards, not test and compare products being standardized. The idea that a SMPTE committee would issue a document on a comparison of codecs as part of their standards work isn’t likely. Did anyone at the EE Times take the time to fact check this article before it was published by at least calling the SMPTE headquarters? As for the test itself having taken place, it’s something I’ve tried to organize using my own materials and found some unbelievable hurtles in getting my video encoded to H.264, let alone finding the equipment that can play it back in real time. I can encode my 1080p/24 and 720p/24, and 720p/60 video to WMV or VC-1 (VC-1 is the name of WMV's SMPTE standard) in my home, as we are doing today for the Professional version of Digital Video Essentials. As for encoding to H.264, one company quoted an hour per frame. My software based VC-1 encoder isn't real time but it works in a time frame I consider real. At this point in time, the VC-1 decoder is far less complex than the H.264/AVC decoding. This should lead to better pricing and more widespread applications for VC-1. It should also result in longer battery life and lower energy consumption, which is important for home and portable devices. In the EE Times article there was speculation that the Blu-ray and DVD Forum groups could delete VC-1 from their specification. The demonstrated reality is that the SMPTE approved the latest VC-1 ballots in November by an overwhelming majority (only about 7 percent voted negatively), and at the last December meeting of the Steering Committee in the DVD Forum, nearly all companies voted positively to confirm the Advanced Profile of VC-1 as mandatory for HD DVD (i.e., 18 voted “yes,” none voted “no,” and two abstained). The most important point for me, and the reason I’m spending so much time demonstrating and providing video in VC-1, is that it holds up to the quality of my masters. I’ve spent a lot of time working with Samsung in helping them build a display device that will let me see what's in the video signal. When we first started looking at Windows Media Video, we saw things going on in the picture we had never seen in the 24 Megabits per second MPEG-2 version of DVE. We traced the problems back to the mastering process, something we hadn't seen on the 31-inch-wide monitors that are used in program production. WMV in combination with the Samsung SP-H700AE projector was good enough to show us we could do better in production, and this after our spending a year getting the production equipment up to an acceptable standard as displayed on a 31-inch-wide monitor. In the Professional version of DVE, we’ve taken the time to go back and re-edit portions of the program to show that, with VC-1, you can see the difference. It’s all covered up in MPEG-2. When the encoder/decoder combination is good enough to show us the differences in accepted production practices, we feel good about delivering in VC-1. The fact that we can get this quality while encoding the video on our computer at home is even more impressive.
You can E-mail Widescreen Review @ mailto:editorgary@widescreenreview.com