AVS Forum banner
941 - 960 of 2128 Posts

·
Registered
Joined
·
2,995 Posts

Quote:
Originally Posted by Mr. Hanky /forum/post/0


???



So the "grain" in the vc-1 shot looks like a natural distribution to you?


I don't how much more simply this example can be distilled. The mpeg2 shot has captured the grain in a natural distribution (macroblock artifacts or not), while the vc-1 shot has attempted to "connect" the grain into weird shapes/blocks and apply blending where macroblock edges would have occured. The end result- the "high frequency" grain pattern has been nuked, and what's left is some low frequency phenomenon, which no longer resembles any kind of natural grain behavior.

And I don't know how much more simply I can put this. There is no additional detail in the MPEG-2 picture in this sample. What you are calling grain is sharp edges of macroblocks. They are artifacts, not detail.
 

·
Registered
Joined
·
4,940 Posts

Quote:
Originally Posted by xradman /forum/post/0


And I don't know how much more simply I can put this. There is no additional detail in the MPEG-2 picture in this sample. What you are calling grain is sharp edges of macroblocks. They are artifacts, not detail.

There is MPEG detail within the blocks (not the edges) where VC-1simply has a one color block. So how is this possible?
 

·
Registered
Joined
·
6,234 Posts
Can you answer this, xradman? What is this, if not "detail"? Are you saying these are "macroblock edges" that exist inside other macroblocks?
 

·
Registered
Joined
·
5,233 Posts

Quote:
Originally Posted by jmpage2 /forum/post/0


Now when will the blu-ray brain trust admit that there is banding and serious macroblocking in some of the darker scenes in POTCTBP despite it being billed as a "flawless" transfer?

Well, where are the pictures of that? And I thought DMC was the "flawless transfer", not Black Pearl.



I think it's hilarious that people defend vc-1 like it's infallible or something. It's a video codec, like mpeg-2 and avc. It can suck too, despite what certain tools tell you.
 

·
Registered
Joined
·
10,954 Posts

Quote:
Originally Posted by Kram Sacul /forum/post/0


Well, where are the pictures of that? And I thought DMC was the "flawless transfer", not Black Pearl.



I think it's hilarious that people defend vc-1 like it's infallible or something. It's a video codec, like mpeg-2 and avc. It can suck too, despite what certain tools tell you.

Ya, sorry, it's DMC that's billed as the flawless transfer. Where are the pics? Right in this thread.


I'm not defending VC1, I'm pointing out that AVC is not God's gift to encoding as some would like to believe.
 

·
Registered
Joined
·
6,234 Posts
...except no one has pushed this idea, hence a strawman argument. Vc-1, otoh...
 

·
Premium Member
Joined
·
9,850 Posts
I'd like to add one obvious point to this discussion:


Yes, the VC1 (grain) example that keeps getting posted here from MI3 does not show grain very well. BUT, one must remember that:


(a) this is part of a background shot where the scene was out of focus. As such, even though the grain would certainly have been present anyway, it was actually not important to this scene that the grain was lost in this area of the picture.


(b) This is one section of one scene of one movie - to say that this represents how VC1 will always encode grain in other scenes or other movies would be a gross misrepresentation. As we know, the grain is preserved with far more detail in many other scenes of MI3 alone, let alone in other movies.


Showing this comparison repeatedly as an implication that Mpeg2 is better than VC1 simply because of the way each handled the grain in a less-important backround scene in one part of one movie is rather simplistic and unfair, IMO. It would be akin to folks picking one section of one scene for a particular Mpeg2 or AVC encode and repeatedly touting it as being typical of those codecs.
 

·
Registered
Joined
·
5,233 Posts
Rdjam, if you actually followed the discussion you'd know there are other examples of vc-1 screwing up when presented with grain. The filtering in We Were Soldiers and the WTC blurry UFO patches are two others and they're pretty hideous. Those examples are certainly not in unimportant areas of the frame.
 

·
Premium Member
Joined
·
9,850 Posts
With maximum respect, We were Soldiers was a terrible master and looked pretty bad on both BD and HD. But, simply, I just re-iterate the points of my post - which are that basically isolated examples do not necessarily make the norm. Mpeg can make a dog's dinner of grain also.


IMO, there is nothing fundamentally weak about VC1 at all - in fact, I would go further and say that it's "best example" encodes (such as King Kong) are still unbeaten by the other codecs for raw encoded detail.
 

·
Registered
Joined
·
19 Posts
U guys are comparing Ferraris and Lamborghinis here... both have their highlights and either one is better than that POS Ford you use to drive.


enjoy the movies!


there im done ranting, continue with color testing and pixel counting ; )


/flameshield on
 

·
Registered
Joined
·
508 Posts

Quote:
Originally Posted by xradman /forum/post/0


And I don't know how much more simply I can put this. There is no additional detail in the MPEG-2 picture in this sample.

No (this of course apply to this particular screenshot).


Mgeg2 has more details, but suffers blocking (due to not high enough bitrate).

I've seen many sources before and after encoding and I can tell you that there are more details in Mpeg2 screenshot.


Which one looks better- Mpeg2 will look better on CRT screens, VC1 on plasma and LCD.
 

·
Registered
Joined
·
9,884 Posts

Quote:
Originally Posted by Andrew_HD /forum/post/0


No (this of course apply to this particular screenshot).


Mgeg2 has more details, but suffers blocking (due to not high enough bitrate).

I've seen many sources before and after encoding and I can tell you that there are more details in Mpeg2 screenshot.


Which one looks better- Mpeg2 will look better on CRT screens, VC1 on plasma and LCD.

MPEG-2 once may have handled interlace better. Is that what you are saying? Because I don't think it is true anymore.


Though we are obviously moving away from interlaced displays anyway.


- Tom
 

·
Registered
Joined
·
508 Posts

Quote:
Originally Posted by trbarry /forum/post/0


MPEG-2 once may have handled interlace better. Is that what you are saying? Because I don't think it is true anymore.


Though we are obviously moving away from interlaced displays anyway.


- Tom


I'm saying that on this screenshot Mpeg2 has more details and is obvious for me .

It doesn't mean that MI3 with VC1 looks worse



Andrew
 

·
Registered
Joined
·
7,205 Posts
Discussion Starter · #955 ·
Coming To America


17.70 GB AVC ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ ............................................................ 18.30 GB VC-1
 

·
Registered
Joined
·
1,540 Posts

Quote:
Originally Posted by Dot50Cal /forum/post/0


VC1 is handling the red's better, check out his tie.



Xylon's blu-ray player software still isn't filtering when doing YPbPr 4:2:0/4:2:2 to 4:4:4 chroma upsampling, but his hd-dvd software is, that is why you can see jaggies around the red tie. This is nothing to do with VC1 vs AVC. Both hd-dvd and blu-ray are the same 4:2:0/4:2:2 format. If both players were using the same chroma up-sampling algorithm I bet the screen shots would look pretty much the same. The Sony PS3 doesn't produce jaggies like that, which I mentioned before when he posted the terminator 2 screen shots, but he has yet to comment. This is swaying a lot of his hd-dvd vs blu-ray screen shot comparisons towards hd-dvd. Personally I think this is enough to invalidate his screen shot comparisons until his blu-ray software is doing the right thing.
 

·
Registered
Joined
·
1,441 Posts

Quote:
Xylon's blu-ray player software still isn't filtering when doing YPbPr 4:2:0/4:2:2 to 4:4:4 chroma upsampling, but his hd-dvd software is, that is why you can see jaggies around the red tie. This is nothing to do with VC1 vs AVC. Both hd-dvd and blu-ray are the same 4:2:0/4:2:2 format. If both players were using the same chroma up-sampling algorithm I bet the screen shots would look pretty much the same. The Sony PS3 doesn't produce jaggies like that, which I mentioned before when he posted the terminator 2 screen shots, but he has yet to comment. This is swaying a lot of his hd-dvd vs blu-ray screen shot comparisons towards hd-dvd. Personally I think this is enough to invalidate his screen shot comparisons until his blu-ray software is doing the right thing.


?????? Or maybe the avc encode sucked? How do you know what his player is doing?
 

·
Registered
Joined
·
1,540 Posts

Quote:
Originally Posted by sound dropouts /forum/post/0


?????? Or maybe the avc encode sucked? How do you know what his player is doing?

every one of his blu-ray screen shots has the chroma bug. it's a problem with his player software. I know it's not the encode because it looks fine on my player.
 

·
Registered
Joined
·
5,233 Posts
Yeah, something is not right. Whatever AVC decoder that is being used has a chroma bug.


The T2 mpeg-2 shots exhibit this as well but they are mpeg-2. It's probably a filter graph issue.
 
941 - 960 of 2128 Posts
Top