About 4 years into DVD's life cycle we started seeing "Superbit" versions of movies where all extras where removed and they studios instead used the space to maximise the PQ's & AQ's of the movie.

I would be interested to know if people think that there will be similar "superbit" transfers of blu-ray movies down the line?

For example, perhaps we would see 24-bit PCM soundtracks instead of 16-bit? Or maybe we could have even less compression on the picture? I'm not sure.

My own thoughts are that if by then Blu-ray titles are being released on 100GB 4-layer discs then there will be no need to remove extras to optimize PQ's, but if we're still typically using 50GB discs then there may be scope for a Superbit re-release.

A BD50 can run full mux for almost 139 minutes.

That leaves 8mbit already reserved for audio.

24bit PCM takes 6,9mbs

So extras, that take up space, but since they dont need to run any movie at full mux, they will have space over for extras.

So there is not much point on releasing Superbit.

dolby tru-hd and dts -master are the super bit of audio.

I hope there won't be any "superbit" (no bonus features) Blu-ray releases. Put the extras on a separate disc if it's going to affect the quality in any way (though they can't do that with PiP commentaries - so any release with PiP commentaries will probably be slightly lower quality maybe)

Superbit didn't really take off with DVD and it probably wouldn't for Blu-ray either.

dolby tru-hd and dts -master are the super bit of audio.
Actually it's PCM that is the superbit of audio as it represents an uncompressed soundtrack whilst the 2 you mention are compression codecs (and people will argue all day about how much AQ you lose with them).

The question really is how high can you do with the number of bits that a PCM audio sountrack has before either capacity or thoughput limitations kick in?

Superbit didn't really take off with DVD and it probably wouldn't for Blu-ray either.
Agree. I think Superbit would only be for the niche market and on only a few select reference titles at that, much like it was on DVD. But I still think that they'll probably do it at some point.

About 4 years into DVD's life cycle we started seeing "Superbit" versions of movies where all extras where removed and they studios...

Superbit is a trademark name used by Sony so there were no other studios and it was just studio. For the most part it was a marketing tactic used very successfully by Sony. It's is amazing how many believe that it was some kind of "special" DVD.

Fact: Many other studios used as high or higher ABR as Sony used on their Superbit line.
Fact: Most titles had plenty of unused space left over that could have contained extras.

Frankly, my eyes are better than my ears. I can live with 16bit pcm if I get more bitrate reserved for video.

But such a disk makes sense only if we have a 4K DI as a master and not some D5's that could conceivably have macroblocks. D5 is roughly 240Mbps(?), and that's not enough for a jpeg-like compression like HD D5 if the goal is to avoid macroblocks in tough frames. (If what people are discussing here is to be believed).

Actually it's PCM that is the superbit of audio as it represents an uncompressed soundtrack whilst the 2 you mention are compression codecs (and people will argue all day about how much AQ you lose with them).

The question really is how high can you do with the number of bits that a PCM audio sountrack has before either capacity or thoughput limitations kick in?

Those people would be silly people, as both TrueHD and DTS:HD MA are lossless codecs and assuming no DNR flags are getting enabled TrueHD=DTS-HD:MA both of which equal bit for bit equal PCM once decompressed.

Those people would be silly people, as both TrueHD and DTS:HD MA are lossless codecs and assuming no DNR flags are getting enabled TrueHD=DTS-HD:MA both of which equal bit for bit equal PCM once decompressed.
Of course the theory of lossless codecs is that you shouldn't lose any quality, but I don't know anyone who has compared soundtracks that hasn't noticed that both TrueHD and DTSHD DO infact lose something. The debate has moved on from whether TrueHD and DTSHD fall short of PCM to "Why do they?". Most people seem to think that it is something to do with the production of the discs during transfer, but I don't know if there is an officially agreed reason for it yet.

Superbit is a trademark name used by Sony so there were no other studios and it was just studio. For the most part it was a marketing tactic used very successfully by Sony. It's is amazing how many believe that it was some kind of "special" DVD.
That isn't strictly true. Superbit releases had higher bitrates for the video encodes. They also mandated DTS or Dolby Digital Soundtracks.

That isn't strictly true. Superbit releases had higher bitrates for the video encodes. They also mandated DTS or Dolby Digital Soundtracks.

The DVD Forum mandates that ANY and ALL DVD's that have DTS MUST also contain a DD and/or a LPCM track (DTS is an optional format). ;) Sony's SB releases had higher bitrates than their Sony standard releases but other studios released DVD's with as high or higher bitrates. Superbit is ONLY Sony and a Sony trademark so no other studio used or could call themselves that.

The DVD Forum mandates that ANY and ALL DVD's that have DTS MUST also contain a DD and/or a LPCM track (DTS is an optional format). ;) Sony's SB releases had higher bitrates than their Sony standard releases but other studios released DVD's with as high or higher bitrates. Superbit is ONLY Sony and a Sony trademark so no other studio used or could call themselves that.

Yes it was more marketing then anything else.

Yes it was more marketing then anything else.

Based on how unoriginal marketing can be, it's possible we will see something like this again.

I think all releases should be this way. Disc 1: pure\perfect master of the movie, Disc 2: whatever extras

The Disney POTC releases are already the equivalent of DVD's "Superbit Deluxe" releases, with the entire 50GB of disc 1 dedicated to the movie and the remaining 25GB of disc 2 dedicated to extras.

Of course the theory of lossless codecs is that you shouldn't lose any quality, but I don't know anyone who has compared soundtracks that hasn't noticed that both TrueHD and DTSHD DO infact lose something.Since they're both lossless codecs (not just in theory but actual practice), what "something" do they lose? Can you be more specific?

Sanjay

Based on how unoriginal marketing can be, it's possible we will see something like this again.

I think all releases should be this way. Disc 1: pure\perfect master of the movie, Disc 2: whatever extras

Well as I written earlier. At full mux you can fit anything under 138 minutes on a BD50.

Since nothing need full mux you vill have plenty to spare for extras.


Adding a second disc would add great cost since you must use copyprotection. I read somewere the cost would be 2,50$ per disc. So you would basicly pay 5$ per movie just in copyprotection. (please correct me if im wrong on this?)

Well as I written earlier. At full mux you can fit anything under 138 minutes on a BD50.

Since nothing need full mux you vill have plenty to spare for extras.


Adding a second disc would add great cost since you must use copyprotection. I read somewere the cost would be 2,50$ per disc. So you would basicly pay 5$ per movie just in copyprotection. (please correct me if im wrong on this?)

Well there are already many 2-disc Blu-ray movies and there are plenty of movies that exceed 138 minutes.

Sign me up. I could care less for extras. That said having the one with the bells and whistles for sale wouldn't be a bad thing for folks wanting all the extras they can stand.

Adding a second disc would add great cost since you must use copyprotection. I read somewere the cost would be 2,50$ per disc. So you would basicly pay 5$ per movie just in copyprotection. (please correct me if im wrong on this?)

I think this is wrong. IIRC, there's a fixed cost per title for AACS. I believe it's $2500 per title. So, the $2.50 per disc example is based on a production run of only 1000 units. I think the original argument was that since AACS was mandatory on Blu-ray, independent filmmakers would probably prefer HD DVD.

I think this is wrong. IIRC, there's a fixed cost per title for AACS. I believe it's $2500 per title. So, the $2.50 per disc example is based on a production run of only 1000 units. I think the original argument was that since AACS was mandatory on Blu-ray, independent filmmakers would probably prefer HD DVD.

Thanks

But besides that you would have enough space for the extras on one disc.

Lets say you have a 100 minute movie. You have 1 PCM track (6,9mbs) and one DD640 track

U have avarage bitrate that is 30mbs for video.

video = 22,5GB
Audio = 5,655 GB

So then you have about 28gb for the main movie and about 22gb left for extras.

Well there are already many 2-disc Blu-ray movies and there are plenty of movies that exceed 138 minutes.

Many 2-disc Blu-ray movies? Are you talking about a movie that is split between two discs? Or are you talking about the movie being on one disc and the extras on another?


And if anyone knows, what is the 138 minute limit based on? There's also movies that have lots of slow movement and I would think a Blu-ray disc could hold much more if there is not a lot of action in a movie.

THe 138 minutes is based on a full muxed BD50. It will run at its peak value the entire time (48mbs)

Of course no movie needs that so it really have plenty of space over.

Since they're both lossless codecs (not just in theory but actual practice), what "something" do they lose? Can you be more specific?

Sanjay

Wrong! In theory YES, in practise NO. This isn't disputed (except by you it seems). What is disputed is WHY this happens when theoretically they are lossless codecs. It obviously has something to do with the process of encryption / de-encryption, but what the floor in the process is has been the subject of much discussion.

Of course the theory of lossless codecs is that you shouldn't lose any quality, but I don't know anyone who has compared soundtracks that hasn't noticed that both TrueHD and DTSHD DO infact lose something.
OK, I will bite :). Which disc has both tracks on it so that I can compare?


The debate has moved on from whether TrueHD and DTSHD fall short of PCM to "Why do they?".
They have? Is there a link to the discussion thread?

Most people seem to think that it is something to do with the production of the discs during transfer, but I don't know if there is an officially agreed reason for it yet.
The only reason is that there is dialog normalization and potential level differences, especially if you are comparing HD DVD to BD as the former allows mixing of the audio so it might be encoded at a lower level to allow headroom for mixing.

THe 138 minutes is based on a full muxed BD50. It will run at its peak value the entire time (48mbs)

Of course no movie needs that so it really have plenty of space over.

We should be getting this now. Full constant bitrate releases for the videophile freaks. :D

OK, I will bite :). Which disc has both tracks on it so that I can compare?




cetk has true hd and dts hd ma but no pcm track.

Of course the theory of lossless codecs is that you shouldn't lose any quality, but I don't know anyone who has compared soundtracks that hasn't noticed that both TrueHD and DTSHD DO infact lose something. The debate has moved on from whether TrueHD and DTSHD fall short of PCM to "Why do they?". Most people seem to think that it is something to do with the production of the discs during transfer, but I don't know if there is an officially agreed reason for it yet.

Nope, those people are all wrong. You end up with the exact same 1's and 0's.

cetk has true hd and dts hd ma but no pcm track.
OK, help me out. I am not as good with movie acronyms as you all are :). Which one is cetk?

And do people think the two sound different?

OK, help me out. I am not as good with movie acronyms as you all are :). Which one is cetk?

And do people think the two sound different?

Maybe CEOT3K was what was meant....as in Close Encounters Of The 3rd Kind.


my guess, at least.

Maybe CEOT3K was what was meant....as in Close Encounters Of The 3rd Kind.


my guess, at least.

Ah, that seems to be the one. I don't have the disc in my library and will be losing the use of BD-30 shotly so won't be able to test it.

cetk = close encounters of the third kind

For 2.35:1 films, I would like to see full res 2.35:1 encoded on the disc, which may take up additional space. The default would still be 16:9 as it covers the vast majority of display, but it would be great to have the option to access the full resolution in a 2.35:1setup. I am not sure if this would require two separate transfers of the film, which would obviously mean additional storage, or one transfer that is just converted/downrezzed for 16:9 output.

Since there's no anamorphic flags then it would have to be a seperate version on the disc. Films scanned at 4k would greatly benefit from the extra vertical resolution but the vast majority of 2.35:1 transfers wouldn't show much improvement at all if any.

dont think there's any point to releasing on superbit version. when sony did it with dvds it added DTS tracks that had higher bitrates than non-superbit dvd's. higher video bitrate prob went unnoticed by most, and i doubt it's much higher since they simply subbed extras for higher bitrate audio.

with HDM tho, pretty much all the discs already come with the highest-end lossless audio and video is no slouch either. superbit will not see any diff just like we dont see any diff in pq between lower-bitrate hd dvd and blu-ray.

Wrong! In theory YES, in practise NO. This isn't disputed (except by you it seems).The output of a losslessly encoded soundtrack is bit-for-bit identical to the orignal encoding master. These have been check-sum tested for accuracy, else they wouldn't be lossless.

You still haven't answered what the "something" is that soundtracks lose when losslessly packed. Since you're convinced they do lose something, could you please describe what it is?

Sanjay

WRT the OP, Blu-ray disc is already like "Superbit" on every release, with high bitrate and large filesize on BD-50 and uncompressed LPCM audio. Blu-ray disc: the "Superbit" of HDM formats. :)

This thread makes my head hurt.

A full resolution 2.35 encode? I assume that's for people who have anamorphic lenses for their 16:9 1080P front projectors. So the horizontal resolution would be the same. If that were to be done, let's come up with an appropriate name for the product. Fox owns the "Cinemascope" name, with its associations with anamorphic lenses and 2.35. I'm a big supporter of using names instead of acronyms.

Remember, 7680X4320 Super HD is only ten years away!

Considering the high capacity of Blu-ray disc there is no need to have a Superbit version of Blu-ray. Superbit on SD DVD offered 8 Mbps max avg bit rate. Check out some of the bit rates offered by normal BDs.

28 Weeks Later = 37 Mbps
Rescue Dawn / Cast Away = 33 Mbps
Man on Fire = 29 Mbps
Black Hawk Down, Starship Troopers, Rocky Balboa = 27 Mbps
Spiderman - 26 Mpbs
Die Hard - 27 Mbps

For LOTR I would request New Line to offer Superbit treatment. They can offer extras on a separate disc.

Yes, LOTR EE had better be just the movie on disc 1 and the extras on disc 2, or I will personally kill 5 kittens in retaliation. It's a good thing New Line does their encodes independently of Warner, considering Warner's miserable track record of starving their encodes for bits like there's a world war going on that requires bitrate rationing.

Warner's bit rate treatment reminds me of the diet chart maintained by Japanese women. Fully starved to death.

OK, I will bite :). Which disc has both tracks on it so that I can compare?
Troy: The Directors cut. Compare the BD version to the HD-DVD version, the BD "sounds" better.



They have? Is there a link to the discussion thread?
There is a great article by Josh Zyber about it http://www.highdefdigest.com/news/show/1233


The only reason is that there is dialog normalization and potential level differences, especially if you are comparing HD DVD to BD as the former allows mixing of the audio so it might be encoded at a lower level to allow headroom for mixing.
I'd agree with that, particularly about the dialogue normalization, and it is pretty much what Josh is saying in the above article that I've referenced, so I'm not sure why your initial doubts? All I said was that the end product CAN (and in some cases HAS) sounded different, whereas in THEORY they should sound the same as it is bit for bit identical.

Troy: The Directors cut. Compare the BD version to the HD-DVD version, the BD "sounds" better.
That may not be a good comparison I am afraid. HD DVD as I mentioned, allows for headroom for mixing. BD versions without interactivity do not. Last I checked, this meant a 4db difference in level which is quite signficant. I will take a listen though to Troy and see where takes me (I think I have that disc in both formats).

My preference would have been a BD disc with both.

There is a great article by Josh Zyber about it http://www.highdefdigest.com/news/show/1233
But he is making my point :). That lossless is lossless but things like DialNorm may change the level, making any A/B comparison moot without compensation. He says nothing about lossless codecs being anything but 1 for 1 same as the source that was fed to them.


I'd agree with that, particularly about the dialogue normalization, and it is pretty much what Josh is saying in the above article that I've referenced, so I'm not sure why your initial doubts? All I said was that the end product CAN (and in some cases HAS) sounded different, whereas in THEORY they should sound the same as it is bit for bit identical.
Fair enough but this is very different than what you said before:

Of course the theory of lossless codecs is that you shouldn't lose any quality, but I don't know anyone who has compared soundtracks that hasn't noticed that both TrueHD and DTSHD DO infact lose something.
If the volume changes, I am not "losing something." Just turn up the volume/ignore dial-norm and you are were the PCM is. The word "theory" makes it sound like in practice lossless is not lossless.

Fair enough but this is very different than what you said before:


If the volume changes, I am not "losing something." Just turn up the volume/ignore dial-norm and you are were the PCM is. The word "theory" makes it sound like in practice lossless is not lossless.

I guess I could have been more specific and said "lose volume", but I without qualifications (as have been mentioned above) that that would also a tad misleading.

Of course the theory of lossless codecs is that you shouldn't lose any quality, but I don't know anyone who has compared soundtracks that hasn't noticed that both TrueHD and DTSHD DO infact lose something.I guess I could have been more specific and said "lose volume", but I without qualifications (as have been mentioned above) that that would also a tad misleading.Lossless encoding doesn't "lose volume". It is DialNorm that causes the soundtrack to play back at a lower level and, as Amir said, just turn up the volume and you're back to the same level as PCM.

As to your earlier claim: The debate has moved on from whether TrueHD and DTSHD fall short of PCM to "Why do they?". Most people seem to think that it is something to do with the production of the discs during transfer, but I don't know if there is an officially agreed reason for it yet.Looks like the big mystery is solved: lower playback volume. But that doesn't mean the sonics "fall short of PCM". False alarm.

Sanjay

Yes, I hope there is a Superbit-style for the enthusiasts. I'll gladly take these over the PiP, dozen language, bandwidth wasting marketing gimmick, cell-phone ring tone selling, filled tripe. That's worth a small (10% or so, maybe as high as 25% for reference level difficult titles). For most titles, Blu-ray already meets these standards, though, as many have mentioned.

If they want to put most of that other stuff on a second disc, they can feel free, of course. I may even get it especially if it's at "normal" pricing.

I love PCM tracks. This would be my track of choice whenever the video bitrate can spare it and much more desired personally than anything like PiP. Dolby TrueHD is just as good on some titles like the Fifth Element. Letters from Iwo Jima had a TrueHD track that I was most underwhelmed by but unfortunately had no PCM for me to compare. I've never experienced full DTS-HDMA, though I hope Sony adds that to the PS3 sooner rather than later.

There is a great article by Josh Zyber about it http://www.highdefdigest.com/news/show/1233
This is funny, your suggested article is actually not even supporting your view???The number of new audio formats on Blu-ray and HD DVD have caused a great deal of consumer confusion, especially with three separate formats (PCM, Dolby TrueHD, and DTS-HD Master Audio) all designed to accomplish the exact same goal -- a perfect replication of the movie's audio master. Apprehensions about lossless compression being inferior to an uncompressed version of the same soundtrack are not borne out by the facts. One methodology may have technical advantages over the other in terms of space savings, but the end result is the same whether the disc you buy has an uncompressed soundtrack or a lossless one. They're both equally good, so sit back and enjoy.

Letters from Iwo Jima had a TrueHD track that I was most underwhelmed by but unfortunately had no PCM for me to compare.Do you believe that the soundtrack would have sounded better (less underwhelming) if the data had not been packed losslessly and instead been left as uncompressed PCM?

Sanjay

Do you believe that the soundtrack would have sounded better (less underwhelming) if the data had not been packed losslessly and instead been left as uncompressed PCM?


Possibly, but I'm not convinced of that. My biggest complaint was on the dialogue, so it might well be dialnorm that was the issue.

Or it could have just been an umimpressing source or possibly botched mix. Fifth Element remastered, CE3k, and 300 all have fine TrueHD tracks, though I almost always listen to the PCM track on TFE and 300.

Possibly, but I'm not convinced of that. My biggest complaint was on the dialogue, so it might well be dialnorm that was the issue.

Or it could have just been an umimpressing source or possibly botched mix. Fifth Element remastered, CE3k, and 300 all have fine TrueHD tracks, though I almost always listen to the PCM track on TFE and 300.

What about the 24bit TrueHD on SM3 versus PCM?

I'm not sure; but I always feel like some Blu-ray folks carry some sort of grudge against lossless codecs especially since it's used almost exclusively for HD DVD (where HD DVD releases have no PCM because of bandwidth and space). It's like saying PCM is better so you can point out Blu-ray has more space and bandwidth.
Me, I think that bandwidth could be used more efficiently for other things instead of wasting it for PCM while you can get the exact same thing using TrueHD and DTS-MA (Dialnorm aside).
The good thing Sony has already started the move from PCM to TrueHD. I hope Disney joins the club as well.

About 4 years into DVD's life cycle we started seeing "Superbit" versions of movies where all extras where removed and they studios instead used the space to maximise the PQ's & AQ's of the movie.

I would be interested to know if people think that there will be similar "superbit" transfers of blu-ray movies down the line?



I doubt it.....Superbit was a marketing term....Superbit titles have less horizontal filtering ( horizontal filtering removes fine detail ) ..... Blu-Ray transfers should not remove that fine detail ( although reports seem to suggest it might on certain movies )

Good article on this here.

http://www.videophile.info/

Bjoern Roy wrote this below.

I sometimes hear complains like "Why would they rerelease title XYZ in superbit, its perfect already!". Well, i have to disagree with that. ALL of the CTHV non-SB titles are filtered horizontally, removing the highest frequencies (finest detail). This results in a less than perfectly focused image on high-end setups. The SB titles on the other hand seem to be horizontally unfiltered and look much more detailed, the non-SB's look 'blurry' by comparison. And no, they not only seem blurry once you have seen the SB and go back. On a good setup, you can easily see the bluriness of non-SB CTHV titles even without a SB counterpart as a reference. They just aren't as detailed.

So, basically EVERY release from CTHV of the past years would benefit from the SB treatment.

This Lord of the Rings review is great.

Anyway, so the first reviews of FOTR started to show up and the euphoria was immense. 'Best PQ ever', 'approaching HD' etc, etc. 'The detail... ', 'Oh...' , 'Ah...'.

I am usually quite skeptical towards most reviews, since few of them are done on the same level of equipement that us videophile's are used to. On TVs and even on many RPTVs, most transfer, from mediocre to sensational, all just look 'great'. Theses reviews ARE very helpful to many, actually 'most' people, since they have the same level of equipment at home! But they are inadequate for 'us' to make buying decisions on, since they fail to differentiate the 'good', 'very good' and 'sensational' on a level that i would ultimately agree on.

Which is a shame, since it means i need to buy everything i am remotely interested in and see for myself. On titles that i really adore, like the LOTR's, i even often buy ALL versions. I don't have any prefilter available in terms of 'would-only-buy-it-if-transfer-is-spectacular', or 'want-to-buy-the-version/region-with-best-PQ'. I must be a movie studios dream-customer. Ka'-ching :-)

So the day arrived where i finally hold the preeeecious in my filthy little hands. The R1 NTSC theatrical version of 'Fellowship of the Ring'. Boahhh! Went straight to watching it that night, of course...

"Eh, wait a moment"... "Whats that?"... "Where is all the supposed detail?"... "Something must be wrong with my setup". So i investigated half an hour, whether everything was in a-ok. And it was. Huh?

To make it short, i found basically all aspects of the transfer to be more or less excellent. Except for the one that has been praised the most: Detail. There just isn't any fine detail on this disc. It looks heavily filtered. Outright 'blurry'. Since detail is what HD is all about, how can this be called 'HD-like'?

"Oh boy, what a let-down", i thought. To double-check, whether my expecations where simply so high that i might have lost any contact with planet earth, i watched a few reference and even a few 'normal' DVDs, but alas, it really is FOTR R1 that is blurry. Buh-huh.

Good thing the 'blurriness' was the only thing that bugged me. As i said, everything else was good to excellent. Not even the slight ringing or the slightly lacking shadow delineation (my 2 most common pet peeves), where issues for me.

Thank god the 'movie' itself is a piece of art that i could even enjoy on a 10-th generation VHS copy. And so i watched it a couple of times, marveling at the unbelievable 'vision' Peter Jackson achieved.... But a disappointment remained about the lack of detail.

There really shouldn't be any need for a Superbit marketing range on Blu-Ray....It has the capacity needed to allow non-filtered titles to be released.

Possibly, but I'm not convinced of that.OK, but I hope you understand that it's like saying a zipped document may "possibly" be less underwhelming a read than the unzipped version. My biggest complaint was on the dialogue, so it might well be dialnorm that was the issue.Can't be, since DialNorm attenuates the entire soundtrack without changing the dialogue level relative to the rest of the audio (which would require a re-mix). There's a really good article describing what DialNorm does (and, more importantly, what it doesn't do) here (http://www.hometheaterhifi.com/volume_7_2/feature-article-dialog-normalization-6-2000.html). Or it could have just been an umimpressing source or possibly botched mix.That seems the most likely. If you level matched a losslessly encoded soundtrack with its PCM counterpart, you'd find them sounding identical (they're already mathematically identical).

Sanjay

OK, but I hope you understand that it's like saying a zipped document may "possibly" be less underwhelming a read than the unzipped version.

Sure, conceptually I have seen other (much more aggressive) encryption algorithms, particularly on secure radio transmission, that cause degradation over realtime the higher it goes. I'm not saying that's taking place here, but it really comes to "trust" on Dolby and DTS that they each and every lossless track is indeed identical to the PCM master and that the man behind the curtain is tweaking things here and there. While I don't yet have a reason to distrust them, I don't have any reason to give them blind faith, either.

Things like dialnorm are merely symptomatic of things happening that end users don't have the ability to switch on or off. Their algorithms are largely voodoo (even restricting ourselves to the technical side and ignoring the marketing mumbo jumbo) to the layperson while PCM is quite simply PCM.

I'm not saying that's taking place here, but it really comes to "trust" on Dolby and DTS that they each and every lossless track is indeed identical to the PCM master and that the man behind the curtain is tweaking things here and there.I can understand some of that possibly happening with lossy compression but not with lossless packing. If you don't trust them, you can always look up the patent information for their lossless codecs. Things like dialnorm are merely symptomatic of things happening that end users don't have the ability to switch on or off.But it's just volume attenuation. You can undo the effect by turning up the volume knob. Their algorithms are largely voodoo (even restricting ourselves to the technical side and ignoring the marketing mumbo jumbo) to the layperson while PCM is quite simply PCM.True, but the layperson doesn't have to understand how something works in order to use it. Or are you trying to say something else? Are you suspicious that the lossless data packing algorithms from Dolby and DTS are not truly lossless?

Sanjay

Sure, conceptually I have seen other (much more aggressive) encryption algorithms, particularly on secure radio transmission, that cause degradation over realtime the higher it goes.
Can you give more specifics? As an old RF guy, I love to know what example you are using here.


I'm not saying that's taking place here, but it really comes to "trust" on Dolby and DTS that they each and every lossless track is indeed identical to the PCM master and that the man behind the curtain is tweaking things here and there. While I don't yet have a reason to distrust them, I don't have any reason to give them blind faith, either.
No trust is necessary. The certification requirement includes running through test suites to guarantee bit-exact decoding. Don't conclude the tests and you don't get to wear the logo.

Their algorithms are largely voodoo (even restricting ourselves to the technical side and ignoring the marketing mumbo jumbo) to the layperson while PCM is quite simply PCM.
Whether something is “voodoo” or not simply is an indication of the level of investment one wants to make in learning them :). It has absolutely no merit whatsoever on the capabilities of the product. So preference for PCM in that regard makes no sense, unless you are the one designing equipment.

Here is a quick summary of what a lossless codec does so that you are not so scared of it anymore :). It uses a set of long filters which attempt to find redundancies over time. The longer the filter, the more the efficiency. But the more CPU cycles it takes to encode/decode them. Once the redundancy is found, it then uses shorter "codes/symbols" to represent them, saving bits.

Since the output rate of the lossless codec is variable and can be whatever it wants to be, the codec is never under any constraint to lose data. Specifically, if it cannot find any way to compress the data, it simply passes through the input to output and thus assuring that it can always do its job.

So the aim of a lossless codec is always the same. Which is to guarantee the same input. The codec has absolutely no value whatsoever if it ever fails this test because losing bits in lossless domain is very ungraceful resulting in significant audio flaws (unlike a lossy codec where you may not even notice if half the bits are gone).

Note that no standards group would ever use a lossless codec if it could not meet this test to their satisfaction. Nor would any content owner accept liability for advertising "lossless" audio if they could not "trust" the codec to act that way.


Net, net, you get in your car, start it and drive away, trusting that it is able to do that and stop on the highway as needed. You don't stand back and say because you don't understand how an engine works, you are not ever going to drive the car. The people who know how to design engines, have done their job so that you don't need to become an engine designer to drive a car :).

About 4 years into DVD's life cycle we started seeing "Superbit" versions of movies where all extras where removed and they studios instead used the space to maximise the PQ's & AQ's of the movie.
Then we saw them learn how to use the tools available and Superbit disappeared because it was no longer needed. Larger encodes are not superior now and will not be needed in the future because of that.

Note that no standards group would ever use a lossless codec if it could not meet this test to their satisfaction. Nor would any content owner accept liability for advertising "lossless" audio if they could not "trust" the codec to act that way.
Once the standards body adopts the format, it is also possible that the encoder can be updated to add new features, since only the decoder is standardized. So the door is open for new tools like this (http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.html&r=1&f=G&l=50&s1=%2220050246178%22.PGNR.&OS=DN/20050246178&RS=DN/20050246178) to be introduced.

Once the standards body adopts the format, it is also possible that the encoder can be updated to add new features, since only the decoder is standardized. So the door is open for new tools like this (http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.html&r=1&f=G&l=50&s1=%2220050246178%22.PGNR.&OS=DN/20050246178&RS=DN/20050246178) to be introduced.


Scalable lossless audio codec and authoring tool
Assignee Name and Adress: Digital Theater Systems, Inc.
...
[0013] In an exemplary embodiment, the audio encoder separates the audio data into most significant bit (MSB) and least significant bit (LSB) portions and encodes each with a different lossless algorithm. An authoring tool writes the MSB portions to a bitstream, writes the LSB portions in the conforming windows to the bitstream, and scales the lossless LSB portions of any non-conforming frames to make them conform and writes the now lossy LSB portions to the bitstream. The audio decoder decodes the MSS and LSB portions and reassembles the PCM audio data.

:eek::eek::eek:

:eek::eek::eek:

:).

I don't understand Roger's post either in citing the DTS patent. In the chance that you are surprised that a lossless codec can be built from a lossy component, that is not odd at all. Here is the generic process.

We take the source and use a lossy system to compress it. We then decode that stream and subtract it from the source. What is left, is the precise mathematical components taken out by the lossy compression. We then *losslessly* compress that difference. We then transmit both the lossy and lossless components and this becomes the sum total of our lossless output.

The receiver decodes the lossy and lossless components and adds them together. By definition, this is the inverse of what the encoder did and the output is the source, totally intact.

Now, there is a complication with above which fortunately, has a simple solution. That is, the decoder must be 100% the same as the same logic in the encoder or the math is not going to reproduce the source due to rouding errors. Since it is routine to allow decoders to not be "bit exact" with the encoder, this can be an issue. Anyone can think what the solution might be? I have already explained it once in the insider thread, many moons ago :).

Back to the point, no new features can be exposed through an encoder. If such features were added, they would not be seen by existing and spec-compliant decoders. So while one is free to add layered coding and such to add new features, they wind up as optional profile type features I talked about earlier and would only play on newer systems with that extra functionality.

In the chance that you are surprised that a lossless codec can be built from a lossy component, that is not odd at all.Amir, are you sure the patent application is talking about the lossy core portion of a DTS-HD MA track? It reads more like they're talking about the capability to momentarily switch to a lossy mode in their lossless encoder. Here are the two paragraphs preceding the one quoted by online: SUMMARY OF THE INVENTION

[0011] The present invention provides an audio codec that generates a lossless bitstream and an authoring tool that selectively discards bits to satisfy media, channel, decoder buffer or playback device bit rate constraints without having to filter the audio input files, reencode or to otherwise disrupt the lossless bitstream.

[0012] This is accomplished by losslessly encoding the audio data in a sequence of analysis windows into a scalable bitstream, comparing the buffered payload to an allowed payload for each window, and selectively scaling the losslessly encoded audio data in the non-conforming windows to reduce the encoded payload, hence the buffered payload thereby introducing loss. Sanjay

Amir, are you sure the patent application is talking about the lossy core portion of a DTS-HD MA track? It reads more like they're talking about the capability to momentarily switch to a lossy mode in their lossless encoder. Here are the two paragraphs preceding the one quoted by online: Sanjay
Actually, nothing I said was in regard to any specific technology. But rather explain how "layered coding" is implemented using a lossy core+lossless extension to arrive at lossless encoding. I don't know that the patent in question is actually about DTS lossless or not.

I have not read enough about the claims to be sure what they are saying. But in the parts you quoted, they could be saying that if forced to stay within a bit contraint, they are able to use scaling and other techniques like that, to gracefully degrade a lossless stream. In the context of HD optical, such a mode is not used. The lossless stream is allowed to have variable rate so that it does not have to resort to lossy methods. After all, if the studios want lossy, they already have it, and in case of DTS lossless, the core is already produced to serve that purpose.

:).

I don't understand Roger's post either in citing the DTS patent. In the chance that you are surprised that a lossless codec can be built from a lossy component, that is not odd at all. My read of Online's response was not that he was surprised that a lossless codec could use a lossy core, but that he was surprised a lossless codec could introduce loss.

in the parts you quoted, they could be saying that if forced to stay within a bit contraint, they are able to use scaling and other techniques like that, to gracefully degrade a lossless stream. In the context of HD optical, such a mode is not used. The lossless stream is allowed to have variable rate so that it does not have to resort to lossy methods.

It might not have to resort to this, but the ability to support VBR does not preclude the option of shaving the peaks--all lossless codecs are VBR, after all. The patent makes a good case that such a tool could be useful when bandwidth gets tight. It happens.

It is not cited to imply this tool is used--I have no idea about that. It was cited as an example that even a certified lossless codec, once adopted by a standards committee, might morph to something else after the fact, yet remain fully compliant with the format standards.

It might not have to resort to this, but the ability to support VBR does not preclude the option of shaving the peaks--all lossless codecs are VBR, after all. The patent makes a good case that such a tool could be useful when bandwidth gets tight. It happens.

It is not cited to imply this tool is used--I have no idea about that. It was cited as an example that even a certified lossless codec, once adopted by a standards committee, might morph to something else after the fact, yet remain fully compliant with the format standards.

Oh I see. So to say another way, you are saying that DTS allows lossy output from their "lossless" codec, if the encoding operator so chooses to save bandwidth rather than produce lossless output? And that this addition was put in, after the codec was ratified?

It is a pretty big bomb if that is what you are saying Roger :).

Oh I see. So to say another way, you are saying that DTS allows lossy output from their "lossless" codec, if the encoding operator so chooses to save bandwidth rather than produce lossless output? And that this addition was put in, after the codec was ratified?

It is a pretty big bomb if that is what you are saying Roger :).
No, that is not what I am saying. I'm saying DTS has developed and applied for a patent on a method for inducing selective lossiness in an otherwise lossless stream for the purpose of containing peak bitrate. This tool was not in operation when the codec was verified for losslessness by the format folks.

No, that is not what I am saying. I'm saying DTS has developed and applied for a patent on a method for inducing selective lossiness in an otherwise lossless stream for the purpose of containing peak bitrate. This tool was not in operation when the codec was verified for losslessness by the format folks.

So doesn't the question become: "What constitutes a 'lossless' stream?" I don't know if there is an industry standard definition, but outsider looking in, I'd be inclined to think that even a small amount of "lossy-ness" in a stream that is supposed to be (considered to be, advertised to be?) lossless would make it, in fact, lossy, wouldn't it? If not, is there a standard for how much "lossy-ness" is allowable before the stream is considered lossy and not lossless? Is it 5% of the bits? 10%? 25%? I know those questions probably sound argumentative, but I don't intend them to be. I just don't know enough to ask anything other than really basic questions.

Also in your reply you say that "This tool was not in operation when the codec was verified . . . " again, I'm not exceptionally knowledgable about this particular topic (although it is interesting to me given how aggressively people here seem to pursue "losslessness"), but how is your quote/statment different than Amirm's statement that " . . . this addition was put in after the codec was ratified?" Is the difference in Amirm's use of the phrase "codec . . . ratified" as opposed to your statement "codec . . . verified"? Again, I'm just trying to understand how all of this works.

My impressions so far from reading here are that: 1) lossless is not nearly as lossless as people want to believe it is, and 2) much of this could be settled by some really good testing, but audio companies and/or C/E manufacturers don't really want to do that sort of testing because it limits the number of logos they can put on the front of their C/E device.

So doesn't the question become: "What constitutes a 'lossless' stream?" I don't know if there is an industry standard definition, but outsider looking in, I'd be inclined to think that even a small amount of "lossy-ness" in a stream that is supposed to be (considered to be, advertised to be?) lossless would make it, in fact, lossy, wouldn't it? If not, is there a standard for how much "lossy-ness" is allowable before the stream is considered lossy and not lossless? Is it 5% of the bits? 10%? 25%? I know those questions probably sound argumentative, but I don't intend them to be. I just don't know enough to ask anything other than really basic questions.
I think the accepted definition for lossless means 0% is lost.

Also in your reply you say that "This tool was not in operation when the codec was verified . . . " again, I'm not exceptionally knowledgable about this particular topic (although it is interesting to me given how aggressively people here seem to pursue "losslessness"), but how is your quote/statment different than Amirm's statement that " . . . this addition was put in after the codec was ratified?" Is the difference in Amirm's use of the phrase "codec . . . ratified" as opposed to your statement "codec . . . verified"? Again, I'm just trying to understand how all of this works.
Amir's statement from which you quote asserts that the tool is in use. I am saying it merely exists.

My impressions so far from reading here are that: 1) lossless is not nearly as lossless as people want to believe it is, and 2) much of this could be settled by some really good testing, but audio companies and/or C/E manufacturers don't really want to do that sort of testing because it limits the number of logos they can put on the front of their C/E device.

Dolby and DTS check to make sure their decoder implementations work correctly. There has been no evidence to the contrary. But I also agree that people get too fixated on lossless. The real benefit of lossless is not in mapping the source's PCM to your DAC with bit-for-bit correspondence. It is to avoid the lossiness of the usual perceptual coder, regardless of the amount of pre- or post-processing the signal undergoes elsewhere in the chain. It removes that one variable from the chain.

Yes, LOTR EE had better be just the movie on disc 1 and the extras on disc 2, or I will personally kill 5 kittens in retaliation. It's a good thing New Line does their encodes independently of Warner, considering Warner's miserable track record of starving their encodes for bits like there's a world war going on that requires bitrate rationing.

HAHAHAHA!!! LMFAO! Now that was a good laugh.:D