Speaker calibration, DTS, DD and Twister (long) (1 Viewer)

OK, it's Saturday morning and since I'm on call-out for work I ain't going anywhere. An ideal opportunity to tidy up all those bits of paperwork - and do a complete speaker level calibration of my system.
I try and get this done every couple of months. Not because anything really changes (unless speakers have been moved for whatever reason), but really just to confirm to myself that the system is singing as it should.
This time, rather than use just Avia, I calibrated using all the tools available to me. Avia, Video Essentials (DVD and LaserDisc) and the internal test tones of the amp, along with the DTS technology setup CD for the Millennium DTS decoder. I wanted to see what differences - if any - using all these test tones would make. I was really hoping that there would be no - or at least very little - discrepancy between them. What I found has left me unsure and really quite sceptical.
Rob's Stuff[*]Denon AVC-A1 THX certified amplifier[*]Millennium 2.4.6 DTS decoder[*]Kef Q65, Q55, 200C loudspeakers[*]Boston Acoustics VR2000 THX certified subwoofer[*]Panasonic A110 DVD player[*]Pioneer CLD-HF9G LaserDisc player[*]Sony KV-32WF1 32" 16:9 TV[*]IXOS, Chord Co. and SonicLink cabling
First of all, the Radio Shack SPL meter was set up in the prime listening position, pointing more or less straight upwards. It was positioned on the back of the central chair in the room in such a way that it could do it's thing hands-free and with me not in the way of the signal. The chair tilts back you park your backside, so the SPL was pretty much in the same position as your head (well, OK, my head).
The calibration level was 74db, simply because I was able to get the needle resting more accurately on this postion that at 75.
The mode used was plain AC-3 and External Input for the DTS decoder. No DSP or THX processing was applied. Also, the Dolby Digital Dialogue Normalization was turned OFF.
[*]Internal test tones
Starting on Front Left, the overal volume was raised until the SPL needle hit 74db dead-on. Overall volume was then left alone.
Results:
Left: 0db
Centre: +1db
Right: +2db
Rear Right: +1db
Rear Left: +1db
Sub: +2db
*NB: The +2db increase on the front right channel is attributable to the room acoustics.
[*]Avia
Same as internal tones (good so far!)
[*]Video Essentials DVD
Left: 0db
Centre: +1db
Right: +1db
Rear Right: +1db
Rear Left: +1db
Sub: 0db
Notice the front right had to be dropped by 1db and the Sub by 2db compared to the internal and Avia tones.
[*]Video Essentials LaserDisc
Left: 0db
Centre: +1db
Right: +1db
Rear Right: +1db
Rear Left: +1db
Sub: -11db
The VE LD obviously does not include the 10db alteration of the LFE channel as required by the Dolby Digital spec. Even so, it still appeared to be 1db higher than the DVD.
Now, the tones on Avia and Video Essentails do not sound the same to the ear. Those in VE have a softer, less harsh spectrum of frequencies - more 'pink', if you will. This may explain the 1db discrepancy in the front right, but the 2db difference on LFE left me completely confused.
Granted, the sub is probably not in an ideal position (between TV and front-right), but due to room limitations, there's not a fat lot I can do about it. As a result, I can never get the LFE test tone to 'sit' with any accuracy. It always likes to bounce that SPL needle around, making the measuring process less than satisfactory. Even so, 2db is a very audible level difference when listening to a film soundtrack.
Finally, the Millennium 2.4.6 DTS decoder was calibrated using the DTS Technology setup CD supplied with it. First of all though, the Milly's own volume level had to be matched to that of Dolby Digital. This isn't really that necessary when just getting the individual levels correct, but it makes cross Dolby/DTS comparisons a lot easier.
The only way I could really get this done was to play an Avia[ front left test tone at 70db dead-on, and then play the DTS front left tone at the same volume level on the amp, but with the Milly's own volume knob adjusted accordingly. The end result was that with the test tone, both Dolby Digital and DTS hit the exact same SPL level at the same given amp volume.
DTS results:
Left: 0db
Centre: +2db
Right: +2db
Rear Right: 0db
Rear Left: +1db
Sub: -3db
The results cannot be compared directly against the Dolby Digital levels because the Millennium has it's own surround, centre and LFE trim controls on the front panel. But, since the DD and DTS test tones now hit the same SPL level, they should be pretty accurate against each other.
So, what we have is Avia giving the same results as the internal tones and Video Essentials giving slight differences with front right and sub level. If anybody can provde any reasonable explanation, please do. I mean - which are you supposed to trust? Since Avia and the internals match, I'm tempted to go with those, but can anyone really say for sure?
But, I suppose the best test is run a soundtrack or two through the system. The ideal candidate was the new Twister DVD as it features an aggressive, envloping soundtrack and allows DTS and DD track selection on the fly (though the amp source still needs to be changed).
I suppose the object of the excercise was really to see if my intial impression of the Twister DTS sub level being lower than the DD was correct. I'm sorry to say, that even after a good hour of calibrating, the same impression was there.
An example: In the opening 1969 scene, we see a shot of young Jo being woken by the sound of the oncoming storm. Just as she turns her head, there is a very deep sounding thunderclap that can really be felt. On the DD track, all was well. One of the shelves holding my discs rattled in sympathy. The same scene in DTS did not have the same bass depth. It was almost as if the sub level had been smoothed out, resulting in a sound heard more than felt.
A little experimentation was required. I switched back and forth between the two formats, raising the DTS LFE level by 1db each time. It was not until I raised it by 3db that the same level of bass was felt.
So - is the Twister DTS track really 3db out or is there a setup problem in my system? I'm really not very happy with the sub calibration for either format, but since moving the woof isn't really an option, I dont see what I can do about it.
This has turned into a rambling post, but if anyone can shed any light on what is going on, or provide any tips, I'm all ears.
------------------

"It's all in the mix"

Rob,
First of all, one thing we don't know about AVIA and VE is at what level were the pink noise recorded at to begin with. Is it mixed to 0dB or something else? (Remember the old dB scale on your tape deck? Max signal for a digital recording is, in theory, 0dB. Higher for analog (tape saturation point)). Is it true wide band pink noise, or a "rough mix" pink noise. (Eq'ed at the 10's: 10hz, 20hz, 30hz, etc) Simple things like that could cause differences. But it is important to remember though, AVIA and VE are really designed to help the the home theater owner dial in their HT room as close as they can. Its not meant to get 100% perfect. The differences you are showing might be considered "close enough" by some.
Secondly, another example of the LFE differences between DTS and DD on the Twister DVD is the "Where's my truck?" scene. In the DD, the yellow pickup falls out of the sky and crashes into the road and you get this real good low level "thump". On the DTS track, its more of a whimpy "thunk". However, after comparing that scene a large number of times, I have a feeling that the DTS track is producing some harmonics that go much lower than my sub can produce. Listening to the sub nearfield, you can kind of hear those real low notes, but just barely. (What I dont know is if those harmonics are recorded quiet anyway, or if my sub is just saying "Sorry, no can do.")
Now, I feel that the new Twister DVD is a good example of the changes that happen to the DD mix when there is no 2.0 track provided. If you read the DVD File this week, they had a good editorial on the down conversion problems to 2.0 due to the fact that the LFE track is dropped during down converting. You have to add bass to the mains for it to stay in the 2.0 down conversion. So what we are doing is comparing a "compromised" track (DD mixed for down conversion) and a "pristine" track (DTS mixed for 5.1) Did they "pump up the bass" for the DD track downmixing? I don't know.
I'm afraid that until we can get a sound engineer who mixes up DVD's in here so we can ask them "How do you do it", these nagging questions will remain.
Steve Peterson
------------------

Whew! I'm tired from just reading about all that work.
I have no light to shed, but I thought of you when I came across this line in Widescreen Review's write-up on the new Twister (July/Aug. 2000 issue, p. 86):

a slightly fuller low-end said:

Quote:

Click to expand...

I, too, would love to hear from a sound engineer, but we'd probably need to hear from more than one. I'm sure that, as in any field, practitioners have different philosophies and use different techniques.
M.

The advantage Widescreen Review have is that they use some serious equipment, in environment more tightly controlled than most of us.
My main problem here is that I am unable to accurately get those LFE levels calibrated. The test tones on the DTS setup CD result in the SPL needle swinging quite erractically and the same applies to Avia and VE, thought not quite as bad.
If I could get a clean LFE measurement, life would be a lot easier. I've tried moving the sub a little (away and nearer to the rear wall), moving the SPL around the room, but the basic characteristic of the tone doesn't change.
------------------

"It's all in the mix"

Robert --

they use some serious equipment, in environment more tightly controlled than most of us

True, but for all the equipment lists in the front of the magazine, they never tell you what system was used to review which discs. You have to take their word that every system is a "reference" system.
M.

Click to expand...

Click to expand...

I hate to confuse you even further, Rob, but after running the DTS test tones on The Ultimate DVD, I found that I had to increase my DTS LFE +7db to match the same SPL output of the Dolby Digital LFE test tone (I normally use Avia).
I have a hard time believing that my DTS LFE has been under-represented by 7db these last few years.
I don't have any hard and fast answers but it definitely makes one wonder how much calibration software and hardware has to do with the differences we hear between the formats.
DJ

Probably a lot more than it should.
What we need is: A disc with calibration test tones in both Dolby Digital 5.1, Dolby Digital 2.0 (for Pro Logic, which often needs different levels) and DTS, all at the exact same level.
But anyway...
Tried a different approach yesterday evening (in between sorting out working problems ):
I set up the SPL meter again but this time attempted to tackle those trim levels on the Millennium decoder. My theory is this: If I can get the Milly's output levels to be 'flat' to each other (i.e. all hitting the same db level) I can then apply the individiual channel calibrations from the amp with more accuracy.
In other words - instead of making a separate set of level changes for DTS in the 'Externeal Input' mode, apply the exact same + or - values as I used for Dolby Digital. Since the channel levels comming from the Milly are the same as each other, it should, in theory give an accuate result.
Well, those trim levels are tricky to get right (small screwdriver job), but I finally got them pretty darn accurate. It seems to have helped the bass level on Twister as the DD and DTS impact is now closer. Still a little off, but not as much as before.
------------------

"It's all in the mix"

This JPG shows the thunderclap waking up little Jo(1969). in the top two traces. One is DTS,one is DD.For all intents, they are indentical. The only pair of traces shows another segment of bass I compared ...about the same here also. In fact,everything I checked showed the bass between the DTS and the DD audio to be very close if not nearly identical.

BTW---twister has some OK bass, but there's nothing jaw dropping or anything....it's mostly >30hz(like most THX DVDs).
TV
TV

Tom, that is excellent - thanks for your work once again.
So, perhaps my system really has been out of whack all this time - i.e. the DTS bass is at a lower level than it should be.
------------------

"It's all in the mix"

Well, I've gotta find something to do when it's quiet

Rob,
One of the advantages of owning a THX processor is a home THX calibrated internal test tone.
THX training instructed us to calibrate a Home THX processor using the internal test tone to achieve THX reference level.
You did not mention the position of the master volume knob during your calibration. It should be at "0" on a THX processor when calibrating each channel to 75 dB.
All THX processors have an input level control. Some are manual and some are automatic. The idea is to have all input sources the same level so when calibrated, all input sources play at the same level in your theater.
------------------
Irv Kelman
We are THX Certified.

Mmm...
Assuming I'm understanding you correctly, I'm not sure how that would work. At 0db, the internal test tones do not hit the SPL at 75db. It's probably around a good 10db higher. Factors such as speaker design, sensitivity, room size and acoustics will all affect that SPL level at the listening point.
Or am I missing something?
------------------

"It's all in the mix"

>>>Tom, that is excellent - thanks for your work once again.
So, perhaps my system really has been out of whack all this time - i.e. the DTS bass is at a lower level than it should be.

Kef Q65 fronts, 200C centre, Q55 rears - not small speakers at all, but I leave the setting on small because the VR2000 takes care of the bass better then they can.

But then again, the bass management is been done by the amp, so unless something really weird is going on, I'd expect it to be the same for both the internal DD and Milly DTS decoder.
What puzzles me though, is that a lot of people here have been commenting that the DD version sounds more bassy. Mmm...
------------------

"It's all in the mix"

Tom,
That plot of Frequency vs. SPL(dB) is interesting. I'm wondering how it was grabbed from the source signal. I assume it is only one INSTANT in time, right? Or is it an average over a certain time period?
-Kieran.
PS:
Anyone,
When I go to the store to buy this Twister DVD, how do I know I'm getting the one with DTS on it? Does it say somewhere? I found what looked like a new Twister DVD in CostCo yesterday for a good price, but I couldn't find the DTS logo anywhere.

On the back of the case, there is a very, very small DTS logo on the bottom line. The new disc has a catalogue number of 18321.