Consumers Must Demand an Increase in Audio Quality for 'HD-DVD', Not Just Video! (1 Viewer)

[ChrisA]

Using Dolby Digital, at least 2.0 is part of DVD spec, nothing anybody can do about it.

Yes, that is what I remember which is why I said you have three options that make sense, mentioned a few posts back:

1) Eliminate DTS track and use the max Dolby Digital rate, allocating the bandwidth to the video.

2) Eliminate the max Dolby Digital track and provide a minimal 2.0 DD track and a 'full' bit-rate DTS track.

3) Eliminate the max Dolby Digital track and provide a minimal 2.0 DD track and raise the DTS bandwidth a little and the video a little

But don't waste (like many DVDs in general) bandwidth on two primary soundtracks... I'm not worried about storage as much as it seems that every sound track takes a chunk of bandwidth, which is the weakest link for standard DVD. I know Sony was asked about why they did not just include a higher bandwidth DTS track instead of a low bandwidth DTS track and max bandwidth DD track...

 

[David Susilo]

About your points:

1. I agree

2. "worms... worms everywhere"

3. more "worms... worms everywhere"

 

[ChrisA]

Well, that is why we need to dedicate a 'large' amount of bandwidth to the audio with HD-DVD, period! And pick only ONE damn primary audio track to 'MAX' out! Last thing we need is a 4Mbit/sec DTS track and a 'new' DD 2Mbit/sec track, etc....

1080p using WM9 at 25(+/-)Mbit/sec and an MLP sountrack using 24/96 for the front 3 and 20/48 for all other channels DD can be used for other languages and accessory tracks

 

[DaViD Boulet]

WM9 on blue-laser DVD would more than afford the bandwidth needed for high-resolution audio.

Let's hope we get it!!!

dave

 

[Chris Farmer]

I like the audio stuff, esp the MLP 24/96 in front with 24/48 in the back, makes sense, rears don't need to resolution of hte fronts anyway, they're not required to do the same things as the fronts are.

I still would prefer MPEG-4 over WM though, fo rreasons mentioned above. And Blu-Ray will never happen with WM 9 anyway, also for reasons mentioned above.

 

[DaViD Boulet]

I like the audio stuff, esp the MLP 24/96 in front with 24/48 in the
back, makes sense, rears don't need to resolution of hte fronts anyway,
they're not required to do the same things as the fronts are.

Chris,

the use of the surround channels is mix-dependent. I don't want my HD-DVD *format* to dictate that the surrounds can't provide fidelity equal to my front channels. Naturally content providers could shift resolution around depending on how they want to encode the audio to maximize perceived quality, but why not do what needs to be done to *allow* 24/96 all way round? Who's to say that an artist might not want to use the surrounds in a way that takes full advantage of it?

BTW, MPEG4 is not as good as WM9...in fact, it's not even as good as MPEG2. Discussions over at AVS have pointed out that MPEG4 is inteded for low-bit-rate video (like internet streaming) but is not designed for high-quality playback. WM9, on the other hand, is designed explicitly to acheive near-transparancy to the uncompressed master even at astonishingly low data rates due to its sophisticated algorithms.

-dave

 

[ChrisA]

BTW, MPEG4 is not as good as WM9...in fact, it's not even as good as MPEG2. Discussions over at AVS have pointed out that MPEG4 is inteded for low-bit-rate video (like internet streaming) but is not designed for high-quality playback. WM9, on the other hand, is designed explicitly to acheive near-transparancy to the uncompressed master even at astonishingly low data rates due to its sophisticated algorithms.

Yes, MPEG 4 and MPEG 2 are both "BLOCK" encoding, and inherently limited in spatial resolution and other limitations, especially when compared to a "WAVELET" type of compression, which I believe is WM9. There was talk that these more sophisticated "WAVELET" algorithms actually require some serious processing power, which is apparent that the WM9 version of "Terminator 2" requires a 3.0 Ghz P4 !! But obviously, a future HD-DVD player using WM9 would have a special processor card designed for WM9 exclusively, which would be much more efficient in decoding/processing WM9.

[rant] MUSIC DVDS: I am becoming disturbed about release of music DVD's with only DD QUALITY... This is another reason we need BLU-RAY. We need to have DVD-Audio quality AND HD-video. These Dolby Digital music releases could be so much better... [/rant]

I'm studying for my physics boards in radiology and I'm surprised at the nice amount of overlap I'm experiencing with film/digital imaging...

 

[Dan Hitchman]

We can stop calling it WM9 (that's the media player software app's name on XP). The name of Microsoft's wavelet codec used is called Corona.

I still believe the Blu-Ray consortium will go with (at least) their own video codec and perhaps not MPEG-2 since they still don't have the specs. for pre-recorded Blu-Ray down pat. I seriously doubt they will put a Microsoft product (a competitor) in their media.

Disney is also in the Blu-Ray camp and Eisner stated they must embrace high resolution media to stay competitive. A very important statement since they were so reluctant to jump on board DVD when it was first created.

We may not end up with it (at least not for movies and concerts), but I have a hunch that Sony and Philips are pushing behind the scenes for lossless compressed DSD (many long playing SACD's use proprietary lossless DSD compression as is) so they can continue to get royalties just like regular compact discs (they also own many of the patents on CD's too). For music only applications they could easily use uncompressed DSD at even higher resolutions and/or more discrete channels then at present.

Dan

 

[DaViD Boulet]

DSD would be very hard to impliment for movie-based content because it must be converted to PCM to do any sort of Digital signal processing.

For movie soundtracks, bass equalization, crossover for the sub content, EQ for THX processing and down-mixing for 2.0 playback or for mixing a phantom center or missing surround channels into the front L/R mains are all digital processes that require PCM.

So point is that even though DSD is great for high-res audio that represents the master recording about as faithfully as we can given any format today...for movie soundtracks it would not provide any real benefit over high-res PCM because it would need to be converted to PCM for the routine DSP algorithms applied almost ubiquitously in HT playback.

-dave

 

[ChrisA]

David really hit the nail on the head as far as the WEAKNESS of DSD is concerned... All the fancy features such as digital crossovers, digital room correction, etc... require conversion to PCM. I own the Meridian 861 and I enjoy an all digital pathway. I am a 'minimalist' person, who likes the least amount in the signal path, however, an all digital pathway is certainly 'minimalist'. Also, you take big hits in extra A/D, D/A conversion... I still see no reason to embrace DSD... there is a lot of BS going on in universal players... I've heard a lot of them, if not all of them convert the DSD to PCM internally anyway... Besides, I like digital crossovers in an all digital pathway... may as well support multichannel 24/192.

And then you have converstaions like this going on in the Surround sound Forum (Major Telarc, David Chesky, and all big time recording artists forum):

This is presumably how you do DSP operations "without leaving the DSD environment" ahem. You simply turn the bitstream into something almost completely indistinguishable from PCM, and then turn it back again afterwards... Which is quite sensible actually, though it begs the question whether it would have been better to use PCM in the first place.

These are the people who do the actual recordings... great insight from the pros... you should monitor the surround sound list...

Personally, I might want to recommend 24/192 for HD-DVD-Audio....

 

[Dan Hitchman]

However, what's stopping Sony/Philips from designing DSP algorithms that work in the DSD domain and then licensing the technology to the various manufacturers, especially if they have any interest in adding it to multimedia applications, not just music-only?

Dan

 

[ChrisA]

DSD 'algorithms' don't work in the DSD domain... as you see in my previous post, DSD is converted into PCM for any processing (even at the recording level using the workstation) and then converted back to DSD, and ultimately may be converted back to PCM for a DVD-audio...

If you ask me, record in PCM and stay in PCM, unless you can prove a succinct advantage to recording in DSD in the first place....

 

[Dan Hitchman]

ChrisA:

At present, they do not on current consumer equipment. On many (if not all) pro grade DSD hard disc workstations and mixing/mastering boards DSD is kept DSD throughout the process. The only widely used audio software production package, that I know of, that is still PCM-only (and requires DSD conversion) is ProTools.

What I was saying is that Sony & Philips would probably work on fixing that flaw in future products licensed to use their DSD technology. I'm sure they're already doing so as we type. Their new ES SACD players add DSP processing and bass management using DSD via a next generation chip, according to the press.

MLP encoded, 6 channel PCM 24/96 all around would be OK by me (as long as they didn't use audio watermarking to muck things up fidelity-wise). Hell, if they could do it, I'd want 8 channel discrete PCM 24/96. For music only apps I could see 8 channel at 24/192 easily because of the very large medium capacity and the higher bitrates.

Dan

 

[ChrisA]

Their new ES SACD players add DSP processing and bass management using DSD via a next generation chip, according to the press.

Dan, as you see in my earlier posts, and what is really going on, is that DSD MUST be converted to a multibit format for any processing next generation chip or not, it is a fundamental requirement. That is the point. You can't add bass management, DSP correction, or digital crossovers (while staying in an all digital path) unless you convert DSD into a form of PCM as mentioned above.

DSD would probably be most appealing for music for audiophiles who don't use subwoofers and don't want any digital processing: someone who just takes the DSD 2.0 or 5.0 directly...

In any case, the point of this thread is that we need to remember:

1) A wavelet (Corona) type of algorithm is apparently the highest quality video compression methodology and far superior to either MPEG-2 or MPEG-4 which uses block encoding.

2) The current lmitiations of DVD techonolgy limiting us to 10Mbit/sec bandwidth (total) are not acceptable because:

a) Whatever video format is chosen that might look 'acceptable' at 7-8 Mbit/sec, will look significantly better at 20 Mbit/sec, regardless of which algorithm is chosen.

b) It is absolutely rediculous to not equate HD-DVD with better audio, which is EQUALLY IMPORTANT AS IMPROVEMENT IN VIDEO, or nonetheless is SIGNIFICANTLY IMPORTANT. It makes no sense to have 720p/1080i/1080p video for HD-DVD and continue utilizing 500-700 kbit/sec bandwidth for DD and DTS... Something significant has to improve. With a format like Blu-Ray, you finally have a format which allows one to have true high defiition video quality and have enough bandwith remaining for a high quality audio track. I see absolutley no reason why we could not AND SHOULD NOT dedicate at least 4Mbit/sec to a PRIMARY audio soundtrack.

I'm not going to continue to debate on whether it be DD, DTS, DSD, MLP, or some new compression format, but one thing is certain, DD is not acceptable in its current form and limitations to be the primary audio standard for any upcoming HD-DVD format.

 

[Richard Paul]

I'm doubtful that Dolby will be releasing a new audio codec in time for Blu-ray so if the Blu-ray commitee is serious about audio quality then there's a good chance that DTS will be one of the mandatory choices. Though Dolby Digital is on almost all DVD's it is not actually a requirement. Instead it is one of the two mandatory audio formats with DD being one and PCM (1 or 2 channel) being the other. All pre-recorded DVD's must include one of the two mandatory audio formats. DTS was unfortunately to late to get their format as one of DVD's mandatory audio formats.

Pre-recorded Blu-ray could have four mandatory audio formats which could include PCM, MLP, DTS, and DD. Though many of the first movie releases on Blu-ray would use DD it would give the option for movies to use DTS and even MLP. This would be the best solution since it creates a range of choices for the quality of the audio without requiring a lesser quality audio format to be included.

 

[DaViD Boulet]

Dan (and anyone else wondering the same thing),

DSD cannot be digitally processed. In a manner of speaking, DSD isn't really "digital" at all...it acts and behaves virtually like a continuous analog waveform...which is why it sounds so good. But you can't do "math" on a wavform made up of a single-bit word sampled a zillion times a second.

All those workstations that do DSP for DSD signals are, in fact, converting to some form of PCM. It's not mathematically possible to process a DSD bitstream (both Pro and consumer gear). I don't know enough about it technically to know whether it ever will be in the future.

Chris...you think some of those conversion horror stories are bad? My friend got a DSD recording that said "DDD" on it and it was a recording from the mid-80's (I'll have to look-up the title). I wondered how this would be possible since obviously the source recording would have been PCM so I called the disc producer. He told me "oh yeah, that was a problem--we played back the master digital tape on the machine that had made recording and fed the analog output to a DSD encoder". In other words...the "DSD" disc is basically replicating the sound of some crummy mid-80's D/A converter! I politely suggested that the "full digital transfer from the master tape" blurb in the write-up might be just a little misleading. I also suggested that preserving the sound of a mid-80's D/A converter forever in DSD isn't exactly the what DSD should be about. He said "oh we had to use that machine's built-in D/A because it was some non-standard sample-rate on the digital master...like 57Hz or something...and we didn't have an outboard DAC that could handle it."

It's just increadible.

 

[Lewis Besze]

DSD cannot be digitally processed. In a manner of speaking, DSD isn't really "digital" at all...it acts and behaves virtually like a continuous analog waveform...which is why it sounds so good. But you can't do "math" on a wavform made up of a single-bit word sampled a zillion times a second.

So then Denon and Sony both lying about their player's ability to do just that?I wouldn't be overly surprised,but it would still "raise my eyebrows".

 

[ChrisA]

Lewis:

You forgot the important part

All those workstations that do DSP for DSD signals are, in fact, converting to some form of PCM.

Again, DSD must be coverted to multibit (PCM) for any digital processing....

In any case, the point of this thread is that we need to remember:

1) A wavelet (Corona) type of algorithm is apparently the highest quality video compression methodology and far superior to either MPEG-2 or MPEG-4 which uses block encoding.

2) The current lmitiations of DVD techonolgy limiting us to 10Mbit/sec bandwidth (total) are not acceptable because:

a) Whatever video format is chosen that might look 'acceptable' at 7-8 Mbit/sec, will look significantly better at 20 Mbit/sec, regardless of which algorithm is chosen.

b) It is absolutely rediculous to not equate HD-DVD with better audio, which is EQUALLY IMPORTANT AS IMPROVEMENT IN VIDEO, or nonetheless is SIGNIFICANTLY IMPORTANT. It makes no sense to have 720p/1080i/1080p video for HD-DVD and continue utilizing 500-700 kbit/sec bandwidth for DD and DTS... Something significant has to improve. With a format like Blu-Ray, you finally have a format which allows one to have true high defiition video quality and have enough bandwith remaining for a high quality audio track. I see absolutley no reason why we could not AND SHOULD NOT dedicate at least 4Mbit/sec to a PRIMARY audio soundtrack.

I'm not going to continue to debate on whether it be DD, DTS, DSD, MLP, or some new compression format, but one thing is certain, DD is not acceptable in its current form and limitations to be the primary audio standard for any upcoming HD-DVD format.

 

[Lewis Besze]

Chris,
i wasn't talking about the recording session/mixing/editing,I was talking about the players doing BM, a far simplier task.

 

[DaViD Boulet]

Lewis,

I know that the Denon "universal" player is a bit of a misrepresentation...it converts all DSD bitstreams to high-resolution PCM. This is how it's able to offer such processing features and also how it's able to transmit hi-res audio digitally via it's proprietary cable to the receiver...Sony will not allow anyone to transmit the native DSD bitstream digitally between consumer equipment right now.

Don't believe the marketing literature. It took me some digging to find out exactly what the Denon did. When it was firt released I was like "COOL!" thinking they had come up with some real miracle universal player that could indeed engage DSP on a DSD signal. Some more intensive research proved that assumption wrong!

BTW, once a DSD signal is converted to PCM all the sonic benefits that ultra-high-res DSD would have afforded are lost. Not that 24/96 or 24/192 PCM is exactly bad...cuz it ain't! But it is, technically speaking, a step-down from the native DSD signal.

DSD is best suited for archiving with the most transparent known recording technique to date, and delivering an unmodified signal of the master to the consumer.

BTW, there's a chance that the Sony player is actually performing bass management in the *analog* domain after DSD-Analog conversion. That's how the outlaw bassmanagement works. It's really the most transparent option for applying any EQ/bass management to native DSD audio signals.

dave