What's new

PQ differences: Codec or Mastering related? (was "AVC better than VC-1???") (1 Viewer)

DaViD Boulet

Senior HTF Member
Joined
Feb 24, 1999
Messages
8,826
This is not an HD DVD versus BD debate. Both formats can and do use VC-1 and AVC video compression CODECs. Both are "advanced video" codecs and their use is studio-dependent since they both can acheive high quality and similar bit-rates.

However, Dan Ramer had noticed a trend that many AVC titles, on average look better than many WB VC-1 titles. The problem, of course, is that it's nearly impossible to know for sure what the original uncompressed master would have looked like to know how faithful the compression is being.

But in this review of Shrek three, there's a clue. He noticed that the special features looked better than the feature film... and only then did he check to see what video codecs were being used in each case:

http://www.dvdfile.com/index.php?opt...=6382&Itemid=3


Now, this still doesn't offer conclusive evidence that the lesser-quality of the feature film is due to VC-1. It's possible that it was slightly filtered to aid in compression. I offer up an example of a film I reviewed here at HTF: The Incredibles on DVD. In that case, the feature film and bonus material was all compressed with MPEG2... yet the clips of the film in the bonus material was NOTICABLY sharper and more detailed than the feature film...which showed that the film had been "filtered" despite all its high praise. It's possible that's what we're seeing here with Shrek 3 as well, though the different video codecs being used in these two cases does invite some questions.
 

RobertR

Senior HTF Member
Joined
Dec 19, 1998
Messages
10,675
I agree with Lou. It's like any other comparison: It's not valid unless you are CERTAIN that ALL variables are controlled.
 

Shane Martin

Senior HTF Member
Joined
Sep 26, 1999
Messages
6,017
I think this debate about AVC vs VC1 has already been decided and the answer was No. Paramount encoded Flags of our Fathers in BR using AVC with a ridiculously high bit rate and encoded the HD version with VC1. The consensus as the time was no difference. If there was one, it's so indistinguishable by eye that it doesn't matter. The only people proclaiming they could notice a difference were blu ray people and you have to know some sort of bias was in place for sure.

The key is the efficiency of the codec, not the bitrate.
 

Carlo_M

Senior HTF Member
Joined
Oct 31, 1997
Messages
13,385
I lost the link, but there was a site out there comparing screenshots of the various HD codecs. In general it was apaprent that AVC or VC-1 was superior to MPEG-2, but comparisons between AVC and VC-1 discs were, in general, hard to distinguish. Most looked identical to my eyes.
 

Jeff Adkins

Senior HTF Member
Joined
Sep 18, 1998
Messages
2,842
Location
Tampa, FL
Real Name
Jeff Adkins
No, even some of the HD-DVD fans said there was a difference but not enough to for them to warrant the extra bandwidth.

My AVC titles seem to look more film-like and have more definition than the VC-1 titles I own. Both codecs can look spectacular, but on average the AVC stuff is superior IMHO.
 

Marc Colella

Senior HTF Member
Joined
Jun 19, 1999
Messages
2,601

I'm assuming you're comparing both codecs on the same title, otherwise you're comparing apples to oranges.
 

troy evans

Screenwriter
Joined
Jul 2, 2005
Messages
1,294
Oh No, Now we're going to discuss codec differences instead of format differences. This should prove interesting. :rolleyes
 

DaViD Boulet

Senior HTF Member
Joined
Feb 24, 1999
Messages
8,826

Still-frames aren't really ideal for comparing compression codecs bcs compression algorithms and bit-rates tend to reveal their artifacts/transparency during motion.
 

DaViD Boulet

Senior HTF Member
Joined
Feb 24, 1999
Messages
8,826

Agreed.

But when you see trends between 100 apples and 100 oranges, though it doesn't point to any definitive conclusions, it does raise some legitimate questions.
 

Douglas Monce

Senior HTF Member
Joined
Nov 16, 2006
Messages
5,511
Real Name
Douglas Monce
I don't think I have seen anything where anyone has been able to show that VC-1 is superior to AVC or the other way around. The only thing I have seen is a huge frame enlargement from Flag of our Fathers where there MIGHT be a little bit more detail in a very small part of the frame in the AVC version. Frankly it might be that AVC handled that particular kind of information better than VC-1. It might be that if you went searching through that same title with a fine tooth comb, you will find other shots where VC-1 does a better job than AVC.

Honestly if you have to enlarge part of the frame 20 or 30 times to be able to see the difference, you might as well give it up and just enjoy the movie.

Doug
 

RobertR

Senior HTF Member
Joined
Dec 19, 1998
Messages
10,675
It would also help if the comparison was done blind, to eliminate any possibility of bias.
 

Carlo_M

Senior HTF Member
Joined
Oct 31, 1997
Messages
13,385
You know, I'm just not a fan of blind, double blind, etc. tests for audio or video.

Why? Because it's so hard to remember what something sounded or looked like when you have to take time to change over to the other audio or video stream. By the time it's done, I've already forgotten what the other stream looked/sounded like. Even in DBTs the mind is a powerful thing that can exert any biases, even when it's below the level of consciousness.

What would be nice would be to have both projected simultaneously, one above the other, or side by side, or if possible splitscreened, by identical source and display devices that were ISF'd to be as close as possible.

Of course, who is going to go to all that trouble? AVC/VC1 both look stunning to my eyes.
 

RobertR

Senior HTF Member
Joined
Dec 19, 1998
Messages
10,675
There are devices used for audio DBTs that switch in a matter of milliseconds, which should be brief enough. I'm not sure how long HDMI switchers take, but it can't take long. This isn't directed at you, Carlo, but I find it interesting that those who are "certain" there are differences never complain about the length of time between nonblind comparisons.
 

Carlo_M

Senior HTF Member
Joined
Oct 31, 1997
Messages
13,385
No doubt what you say is true Robert, but to the vast majority of us who are mere mortals, these devices do not exist.

My Sony KDS-60A3000 takes at least 3-5 seconds to go between inputs, so when I try to show people even the difference between DVD in component vs. BD on HDMI, it's nearly an impossible comparison because people are staring at a black screen for 3-5 seconds which serves, unbelievable as it sounds, to blur even short term memory.

When I tried several years ago to DBT my MB-Quart vs. Energy speakers with 3 friends, it was a bear of a time for me to level match and then try to go A/B with them, because level matching on the fly was impossible with just one receiver. The impedences differed so I had to up one by 3 db while switching back and forth. It was a hassle and of course, I wouldn't put too much stock in the results if they were challenged "scientifically".

So what am I saying? Or trying to say?

I guess it's this: true, easy, accurate DBT'ing sounds good in theory, but for most mortals who don't have the equipment [and I don't even know what equipment you refer to, nor would I be likely to buy it simply because I only have so much money and I'll reserve that for upgrading my system or buying discs rather than buying DBT equipment], is nearly an impossibility from a practicality standpoint. I know, I've tried. And any result I've gotten from DBT I wouldn't put significantly more stock in than just showing (visually or aurally) someone one thing, then changing to another. But this is a topic for a whole other thread, so I'm going to try and gracefully bow out.
 

Douglas Monce

Senior HTF Member
Joined
Nov 16, 2006
Messages
5,511
Real Name
Douglas Monce

If the differences are so insignificant that you can't remember them, I think we should probably just call it a draw.

Doug
 

Mark Zimmer

Senior HTF Member
Joined
Jun 30, 1997
Messages
4,318
Wasn't there discussion at AVS about the AVC codec process automatically laying on filtering? That would be consistent with what I'm seeing, where there's excessive filtering on AVC titles, with edge enhancement often being slapped on top in order to compensate. Kind of the worst of both worlds. Or it could just be the people doing the AVC encodes are filter-happy and need to learn to keep their paws off the knobs. Part of that is the trend by all too many "reviewers" to be critical about film grain that's supposed to be there, so the studios are making the mistake of eliminating it--and a lot of the fine detail in the process.
 

Douglas Monce

Senior HTF Member
Joined
Nov 16, 2006
Messages
5,511
Real Name
Douglas Monce

If Close Encounters is any indication there is no automatic filtering going on with AVC. There is very natural looking film grain through out the film.

Doug
 

ppltd

Senior HTF Member
Joined
Aug 30, 2006
Messages
3,041
Location
Phoenix
Real Name
Thomas Eisenmann
Couldn't agree more. Even MPEG2 done properly can look stunning. Comparing video codecs seems to be a bit silly and impossible to validate one way or the other. So many other variables play apart in the final look of a release, the codec used is of minor consequence.
 

Carlo_M

Senior HTF Member
Joined
Oct 31, 1997
Messages
13,385
110% agreed. (to quote an overused sports cliche)

Heck, even MPEG2 on DVD is highly impacted by other factors. I can tell this after buying my new 1080p TV and using my PS3 to upconvert DVDs to 1080p. Obviously commercial DVD all use MPEG2. But some transfers look only marginally better than LD (and some straight to video DVDs look like VHS!), while some transfers are so good that when upconverted they give Blu-ray a run for its money! All while using the same CODEC.

So while I can appreciate that there is undoubtedly a difference between AVC and VC-1, a well done transfer from a great source should yield very little difference to the naked eye.
 

Users who are viewing this thread

Latest Articles

Forum statistics

Threads
356,814
Messages
5,123,657
Members
144,184
Latest member
H-508
Recent bookmarks
0
Top