300: HD DVD edition - Best hidef disc, period? A review. - Page 3

RobertR · 07-27-2007, 10:10 AM

Quote:

--------------------------------------------------------------------------------

The grain haters will be all over this release.

Which raises the question: Why are they interested in buying a movie the inherent look of which they hate?

DaViD Boulet · 07-27-2007, 10:18 AM

Quote:

he is writing bitrate for PCM 6.9mbit 16bit, as far as I know 6.9mbit is 24bit

(edit)

I think you're right, I think 6.9 mbps is usually what we see with 24-bit 5.1 PCM. I'll check the bit-rate on Chicago to see if it matches (and on T5E which is 16-bit PCM).

Quote:

Warner has supplied both next-gen editions with matching Dolby TrueHD 5.1 surround tracks (48kHz/16-bit), but this Blu-ray is also graced with an additional PCM 5.1 surround option (48kHz/16-bit/6.9mbps). Right upfront, the PCM sounded a bit louder, but after some level matching, a direct A/B comparison of several scenes revealed only slight differences. Although I'm sure this disc will stir up the whole TrueHD vs PCM debate, either way you slice it, the action scenes in '300' deliver the kind of demo-worthy audio that should be pure nirvana for any home theater enthusiast.

Sounds like Dialog-Normalization is applied to the TrueHD tracks. Big surpirse, coming from WB.

Paul Hillenbrand · 07-27-2007, 11:01 AM

Quote:

Originally Posted by DaViD Boulet

Sounds like Dialog-Normalization is applied to the TrueHD tracks. Big surpirse, coming from WB.

David, can you explain to me again why dial norm is a problem?

On SMR Forums Lexicon section, I was told the following:

Quote:

Paul,

Dialnorm is in the metadata, like downmix coeficients or EX flag, not in the audio/sound portion of the bitstream. You still get the studio master (all 24 bits).

I guess one could complain that dialnorm attenuates the volume level by 4dB, but this can be easily rectified with a small turn of the volume knob. Also, if you calibrate using Video Essentials or Avia, then you won't be 4dB low because both those popular calibration discs use dialnorm.

Have you read the Secrets article about dialnorm? If not, take a look. It dispells most of the myths surrounding dialnorm.

Feature Article - "Dialogue Normalization: Friend or Foe" - June, 2000 (Updated August, 2001)

Thanks,

Paul

DaViD Boulet · 07-27-2007, 11:20 AM

Hey Paul,

the problem with DN isn't that it lowers the dynamic range (volume) by 4 (or more) db... the problem is HOW it does it.

It does it by digitally recalculating the audio data to digitally down-scale the waveform.

That means that you will NEVER get bit-for-bit accuracy from a TrueHD stream that's been flagged with dialog-normalization.

And it can't be bypassed because Dolby won't allow any in-spec consumer gear to let the user avoid it! It doesn't matter that it's just a meta-data instruction flag: if Dolby REQUIRED that it's processing instructions be followed by your decoder, then the fact that the original PCM lossless data was represented prior to DN processing is moot.

Dolby can claim that the process is transparent. But I find it less-than-ironic that when Dolby TrueHD streams on HD DVD didn't sound as good as their PCM counterparts, the first reason offered by industry sources involved with the mastering were "probably because of the Dialog Normalization on the TrueHD".

Also, when I heard a Dolby track on DVD (Lion King) that sounded, to my ears, "like DTS" so much so that it stunned me because I had never heard a Dolby track sound so good, I contacted the audio engineer to ask why. The reply? "Oh, we didn't set the Dialog Normalization flag for that soundtrack". I've since been told by SEVERAL mastering egineers that they personally feel that most of the sonic benefits ascribed to DTS over DD are really just because of the slight sonic degredation from the DN on the DD tracks.

DN is fine for commercials on TV where you're interested in level matching across multiple programs. BUT IT HAS NO PLACE ON A HIGH-FIDELITY SOUNDTRACK FOR A PACKAGED PRODUCT LIKE AN HD DVD OR BD MOVIE SOUNDTRACK. If multiple tracks need to be "matched" for streaming together, like commentary, then the secondary tracks should be mastered with DN to match the "reference" primary dialog level of the master/reference recording. You should never dumb-down the primary audio soundtrack to level-match to menu-clicks and over-compressed commentary.

I find this humorous (from your link):

Quote:

There is more to Dolby Digital than just raw audio information. That digital stream flowing into your decoder has other bits of information - called "Metadata" - along for the ride. Metadata is information your decoder can use to do certain things better, like downmixing the soundtrack to two channels or changing the dynamic range. One piece of Metadata which has gotten a pretty bad rap is Dialogue Normalization. In their 30 year history, Dolby has never been known to add something that would deteriorate audio quality, so I started to investigate. Like so many things in this world, Dialnorm, as we’ll call it for short, is somewhat misunderstood.

I consider a 10:1 data-reduction algorithm used for lossy tracks to degrade sound quality. Don't you? Seems that Dolby's entire *business* has to do with compromises that were necessary in order to work within the confines of practical boundaries. I doubt that dynamic range compression or down-mixing algorithms enhance the sound quality of the PCM master much either.

Note: Lexicon is a high-end digital decoder and may perform its own volume control in the digital domain. However, the sophistication and transparency of Lexicon's algorithms are superior to what comes programmed on "stock" eproms like Dolby Digital decoding chips. Most high-end processors still perform volume adjustment in the analog domain to avoid problems with degredation from digital-processing to reduce amplitude.

Grant H · 07-27-2007, 11:36 AM

Quote:

Originally Posted by DaViD Boulet

Hey Paul,

the problem with DN isn't that it lowers the dynamic range (volume) by 4 (or more) db... the problem is HOW it does it.

It does it by digitally recalculating the audio data to digitally down-scale the waveform.

That means that you will NEVER get bit-for-bit accuracy from a TrueHD stream that's been flagged with dialog-normalization.

And it can't be bypassed because Dolby won't allow any in-spec consumer gear to let the user avoid it! It doesn't matter that it's just a meta-data instruction flag: if Dolby REQUIRED that it's processing instructions be followed by your decoder, then the fact that the original PCM lossless data was represented prior to DN processing is moot.

Dolby can claim that the process is transparent. But I find it less-than-ironic that when Dolby TrueHD streams on HD DVD didn't sound as good as their PCM counterparts, the first reason offered by industry sources involved with the mastering were "probably because of the Dialog Normalization on the TrueHD".

Also, when I heard a Dolby track on DVD (Lion King) that sounded, to my ears, "like DTS" so much so that it stunned me because I had never heard a Dolby track sound so good, I contacted the audio engineer to ask why. The reply? "Oh, we didn't set the Dialog Normalization flag for that soundtrack". I've since been told by SEVERAL mastering egineers that they personally feel that most of the sonic benefits ascribed to DTS over DD are really just because of the slight sonic degredation from the DN on the DD tracks.

DN is fine for commercials on TV where you're interested in level matching across multiple programs. BUT IT HAS NO PLACE ON A HIGH-FIDELITY SOUNDTRACK FOR A PACKAGED PRODUCT LIKE AN HD DVD OR BD MOVIE SOUNDTRACK. If multiple tracks need to be "matched" for streaming together, like commentary, then the secondary tracks should be mastered with DN to match the "reference" primary dialog level of the master/reference recording. You should never dumb-down the primary audio soundtrack to level-match to menu-clicks and over-compressed commentary.

I find this humorous (from your link):

I consider a 10:1 data-reduction algorithm used for lossy tracks to degrade sound quality. Don't you? Seems that Dolby's entire *business* has to do with compromises that were necessary in order to work within the confines of practical boundaries. I doubt that dynamic range compression or down-mixing algorithms enhance the sound quality of the PCM master much either.

David, you remind me of my early high school days, when digital systems were just entering the fray in theaters. For the first time, you had your choice between Dolby Surround and Dolby Digital. It was a while longer before I found a DTS theater not far from me and was in heaven. Suddenly I was driving a lot farther to see films since the sound in my local theater was truly pathetic, like hook your VCR up to a very old, low quality stereo pathetic.

My buddy, studying the sound formats had summed it up: So Dolby compresses the sound so they can have more channels, and then applies noise reduction so you don't realize it sounds crappy.

I always remember that, because it's essentially true, especially when you're talking the original DD that hit theaters, as opposed to the higher quality version we got at home.

RobertR · 07-27-2007, 11:56 AM

Quote:

For the first time, you had your choice between Dolby Surround and Dolby Digital. It was a while longer before I found a DTS theater not far from me and was in heaven. Suddenly I was driving a lot farther to see films since the sound in my local theater was truly pathetic, like hook your VCR up to a very old, low quality stereo pathetic.

Interesting. A friend of mine, who has worked as a projectionist, once offered to have Gary Reber (whose fondness for DTS is very well known) listen to a movie with a DD soundtrack and a DTS soundtrack under blind conditions, and say which one he preferred. So which soundtrack did the well known DTS cheerleader pick? The DD one.

Which tells me that your preference for the "DTS theater" had essentially nothing to do with DD vs. DTS, but the particular theater setup.

Robert George · 07-27-2007, 12:04 PM

Quote:

the problem with DN isn't that it lowers the dynamic range (volume) by 4 (or more) db... the problem is HOW it does it.

It does it by digitally recalculating the audio data to digitally down-scale the waveform.

That means that you will NEVER get bit-for-bit accuracy from a TrueHD stream that's been flagged with dialog-normalization.

You are wrong.

DaViD Boulet · 07-27-2007, 02:58 PM

Robert G,

Roger Dressler has confirmed what I've said is accurate in several discussions at AVS.

DN flag (set to anything other than -31) causes the digital filter in the DD decoding engine to then recalculate the audio data (which *is* bit-for-bit accurate from TrueHD to start with) to perform digital-level-reduction based on the value. You no longer have bit-for-bit accuracy to the original data, just like running your audio data through a sampling rate converter or noise-shaping filter would re-write data.

That's how it works. Even Dolby doesn't challenge this. The only thing Roger Dressler had to say was that he felt personally that the re-writing of the data didn't cause any serious sonic degredation to the signal (a natural point of view for him to take). But he confirmed 100% that the process I've described above is exactly how DN is applied.

Quote:

Which tells me that your preference for the "DTS theater" had essentially nothing to do with DD vs. DTS, but the particular theater setup.

Possibly. But one person's impression in a double-blind doesn't mean that same impression will be shared by another listener. Some listeners prefer the sound of 16-bit audio 20 bit audio, for instance-- not because it's more accurate (which it isn't), but because they like the "bite" it adds to high-frequencies becuase they prefer that sonic signature. The smoother, more accurate sound of the 20 bit sounds too "tame" to some listeners who like the more agressive mid-range of 16-bit sound.

RobertR · 07-27-2007, 04:09 PM

Quote:

one person's impression in a double-blind doesn't mean that same impression will be shared by another listener.

Grant's post clearly indicated that it's not a matter of an "impression". He described the non-DTS local theater as "pathetic". Given that Reber failed to pick the DTS soundtrack despite his often stated preference for DTS (which would be essentially impossible if DD is truly "pathetic" compared to DTS, and it makes no sense to say that Grant has a stronger preference for DTS than Gary Reber), it's clear that mechanisms other than the encoding are responsible for the perceived differences.

I'm curious if you've ever participated in a proper double blind listening test between PCM and Dolby HD.

Quote:

Some listeners prefer the sound of 16-bit audio 20 bit audio, for instance-- not because it's more accurate, but because they like the "bite" it adds to high-frequencies becuase they prefer that sonic signature.

The Reber experience is an example of the opposite happening--someone picking the 20 bit audio in a blind test despite a stated preference for 16 bit, which would once again call into question the source of the preference.