Dts On Classics (1 Viewer)

[Lew Crippen]

I do think by now that Robert (the one who started this thread) should have a good understanding of why many on this forum approach any DTS discussion with trepidation.

 

[DaViD Boulet]

:b

I admit that Robert makes some excellent points and I won't dispute the merits of double-blind testing. It's just that (coupled with being in an argumentative mood...having a bad day) I know that I heard what I heard, though I cannot expect anyone else to accept that for themselves.

 

[RobertR]

Well, I for one am glad that you kept the discussion civil, Dave, and didn't resort to the "I don't need no stinkin science" response I see too often in these debates. I'll still quibble with your volume methodology, but I sure can't argue with your desire for tecknical excellence.

 

[Lew Crippen]

I think that both you guys were quite civil and reasoned.

 

[Ed St. Clair]

Talk about apples & oranges!!!
I am address compression ratio's!
Your talking about sample rates (I believe, correct ME, if I error).
The number's may be off, DTS may be 3:1 lossy.
Which would make you confront DTS being FOUR times less lossy compression than DoubleD.
All the best.

 

[RobertR]

I'm not talking about sampling rates. I'm saying that one codec is more efficient than the other, therefore you can't say that codec A is "3 times better" or whatever (or even better at all) than a completely different codec.

 

[Ed St. Clair]

RobertR,
Both these quotes are yours.
And, you are talking about sampling rates (I believe, or you or anyone else please tell me if I am wrong in identifying these figures).

So, please, direct me too your point.

 

[DaViD Boulet]

Agreed Robert. Think of video codecs. WM9 can acheive higher-fidelity at less than one fourth the bit-rate that MPEG2 needs to just get by. In such a case, merely comparing data-reduction ratios to determine absolute image quality would be misleading.

Of course, all things being equal, the more bits that any given codec has to worth with, the theorectially higher-fidelity the encoded signal will become. Some codecs even reach a "golden" point where they become lossless. I know that DTS has proposed their codec running in such a mode for multi-channel high-res on HD-DVD.

Ed, he's not talking about sampling rates. That's an element of the uncompressed PCM source which is usually 48 kHz for audio on DVD. From there, the compression ratio is the bandwidth of the original uncompressed linear PCM bitstream compared to the bandwidth of the data after compression. For example, a codec that compresses the linear PCM datastream to 1 fifth its original bandwidth is giving you a 5:1 compression ratio (about what ATRAC on MD does, for example). Dolby Digital on DVD with 5.1 channels at 384 kbps operates at about a 10:1 or 12:1 data-reduction ratio. I think that DTS running at the max bitrate on DVD gives you about a 4:1 datareduction ratio. Given bit-space and bandwidth limitations on a format like DVD, the goal is to acheive the highest fidelity with the greatest compression/datareduction ratio possible.

In any case, bit resolution and sampling rate of the original signal haven't changed...just the fidelity of the waveform inside those paramaters.

-dave

 

[Ed St. Clair]

RobertR said;

 

[DaViD Boulet]

Ed, just tryin' to stir up trouble? Relax...

 

[Ed St. Clair]

When i did a comparison there was NO difference

This is a valid statement.
I can not tell the poster: Yes, you did!
However, if someone post DTS sounds better to them...
LOOK OUT!!!

I hope that we can accumulate some support on this post

Well...
How's it going so far?

 

[Ed St. Clair]

Please list the DD DVD's that you think sound as good, or better than the LPCM LD counterpart.

 

[Ed St. Clair]

The "754 k DTS" & "448 DD", are NOT sample rates, correct?
And 384 kbps is NOT a sample rate, correct?
What is it?
Thanks.
It's fun too go back & forth on these things.
However, it's even better when you LEARN something!!!

 

[RobertR]

It's a data rate--the number of bits per second.

 

[DaViD Boulet]

Exactly.

I'm sure Ed isn't the only one wondering what these "rates" are about.

When we use the term "sampling rate" it means the number of times per second that the analog-to-digital converter "sampled" the waveform to measure it. Think of it like making a movie of a waveform...how many "frames" per second you actually film will dictate how smooth the motion appears and what the fastest transitions are that you can record. Anything that moves faster than 24 frames per second wouldn't record well (it might stutter) and in fact movie directors are careful when they film motion to try to avoid such "judder" artifacts.

So we're supposed to be talking about audio. Back to sample rates for digital audio. Again, it's literally the number of "snap shots" per second that are taken to measure the waveform moment by moment. In the case of digital audio...you need at minimum two "up/down" points to describe a sine-wave. So we typically say that the highest frequency you can record is 1/2 of your total sample rate. In the case of 48 kHz, that would be about 22 kHz of frequency response you could record. In practice, it seems that higher sampling rates make things sound better even if you don't have ultra-high frequencies to record...my guess is that the more sample points allow for more natural variation in waveform reconstruction during d/a conversion since natural musical waveforms are very complex and not necessarily simple sine-like in shape.

Each "snap shot" per 48-thousandth of a second is measured. This becomes the "bit word" that describes the amplitude of the waveform at that moment in time. CDs (which are sampled at 44.1 kHz, BTW) store audio data with a resolution of 16-bit words. On paper, this describes a very wide "dynamic range" so lots of folks think that CD audio is basically as good as it needs to be. In reality, however, your ear can hear the difference between a 16-bit quantized (measured) waveform and one that's converted to digital at a higher resolution like 20 or 24 bits. That seems to be the point at which going any higher doesn't yield any improvement to the human ear. The higher number means that the amplitude of the waveform can be measured more accurately, which preserves more subtlety in the sound especially during quiet passages and musical "decays" etc., but I've found it also helps large-scale waveforms like loud bass notes (they just sound more "real" at higher resolution). This, BTW, is why DVD-audio allows for 24/192 resolution audio (24 bits resolution, sampled at 192 times per second).

Now, linear PCM like this takes up a *lot* of space to store and also a lot of bandwidth to transfer in real-time. DVD has limitations on both counts, so compression algorithms are brought in to try to "shrink" the space that that audio needs to use on the disc and also shrink the diameter of the "pipe" that it requires when being read off the disc in real-time during playback (bandwidth). When we say that the 5.1 DD soundtrack was encoded at 384 kbps what we're saying is that the compression algorithm took the linear PCM and "shrunk" it to fit into a space that could be read in real time using only 384 kilobits per second. I can't remember the exact number that 5.1 channels of 16/48 linear PCM audio would require, but it's about 10-12 times that rate. That would make it impossible to use with a "movie" because DVD has a max bandwith of about 10 megabits per second and that would take just about all of it just for audio (fine for audio CD, BTW) leaving just a few bits left for video. This is why we have compression to begin with...it's a compromise not only for raw storage space, but also so that everything can fit through the "pipe" of 10 mbps of DVD playback.

 

[Benjamin.D]

To return to the original intent of the first post...

Can anyone discern a difference between the Dolby and DTS 5.1 tracks on Airport in the Terminal Pack? This DVD was an odd duck as far as classic movies go in audio treatment.

Ben

 

[Adam Barratt]

David, as usual it is fascinating to watch your arguments. With such fancy footwork and constant shifting and weaving (unintentional, I sincerely hope; doing so deliberately would be very unsportsmanlike) you would make an excellent prizefighter, I'm sure.

Adam

 

[Ed St. Clair]

BUTT, 44.1 is NOT a sample rate.
Interesting!

 

[Ed St. Clair]

LPCM fan club, is aware of the limits too reproduce a 5.1.
We are hoping/wishing for mono & stereo soundtracks, in LPCM.

 

[DaViD Boulet]

Yes it is. The "sampling rate" for CD is "44.1 kHz" or "44.1 thousand times per second", whichever way you'd like to say it. It's literally how many times per second the analog waveform was "sampled" and measured when converted to digital.

that's what sampling rate means...it tells you how many times per second the analog waveform was measured when it was first converted to digital. Basically, it's the "temporal resolution" of the waveform.