I recently bought some DVD-A/SACD MC titles. In particular, I got Steely Dan's 'Two Against Nature'. I've always been a SD fan. I like the tunes, lyrics and, of course, the extremely high production quality of their albums.

Since the disc has lossy versions (DD and DTS, 5.1), I decided to see if I could find any evidence to support various things I've read (and some things I've had rather obnoxiously shoved down my throat) as to why many people prefer the DTS lossy version when given the choice (me being one of them).

I chose the first 1 minute of track 5 (Janie Runaway) for this 'test'. BTW, this track has the best sounding, though played in the 'less is more' style, studio electric bass tracks through a Fender Jazzbass (my guess) that I've ever heard.

I set my SPL meter on a tripod at the listening position and set it to C weight, slow. The player is set to all speakers 'large' and sub 'yes'. The 950's digital BM is set to all speakers 'small', sub to 'on' and all 3 crossovers are set to 80 Hz. I use the the 950's calibration menu for tones and with master volume set to '00', I calibrated the 5 sats and the redirected bass sub to 75 Db. I then calibrated the LFE sub by using a disc that I know contains an exact copy of the sats bass and leveled it to the redirected bass sub (I can switch them back and forth on-the-fly).

Since my low freq system, which consists of 2 subwoofers, 1 for redirected bass and the other for discrete LFE only playback, allows me to easily isolate any of the 5.1 channels for monitoring/testing, I found the following:

I began the test with the master volume set to -15.

I played the DD version first. With the SPL meter at the LP, I measured the LFE signal only and wrote down the reading for every note of the first four bars. This took 12 replays, so I was sure of the numbers being correct relative to the meter's limitations.

I then isolated each of the 5.1 speakers, 1 at a time and, with the SPL set 1 meter in front of the speaker, repeated the above scenario.

For each of the seven speakers, I switched to DTS and repeated the numbers game, so that the tripod would be exactly in the same spot for each format before I moved to the next speaker.

I then set the tripod back to the LP and raised the Master volume to -5 with all channels playing. I, finally, raised the volume to '00', reference level. Both times I wrote down the numbers using C weight, slow, C weight fast and C weight MAX, looking for both average SPL and transients.

The results are strange and this is why I'm posting here, so that any may chime in and offer any thoughts.

1. The 'BoB's ears' test. I listened to both versions of the song I tested. Hands down I like the DTS version better.

2. With the Master volume set at -15 Db:
a. The LFE was 4 Db louder with DTS.
b. The center channel was 2 Db louder with DTS.
c. The FL/FR and redirected bass sub were 1 Db louder with DTS.
d. The surrounds were identical with both.

3. With the master volume at -5 and again at '00', the volume was precisely, exactly the same in DD and DTS. All 7 channels (remember, I count the redirected bass subwoofer as the 7th 'channel' added to the 5.1) AND the overall soundfield. This volume setting range was important to me because it's the range I almost always listen at with all formats (which are exclusively multichannel).

I would describe the bass to be fuller-range and the vocals to have more depth with DTS. These are the 2 I focused on, bass guitar and vocals. I personally feel the added depth (though it's a subtle difference) may be because the reverb, delay, etc. effects are slightly more discernable.

I'm baffled as to why there is a difference in volume at lower Master volume, but absolutely no difference at higher volume.

Side note: Disc of the month...Al Dimeola's 'Flesh On Flesh'MC SACD. Excitingly different musical pieces. Good pianist, percussion and bassist. Great guitar work. I dig it.

Does the 950 apply any kind of loudness compensation? That could explain the differences in levels at lower volume.

Heck, it's also possible that DTS is the one applying a loudness compensation.

Jeff

I believe for music dvd's the dts track is encoded with the standard bitrate, not the half or whatever is normally put on regular video dvd's... so that could explain why it sounds better...

BoB,

Any way you can set up a double-blind, level-matched test? That would be really informative.

Jeff

Jeff,

I would definitely like to do that. I have some guests staying from out of town soon. None of them (ages 13-70, 6 males, 5 females) have any experience with or own a HT or MC audio system.

Can you please suggest the criteria? This would help alot.

I've done informal but blind tests with DTS and DD. The DTS was clearly identifiable. But I think it may have more to do with whatever engineering skills the DTS are bringing to the mastering, rather than any simplistic techno difference between the compression schemes.

More interesting, I think, would be a comparison between the non-lossy and the lossy tracks. In my comparisons, the differences are heartbreaking. The lossy technologies throw so much of the sound away.

I’m working and you guys always bring in all this intriguing stuff to think about, whereupon if I were to spend the time to look up the basic parameters would kill a workday.

But some questions, What are in the decoders themselves? What kind of detail in control are they capable of? If I assume (magic instant answers), that you try any pre-pro in this test, and get the same results.
Increase in SPL levels for selected channels at one dB settings, with a unity gain (?) of SPL at a higher dB output, something’s going on.

Decoders are propriatorial, and set up to do what they do, per each standard. Are decoders themselves capable of details of effect in the system to an extent that could be, - say a expander type setup in the DD decoder if theoretically the dynamic range possible in DD is a little less. (or is it a theoretically possible but argued not real life higher sample rate of DTS which supposedly gives it greater DR? (can’t remember) regardless, - if I understand expanders correctly at a lower gain level, this would create a reduction in dB at lower input with a more uniform gain level across the board. When lower input levels is received?
With something else going on in how DTS ‘compands’ the signal at given input level. Or since this is off the cuff without understanding some other decoder based: unity/gain (or) compression/expansion (or) saturation of digital bit processing speed (but dB level shouldn’t effect that – I think) control per signal input to output?

Anyway very interesting find, I knew there were SPL differences between channels in DD and DTS, but had never noticed that a higher volume setting would level those discrepancies. One of my arguments for the heck with all arguments surrounding DD/DTS, - was that, - if I like the higher SPL at a given dB setting that DTS starts out. Why am I hearing any differences between the decoders? If theoretically I want to level the playing field I should just be able to increase the volume when playing a DD track to equalize my bose ear pychoacoutical phenomenon of “louder is better” In other words I should be able to just turn up a DD recording louder to like it [/I]better[/I]. Your test results bring a new twist to what’s going on between the two. Nice documentation of your experience it doesn’t take a doublebindlevelmatch to record the SPL but I guess it does for your opinion of what you thought you heard given these SPL level readings at equal dB, (as regards preferences). Jeff your so hardcore as to all that, when it comes to preferences. Jeff how ..did you..ever pick a wife? Its such an unscientific messy uncontrolled Lust blinded, unleveled playing field.

Lossy Vs Lossless now that’s a difference I hope they really keep working towards improving. But then you have all those ‘buts’ to understand like string entropy. I must have more time…to grasp all this stuff!

Lena,

First, I'm a good dancer. Second, I own two tuxedos. And third, we're not married yet. [ ]

Jeff

And third, we're not married yet Does that mean the double blind came in at 50/50?

First, I'm a good dancer. Second, I own two tuxedos Lets see,…in the computer industry, you don’t happen to moonlight for the govt?…..For some reason (after that sentence) my minds eye will now always envision you as Arnold Schwarzenegger in "True Lies". Do you wear a wet suit enroute to your tuxedo funtions?

"Ever kill anybody?"

"Yes, but they were all baahd."