Harry Potter and the Order of the Phoenix: The ADM EE

[dark jedi]

OK dudes I have it, and made a SL version of it, I am going to watch it tonight and probably start upping tonight as well.

DJ

 

[AvP]

Thanks DJ

 

[DoctorM]

Um, this wasn't an official DVD5 version from ADM?
What method did you use to compress it?

 

[dark jedi]

ADM and I both use Nero recode,but I will be upping the DL version for you guys to compress for yourselves,because I know some of you may cry about recode,but we like it,and it is alot faster than DVD RB Pro(although I do use this sometimes myself).

I don't want to spend 4-5 hours compressing,and then another 1-2 hours rarring,just to get to upping.

I have alot of stuff on my plate at the moment and it will probably take me a couple of weeks just to get caught up.

DJ

 

[DoctorM]

Not wanting to get into a debate on this, but DVD Rebuilder's One Pass VBR can be quite fast and as good or better in quality than some multipass encodings.

 

[SwordStone]

Interesting, never thought of doing that.

 

[joebshmoe]

you want to do at least 2 passes for quality video encoding.

the first pass is just a guide for the actual encoding process to get to a preferred size while still looking good.

in this way you don't waste bits on areas not needing as much per frame and give it to other areas that otherwise would look worse.

 

[DoctorM]

Which would be true if you didn't understand how OPV works.

 

[joebshmoe]

then i guess i need to learn more about OPV

* goes googling *

edit:

how do you think OPV is working differently?

everything i'm seeing on OPV supports my understanding of VBR and multiple passes related to video encoding.

i will admit that most of my knowledge on compression comes from audio rather than video and sometimes the "definitions" are a tad different in implementation.

edit2:

ok so the CCE implementation of OPV actually does an analysis/prediction (but not 100% like in multipass) as a first pass so "OPV" is kinda misleading. it is using an incomplete first pass to predict the desired settings of the encoder.

still, the difference between OPV and a true 2-pass VBR encode would be:
* the first pass with 2-pass VBR was complete
* final output would be closer to desired size with 2-pass VBR
* quality will be better, but not always obvious with 2-pass VBR

the only positive outcome of OPV is time not used to encode (so you save 95% of the time of that "first pass").

so... i'd still recommend a 2-pass VBR encode for the above reasons

 

[DoctorM]

OPV is done as a single pass. No analysis or prediction is done by CCE. The best description is that it's constant quality (though constant quantization is more technically correct).

For a given Q value the quality should be unchanged for the entire movie. Distribution of bits is unimportant since it will use as much as it needs to hit the set Q value.

Software like DVD-Rebuilder uses analysis/prediction on test clips only to extrapolate a final file size for different Q values.
Once it hits on one that will fit on a DVD5 it encodes with that.
One pass, constant quality.

Most experts (read: the geeks on Doom9) agree that it is equivalent in quality to 3-4 multipass VBR.

To that end I'm using DVD-rb on this. The Q value estimated is 3!
Lower is better. 1 is suppose to be almost indistinguishable from the source.

And if the analysis idea still haunts you, it only took my elderly PC 5-10 minutes to figure out the Q.

 

[dark jedi]

DoctorM, could you post exaclty how to set up DVD RB Pro for those settings,I would love to give that a shot,if it truly is better than Recode,I am not above switching.
also can you add how much you let DVD RB Pro take from the extras,I mean all settings would be nice,I want to try it exactly the way you do it.

Thanks

 

[ADigitalMan]

DJ, ya took the words right out of my mouth. I have DVD RB but use it sparingly because of the time it takes, and don't know how to get it above 2 passes. Nero gives such good results that I haven't lost any sleep. I used to use Pinnacle back in the day which was great, but slow. I moved on from it about three years ago for Nero 6, which I'm still using.

 

[Ghostcut]

I remember doing 5 passes on a DL DVD with DVD Rebuilder and CCE, which came out crystal-clear quality, but took a full day to do it on my lovely old laptop.

 

[joebshmoe]

DoctorM>

i need to understand this more than i do now, but my gut reaction is avoidance, LOL

thanks for the info, etc.

ah i found a great post by the author of dvdrb from doom9:
http://forum.doom9.org/showthread.php?t=122724&page=3

The OPV quality is going to be similar to the multiple pass encoding -- no matter what Q is selected. The Q is selected based upon what is needed to meet output size requirements. What you'll find is that if a Q is high, the quantization in a multipass encode will also be high.

The downside to OPV is it's limited output size accuracy. When you do a prediction, you are doing a "best guess" as to what Q is needed to meet sizing requirements. It works a lot like multipass -- but multipass does a 100% first pass (making the output prediction very exact) -- OPV does a series of 2% passes, making output prediction less accurate. Where OPV gets tough is in the granularity of setting Q factors. When Q factors get smaller, single increments can make extreme differences in sizing... so it may not be physically possible to get close to the target size. In that case, multiple passes will probably do better -- because it utilizes that lost space. Some people get around that by doing a different Q prediction for each segment -- making sizing accuracy a little better. But in my opinion that also detracts from the quality of playback -- because you will notice quality differences between the segments. Ideally the entire movie should be encoded with the same Q.

It's a trade-off... accuracy in sizing for encoding speed. For those people who get wrapped around the axle over a couple hundred wasted megabytes -- OPV is probably a bad idea. It isn't unusual at all to get a 4.10GB output size for an OPV DVD. Frankly, though, the true visually noticable effect of that unused space is pretty insignificant.

so... it is a bit contradictory to what you are saying (there is a prediction first pass with CCE, called "analysis" that OVP does partially that a multipass would do completely).

from other formats (xvid, for example) it is DESIRABLE to have a varying quantizer selection (if that quantizer selected is lower - i.e. higher quality - in sections to achieve a closer size output i.e. 4.4GB for an SL DVD).

so... i still need to look closer at OPV but it still looks like a 2-pass VBR encode will be better. really, it's almost technically impossible for a 1-pass to be higher in quality to a 2-pass (when you consider a size constraint!). without a size constraint than a 1-pass, true quality based encoding would be obviously higher quality (by using the correct amount of bits per sample).

and it's not that the analysis "haunts" me, it is that it is just another term for a "quick pass". and i'd rather that pass be complete.

that's how i understand it, anyways.


fun discussions today, huh?

 

[DoctorM]

I swore I said something to the effect I wasn't going to debate this.

A bit dated, but still a useful guide: http://www.afterdawn.com/guides/archive ... torial.cfm

As far as the test passes. It's just a piece of the film encoded to get the idea for how compressible it is. Usually something like selectrangeevery or something is used. After the calculations are made it's discarded, it isn't like a .vaf file is kept afterwards.
Anyway, that's really all I'm going to say, we've gotten too far off topic anyway.

Except in rebuttal to the author of the program: although the Q value getting higher may produce similar results to a multipass, frankly once the Q goes TOO high, Nero Recode/DVD Shrink can produce better results.
DVD-rb shines for movies that give low Q's at high compression (50-70%).

 

[joebshmoe]

well if you thought this was a debate i'm guessing you'll be winning this one!

but seriously, from what i've just read today OPV looks to be a good strategy for sure. it has obvious pros (time), a debatable "pro" with constant quality and only slight cons (accurate size output).

...i'm just not convinced that it is BETTER than 2-pass VBR (which takes about 2x longer, has a near-exact size output and best quality per size given).

edit:

i just found a pretty final word on the subject here:
http://dvd-rb.dvd2go.org/modules.php?na ... opic&t=114

You often hear the comparisons of OPV to multipass and which one is better to use. The bottom line is: It depends. It is safe to say that OPV is never "better" than multipass. That's because in multipass the encoder looks at the entire stream and determines what is the best quality (or Q in CCE terminology) that can be attained within the space available. It then on second and subsequent passes distributes bits accordingly.

But it is true that OPV can be just as good as 5 passes... DVD-RB does a prediction pass analyzing 1% (by default) of the source and decides which is the highest Q that can be used with the space available. OPV then holds that Q for the entire encode using whatever bitrates it needs at the places it needs them. If that Q is highest you can get -- it's the highest no matter how many passes you do.

OPV's weakness is that the accuracy of the DVD-RB prediction is limited, and you could end up undersizing. Also, OPV's Q setting is an integer value limiting its granularity. So if it comes up with a Q of, say, "5" -- then the stream will be encoded at a Q of 5. That might make the output a little small, but is the best that can be done because a Q of 4 may make it too big to fit on a DVD-R. Multipass, on the other hand, allocates bits based upon demand across the entire stream, so it's entirely possible it could create the equivalent of a Q of 4.795 (or any other value between).

I personally use multipass and do 2 passes, and for sources that are challenging (e.g. long movies) I do 3 passes. Every pass above 3 has continually shrinking effect. Cinemacraft says 4 passes is pretty much the end of any noticable improvement. But when you have lots of time and want to make a perfect disc... no harm in doing as many as you want.

I use OPV when I'm in a hurry or when the movie is obviously not very difficult to compress. On my computer a complete DVD-RB OPV run takes about an hour and a half.

Hope that helps.

it's not that i don't believe you or something. the problem (for me) is that what you are saying contradicts what i know of compression so i needed learn more about what you are talking about. what was "contradictory" was just terminology.

...after all that it's clear that OPV isn't BETTER than 2-pass VBR (sometimes AS GOOD) and you give up those additional benefits of multi-pass for time savings of OPV.

 

[ThrowgnCpr]

And here are cover and disc arts for ADigitalMan's Extended Editions of the Harry Potter films. This will fit a 7-disc case with a 1" spine. Each disc has a different logo (the 4 house shields, the dark mark, Hogwarts shield, and Harry's scar)

DOWNLOAD COVER ART

warning: download is 29MB

 

[boon23]

this is WOW!

 

[ADigitalMan]

Beautiful work. I recommend creating an eighth disc and size the case to match. A little early spoiler: there is a different version of film 1 you'll be hearing about in the next week or so. Think about it an the answer will come to you all I'm sure.

 

[ThrowgnCpr]

cool. The size of the disc wont change, but it will be easy to whip up another disc, and add the text on the back