DTS/DD vs. CD Audio quality-opinions?

[kexodusc]

Three of my friends and I performed a little test this weekend, comparing 5 albums of various music types. We used my friend's setup: a Yamaha RX-V1400, with Paradigm speakers - 6 Studio 20's, a PW-2100 and the CC-470 for playback. Older Yamaha CD player and Panasonic DVD player were used. The goal was to determine if DD/DTS audio tracks sounded better than CD audio.
Here's what we collectively came up with:
Metallica (Black Album) - Definitely better in DD than CD audio in stereo
Diana Krall - Live in Paris - Again, DTS better than the CD audio version
Pink Floyd - Dark Side of the Moon - 5.1 DD has won the test already.
Neal Young - Harvest - Don't know if it was the recording or what, but 2-channel stereo won out over 5.1 audio
The Eagles - Hell Freezes Over - 3 votes to 1 in favour of DTS over CD Audio.

Now, is there some technical information I can use to support the results of this admittedly subjective test? Is it bit rate/sec, sampling rates, 16 vs. 24 bit "word length"? Or do we just subjectively prefer the added ambience a 5.1 music source delivers?
Any comments?

[mtrycraft]

Did you compare stereo of the CD and multi channel for DD/DTS? Multi channel will or should sound better. That is one of its benefits:)

[kexodusc]

We did compare the multi channel, but we were listening for detail. We also compared 2.0 Dolby Digital, on a few which still clearly sounded more detailed and natural. Maybe this is the norm, but I was previously under the impression that Dolby Digital and DTS were lossy forms of compression that didn't quite measure up to CD (which is limited in itself).

[recoveryone]

Three of my friends and I performed a little test this weekend, comparing 5 albums of various music types. We used my friend's setup: a Yamaha RX-V1400, with Paradigm speakers - 6 Studio 20's, a PW-2100 and the CC-470 for playback. Older Yamaha CD player and Panasonic DVD player were used. The goal was to determine if DD/DTS audio tracks sounded better than CD audio.
Here's what we collectively came up with:
Metallica (Black Album) - Definitely better in DD than CD audio in stereo
Diana Krall - Live in Paris - Again, DTS better than the CD audio version
Pink Floyd - Dark Side of the Moon - 5.1 DD has won the test already.
Neal Young - Harvest - Don't know if it was the recording or what, but 2-channel stereo won out over 5.1 audio
The Eagles - Hell Freezes Over - 3 votes to 1 in favour of DTS over CD Audio.

Now, is there some technical information I can use to support the results of this admittedly subjective test? Is it bit rate/sec, sampling rates, 16 vs. 24 bit "word length"? Or do we just subjectively prefer the added ambience a 5.1 music source delivers?
Any comments?

I would at first suggest expanding your (Various) music choices to include some jazz, classical and R&B. IMHO slower type music gives you a chance to really evaluate the true sound of any disk (when you can hear each instrument). I remember back in 84 when CD's were coming out and I went to a local HiFi shop in the New England area. The sales person put in a Classical CD and my ears and jaw just hit the floor. I don't really care for this form of music, but hearing every little sound so clear (by those standards then) was just awesome. Now with High Rez form of music out there (192/24-bit, 96/24-bit & 48/24-bit) the sound is much fuller and rich almost like sitting in on a studio session. IMHO live music (concerts) is not all that good in this format and that may be why I prefer the jazz style in High Rez. You hear each and every note from all insturment. I would ask you to try Donald ***en Nightfly DVD-A and play it on the DTS mode and sit back and close your eyes. And you should not have to play any of the test disk at high volume to hear the difference

[kexodusc]

Admittedly there was no classical music sampled...we just didn't have one.
However plenty of jazzy bluesy stuff in Diana Krall and slower music with Neil Young. Truth be told, the most spectacular of the bunch was the somewhat progressive passages in Metallica's album.
I do agree that the slower jazzy selections would offer a more noticeable difference but that being said, if a format is suppose to be better, it should noticeable regardless of the type of music being played.
I have no doubt DVD-A is superior to CD. But our test was to compare basic DVD-video DTS and Dolby Digital recordings to the same recording on CD, not DVD-audio hi-res 5.1 formats.
We generally found the Dolby Digital (448 kbps) to be superior to CD Audio's sound despite giving up bit rate. I'm just trying to find an easy answer as to why? I'm assuming bit-rate isn't the all important factor.
For MP3's and similar formats, generally, the better the bitrate, the better the sound. Why doesn't this hold true for the CD vs. Dolby Digital or DTS? I can only assume it has to do with sampling rate or 24bit "wordlength" producing more resolution.

[mtrycraft]

We did compare the multi channel, but we were listening for detail. We also compared 2.0 Dolby Digital, on a few which still clearly sounded more detailed and natural. Maybe this is the norm, but I was previously under the impression that Dolby Digital and DTS were lossy forms of compression that didn't quite measure up to CD (which is limited in itself).

CD is not a lossy format, DD and DTS are. CD is limited to human perception capability, or the vast majority of it.
Perhaps your listening protocol has something to do with this? One may have been a little louder than the other?

[kexodusc]

We had to adjust volumes, as the DVD's generally seemed louder than CD source.
I understand the lossy encoding nature of DTS/DD. I wonder, since DTS has a higher bitrate than CD audio, and both have higher sampling rates and use 24 bit wordlengths, does this over come the "lossy compression" set back?.
Is it possible that compressing a 24 bit, 48 kHz audio track could sound better than CD's 44.1 kHz, 16 bit signal without compression?
What I'm really looking for is a decent article or link comparing the three formats, I haven't been able to find one that compare sound quality with fundamentals to support the analysis.

[mtrycraft]

We had to adjust volumes, as the DVD's generally seemed louder than CD source.
I understand the lossy encoding nature of DTS/DD. I wonder, since DTS has a higher bitrate than CD audio, and both have higher sampling rates and use 24 bit wordlengths, does this over come the "lossy compression" set back?.
Is it possible that compressing a 24 bit, 48 kHz audio track could sound better than CD's 44.1 kHz, 16 bit signal without compression?
What I'm really looking for is a decent article or link comparing the three formats, I haven't been able to find one that compare sound quality with fundamentals to support the analysis.

DTS cannot be higher bit rate than CD as it is perceptually coded, information is discarded that is deemed inaudible due to masking. Same with DD or any perceptual coding. The bit rate is not everything but the algorythm used is. That is what calculates and discards inaudible signals.
DTS/DD has 5.1 channels too, so that may explain what you may be thinking.

I am not aware of any reliable DBT comparisons on the three.

[Woochifer]

DTS cannot be higher bit rate than CD as it is perceptually coded, information is discarded that is deemed inaudible due to masking. Same with DD or any perceptual coding. The bit rate is not everything but the algorythm used is. That is what calculates and discards inaudible signals.
DTS/DD has 5.1 channels too, so that may explain what you may be thinking.

I am not aware of any reliable DBT comparisons on the three.

Impossible to do a DBT like you specify given that there are no soundtracks I'm aware of that have two-channel 16/44.1 PCM, DD, and DTS tracks that were encoded simultaneously.

[Woochifer]

Three of my friends and I performed a little test this weekend, comparing 5 albums of various music types. We used my friend's setup: a Yamaha RX-V1400, with Paradigm speakers - 6 Studio 20's, a PW-2100 and the CC-470 for playback. Older Yamaha CD player and Panasonic DVD player were used. The goal was to determine if DD/DTS audio tracks sounded better than CD audio.
Here's what we collectively came up with:
Metallica (Black Album) - Definitely better in DD than CD audio in stereo
Diana Krall - Live in Paris - Again, DTS better than the CD audio version
Pink Floyd - Dark Side of the Moon - 5.1 DD has won the test already.
Neal Young - Harvest - Don't know if it was the recording or what, but 2-channel stereo won out over 5.1 audio
The Eagles - Hell Freezes Over - 3 votes to 1 in favour of DTS over CD Audio.

Now, is there some technical information I can use to support the results of this admittedly subjective test? Is it bit rate/sec, sampling rates, 16 vs. 24 bit "word length"? Or do we just subjectively prefer the added ambience a 5.1 music source delivers?
Any comments?

Very interesting results, although I would caution that you're comparing surround tracks with two-channel tracks, and there are lot of other variables at play other than the formats themselves. For example, in order to put together a 5.1 soundtrack for an older recording, the mixing engineer needs to actually go all the way back to the multichannel master tape, which might have significantly different tonal characteristics than what ultimately made its way down to the original two-channel master tape (a lot of vinyl masters had EQ and/or compression applied, and the CD transfers were done from these without compensating for this). Also, keep in mind that a lot of CDs have had multiple remasterings done, so different versions of the CD releases can also sound different (a lot of the early transfers were not done properly).

I've done similar comparisons with all sources folded down to two-channel, but even that method is flawed given that the channel balances between two-channel and repurposed 5.1 mixes can be very different, and create a dissimiar sound overall even in two-channel. In general, I've noted that DD can come very close to the audio quality of the CD versions, and can be superior in some cases (these were all two-channel direct comparisons). These differences could have more to do with the mastering or other source differences. In cases where I know that the DD and DTS tracks were encoded simultaneously at identical levels, the DTS track typically sounds subtly better.

I think for now, it's a good subjective comparison for the various versions of those specific recordings. Whether or not those listenings can be applied more widely is where some caution's in order.

[kexodusc]

DTS cannot be higher bit rate than CD as it is perceptually coded, information is discarded that is deemed inaudible due to masking.

mtrycraft: I admit to being very inexperienced with the technical details of digital formats, so I have to ask a question here. Every source I've been able to find tells me CD audio quality's bit rate is 1.4Mb/s, whereas DTS bit rate is 1.5 Mb/s. Masking and encoding aside, my simple math gives me a +0.1 edge to DTS over CD. What am I doing wrong here?
How is 1.5 not bigger, or more than 1.4???
My head is starting to hurt...
I do agree with you, bitrate isn't everything, one has only to compare MP3 with WMA files to realize that.
Still, my admittedly subjective test left me somewhat embarassed, as I previously had been telling my friends that there was no way a DVD concert's audio quality could top that of a CD...boy was I wrong.

[maxg]

It may be that you have proved that your system is better at surround sound that at stereo. Repeating the experiment in a 2 channel only environment might yield very different results (it did for me).

Also worth mentioning that when I have compared DVD's to CD's using the 44.1/16 bit soundtrack on the DVD the CD has generally sounded way better. case in point Elton John double album of greatest hits which I have on DVD, CD and vinyl. The DVD is so bad in comparison to the other 2 it aint funny.

I have other examples (but cant remember them as I type):

As for your DTS throughput figure - that is for 6 channels of audio as compared to CD's 2 channels. DTS does seem to sound better than the equivalent DD5.1 tracks but I am not sure for music it beats CD (in 2 channel format).

[kexodusc]

Maxg: You are very much right. The CD in 2 channel sounded a way more real than Dolby Digital in stereo, I wouldn't compare those two formats. 2-Channel Dolby Digital seemed to have way too much bass output at the sub, had to be turned back a bit. Might just be the setup we used.

As far as my "test" results go, I don't think I've found much at all, nothing of any reasonable value to anyone at least. If everyone would go back and read the original post, you'll see I was looking for any kind of scientific support to reinforce our very subjective test. Like any subjective test, 4 different people could use the same equipment, the same sources, and have the opposite results too. I was just wondering if there was any science out there to support our result...don't think I'm going to find it now.

At this point, I might consider doing the test again, but with a soundmeter to more accurately measure volumes. I also wonder how much the studio engineering influences the results...maybe more detail was given to the 5.1 tracks than the stereo tracks? I don't know.

Thanks to all who have posted, by the way...some food for thought

[mtrycraft]

mtrycraft: I admit to being very inexperienced with the technical details of digital formats, so I have to ask a question here. Every source I've been able to find tells me CD audio quality's bit rate is 1.4Mb/s, whereas DTS bit rate is 1.5 Mb/s. Masking and encoding aside, my simple math gives me a +0.1 edge to DTS over CD. What am I doing wrong here?
How is 1.5 not bigger, or more than 1.4???
My head is starting to hurt...
I do agree with you, bitrate isn't everything, one has only to compare MP3 with WMA files to realize that.
Still, my admittedly subjective test left me somewhat embarassed, as I previously had been telling my friends that there was no way a DVD concert's audio quality could top that of a CD...boy was I wrong.

I am pretty sure that DTS bit rate is for all channels, not one channel as in CD.
That would make DTS bit rate of about 7.5Mbits/sec+. I seriously doubt that.
Why not ask Sir Terrence at HT. :)

[kexodusc]

mtrycraft: That would better explain the relative file sizes between CD and DTS tracks, good thinking man...
Curious, if theres a bit of information being translated back into audio information, does it matter what channel it is emitted from as long as the sound is being reproduced?
Certainly each channel in stereo would have more detail than its counterpart in 5.1, but the cumulative effect...never mind...I'm obviously in way over my head...sigh...getting old...
Well, I've opened a whole can of worms with this thread. Curious as I am, I have to know how and why everything works, sometimes at the expense of enjoyment.
I'm going to have to learn to get past that.
Things were so much simple with vinyl...

[maxg]

Curious, if theres a bit of information being translated back into audio information, does it matter what channel it is emitted from as long as the sound is being reproduced?
Certainly each channel in stereo would have more detail than its counterpart in 5.1, but the cumulative effect...never mind...I'm obviously in way over my head...sigh...getting old...
Well, I've opened a whole can of worms with this thread. Curious as I am, I have to know how and why everything works, sometimes at the expense of enjoyment.
I'm going to have to learn to get past that.
Things were so much simple with vinyl...

No it doesnt matter which channel is being reproduced. It does matter, however, when there are 6 channels being reproduced together.

Vinyl has its own complexities but ultimately is probably more simple to understand. Another reason I have re-embrased it....

[Sir Terrence the Terrible]

I am pretty sure that DTS bit rate is for all channels, not one channel as in CD.
That would make DTS bit rate of about 7.5Mbits/sec+. I seriously doubt that.
Why not ask Sir Terrence at HT. :)

The Dts bit rate is spread over the 5.1 channels. For DVD the bitrate is 1509kbps. To understand how Dts(and DD potentially)can sound better than 16/44.1 you must understand how the encoding process works. To put in in simple terms Dts encoding discards signals that are masked and rendered inaudible by louder adjacient signals(precedence effect) Once those signals are discarded the encoding process breaks down these signals by frequency. These signals are encoded into the Dts bitstream at 20 bit depths. So what you hear is 16/44.1khz signals at 20/48khz resolution. Because more bits are used, and a slightly higher sample rate, you get lower noise and greater dynamic range. Where Dts has the edge over DD lies in Dts's ability to encode not only the primary signals, but it also codes the sub band frequencies as well because it has the bits to do so.

One would think that uncompressed 16/44.1khz PCM should sound better than compressed 20/48khz. But 16/44.1khz contains signals that are heard, and unheard so in theory because of the masking effect of louder signals in relationship to the softer unheard signals, you are not hearing more detail. Dts and DD more efficiently code signals that ARE heard with more resolution, lower noise and greater dynamic range(CD has a dynamic range of 96 db and Dts has about 120db)

[mtrycraft]

. So what you hear is 16/44.1khz signals at 20/48khz resolution. Because more bits are used, and a slightly higher sample rate, you get lower noise and greater dynamic range.

This can only be true if the master had the 20 bit depth. If your source is only 16 bits, that is all you can hope to ever get. But I would defer this to better experts in th efield, beyond both of us. Sorry.

But 16/44.1khz contains signals that are heard, and unheard so in theory because of the masking effect of louder signals in relationship to the softer unheard signals, you are not hearing more detail.

Yes, :)
That is exactely why perceptual coding works so well. :)

[kexodusc]

Wonderful explanation Terrible T, simple, logical, and easy to follow. At least now I have a better grasp of the variables involved and their impact on sound quality.

[Sir Terrence the Terrible]

. So what you hear is 16/44.1khz signals at 20/48khz resolution. Because more bits are used, and a slightly higher sample rate, you get lower noise and greater dynamic range.

This can only be true if the master had the 20 bit depth. If your source is only 16 bits, that is all you can hope to ever get. But I would defer this to better experts in th efield, beyond both of us. Sorry.

But 16/44.1khz contains signals that are heard, and unheard so in theory because of the masking effect of louder signals in relationship to the softer unheard signals, you are not hearing more detail.

Yes, :)
That is exactely why perceptual coding works so well. :)

Well Mtry, its not as simple as 16 bits is all you get when you speak of encoding to Dts(or any high bit perceptual encoder). Remember, during the coding process you are getting rid of unheard signals. That includes the floor noise, and any signal that have louder adjoining signals. It is then re-encoded at 20bits at 48khz sampling. While the sampling doesn't necessarily raise the cutoff frequency(the 16bit audio has already limited that to 22.050khz) you will lower the noise floor and get an increase in dynamic range as a result. The audio is also reclocked(which would reduce any process inducing jitter) which aids in smoothing the more complex instrument textures(strings, cymbal crashes glocks and high brass). So you may only have a 16 bit master, but there are some 20 bit performance features you get from coding the information at 20 bits

[hertz]

2 channel redbook audio cd if mastered properly will sound better than DD or DTS anyday for the simple reason that both DTS and DD use lossy compressions.The test that you did proves any of the following:
1.Your audio setup is better for movies than stereo.
2.Some audio cds are mastered very badly.There is a chance that the newer DTS or DD version could have done a better job.THIS DOES NOT PROVE THAT DD OR DTS FORMAT IS BETTER.

[av_phile]

2 channel redbook audio cd if mastered properly will sound better than DD or DTS anyday for the simple reason that both DTS and DD use lossy compressions.The test that you did proves any of the following:
1.Your audio setup is better for movies than stereo.
2.Some audio cds are mastered very badly.There is a chance that the newer DTS or DD version could have done a better job.THIS DOES NOT PROVE THAT DD OR DTS FORMAT IS BETTER.

There are good and bad recordings in every formats. How a performance is captured and how well it is encoded in LP, CD. or DVD is what matters. The recording engieners or the studio has lot to do in making excellent or lousy recordings in whatever media. The exception being the cassette format which is really technically inferiror as an analog medium.

But I tend to agree with Sir Terenece's brillant laymanish explanation of the difference between CD and DD/DTS. At first glance you would think that because DD/DTS uses lossy compression versus uncompressed encoding in CDs, that the former is inferior. But psychoaccoustic-based compressions simply discard what are unnesessary in the encoding process, thus leaving more room for the more audible segments of the musical passage to get more bits. CD's uncompressed format allocates the same 16-bit word length even for complete silence. I consider that an unnecessary waste.

Hence, DD/DTS sound works wonderfully to make prominent instruments sound more detailed. I, too, have experienced hearing some sonic detailing and spatial depth in DD/DTS that I did not hear with CDs of the same title. Even the instrumental balance is different. I cannot say outright that DD/DTS is better, Some are. Like I said there are good and bad titles in every format. DD/DTS definitely sound dfifferent in a way that gives a totally new listening experience. I can only assume that the reocrding engineers who revisited the old multi-track masters did a good job at remixing into multi-channel and DD or DTS format. I can speak of the George Benson Breezin' reocridng which I had as an LP, then CD and now in DVD-Audio where I only play the DD tracks as I still don't have a universal player. Same with Fourplay, I have the CD and the DVD-A. Much as I am a stereophile, I must admit I get more instrumental detailing and depth in multi-channel playback of these recodings.

However, I must admit that not all DD/DTS transcriptions are an improvement over their CD counterpart. I find The King and I DVD soundtrack no better that the CD soundtrack. Same with the Fiddler On The Roof. However, the My Fair Lady DVD sound has more body than the CD soundtrack of the same. But that's just me.

[Sir Terrence the Terrible]

2 channel redbook audio cd if mastered properly will sound better than DD or DTS anyday for the simple reason that both DTS and DD use lossy compressions.The test that you did proves any of the following:
1.Your audio setup is better for movies than stereo.
2.Some audio cds are mastered very badly.There is a chance that the newer DTS or DD version could have done a better job.THIS DOES NOT PROVE THAT DD OR DTS FORMAT IS BETTER.

This information isn't quite correct. All a lossey codec does is pack signals more efficently. Any signals removed are below the threshold of hearing anyway, so it doesn't matter that there are discarded. 2 channel redbook CD does not have the dynamic range or the spatial deminsionality of either Dts or DD. Redbook CD does not have the flexibilty of bitrate, bit depth, or sample rate that Dts has. Redbook CD does not have the LFE channel which can be used for more bass impact. Dts(and DD for that matter) can have a lower noise floor than redbook CD.

A well mastered Dts CD with the highest bit rate can sound identical to the master tape. It also has the benefit of multichannel to properly place ambience where it belongs, to the sides and rear. It can operate on the fly, and transparently in a lossey, or lossless mode depending on the complexity of the original PCM signal.

So to say the redbook CD will sound better than Dts and DD based on the sole fact that they are lossey is not factual and just shows a lack of understanding of lossey codecs, and how they encode audio.

[mrkiko]

I agree with most of the posts here, just want to add something.
1) Perfectly agree with the fact that it all depends, first of all, on the mixing. If the music is not recorded and mixed with quality, there is nothing no good raw material to encode. And unfortunately this happens too many times.
2) It depends on your gear. Using a cheap reciever, not really good speakers, and using junk cables, one probably has nothing to worry about these differences.
3) I suppose that you meant PCM vs DD or DTS. CDs can be encoded both in PCM (Pulse code modulation) AND in DTS.
PCM requires lots of space, because it has a lot of data. (used either for silence, so there is really alot of infomation, non compressed, but is not the music you hear (or your gear is reproducing), it's just a lot of data. DD and DTS has less data, but more "audible" ones.

PCM doesn't sound good "because" it is a better tecnology. The same music, if recorded both in PCM, DD, and DTS, with the same mixing effort (including artistical capacity, that you can't measoure) will sound well in PCM, some better in DD, and much better in DTS. (The best would be in SACD.)

Unofrtunately there are really few recordings done in all these formats. What's more, i would add that is quite frustrating to have a +20000 system, to spend more tha 1k bucks on cables, and to find out that 90% of the dvd's are not much better then your old stereo VCR. There is a lot of tecnology, but unfortunately we are lacking artists. Like having a lot of painting, but no designers or painters.

Now some tech data.
DD compresses a 5.1 soundtrack to 448k. It is capable to 640, but being part of the DVD standard, it is limited to 448.
DTS uses 1.4Mbps for CD and 1.5 for movies.
Dolby sais that they have a better compression pattern. Maybe. But not 3.4 times better. This is why the audio of the DTS movies (concerts etc) souns better. When the original recording is better.

By standard (LAW) every DVD has to have a PCM (stereo, what you would got on the cd) track, a DD track, and some a DTS track.
Try DTS. Good sound (if the recording is good). Swithc to DD. You will hear that is worst. Swithc to PCM. Crap.

[sslabs]

First I must say that while this is kind of interesting to compare all this, it really is a BAD test for so many reasons.

First to be more accurate in comparing these codecs, stereo versions of Dolby, DTS, and CD at max bit rates would be more fair. That isn't possible, so it's a funky way to compare everything at best.

Next, it wasn't clear just what kind of Dolby and DTS tracks were compared. There are a lot of numbers being thrown around, but they are not set in stone. Yes DTS has a max bit rate of 1.5mb, and DD 448, (the higher one is never used). But for movies the 1.5mb rate is RARE! The rate most commonly used for DTS video is the 754K rate, not the 1.5mb. The max DTS bit rate on a movie would absolutely smoke up space like you wouldn't believe.

Next, the max for DD is 448, but the 348? rate is quite common as well.

When comparing DTS to anything it should be made clear what kind of DTS disc is being used. If it's a concert DTS title, it might run at the 700+ K rate, not the max. A DTS music CD does run at over 1.4mb and that rate is fixed.

Also, what dark side of the moon in DD 5.1 is that? I know the DSOTM 'making of' is dolby stereo only isn't that correct? So what version is this?

As for the compression techniques used, it goes far beyond simply being technically lossy. There are other tricks going on to compress the data while making the fewest changes to the sound (that can be percieved). I bet most people don't know that DD is no longer stereo past 15kHz. To save space, sounds above that become mono. Many would argue that at 15k sounds aren't as noticable. DTS has many similar tricks as well.

I tend to shy away from DTS, MP3 etc based on theory alone. The very basis of these codecs is to CHANGE the sound to make other things possible. Whether or not I or anyone else can tell is kinda beside the point. For DVD movies DD and DTS work just fine. But in the case of music we now have SACD and DVD-A, so the CD vs DTS thing is Soooo late 1990's. Sorry I couldn't resist ;0)

As for CD not being lossy, the isn't entirely true. PCM by its very nature is lossy if you really think about it. Without decimation, a CD would hold like what? 3 minutes of music?

Even having said all that, I have come across plenty of DD, DTS and CD that sound really good. But comparing them in this fashion is just plain weird. The dolby 2.0 tracks run at about 180+K? And DTS stereo doesn't really exist, except maybe as an artificial down-mix.


..... more thoughts to ponder

- Tony



www.StrangerSoundLabs.com

[Sir Terrence the Terrible]

First I must say that while this is kind of interesting to compare all this, it really is a BAD test for so many reasons.

First to be more accurate in comparing these codecs, stereo versions of Dolby, DTS, and CD at max bit rates would be more fair. That isn't possible, so it's a funky way to compare everything at best.

Tony, in theory you are correct, however if you were to encode only two channels of Dts at full bit rate, it would be lossless when compared to the CD max bit rate. Dts carries a bit rate of 1.411kbps for the 44.1khz platform divided between six channels. CD also carries a maximum bit rate of 1.411kbps but with only two channels. With only two channels in operation, that would make Dts a coder, instead of a encoder. Dts could make a bit for bit copy without any loss whatsoever. Dolby digital in its current form is fixed at 192kbps per channel(384kbps total) in its two channel mode. So there is really no way in this case to make a fair comparison without compromising something in the process.


Next, it wasn't clear just what kind of Dolby and DTS tracks were compared. There are a lot of numbers being thrown around, but they are not set in stone. Yes DTS has a max bit rate of 1.5mb, and DD 448, (the higher one is never used). But for movies the 1.5mb rate is RARE! The rate most commonly used for DTS video is the 754K rate, not the 1.5mb. The max DTS bit rate on a movie would absolutely smoke up space like you wouldn't believe.[\quote]

You are correct about the variable data rates used on DVD. However depending on the length of the program, and how many audio tracks, Dts does occasionally use the full bit rate of 1509kbps for music video DVD's. DVD's that have classical music with video almost always use the highest bit rate.

No studio that I know of uses the 384kbps data rate for DD movies anymore. All the studios at the behest of Dolby use the 448kbps data rate now. Yes Dts full bit rate is a space hog on DVD's, but that can be conquered by using DVD-9, 14, 18, and a higher expense of course.

[quote]Next, the max for DD is 448, but the 348? rate is quite common as well.

Actually its the other way around as I stated above.


When comparing DTS to anything it should be made clear what kind of DTS disc is being used. If it's a concert DTS title, it might run at the 700+ K rate, not the max. A DTS music CD does run at over 1.4mb and that rate is fixed.[/qoute]

When you speak of concert titles, Dts could run at both rates. I know a few of their classical music videos do run at 1509kbps.

[quote]As for the compression techniques used, it goes far beyond simply being technically lossy. There are other tricks going on to compress the data while making the fewest changes to the sound (that can be percieved). I bet most people don't know that DD is no longer stereo past 15kHz. To save space, sounds above that become mono. Many would argue that at 15k sounds aren't as noticable. DTS has many similar tricks as well.

DD is no longer multichannel above 15khz, not stereo. But an argument could be made that its bit sharing process doesn't exactly make is discrete in the true since of the word. The only "tricks" to be found in the coherent acoustics codec is the roll off above 15khz at the 754kbps data rate to save data space. For film sound, that is not a problem at all. But for music this could present a problem for instruments with lots of high frequency harmonics or overtones.

[quote]I tend to shy away from DTS, MP3 etc based on theory alone. The very basis of these codecs is to CHANGE the sound to make other things possible. Whether or not I or anyone else can tell is kinda beside the point. For DVD movies DD and DTS work just fine. But in the case of music we now have SACD and DVD-A, so the CD vs DTS thing is Soooo late 1990's. Sorry I couldn't resist ;0)[/qoute]

You have a huge misunderstanding about how codecs work if you think they are designed to "change" things. The purpose of an audio codec is to more efficiently encode the audio, with little or no change in audio quality. Some audio codecs are better than others at reaching this goal. Dts IMO is the best because at its highest data rate, it A/B's with complete transparency when compared to the printmaster tape. No other codec(including half rate Dts) can do this. Both DVD-A and SACD still do not have nearly the market penitration of CD, DD, and Dts, so its not really soooo late 1990's to make this comparison.

[sslabs]

All points taken, but I still hold to the fact that DD, DTS and MP3 change the sound, in fact they all do, and the results can be measured, and in fact they have been discussed at great length many times in magazines and on the net, whether we can tell is another story. To say that I don't understand how codecs work, and that they merely pack the data more efficiently seems a little defensive rather than anything based on fact. MLP for DVD-A is a way to pack data more efficiently without currupting it. DD, and DTS are not.

By just pointing at how DD and DTS go from stereo to mono at certain freq ranges proves that the original input sound has been manipulated to something else. I need not point to another fact to make that point. But I'm not saying that it means it will sound horrible, just that it goes against what everyone here (I think) is about, being an audiophile.


Not trying to start a flame war here, just my honest feelings.

As for the 192K stereo rate for dolby, I believe you are incorrect to double the number.

When the data rate is displayed at say 348 that's for all six channels (5.1) when that rate is displayed for stereo, (192) that should be the total, or 96k per channel, isn't that correct? I could be wrong but that would seem to be right.

As for market penetration, DTS shouldn't even be in the argument. There are more DVD-A and SACDs floating around out there than DTS music CDs.


Anyway, great conversation starter, but an uneven test at best. But I must admit that in this case it's all about the 'conversation' otherwise nobody would be hangin' out here at all right?






- Tony

[jeskibuff]

Someday, if I ever get the time (seems to be in short supply these days), I plan on dubbing some hi-res DVD-A and SACD stereo tracks to a recordable CD. Then, play the original source and the recording back simultaneously, volume matched to provide a good A/B comparison. This seems about the best way to compare the difference between the formats, as we can't rely on the recording companies to provide us with equivalent recordings in various formats.

Sure, there should be an expected degradation of the signal due to the reconversion of the analog signal back to digital (for recording), then back to analog again during playback, but will that be more severe than the supposed "graininess" of the PCM format as described by vinyl afficionados? I anticipate that there really won't be a big difference, but I think it will be a good experiment. Has anyone tried this yet?

[Sir Terrence the Terrible]

All points taken, but I still hold to the fact that DD, DTS and MP3 change the sound, in fact they all do, and the results can be measured, and in fact they have been discussed at great length many times in magazines and on the net, whether we can tell is another story.

I will give you the fact that MP3 and DD do alter the sound, and yes its very audible. But I know for a fact the Dts at its highest bit rate does not audibly alter the sound at all. While the encoding process does alter the original wave form to a degree(that's what happens when you remove inaudible data), its only when it degrades the audio enough to be heard that it becomes an issue.



To say that I don't understand how codecs work, and that they merely pack the data more efficiently seems a little defensive rather than anything based on fact. MLP for DVD-A is a way to pack data more efficiently without currupting it. DD, and DTS are not..

I think you are misusing the word defensive. I am not defending anything. I am stating facts, and only facts. Now let's look at the facts; DD and Dts(to a lesser degree) does remove noise, redundant and inaudible signals, stuff we wouldn't hear anyway. Dts encodes the data that's left at 20bit resolution, DD and about 18bits. So I am correct in saying it more effeciently packs data. And by the way, it doesn't "corrupt" the data as you insinuate. MLP doesn't even belong in this discussion because it is not a codec, you can't "play" anything on MLP.


By just pointing at how DD and DTS go from stereo to mono at certain freq ranges proves that the original input sound has been manipulated to something else. I need not point to another fact to make that point. But I'm not saying that it means it will sound horrible, just that it goes against what everyone here (I think) is about, being an audiophile. ..

Dts does not at any stage go to mono. It is a true discrete system. Also, it might be a stretch to say encoding audio into DD and Dts is manipulating the signal into something else. You room acoustics do it, do you stay away from it?
The same notes are being played, encoding doesn't alter pitch or time, or change any performance aspect whatsoever. All they do is remove what you already cannot hear. If you cannot hear it, then the process is benign. What degradation that does occur(mostly with DD or Dts half bit rate to a lesser degree) is minimal, and room acoustics will probably mask that. I fear room acoustics more than I fear what is coming from the output buffer of a Dts decoder.



Not trying to start a flame war here, just my honest feelings.

No flame war here, feelings are good. Facts are better.


As for the 192K stereo rate for dolby, I believe you are incorrect to double the number.

No, I am very correct. Since I own a professional encoder and decoder from both DD and Dts, I am well aware of all bit rates supported by both formats


When the data rate is displayed at say 348 that's for all six channels (5.1) when that rate is displayed for stereo, (192) that should be the total, or 96k per channel, isn't that correct? I could be wrong but that would seem to be right.

You are wrong. When the data rate says 448kbps for DD, that is for six channels(studios do not encode film soundtracks at 384kbps anymore) For stereo, its 192kbps for each channel for a combined total of 384kbps. DD is scalable from 32kbps(for voice only) to 640kbps(no current decoder supports this data rate)


As for market penetration, DTS shouldn't even be in the argument. There are more DVD-A and SACDs floating around out there than DTS music CDs.

Wrong again. There are over 200 million Dts decoders out there to support not only movies with Dts encoded soundtracks, but those DVD-A, music video's and CD that contain Dts tracks. Last time I checked, there were only about 500,000+ DVD players sold world wide that are capable of playing either SACD or DVD-A disc. That makes it still a niche product. The software end does not drive the hardware, the hardware drives the software.

[sfunches]

i know this is an older post,but ive just got a chance to read it. in each response i have heard nothing of how the music was recorded in the frist place.each recording is recorded differently some use aad, ada ,etc. some of my favorites were in ddd and add. telark make some of the finest recordings around even their sacds have a better sound better than most.george duke records his cds in ddd and it is spetacular. stealy dan(donald f) often uses aad or add,which makes his recordings so warm and detailed. cassandra wilson recoreded a cd called (full moon daughter) that will test what you have, it is a excellent recordng. so sacd,dvda,dts and dd doesnt mean anything if the original is not a good recording. oh, if you have heard the fleetwood macs greatest hits, the sound is very warm and detailed because it derived from an excellent analog recording transfered to digital,but still staying faithful to the analog recording.

[Francois1968]

Just the other day I compared some DVD stereo recordings with the good old CD's and found the CD sounded way better. I played the DVD's on the Arcam DV88+ en the CD's on the Arcam CD192, both components were connected analogue.
DVD's played in DTS were louder than the CD playback, but not better.
In my opinion DVD and DTS is not really an improvement. In terms of a pure audio sound I prefere CD's. :)