[woodyak]

What is the difference? Which is better? For some reason I can't get my receiver to recognize my DVD player's DTS DVD's. Both reciever and player are DTS compatible and the DVD (LOTR) had DTS. I think the DDolby sounds incredible but I can't help but wonder if the DTS sounds better.


Forgive me if this has been discussed to death. I've tried searching on this site but it's always too busy to do a search.

 

[John Kotches]

Woody,


You asked:

What is the difference?


DTS and DD use different schemes to do lossy compression at a fixed data rate. Both use research in psychoacoustics to minimally impact the fidelity.


DD has the ability to eke more compression out, and does so several ways, most notably by having the spectrum spread into more frequency bands (I think it's 256 bands), with DTS having only 10 (or so) frequency bands.


In addition, DD can define a larger window of time so that more common information can be correlated.

Which is better?


Yes.


That didn't answer your question I realize, but the problem is that on many movies that have both soundtracks, the individual channel levels are radically different, meaning you will automatically favor the louder of the two based on psychoacoustic principles.


Once levels are matched, the differences become less obvious.


Regards,

 

[Raistlin_HT]

Oh boy ... the can of worms has officialy been opened!!

 

[LordHz]

I would have to agree with John Kotches, there is little measurable difference between the methods.


Either is fine..

 

[lwang]

Quote:

Originally posted by LordHz I would have to agree with John Kotches, there is little measurable difference between the methods. Either is fine..

There is plenty of difference that could be measured, but as far as it being audible, that is another issue.

 

[Morbius]

Quote:

Originally posted by John Kotches DD has the ability to eke more compression out,

i.e. DD throws away more information than DTS.


Whether that information is necessary is subject to debate.


That's why Raistlin_HT is correct - it is a "can of worms."


However, both give an acceptable audio experience.


Dr. Gregory Greenman

Physicist

 

[Morbius]

Quote:

Originally posted by John Kotches That didn't answer your question I realize, but the problem is that on many movies that have both soundtracks, the individual channel levels are radically different, meaning you will automatically favor the louder of the two based on psychoacoustic principles.

John,


Additionally, the DVD standard *requires* a DD soundtrack.


However, the DVD may not have room for both a DTS 5.1 soundtrack and a DD 5.1 soundtrack.


So in order to legally comply with the DVD standard, the author of the DTS version of a DVD will put a DD Pro Logic

soundtrack on the DVD.


So if you are going to compare DTS 5.1 to DD 5.1 for the same movie - get 2 different discs. The DTS 5.1 disc may

not have the DD 5.1 track on it - only a DD Pro Logic.


[DD and DTS are codecs - standards for writing/processing the content. They are "independent" of the content - that

is just because you have DD doesn't mean you have DD 5.1, you may only have a Pro Logic signal compressed by DD ].


Dr. Gregory Greenman

Physicist

 

[Colm]

The Dolby site has a discussion of the differences. So does the competition.


There is sometimes definitely an audible difference to me between DD and DTS soundtracks on the same DVD. IMHO it is primarily due to differences in mastering the recording rather than inherent strengths and weaknesses of the technologies. Any differences are really a moot point if you have a receiver that can decode both formats. Just listen to the one that sounds best to you, or both if you watch the movie twice. If your equipment can only DD, don't give it another thought. You already have great sound.

 

[John Kotches]

Greg,


By having 16x the frequency regions more redundant information between channels can be eliminated, allowing for full restoration of the original data.


If it is redundant information that can be recovered on playback, you've increased compression without sacrificing your audio quality at all.


The same is true of the window sizes, where DD can use up to (I think) a 30ms window, vs a 10 to 15ms window for DTS. This being the case, with a larger window, you can (once again) increase compression without deleterious effect on the results.


One cannot say generically "more information is thrown out" without analyzing the input to determine the level of redundant information (ie common content) between channels. With high redundancy it is possible to discard less information where greater compression opportunities lie.


Regards,

 

[John Kotches]

Greg said this in his second reply:

Quote:

However, the DVD may not have room for both a DTS 5.1 soundtrack and a DD 5.1 soundtrack.

Current practice is to use 1/2 rate DTS @ 768K/second, which leaves room for a 448K DD track in 5.1. This is the case with most (but not all current releases.

Quote:

So in order to legally comply with the DVD standard, the author of the DTS version of a DVD will put a DD Pro Logic soundtrack on the DVD.

Correctly termed, it is DD2.0/Dolby Stereo, not DD Pro Logic. The results are the same, assuming sufficient DSP cycles, the PL or PL/2 algorithm can be applied to the Dolby Stereo track to derive a matrix based surround mix. As I said above the majority of DTS soundtracks are half rate @ 768K, and current practice is to include DD5.1 @448K/second + DTS @ 768K/second and sometimes DD2.0 @ 192K/second.


Full bit rate DTS is too much of a space hog @ 1.5Mbits/second + minimal requirements..

Quote:

So if you are going to compare DTS 5.1 to DD 5.1 for the same movie - get 2 different discs. The DTS 5.1 disc may not have the DD 5.1 track on it - only a DD Pro Logic.

Give me a little bit of credit here. Do you think I'd type a comment in such as I did, and NOT use a soundtrack with both DD 5.1 and DTS5.1 included?


Examples include, but are not limited to:

Shrek, Steely Dan Two Against Nature, Diana Krall Live in Paris, Unbreakable and numerous others.


Flat out, at the same volume setting, the SPL levels do not match. It is most apparent that the LFE channel, and the surrounds do not match between mixes.


Is this because the encoders are fed different mixes or because the encoders produce different results from the same in put. I sincerely hope its the former, because if its the latter something is seriously amiss with one of these two codecs. I have no way of knowing, because I'm not there seeing the inputs to the two encoders.


Quote:

[DD and DTS are codecs - standards for writing/processing the content. They are "independent" of the content - that is just because you have DD doesn't mean you have DD 5.1, you may only have a Pro Logic signal compressed by DD ].

See above.


Regards,

 

[John Kotches]

Greg also said this, and I missed it:

Quote:

Additionally, the DVD standard *requires* a DD soundtrack.

Read the standard again.


The real standard is either a DD soundtrack or Linear PCM stereo, sampled as high as 24bits/96kHz.


The DAD discs from Chesky and Digital Compact Classics (I think) have only Linear PCM @ 24/96K, there is no DD track at all.


Regards,

 

[sdurani]

John,


You forgot your usual example: the 'Hell Freezes Over' DTS DVD, which contains 5.1-channel DTS and 2-channel PCM but no DD track.


Best,

Sanjay

 

[John Kotches]

Sanjay,


Joni Mitchell's Painting with Words & Music is another example.


Regards,

 

[Morbius]

Quote:

Originally posted by John Kotches G Give me a little bit of credit here. Do you think I'd type a comment in such as I did, and NOT use a soundtrack with both DD 5.1 and DTS5.1 included?

Good Grief!


John, the comment that you are referring to was actually directed to the casual reader!


Your original post could have been interpreted as saying that one could compare the DTS and DD tracks on the same

disc that did not have both 5.1 tracks - an admonishment to the contrary not forthcoming.


Dr. Gregory Greenman

Physicist

 

[Morbius]

Quote:

Originally posted by John Kotches Flat out, at the same volume setting, the SPL levels do not match. It is most apparent that the LFE channel, and the surrounds do not match between mixes. Is this because the encoders are fed different mixes or because the encoders produce different results from the same in put. I sincerely hope its the former, because if its the latter something is seriously amiss with one of these two codecs. I have no way of knowing, because I'm not there seeing the inputs to the two encoders.

John,


The average SPL levels can be different for another reason.


The encoder that preserves a greater dynamic range will tend to have a lower SPL level. That's just efficient management

of the "dynamic range" of the available bit-space.


The average level of the digital data has zero significance. What one wants to do is get the best representation of the

signal with the available bit patterns. If this means the average level is lower - so be it - and use a little more

gain from the amp.


If someone wants to do comparisons, they need to level match, as you correctly point out, in order to get meaningful results.


Dr. Gregory Greenman

Physicist

 

[Dougofthenorth]

I read the series of "shootout" papers that DTS & Dolby fired back & forth.

DTS's papers can be found at http://www.dtsonline.com/ I lost the DD link.


I think & mention with all due respect, (re the posts in this thread) that we must also consider that some DVDs are direct transfers from the original movie soundtracks to DVD. This is mostly found with Dolby. The parameters of a soundtrack for a Cinema are vastly different than one for a HT. DTS on the other hand is often "translated" for HT from the cinema master then transfered to DVD. It is then EQ'd for home theatres. If one views the extras on the DVD Se7en, one can find an extra feature which discusses the "tailoring" of the sound track of this movie for HT specifically & discusses some of the issues I am mentioning here.


I might also note that some HT recv'rs have a DD & DTS EQ setting menu to tailor (tweak) the sound to more suit HT.


Dougofthenorth

 

[sdurani]

John,
Quote:

Flat out, at the same volume setting, the SPL levels do not match. It is most apparent that the LFE channel, and the surrounds do not match between mixes.

Here's one example where you can compare the LFE channels of three formats that were encoded from the same master. It's the DVD-Audio disc of 'Germerica'. Note the similarities between the MLP and DD tracks, versus the DTS track.
Quote:

Is this because the encoders are fed different mixes or because the encoders produce different results from the same input. I sincerely hope its the former, because if its the latter something is seriously amiss with one of these two codecs. I have no way of knowing, because I'm not there seeing the inputs to the two encoders.

On this particular example, I don't think the encoders were being fed different mixes; the discrepency seems to be pointing to the output produced by the DTS encoder.


Best,

Sanjay

 

[John Kotches]

Sanjay,


Thanks for the link. I knew it was out there somewhere.


What we see in this example is that the DTS encoder/decoder pair has failed at its primary mission, as it has added to the LFE content at the bottom end, and subtracted at the top end in comparison to both the original input (MLP) and the competition DD.


Kind of dampens the while "more bits is better" argument.


It does however lend creedence to my earlier point, which is "louder is better" perceptually.


Care to comment Dr. Greenman?


Regards,

 

[John Kotches]

Quote:

The average SPL levels can be different for another reason. The encoder that preserves a greater dynamic range will tend to have a lower SPL level. That's just efficient management of the "dynamic range" of the available bit-space. The average level of the digital data has zero significance. What one wants to do is get the best representation of the signal with the available bit patterns. If this means the average level is lower - so be it - and use a little more gain from the amp. If someone wants to do comparisons, they need to level match, as you correctly point out, in order to get meaningful results.

That's all well and good, but my comment was that DTS had a higher SPL level from LFE and surrounds (average and peak) in a calibrated environment.


If you want specific examples, try Chapters 1-4 of Unbreakable and measure in the fashion that HFR did with Germerica. You will find that for LFE, the DTS program levels average higher and peak higher than the DD levels do.


In this case we don't know what the encoder input was, which is a big unknown in this case.


I've seen this effect on multiple manufacturers of DSPs (Motorola, Crystal Semiconductor, Analog Devices) so it is not a matter of codec interpretations.



Regards,

 

[sdurani]

John,
Quote:

Thanks for the link. I knew it was out there somewhere.

You're welcome. It's a handy one to keep bookmarked for when discussions like this come up. I just wish there were more examples like it.
Quote:

Kind of dampens the whole "more bits is better" argument.

That argument never held much water for me because DD always seemed to be a more efficient codec than DTS. Imagine DTS at 448K, let alone 384K (old DD rate on laserdisc).


Best,

Sanjay