The official "Theta" thread - Page 79

I understand your very interesting points - thanks. Some more questions please:

1. Say the studio that masters the film does not use room correction, and I do. Were I to use say a tilt up bass curve, wouldn't I be listening to something that the studio engineer does not intend?

2. I assume all room correction is done in digital domain; I wonder if Dirac Live processing is done before the conversion to analog by Theta DA processing? It would mean one less D-A conversion step, much cleaner? Otherwise what's the point of including it in the Theta SSP; might as well get a stand alone Dirac?

3. Since we have talked about correcting for the room, and correcting for the mic. Would the future brings a unit that takes into consideration the owner's hearing test - his hearing frequency curve - as well?
As we get older, there is high frequency loss, so perhaps the tilt up bass response is not necessary for older folks? So many interesting issues to think about!

The recording engineer likely mixed the movie in a dubbing stage that seated over 100 people, had vastly different reverberant characteristics than your comparitively tiny room, use pro gear instead of your consumer electronics, used commercial horns rather than your audiophle speakers, had long arrays of surrounds instead of your 2 or 4 surrounds, had more amp power per channel than you likely have in your entire system, had at least a dozen subwoofers instead of one or two, and he was listening maybe 50 feet from the speakers instead of in the nearfield, and listening at reference level instead of a comfortable volume for you. Nothing, absolutely nothing, in your home system is what the mixer intended. So don't worry about tilting up the bass a little. Won't even be noticeable compared to ALL the other things that are different.

Your audio system exists for one sole reason: to give you pleasure. Dial in your system so that it sounds good to you. Once you get comfortable with the Dirac software, you may well be editing the target curve to your personal preference. Doesn't matter how smart the folks at Dirac are, they're not psychic, so they don't know what your tastes are like. You bought speakers because you liked how they sounded. Down the road, don't hesitate to adjust Dirac's results for the same reason.
Correct; unless you're using an outboard room correction box or outboard equalizer, everything inside your processor (including room correction) will happen while the signal is still in the digital domain. There's only one D-A conversion, and that happens after everything else is done and the signal is about to come out of the analogue connections. All receivers and pre-pros with room correction handle it this way.
That's possible, but I haven't seen interest for anything like that from the industry, since it's not something they can objectively measure (hearing tests rely on you verbally confirming what you can and cannot hear). Meanwhile, if you're getting older and starting to lose hearing in the high frequencies, there's always the treble knob. Again, don't think of it as messing with the signal, but simply compensating for your hearing deficiency.

Sanjay, I just now realize the significance of the word you used in your first post to me: "prefer." No where in your answers have you mentioned "accurate," or "reproduction of original event," the stuffs that audiophiles like me argue over and spend our hard earned $$$ trying to produce?

1. So the basic premise of the Floyd Toole paper is that people "prefer" this sound, so therefore we are correcting the sound to this preference?

2. All room correcting devices on the market are based on the above premise of Floyd Toole's paper?

3. You seem to like Dirac Live very much? I don't understand the significance of their web site's claims - I can't tell whether they are real break-through's or just marketing materials designed for newbies like me. In your opinion, is Dirac Live for some reason better than other competing technologies, like Trinnov or Audyssey? Better algorithm? Better curve? A significant break-through technique?

You should be able to calibrate a flat frequency response.

"No where in your answers have you mentioned "accurate,"

Sure he did, that is what a perceptually flat frequency response is.

" or "reproduction of original event," the stuffs that audiophiles like me argue over and spend our hard earned $$$ trying to produce? ""

Were you at the original event?

If not how do you know if you are reproducing it or not? It just boils down to what you consider a 'reproduction of the original event.' IOW... it just boils down to what you prefer.

There is nothing wrong with that.

Shawn

Come on .

Correct. Pick any CD on your shelf. Were you in the studio when this CD was mixed? If not, then how would you know what "accurate reproduction of the original event" is?
Correct. ALL the testing that Harman does is based on correlating measurements to listener preference. Whether they're testing loudspeakers or room correction or whatever.
No, each company has their own approach. As mentioned earlier, Audyssey uses a measured flat target curve whereas Dirac targets something closer to perceptually flat.
Not a question of like or dislike, since I've never heard it. I was trying to explain the graphs on their website in order to show you that the information there shouldn't be over your head. In doing so, I also wanted to point out some of the things Dirac was doing well. And in order to do that, I had to give you a basic understanding of listener preference.

So now you know what people like when it comes to frequency response (smooth and perceptually flat) and why (so they can hear all the sounds in the recording evenly from bass to treble). Nothing complicated there. You also know what people prefer when it comes to reflections (side walls yes, front & back walls not so much) and why (spaciousness & envelopment are pleasing, muddied imaging & soundstage are not). Again, nothing difficult to understand.
The graphs on Dirac's website and their white paper show that they have a good understanding of the things mentioned above, as explained in my original post. And that's definitely reassuring. So what difference does it make whether their approach is a unique "break-through" or whether other companies know the same basic things about human hearing and preference? Theta is getting Dirac, not Trinnov or Audyssey.

Do we need studies to conclude that the general public does not like flat bass? I certainly know it. About 6 months ago, one of my friends visited another to hear his high-end speakers. His conclusion was that my friend had wasted a lot of money on those speakers because they were weak in bass response. He invited the other friend to his home to hear his Cerwin Vegas. True story.

I do not like to make assumptions and would like to see the research. For the sake of discussion, I will make a few. My first assumption is the the research objective is to enable product design for the Harmon market. The audiophile market is a fringe market so I will assume the population from which the experimental group was selected was not selected for using any criteria other than basic demographic of potential market? I think you will find audiophile criteria much different from the population in general. and the group that buys the Casablanca different than owners of other processors.

Secondly, I would point out that there is a fundamental difference in philosophy between Theta Casablanca owners and say Lexicon owners. For many years the debate centered on the superiority of Logic 7. The basic premise seemed that if you understood the argument,how great Logic 7 was, you would agree? However, I think most Casablanca owners fully understood the concept of DSP modes and further ,they understood Logic 7 but had no interested in obtaining the technology. It seem the argument progressed further down the line to point , that you are just rationalizing because you can't have Logic 7. NO we understood fully and just didn't fracking want to play around with DSP modes.

Now we find ourselves in a similar situation of philosophical opposition. I have no problem that some want to manipulate the signal to their own personal preferences. Can we have some respect that some on the other side seek to as little as possible alter the recording engineers intentions?

Now where, speaking only for myself, there is common philosophical ground, is that by not addressing the room you are already altering the intentions of the recording engineer with the "distortions" the room produces. It is my belief that Casablanca owners will never be swayed by arguments of how much of a benefit DSP processing is because it allows one to alter the intentions of the artist to personal taste. To the purist, this is profane. Again before any of you take a single line out of a paragraph out of context as has been the history, I AM NOT AGAINST ROOM CORRECTION. However the intention should be to faithfully reproduce the information on the disk for many Casablanca owners. If you disagree fine. That's likely you why own a processor with many DSP modes while this group chose a processor with almost none.

The Dirac Live will allow both camps to further realize their goals. For the purist, room correction can help factor out the room. For the enthusiast who seeks to manipulate the signal to personal preference, they are free to do it. Dirac Live is a tool, and nothing more. I think is it is a valuable tool.

In the past the another argument has been that the room matters much more than the quality of the electronics. I submit both are important. My personal experience is that room correction has reach the point of maturity where should be considered. "Cake and eat it too," comes to mind.

I personally prefer the bass with boobs!!!@@

Seriously, my subs are all pretty flat where they're placed (the surround sub thanks to
a Velodyne SMS-1). Yet I find that for both movies and music it sounds just right when I set the sube 3 dB higher than flat in terms of being level matched with main speakers.

Been on the road and just getting back to setting up my new family member.

Anyway, I assumed the regenerator process stabilized the output voltage. If not, then perhaps a J-Type UPS/Filter/Conditioner (like the APC AV J10BLK) would be a better choice for the Thetas than either the Panamax or the PS Audios whether up or downstream a straight UPS, such as the APC you suggested. Thoughts?

I thought with the Theta HDMI solution, a lot of folks would be selling there SixShooters on EBay or AudioGon, yet I've only have seen 2. And lost out on both. Any idea what a fair offer would be?

Here is the most recent one I've seen (2009), where Sean Olive did a room correction comparison that looked for correlation between objective measurements and subjective preferences:

The Subjective and Objective Evaluation of Room Correction Products

In particular, compare slide 24 (measured response) to slide 25 (perceived response). Notice the two room corrections that scored on top had a noticeable downward tilt when measured, but their perceived response was flat. Also notice that the top measurement shows a relatively smooth response (no major peaks and dips). You can see both these qualities in the Before/After graph on the Dirac page you linked to.

Yes, it does indeed. But how long the output continues after the AC drops is not specified by PSA.


If it is primarily your aim to solve the power glitches caused by the turn on surge of the amp, then any adequately spec'd APC will do the job. The J-type looks the business, though.

I have not followed the whole power conditioning discussion, but if your issues are related to a current surge / AC drop when powering on heavy duty poweramps, be aware that the P5 and P10 have and an "in rush" protection cycle on their high current outlet. This basically phases in to full voltage in a few seconds, avoiding the spike. I use it myself, having been advised by my technician my amps draw an inordinate amount of current upon powering up, potentially causing all sort of problems (some of which have actually materialized).

That it does. The question is whether a unit such as the APC AV J10BLK can be relied upon as a stand-alone power solution, or are the UPS and conditioning duties better split between two appliances?

Since a) UPS is a known need and APC is a proven solution, and b) power conditioning (noise reduction), or the lack thereof, has not manifested itself as a factor in your system's performance, then yes, the APC is sufficient as a complete solution.

You mentioned having access to some APC units. Why not try one for a few days?

Indeed. I plugged an APC BR1500G into the Panamax and plugged both the Compli-Blu and CBIIIHD into the APC. Powering up the Perreaux in this configuration created no noticeable problems with the Theta twins, both of which appear to be performing just fine. Mission accomplished, or so it would seem.

Of course, I can't speak to any performance degradation without an extended A/B session. I'm still deciding whether or not to lose my analog cables and go strictly HDMI for SACD, so I'm trying to minimize the variable spread.

This particular APC, while a nice little unit for computer equipment, is not one normally recommended for HT duties. That's why I wonder whether anyone in Thetaville has had any experience with an integrated UPS/Conditioner the likes of the J10. Everyone raves about their PSA's, but it's rare to read from anyone wowed by their HT UPS.

My power amps are all plugged either into the wall or into PSA Ultimates, so we're talking about my always in flux front end only: currently the Theta twins, my cable DVR, my Polk XM Receiver, and my Wadia I171i.

I was considering picking up a Power Plant Premier to replace the Panamax, but now I'm wondering whether the J10 might be a better choice.

Just switched from AMX landmark control for CB3HD to Crestron, CB3HD now lock up at startup with CB3HD lite on display and locked. Intergrator using CB3 Crestron Module. CB3HD is controlled vis RS232

Unit also started locking up more frequently on startup prior to this and required re-boot from back of unit, even before Crestron upgrade.

I believe the output from the J-models APC is an approximation of a Sine Wave. If you went APC I would go with the S models. That being said I am cycling out my S models to family members and replacing them with the new Richard Gray UPS22 models. They are better spec'd and have better output waveforms vs. APC. All of my APCs failed for faulty circuitry within 1-2 years. Some are bad out the gate. Check all outlets if you get one.

Perhaps it is what Les discovered.

I use Crestron with my CB3HD - my CB3HD was amongst the first five upgraded. I got the installer to install a hard switch in my middle Atlantic rack with which I can cold boot my CB3 HD without pulling out the rack anticipating lockups.
I had Zero lock ups on my CB3HD since i got the upgrade over the past months usage.

It is intriguing though I have some reservations about the study. Does the "body of work," support the same conclusions about preference? As long as Dirac live can be set-flat, and I am told it can, I don't see a problem. I am a little unusual in that I married into a family of Jazz musicians. My reference may be a bit difference and therefore my preference but that remains to be seen. Funny thing is that one of my wife's uncles, well known local jazz musician, owned a Sony built recording studio. He was anything but a "purist." I remember "sheepishly" inviting him over to hear my set-up after we first married. He loved it and the song. However, he grinned and told me this stuff I like was not properly mastered. I had Mobile Fidelity Sound Labs recordings so could not fathom his comments. Later I pressed him and he just said recordings had to be mastered so that they sound good on all electronics not just high-end high powered stuff like I was using. My amp at the time produced 1000 watts into 4ohms. I dropped it but he later elaborated. What he meant was that my recordings were not compressed like is typical in popular music. I believe I loosed my purist position some when I realized that it was likely that many of the recording that I enjoyed were not produced to be a "window on the source," but rather to satisfy the preference of the masses.

This has held the test of time. I've only had a soft lock or two in the intervening period. Working theory is the IIIHD does't like waking up to a non-functional source but that's been inconsistent enough to go unproven. My III was similar.

Note that it is only between different commands that I found the pause needed. Having the same pause in between something like volume up commands is more than clunky. This made implementing the pause a little of a hassle as it has to be done in each and every macro that addresses the IIIHD instead of the driver level.

Must make family get togethers interesting. Love jazz.

I've seen that come up more than once in artist interviews. It's got to be mixed so it sounds good in a car, it's got to sound loud etc. And the systems they own more often than not made a PC with a couple plastic speakers look upscale.

Sounds good in a car means compression, though not to the lifeless point most radio stations go. If you keep the original dynamic range you'll be riding the volume to hear the quiet parts and keep you're ears from bleeding on the loud ones.

Aside from the Harman study, there hasn't been much research comparing various room correction systems in order to discover which aspects listeners prefer. Jan Pedersen of Lyngdorf did some research trying to find out what listeners consider "natural timbre", in order to help them design the default target curve for their RoomPerfect system:

Fully Automatic Loudspeaker-Room Adaptation

You can see a couple of examples on page 4. One page earlier, Pedersen explains: Again, notice the similarity to Dirac's target curve, as well as an understanding that some reflections have a positive effect (like in the Dirac white paper).

What Pedersen is saying is that if you take a loudspeaker that measures ruler flat in an anechoic chamber, and plop it in a typical room, its response will have a downward tilt. Why? Because the high frequencies will beam like a flashlight towards the listening position (you'll hear mostly direct sound, without much additional room reflections), the mid frequencies will fan out (lots of reflections added to the direct sound), and the low frequencies will be omni-directional (you're hearing more reflections than direct sound).

So the lower in frequency you go, the louder it gets, as more and more reflections are added to the direct sound. This is what tells our hearing that we're listening in a room. Take that away, by flattening the response, and it no longer sounds natural. Our human hearing will let us know reflexively that something is wrong. Which is why RoomPerfect preserves room gain rather than EQing it away.

It's no wonder that the Lyngdorf room correction box scored well in the Harman comparison. I'm guessing a product that used Dirac would have done similarly well. I don't know what you would consider the "body of work", but at least there is concensus between three companies (Dirac, Lyngdorf, Harman). Bottom line: ATI made a good choice of room correction for the CB3HD (and the B&K stand-alone EQ box).

Sanjay,

Don't forget the DTS room correction paper, too, which came to the same conclusion. JJ's explanation in there was that the ear's integration windows --- the time period over which the ear thinks of the arrival of sounds as one thing --- get shorter as frequency rises (see below for a slightly less simplified explanation). That means fewer reflections affect the perceived frequency response as frequency rises.

But a measurement with a mic and a regular measurement system has equal time windows across frequency and will overestimate high frequency contribution, and be tilted up. That's why the flat measured response needs to be tilted down for perceived flatness.

More complex explanation: the bandwidths of the ear's critical bands approximately rise as frequency gets higher. This means that they cover more Hertz as frequency gets higher. What's long in frequency domain is short in the time domain. That means the impulse response of the ear gets shorter for higher frequency sounds, so the time window with which the ear reacts to higher frequency sounds gets shorter.

Sorry to hear of your problems with the S models. I have a pair of S-15s that are still flawless after 5 years of use.

Right, the target they used had a 10dB downward tilt from 20Hz to 20kHz. More concensus.

That is debatable. While it may not sound the way expect small rooms to sound, isn't the goal to simulate the sound of the original recording and recording site? Of course, this depends on how the source material was mixed and mastered. Nonetheless, for classical acoustic music, I do not want it to sound like the music is being performed in my room but rather as it sounded in the concert hall.

Kal

Actually, research supporting that curve dates many yrs back. I remember DBX had an EQ that also used a similar curve in its auto eq in the early 90s (probably dated to the late 80s), and it was based on research in psychoacoustics. Going back to the research at the NRC, speakers were recommended to attempt to measure flat, but it was never expected that they measure flat at the listening position since it would sound unnatural in the room (i.e. the speakers curve would have been req'd to be tilted upwards to give a "flat response" at the LP).