Originally Posted by Richard Paul
I am merely pointing out some facts and expressing an opinion. And my opinion on the video codecs has always been that both MPEG-4 AVC and VC-1 are capable of reference quality encodings on Blu-ray. Also just to mention this but based on the information from the blu-raystats website
(which has detailed information on most Blu-ray titles released by the major studios) over 90% of Blu-ray titles released in the United States last year were encoded with MPEG-4 AVC.
From what I heard, Microsoft disbanded the group that was advancing the professional VC-1 encoder which post houses were using after I left. Without support, companies gave up on using it and shifted to the only option available to them with was AVC.
That Dr. Gary Sullivan used to work on your team isn't what I was referring to.
Nothing strange about it. It shows that we knew quite well the capabilities of MPEG-4 AVC as we were developing VC-1. Indeed, Gary reported to the same team that developed VC-1. And the VC-1 team created technology that became part of MPEG-4 AVC. Importantly, Microsoft became a patent holder as a result, and benefited from its broad adoption.
It is just my opinion but what I find a bit strange is that you point out that he worked on your team when someone questions you on HEVC but that I remember that you had the opinion that MPEG-4 AVC was fundamentally flawed. Of course those posts were made several years ago and opinions can change with time.
My opinion has not changed. And you left out the last chapter in that discussion where I showed with full blown simulations that Ron was wrong. Saving others from reading all that, the debate was around the de-blocking filter in both codecs and how they could soften video. I said AVC did a lot of that and Ron said that was not the case in the response Richard provided. What Richard left out was my super detailed response from me back to Ron's post:http://www.avsforum.com/avs-vb/showt...9#post10029559
And the post after it: http://www.avsforum.com/avs-vb/showt...9#post10029569
Both showed that what Ron said was incorrect. That the MPEG-4 AVC did indeed overfilter pixels and hence, produced softer images:
"VC-1 pixels filtered: 54234 (3%)
AVC pixels filtered: 326148 (16%)
VC-1 pixels filtered more than once: 54 (0.00%)
AVC pixels filtered more than once: 115181 (5.51%)
Ratio of AVC to VC-1 pixels being filtered: 6.01:1
AVC pixels touched twice: 102594
AVC pixels touched three times: 10686
AVC pixels touched four times: 1901
Ratio of AVC pixels touched more than once compared to VC-1: 2133:1"
I can't find Ron's (dr1394's) answer now but I distinctly remember him putting down his sword saying he could not counter because he lacked the analysis tools we used to see what is going on in the two codecs.
Anyway, you are helping me with the argument relative to this thread. If folks follow those links, you can see that if I am uninformed about video compression, I sure know how to write a few books about it here
You may also want to know Richard that post leaving Microsoft, I have had countless meetings and cordial discussions with the blu-ray studios. So while you may have trouble letting go of the format war, none of us in the industry have that issue.