I started this thread to see if there would be support for an idea I have. I will say up front this idea serves two (2) goals for us.

1 - To allow everyone to audition for themselves the difference between MPEG-2, AVC, and VC1 Video Encodes as well as Dolby Digital Plus verses DTS-HD and Dolby Tru-HD verses DTS Master Audio Lossless. I would also demo pre-processing encodes and post processing encodes....meaning before and after digital cleanup and enhancement.


2 - To help me drive more sales of my HD-DVD and Blu-ray Discs.


The idea is to do a fair evaluation of all these audio / video formats and put them in a disc along with some refrence demo content and maybe a few basic A/V setup tools.

Now I don't have Disney or Sony or Warner content to work with...but I do have some very nice material to work with. However, I would be happy to get support from any industry folks with respect to content, lab services, or anything else that could help support this project.

My idea is to either sell this as a stand alone product for technical folks like us, or bundle it with other HD Releases to help drive sales via the value added content.

If not supported by industry folks on AVS, I would do all of the video encodes at top labs such as Technicolor and i-Candy Interactive and ask them all to submit several versions of the video encodes. I want to get a fair "real world" comparison for this test. We do all of our audio encodes at a lab we work with that uses a process we specify.

The film and pre-encode processing would be done at DTS/Lowry Digital Images. These folks are the very best that I know of in the industry.

I would like to know what AVS Forum Members think of this idea.

Should it be sold as a separate item? Would anyone buy this disc? What price should it be?

Or...

Should it be bundled with other products we are selling?

I am not 100% sure I am doing this yet, but I am curious as to the response this might get if I released such a disc.

Rich

I would buy it. Will you make it a 2 disc Blu-ray HD DVD set or sell it in separate Blu-ray and HD DVD version? I would prefer a 2 disc Blu-ray HD DVD set, and I would pay up to $30 retail (not MSRP) for this. If you are a real pal, please include basic color bars and pluge patterns in both formats.

Thank you,

It has to be done by those who routinely use all three codecs and both disc formats? -so the results aren't skewed by unfamiliarity? Could be tough. Otherwise, get the best from any source. It would be good to compare HD and BD max bit rates too on the BD version, and low and high bit rates on the HD version.

I would be interested to buy a combo of both format.

A couple of issues with this idea:

AUDIO: No commercially available player currently decodes DTS-HD or DTS-MA, which means we'd be comparing Dolby TrueHD vs. 1.5Mbps DTS Core (though that in itself might be an interesting comparison).

VIDEO: I think the individual strengths of the encodes have a lot to do with the complexity of the source material and the allowed bandwidth. For instance, as you allow the bandwidth to rise higher, your compression loss decreases and converges for all encoding formats (VC-1, AVC, MPEG-2).

A more interesting shootout condition would be to force the encode to be limited to various commonly encountered bitrates for A/V.

For instance:
Take some very complex(high contrast and/or fast animation) material, and force the video compression to drop down to various bitrates (10Mbps, 20Mbps, 30Mbps, 40Mbps). The 40Mbps is only going to be possible on a Blu-ray disc, though.

Then maybe repeat with some more sedate material and see what the difference in PQ are to simulate "typical" movie material?

It is OK to agree to disagree....I am very old school. Any processing is bad processing...any alteration of the original...is a change form the original. I think along the lines of the old "straight wire" theory so to speak.

Perhaps I am too old in age and too outdated in my thinking for all these youngsters in here...I am just a dinosaur with a passion for high tech ...:)

BTW - I am from NY so I have thick skin and thrive on logical and well argued debates...I have just noticed heavy VC1 / HD-DVD favoritism on this board...at least that is my take. I am smart enough to know when I am outnumbered.

Bedsides, I really don't have a vested interest in any of these formats or encoder technologies. Also, I am not interested in changing anyone's mind about it either....I just state what I know...and I don't know everything...unlike some people on this board...who seem to know it all..:)

However, with that said, if I get riled up enough...I may just put out that disc I mentioned in a new posting under "Ultimate HD Shootout" disc. Even if I release that disc, nobody will agree ... and probably wrongly accuse me of something or other I did or did not do to their liking. Of that, I am sure. :)

Rich

Rich,


the "heavy VC1 / HD-DVD favoritism" on this board is the result of months and months of heavy "viral marketing" and propanda. It is just part of MS business plan and it works - they succeeded in building up the VC-1 HYPE. Do we all like this? Certainly not. But what to you expect when everybody here gets "brain washed" by hearing just "one side of the story" and selective facts. There is certainly no balance and "thruth" to be found (especially in the insider Q&A thread) when almost any single post has the sole purpose of gaining competitve advantages.

If a company has high ranked employees (with considerable income) spend hours and hours posting statements here you can do the math how keen its interest is in influencing general public and opinion leaders.

YOU ARE LIKE A BREATH OF FRESH AIR. You don't hesitate nor are you "afraid" to give us unfiltered, unbiased information WITHOUT strategic intend.

I personally thank you for that!

btw I would buy this disc in a heartbeat!

I'd definitely buy it if had some even rudimentary calibration items (like THX optimizer, etc), but wouldn't necessarily have to be a full blown affair like DVE or Avia (Bonus points if you hit somewhere in between, though). :D

If it was included in another release, then just short clips would be okay...

If it was stand alone, I'd hope for some sort of useful calibration charts and such.

Either one of those would grab my attention.

I support this idea, and would definitely buy a disc that would show me differences between audio/video codec performance.

However, I must say that this test will only be benefical to enthusiasts and enthusiasts only, i.e. just about everyone on this board. Whether we like it or not, the success of either format (or codec, depending on how you look at it) will depend entirely on adoption by the masses, J6P and all. Until this day, I know NO ONE outside of this board who would prefer to purchase a certain version of a given movie based on codec alone. The main deciding factors are variety, cost and vaue, NOT VC-1, MPEG-2, or AVC, as some insiders would have you believe :) (no offense ;) ).

That said, this project will only help those of us who would like to see which specification of audio/video suits our equipment, so that we could make an educated decision from there.

Thanks in advance for your time and effort :) !

If you add test and alignment patterns to this set, I would buy it in an instant.....

Vern

I'd definitely buy something along these lines purely out of idle curiosity. As suggested by others I'd love to see a comparison not just of CODECs but of various bitrates for each CODEC. The audio comparison would be at least as interesting as video - and I'd emphatically disagree with the prior posting regarding DTS HD MA. We will have players that support it eventually, and I'd love to have the content waiting.

Given the two options you describe (standalone product vs. value-add for other products) I'm sure my intuition can pick the more viable route from a business standpoint. The market for the standalone product isn't huge unless you take on an Avia/DVE-like exercise. Having something studios could include on their discs for basic calibration sounds compelling, but I'm not sure they'd devote the space for meaningful clips using different CODECs, bitrates, etc. Presumably they pay royalties on a per-disc basis for use of the encoders and this could muddy the royalty situation, etc.

If you pursue the standalone approach this is perhaps the only disc that I'd actually like to see use TotalHD. Normally I'd prefer label artwork and wouldn't want to pay two sets of royalties, etc. but this is one project that positively screams out for dual-format support.

There was an attempt on doom9 to do the video part of such comparison using the "Elephant's Dream" movie and compress using all 3 codecs.
As all things on doom9, it was very quantitative, but for some reason was never finished.

Diogen.

There was an attempt on doom9 to do the video part of such comparison using the "Elephant's Dream" movie and compress using all 3 codecs.
As all things on doom9, it was very quantitative, but for some reason was never finished.

The guy making up the specs didn't have access to the actual HD DVD or BD specs, so the actual proposed requirements wound up not being that relevant.

At some point I got the impression they agreed on using something "resembling" a reasonable compromise.

Diogen.

At some point I got the impression they agreed on using something "resembling" a reasonable compromise.

Not close enough to be useful last I looked at it.

For example, they were mandating a IBBP GOP, when VC-1 is best with an adaptive GOP defaulting to IBP. This also forced us to a 12-frame GOP instead of the 14 we usually use.

sounds interesting in concept. But I would guess impossible to implement in a reasonable way.

How would you pick clips (A/V). Obviously a totally silent part in a movie would be the perfect example of why DD = DD+ =DTS= DTHD = DTS HD MA = PCM or do you pick the most complex audio you can cretae and then show the differences just to get a response that "it does not matter if everyone can hear a differences 99% (used by some to mean most) of movies won't have something that complicated.

do you pick a video that will be better for MPEG AVC or VC-1 (each has its strength and weaknesses) how do you pick bitrate?

----
I guess what I am trying to say is that it is a sad fact but chances are people will find reasons to dismiss your test because it won't show what they want.

----

as for buying. Curiosity has me interested but what will be on the disk and how much it costs will be a factor.

How would you pick clips (A/V). Obviously a totally silent part in a movie would be the perfect example of why DD = DD+ =DTS= DTHD = DTS HD MA = PCM or do you pick the most complex audio you can cretae and then show the differences just to get a response that "it does not matter if everyone can hear a differences 99% (used by some to mean most) of movies won't have something that complicated.
Probably best to have different audio and video test sections, so one can mix and match the experience, with clear documentation on ABR/PBR for each. The audio tests would have as good as video as can be done with remaining bits, and vise versa for video tests.

do you pick a video that will be better for MPEG AVC or VC-1 (each has its strength and weaknesses) how do you pick bitrate?
A half-dozen decent test clips of different types should be able to be agreed on. Like:

Good 35mm master
Good 35mm master with DNR
Good 65mm master
Grainy 16mm master
CGI
Cel animation

I guess what I am trying to say is that it is a sad fact but chances are people will find reasons to dismiss your test because it won't show what they want.
Likely true, but with proper planning, it could be made interesting and relevant. Get all the regulars to sign off on the specs BEFORE the disc is made, so they can't claim it was biased after the fact :).

Probably best to have different audio and video test sections, so one can mix and match the experience, with clear documentation on ABR/PBR for each. The audio tests would have as good as video as can be done with remaining bits, and vise versa for video tests.

agree, then again you don't need them together.

Likely true, but with proper planning, it could be made interesting and relevant. Get all the regulars to sign off on the specs BEFORE the disc is made, so they can't claim it was biased after the fact .

Now you are just naive. You reminded me of last year around this time when, I think it was Nataraj, said can't wait for content to come out because then the fights will stop.

Now you are just naive. You reminded me of last year around this time when, I think it was Nataraj, said can't wait for content to come out because then the fights will stop.
The heat is inevitable - but I'd like there to be as much light as possible in there as well :).

Hi Anthony,

Not impossible, but not easy either.

I am open to ideas, but right now I am thinking of content that I know will test encoders such as complex animation with a lot of polygons and fast motion, good 65mm Film transfered to 1080/24P with a lot of detail, light & shadow areas, and darker areas with subtle shade and tone variations and also HD Video with CGI Overlays that have very complex detail and small moving CGI foreground elements layered against a 1080i HD Video Background.

I will probably use 192/24 and 96/24 soundtrack elements for the audio encodes. It will not be complex as in busy....it will be music with great dynamic range, tight bass, and clean highs with great depth, staging, and imaging in the mix. This is a good test for us since we are music guys and this is where the real differences can truly really be heard. However, we probably need to throw in a few things blowing up as well for the cinematic experience that most people are expecting. :)

We will max out the Birates for starters, even if we have go MOS (without SOund) and work our way down to real world numbers in the 22mbps to 24mpbs range, We might also do some tests at lower bitrates in the 13mbps to 16mbps range.

You are correct. No matter what, certain people will dismiss the tests and argue that the results are not valid...especially if it shows something they do not want to believe or accept. That is OK...and we are posting in this forum to minimize the chances of this...taking everyone's input and incorporating what we can that makes sense.

We will also mark the disc Version 1.0 so we can update and revise the disc as technology improves and as we get feedback from members.


I am still not sure if we are doing this yet, but I am leaning towardds doing it.

Rich

sounds interesting in concept. But I would guess impossible to implement in a reasonable way.

How would you pick clips (A/V). Obviously a totally silent part in a movie would be the perfect example of why DD = DD+ =DTS= DTHD = DTS HD MA = PCM or do you pick the most complex audio you can cretae and then show the differences just to get a response that "it does not matter if everyone can hear a differences 99% (used by some to mean most) of movies won't have something that complicated.

do you pick a video that will be better for MPEG AVC or VC-1 (each has its strength and weaknesses) how do you pick bitrate?

----
I guess what I am trying to say is that it is a sad fact but chances are people will find reasons to dismiss your test because it won't show what they want.

----

as for buying. Curiosity has me interested but what will be on the disk and how much it costs will be a factor.

I am open to ideas, but right now I am thinking of content that I know will test encoders

sounds good so far. The only things I would add

Audio: Is lower quality audio 16/48, 20/48, 24/48 for comparison reasons (lossless/lossy). Most movies are 16/48 and some 24/48, I think you are the only one that has released with something better. This would compare the coidecs at the unfortunate level that we will get

Video: I think it might also be cool (for the BD disk) to have a longer clip that is VC-1 and AVC at 20mbps (or less –toatl <40) that are together on the disk and where the user can flip between one and the other so he starts it in X and then pushes a button to continue in Y then back to X…..

Would it not be possible to get content like trailers from the studios? Given that they're essentially adverts, surely they'd be happy to give them away? There is a possibility that they would have 4k masters that you could work with, that would make things easier right?

I will buy it at whatever cost! :D

And don't forget some (high quality) VIDEO content, besides progressive (film) source at whatever frame rate.

Thanks in advance! :)

regards,

Li On

sounds awesome! :)

...

Or...

Should it be bundled with other products we are selling?

I am not 100% sure I am doing this yet, but I am curious as to the response this might get if I released such a disc.

Rich

Hi Rich, I'd love to see this disc as a bundled solution, where full length material (e.g. anime, comics, short movies, video) is contained with the value add that they are present in different versions (e.g. video and audio codecs). My motivation to buy it will probably be not very high if I know there are "only" short scenes or crippled movies for the purpose of comparing bits and bytes.

On a side note: do you think it's possible to add some material which is coded in 25p? This would be, to my knowledge, the first disc which contains 25 fps film, and would help to see whether players can handle this as well.

Thanks, Torsten

I agree...and would bundle this as content if it does not have A/V Tools for setup....

If I do add A/V Tools. I would probably sell it as separate item for a very reasonable cost as well as bundle it.

Thank you for the input..:)

Best,

Rich

Hi Rich, I'd love to see this disc as a bundled solution, where full length material (e.g. anime, comics, short movies, video) is contained with the value add that they are present in different versions (e.g. video and audio codecs). My motivation to buy it will probably be not very high if I know there are "only" short scenes or crippled movies for the purpose of comparing bits and bytes.

On a side note: do you think it's possible to add some material which is coded in 25p? This would be, to my knowledge, the first disc which contains 25 fps film, and would help to see whether players can handle this as well.

Thanks, Torsten

I am not sure if studios would agree to having their content use in this way. They all seem to favor one format or codec or another for a variety of reasons....which I won't get in to here.

I will be doing my best to select content that challenges encoders ... and if any Industry Insiders whish to contribute content, I am open to g it on the disc.

Again, I have not decided whether or not I will do this as of yet. A lot depends on the feedback I get from this forum.

Rich



Would it not be possible to get content like trailers from the studios? Given that they're essentially adverts, surely they'd be happy to give them away? There is a possibility that they would have 4k masters that you could work with, that would make things easier right?

I can include lower resolution content....but I will have to struggle a bit to find it. We don't have anything lower than 48/24 in our library that we own....so I will have to "borrow" some content from an associate for this test.

The results are noticeable but not as dramatic when you start going to lower resolution audio.

I think going to extremes with the video codecs is a good idea as well. I would plan on doing that as it is interesting to see which codecs perform best at which bitrates.

Rich

sounds good so far. The only things I would add

Audio: Is lower quality audio 16/48, 20/48, 24/48 for comparison reasons (lossless/lossy). Most movies are 16/48 and some 24/48, I think you are the only one that has released with something better. This would compare the coidecs at the unfortunate level that we will get

Video: I think it might also be cool (for the BD disk) to have a longer clip that is VC-1 and AVC at 20mbps (or less –toatl <40) that are together on the disk and where the user can flip between one and the other so he starts it in X and then pushes a button to continue in Y then back to X…..

Good point....we do not want this skewed either way. I just want a real world test...meaning if I use a sampling of the top Authoring Labs, the same labs the studios use, I would tend to believe I would get results similar to what a real client might expect.

I could include a "Best Example" from someone like NS for VC! who has a vested interest as well as a "Real World" example from what the labs are actually turning out. However, I believe this may fuel the debate even further.

As you probably know, anyone can spend time to make an encode look really good, if they have the skills, time, and patience. However, I want to know what an average result would be like verses what an ideal result.

Make sense?

Rich


It has to be done by those who routinely use all three codecs and both disc formats? -so the results aren't skewed by unfamiliarity? Could be tough. Otherwise, get the best from any source. It would be good to compare HD and BD max bit rates too on the BD version, and low and high bit rates on the HD version.

You are correct on the Audio Comparisons...there are no players that support DTS Master Audio Lossless..yet.

You are correct again, but I believe that AVC and VC1 have a ceiling or point of diminishing returns where MPEG-2 does not.

I would include common bit-rates as well ranging from an atypical low of 10mbps to an almost impossible to achieve high of 30mpbs to 40mpbs. We may need to eliminate audio tracks an go MOS for the highest bitrate video encodes.

We need headroom so I am not sure we can get a 30mbps encode out of HD-DVD...maybe 25mbps or slightly higher ...but that is a guess. Same with BD, I am guessing that we can get maybe a 35mbs to 38mbps encode. I won't know for sure until we try. I do know that we always have trouble achieving the stated max bitrate for DVD due to the Scenarist authoring system's refusal to create a final master based on us pushing encode rates too close to the max. I am expecting the same scenario here as well.

If I do include an audio track, it will be a very basic Dolby Digital track for reference and it will be the same for every encode test. We do not want the "experience" to feel any different due to audio performance....that has to remain a constant.


Rich


A couple of issues with this idea:

AUDIO: No commercially available player currently decodes DTS-HD or DTS-MA, which means we'd be comparing Dolby TrueHD vs. 1.5Mbps DTS Core (though that in itself might be an interesting comparison).

VIDEO: I think the individual strengths of the encodes have a lot to do with the complexity of the source material and the allowed bandwidth. For instance, as you allow the bandwidth to rise higher, your compression loss decreases and converges for all encoding formats (VC-1, AVC, MPEG-2).

A more interesting shootout condition would be to force the encode to be limited to various commonly encountered bitrates for A/V.

For instance:
Take some very complex(high contrast and/or fast animation) material, and force the video compression to drop down to various bitrates (10Mbps, 20Mbps, 30Mbps, 40Mbps). The 40Mbps is only going to be possible on a Blu-ray disc, though.

Then maybe repeat with some more sedate material and see what the difference in PQ are to simulate "typical" movie material?

I would definitely be interested in a release like this. Wouldn't hesitate for a second to buy it.

You are correct again, but I believe that AVC and VC1 have a ceiling or point of diminishing returns where MPEG-2 does not.

Just curious - are you saying that MPEG-2 hits is quality threshold at much higher bitrates, or that MPEG-2 has a higher quality threshold than the advanced codecs.

If it's the latter, I should point out that VC-1 can be viewed as a superset of MPEG-2. By turning off advanced features like loop filter, variable block size, etetera, you can wind up with are core that maps to MPEG-2 very closely - there shouldn't ever be a case where we can't beat MPEG-2 at a given bitrate (MPEG-2 has some architectural limitations that can keep it from being truly transparent in some cases, regardless of bitrate).

Rich,

This sounds like a great idea. I would certainly be interested in purchasing this.

My two cents on what I'd like to see:

1) Most of us would like an answer to the question regarding whether or not AVC or MPEG 2 at much higher bitrates provides a superior image than VC-1 at the bitrates they usually use for HD-DVD. For example, many around here complain that Warner is not using BD to its potential by simply cloning their HD-DVD discs. Those on the VC-1/HD-DVD side say the extra bandwidth won't produce any better results. It would be nice to see how a higher bitrate AVC encode compares to a lower bitrate VC1. We may get to compare this when the UK version of The Prestige gets released, but any differences will probably be attributed to a different master being used.

2) The other controversy regarding the "smoothing" effect that many see with VC-1. The only problem with this is that there isn't a real way to determine whether or not that the releases in question have had this "smoothing" added on purpose. In other words, VC-1 could be fully capable of producing proper film grain but the releases in question were purposely encoded with a softer, glossier image in order to keep bitrates down for HD-DVD.

In any event, it sounds like a great idea. I look forward to purchasing it if it happens.

2) The other controversy regarding the "smoothing" effect that many see with VC-1. The only problem with this is that there isn't a real way to determine whether or not that the releases in question have had this "smoothing" added on purpose. In other words, VC-1 could be fully capable of producing proper film grain but the releases in question were purposely encoded with a softer, glossier image in order to keep bitrates down for HD-DVD.

I feel compelled to point out once again our codec and entire PEP toolchain doesn't include any way to do this kind of intentional softening like you talk about. Were it to be done, it'd have to happen upstream of encoding entirely.

To the best of my knowlege, the titles often suspected of being softened on purpose simply reflect the master. There are some titles that have been run through a noise reduction pass prior to encode, but no one has figured out which those are :).

Hi,

Yes, the one size fits all approach is not one we subscribe to...but it is part of the marketing concept for the newer encoding technologies.

I am not sure what others are doing in the way of pre-encode processing, but we will do our best to demonstrate both processed and unprocessed encodes if we do this disc.

Thank you for your input.

Rich


Rich,

This sounds like a great idea. I would certainly be interested in purchasing this.

My two cents on what I'd like to see:

1) Most of us would like an answer to the question regarding whether or not AVC or MPEG 2 at much higher bitrates provides a superior image than VC-1 at the bitrates they usually use for HD-DVD. For example, many around here complain that Warner is not using BD to its potential by simply cloning their HD-DVD discs. Those on the VC-1/HD-DVD side say the extra bandwidth won't produce any better results. It would be nice to see how a higher bitrate AVC encode compares to a lower bitrate VC1. We may get to compare this when the UK version of The Prestige gets released, but any differences will probably be attributed to a different master being used.

2) The other controversy regarding the "smoothing" effect that many see with VC-1. The only problem with this is that there isn't a real way to determine whether or not that the releases in question have had this "smoothing" added on purpose. In other words, VC-1 could be fully capable of producing proper film grain but the releases in question were purposely encoded with a softer, glossier image in order to keep bitrates down for HD-DVD.

In any event, it sounds like a great idea. I look forward to purchasing it if it happens.

I think the best idea for this "shootout" is to put some of the source material online and give the various compressionists out there a chance to give them their "best shot".

The winning encodes submitted in each codec format category (ie. AVC, VC1, MPG2, DD+, DTHD, DTS HD, Etc) should then be included on the disc offered for sale.

This will allow a fair comparison, as some people will be more experienced or "committed" to one codec or another and will allow the best-possible real-world results to be evaluated against each other.

The main rules are that all entries must used commercially available encoding tools (ie that would be available to studios, not necessarily consumer hobbyists).

Does this sound fair?

I feel compelled to point out once again our codec and entire PEP toolchain doesn't include any way to do this kind of intentional softening like you talk about. Were it to be done, it'd have to happen upstream of encoding entirely.

To the best of my knowlege, the titles often suspected of being softened on purpose simply reflect the master. There are some titles that have been run through a noise reduction pass prior to encode, but no one has figured out which those are :).

Ben,

I was simply pointing out that this topic has been brought up time and time again on this forum. I don't claim as a matter of fact that VC-1 softens the image. I just thought it would be nice to take a highly-detailed source with lots of film grain and see how the different codecs handle this.

I was simply pointing out that this topic has been brought up time and time again on this forum. I don't claim as a matter of fact that VC-1 softens the image. I just thought it would be nice to take a highly-detailed source with lots of film grain and see how the different codecs handle this.
Yes, I'd love to see some grainy source in there as well.

I think the best idea for this "shootout" is to put some of the source material online and give the various compressionists out there a chance to give them their "best shot".

The winning encodes submitted in each codec format category (ie. AVC, VC1, MPG2, DD+, DTHD, DTS HD, Etc) should then be included on the disc offered for sale.

This will allow a fair comparison, as some people will be more experienced or "committed" to one codec or another and will allow the best-possible real-world results to be evaluated against each other.

The main rules are that all entries must used commercially available encoding tools (ie that would be available to studios, not necessarily consumer hobbyists).

Does this sound fair?
Yes, it would be a huge win for the industry to have a good variety of HD source content to use for testing. There's some stuff out there (like Elephant's Dream), but I'm not aware of any 1080p24 film-shot content.

For example, they were mandating a IBBP GOP, when VC-1 is best with an adaptive GOP defaulting to IBP. This also forced us to a 12-frame GOP instead of the 14 we usually use.GOP? You support the republicans? Huh? :confused:

Sorry.... couldn't resist making a little fun of the geekspeak! ;)

AUDIO: No commercially available player currently decodes DTS-HD or DTS-MA, which means we'd be comparing Dolby TrueHD vs. 1.5Mbps DTS Core (though that in itself might be an interesting comparison).Is this statement correct? I could have sworn I've played DTS-HD on my XA1... I think it was on Harry Potter IV?

DTS-HD Master Audio Lossless is what BD Player's do not not support yet. They will eventually. DTD-HD, which is a lossy encode, is something I believe they all support.


Is this statement correct? I could have sworn I've played DTS-HD on my XA1... I think it was on Harry Potter IV?

Maybe...I am not hands on tech....but we have seen great encodes using MPEG-2 when we raise the bandwidth limitations. Again, the new and improved VC1 and AVC encoders are capable of great encodes. That was not the case a year ago.

I am pulling together all of the content I can do try and pull off this shootout disc. I am happy to have MS Contribute encodes.


Just curious - are you saying that MPEG-2 hits is quality threshold at much higher bitrates, or that MPEG-2 has a higher quality threshold than the advanced codecs.

If it's the latter, I should point out that VC-1 can be viewed as a superset of MPEG-2. By turning off advanced features like loop filter, variable block size, etetera, you can wind up with are core that maps to MPEG-2 very closely - there shouldn't ever be a case where we can't beat MPEG-2 at a given bitrate (MPEG-2 has some architectural limitations that can keep it from being truly transparent in some cases, regardless of bitrate).

GOP? You support the republicans? Huh? :confused:

Sorry.... couldn't resist making a little fun of the geekspeak! ;)
GOP=Group of Pictures. An I-frame and all the B and P frames associated with it. In HD DVD with VC-1, we normally use a 14-frame GOP, with the pattern:

IBPBPBPBPBPBPB

DTS-HD Master Audio Lossless is what BD Player's do not not support yet. They will eventually. DTD-HD, which is a lossy encode, is something I believe they all support.OK, that is what I thought. Is there some confusion over the acronyms of those two codecs? I was using DTS-HD for the lossy codec and DTS-MA for the lossless codec, but I see a ton of variances which I think is causing even more confusion.

GOP=Group of Pictures. An I-frame and all the B and P frames associated with it. In HD DVD with VC-1, we normally use a 14-frame GOP, with the pattern:

IBPBPBPBPBPBPBThanks, I was actually trying to follow your technical dialog with TrevorS on the insiders thread earlier this week. I appreciate the time you take to explain how the compression encode and decode theories operate.

Ben, I re-read this and it peaked my interest. Are you saying there is a way to raise the ceiling so to speak on VC1 encodes? I would be very interested in seeing the results of this if that is the case. All of my compression and authoring labs are telling me there is a limit and point of diminishing returns on both AVC and VC1.

I would be very interested in being able to push my encodes to the maximum if it actually yielded better results. Most of my content is between 40 and 50 minutes, so I have a little bit of room to work with on an average disc compared to a major studio. Also, I do not mind paying the additional cost of going to a higher capacity disc if required to push my encode rates higher.

Rich


Just curious - are you saying that MPEG-2 hits is quality threshold at much higher bitrates, or that MPEG-2 has a higher quality threshold than the advanced codecs.

If it's the latter, I should point out that VC-1 can be viewed as a superset of MPEG-2. By turning off advanced features like loop filter, variable block size, etetera, you can wind up with are core that maps to MPEG-2 very closely - there shouldn't ever be a case where we can't beat MPEG-2 at a given bitrate (MPEG-2 has some architectural limitations that can keep it from being truly transparent in some cases, regardless of bitrate).

Is this statement correct? I could have sworn I've played DTS-HD on my XA1...

you don't. The difference is that DTS created their newer codecs to include a DTS core. IN essence every DTS stream is DTS with an error correction applied. Any DTS decoder will take a DTS-HD MA/HR and if it can do MA/HR then it will, if not it will just decode the lossy DTS part

I will be doing my best to select content that challenges encoders ...

Richard -

As much as many like to measure encoders by how well they reproduce grain it would also be nice to measure how well they reproduce images. So, if you can, please include some very clean, low grain, highly detailed unfiltered sections, just for the wow effect.

After all, as film makers realize grain does not encode well on electronic media they may start using less of it, in favor of other reality-obscuring techniques that compress better.

And as a demo disc this effort should probably have at least some sections that look good without taxing anybody's encoders, just to show how good the format can look if it was really planned for from the beginning of the chain.

For instance, a 4K scan of some clean nicely illuminated slow moving 65mm master properly Nyquist filtered and downsampled to 1080p might be optimum. Surely there is some available somewhere.

It doesn't ALL have to be a stress test. And possibly we might see encoder differences here too, even at high bit rates.

- Tom

OK, that is what I thought. Is there some confusion over the acronyms of those two codecs? I was using DTS-HD for the lossy codec and DTS-MA for the lossless codec, but I see a ton of variances which I think is causing even more confusion.But dts-HD HR (lossy) is still more than legacy 1.5 Mbps dts since it can have 7.1 channels and higher bitrate (3.0 Mbps on HD DVD, 6.0 Mbps on BD):
http://www.dts.com/dts-hd/dtshd-high-resolution-audio.php

all three types of dts on hi def shiny disc:
http://www.dts.com/dts-hd/dts-on-bluray-and-hddvd.php

Push! :D

regards,

Li On

Wow. Great idea man! I really hope this bears fruit. I will definitely buy this in a heartbeat if this shootout is done right. Even though I don't own a HD DVD or Blu-ray player yet.

About the whole audio codecs thing. I'm aware not all the formats can be decoded yet. But you should still include it. DTS-HD MA decoders may not be available now but it will in the future.

And I will *heart* you if you are able to sneak in 24-bit/192KHz in 5.1 surround.

[...] I am happy to have MS Contribute encodes.
Richard, are you going to do the AVS and MPEG2 encodes in your lab, or will this work being done by third parties? It would be great to see the encoding work done by the parties behind the encoders (Microsoft for VC1, AVS guys for the AVS encodes, MPEG guys - whoever this might be - the MPEG2 stuff). As you have had very good experience with MPEG2 and seemed a bit biased regarding encoders --- no offense intended!!! --- it would be great to see the work done by independent parties, who I expect to know how to squeeze the last bit of quality out of their most recent encoders. This would give you the possibility to comfortably step back after the disc is ready and state "I just pressed the disc, look at the results, all encodes were done by their supporters and specialists", and would prevent from statements I did earlier in this post.