Room correction - what device works best?

Some clarifications to some of the comments in Gammajo's post just above, for the benefit of others who may read this thread and want to consider DEQX:

Seems intimidating to have to take the speakers (mine about 300 lbs each) into the yard....

While it is theoretically ideal to do this, it is not necessary. That was clearly stated by experienced DEQX users in the "Game-Changer" thread I linked to.

The goal of the DEQX speaker calibration (aka "speaker correction") process is to make the speakers time-coherent (and therefore also phase coherent) at all frequencies above a lower limit falling in the area of roughly 200 to 500 Hz (our hearing mechanisms being most sensitive to phase and timing issues at mid-range and treble frequencies).

If the speaker calibration process is performed outdoors, so that there are no room reflections to degrade the accuracy of the process, that lower limit of the range of frequencies for which the process is effective will be as low as possible, within that area of 200 Hz to 500 Hz or so.

If the speaker calibration process is performed with the speakers moved to the center of the room, as far away from reflective surfaces as possible, that lower limit will be a bit higher than if the process was performed outdoors.

If the speaker calibration process is performed with the speakers in their normal position, that lower limit will be somewhat higher still.

So which of those three alternatives is chosen will simply affect how low in frequency the benefits of the process will extend. The mid-range and treble will benefit regardless of which alternative is chosen.

...and then spend month dialing the thing in to get the best from it.

My understanding is that if the DEQXpert service is used, or the comparable service provided by Acoustic Frontiers, the speaker and room calibration processes can be accomplished in a matter of hours. Or at most perhaps two sessions of a few hours each.

In my own case, I don't plan to use those services. And working at my own deliberate pace, having no previous experience with the product, and with the time I can devote to it limited to some extent by other activities and by the need to not perturb my wife's activities by playing test tones and loud music through the speakers (I use headphones when I want to listen at those times), I envision taking perhaps a month or so to get it fully dialed in. But that's just me.

My speakers are designed for excellent time and phase response across a wide axis.

Regardless of what manufacturer literature may say about time coherence, time alignment, phase response, phase linearity, etc., if the speaker has a crossover and if the crossover is not first order (meaning 6 db/octave), the speaker is not time coherent. If the speaker has a crossover and is not made by Vandersteen, Thiel, Green Mountain Audio, and perhaps one or two others, it is highly probable that its crossover is not first order, and that the speaker is therefore not time coherent.

For background on the benefits of time coherence (which in turn automatically implies phase coherence), see the "Sloped Baffle" thread I linked to earlier.

The Lyndorf idea of not changing the character of the speaker and amps etc, assuming people selected them because they liked them, is also logical to me.

Of course, with DEQX one could choose to simply not do the speaker calibration process, and just use some or all of its other functions (room correction, equalization, DAC, preamp functionality in some models, USB interface in some models, etc). But one would be losing out on what with many speakers in many rooms may be its most important benefit, that is not provided by most competitive products.

Regards,
-- Al

05-08-15: Gammajo
Looking at room correction and all the threads I found seem old.

The "Is DEQX A Game-Changer" thread is essentially current, and is on-going. If you want to consider DEQX, and you haven't already seen that thread, I would consider it to be required reading.

Also, for background on loudspeaker time and phase coherence I would commend last year's "Sloped Baffle" thread to your attention. (Bob R., as you'll realize speaker time and phase coherence, as defined in this post, are unrelated to absolute phase (i.e., polarity), to which you were apparently referring in your post. And time coherence is something that only a very small minority of speakers inherently achieve, and is something DEQX attempts to correct via DSP prior to implementing room corrections).

05-08-15: Zd542
I do base my opinion on actual experience, and not just guessing. I have a Behringer 24/96 digital EQ that destroys the sound of my Wadia CD player. I also have a dbx analog EQ that ruins the sound of whatever preamp it gets used with. So, in my opinion, the OP's Ayon CD player is just to nice a player to be run through processing.

ZD, I would not extrapolate anything regarding DEQX or many of the other processors that have been mentioned from your experience with the Behringer or dbx units. As someone having considerable proficiency with computers, I'm sure you realize more than most that DSP and computer technologies have advanced by leaps and bounds since those devices were introduced. And although I'm usually among the first to caution people that in audio price and performance don't necessarily correlate, I think it says something that the flagship DEQX HDP-5 model costs well over 20 times as much as the Behringer.

I have an HDP-5 on order, BTW, which I will be receiving soon, and which will replace my preamp in addition to providing speaker and room correction functions. I ordered it from Nyal Mellor of AcousticFrontiers.com, who has a great deal of directly relevant expertise, offers a free webinar/walk through on the use of the associated software, as well as unlimited phone and email support. And both he and the DEQXpert service Roscoe mentioned can perform the entire correction process on a paid basis, via Skype. Nyal also provides 30-day return privileges, which I have no expectation of having to utilize. I will be chronicling my progress and findings in the "Game-Changer" thread.

Inputs to that thread from several A'gon members having both extensive DEQX experience and very high quality systems, including Psag who posted above, were among the major factors which have given me sufficient confidence in both the transparency and the effectiveness of DEQX to proceed down that path, given that a hardware solution best suits my particular circumstances. Kal's (Kr4's) review in Stereophile was also helpful.

Regards,
-- Al

Thanks very much, Lewinski. Yes, I've followed the posts and threads related to your project, and I'm happy to comment whenever I can contribute anything meaningful.

As you realize, the three independently configurable pairs of output channels provided by the HDP-4 and HDP-5 can support active biamping or triamping, but I have no plans to do that. In part because doing so would entail what to me would be desecration of the fine craftsmanship that went into the construction of my speakers; in part because the innards of my speakers are not readily accessible as practical matter; and in part because in the absence of any knowledge of the design of the speaker's internal crossover I would by no means rule out the possibility that the sonic results would end up being a downgrade rather than an upgrade. And of course there would be a lot of expense for additional amplification.

Based in part on listening comparisons with my Stax electrostatic headphones I'm pretty much convinced that the weak link in my system is presently some combination of room effects (which I can't address with conventional room treatments since it is my living room), and speaker time incoherence. So I'm just looking for some modest improvement in those areas, and in the process perhaps I'll also benefit from substitution of the DEQX for my preamp, and from utilization of the DEQX's DAC function in place of the one that's internal to my CDP.

Yes, as you indicated these are fun projects, and also highly educational. Those are the main reasons I am intending to do it all myself rather than utilizing the DEQXpert or equivalent services. And if it takes a month or even two, I'm fine with that.

Best regards, and continued good luck with your project.
-- Al

Lewinski, +1 to Roscoe's response. Some time ago I had read through the 143 page calibration manual and the 36 page user manual for the HDP-4, and I found many indications of various kinds that are consistent with it being able to perform timing corrections on single-amp'd speakers having their internal crossovers in place, and nothing that would indicate to the contrary.

Also, this post by member Drewan77, who as you've seen is extremely knowledgeable and experienced with DEQX, appears to have provided further confirmation with respect to my specific application.

Best regards,
-- Al

ZD, with respect to frequency response equalizations that may be introduced in the speaker calibration process the answer to your good question is certainly yes. I believe the same holds true for equalizations that are introduced in the room correction process, or subsequently by the user, but I can't say that with certainty. Perhaps one of the experienced DEQX users will chime in and confirm that.

Best regards,
-- Al

Michael (Swampwalker), here are my thoughts in response to your post:

Regarding the last question, about flat frequency response, see these paragraphs from the DEQX FAQ. Makes sense to me, and note that DEQX provides a goodly amount of flexibility in that regard, for both the speaker and the room calibration/correction processes. Also, keep in mind that a large number of user-created equalization settings can be set up in the DEQX, and selected among via the remote control.

Regarding the speakers, my instinct would be that the sonics of the 5 inch mains you are presently driving full-range would be likely to benefit from having deep bass frequencies kept out of them. Provided that the plate amps on the subs can accept line-level inputs, you could accomplish that by taking advantage of the DEQX subwoofer integration features.

If you were to do that, and if you were to place the DEQX ahead of the preamp in the signal path as you described, to keep the volume levels of the mains and the subs in sync you would use the DEQX's volume control to adjust volume, with the preamp's volume control set to a fixed position. DEQX volume can be adjusted via both the remote and the front panel.

Regarding item 4, a point to be aware of is that John Atkinson's measurements of the DEQX PreMate that were provided with Kal's review in Stereophile indicated that the DEQX downsamples 192 kHz data to 96 kHz, and 176.4 kHz data to 88.2 kHz. In the comments section following the review JA said the following, which makes sense to me:

Running powerful DSP at 4Fs sample rates is very consuming of resources, so this compromise is not uncommon. It is likely that the benefits of the DSP correction outweigh the potential drop in sound quality due to the downsampling.

(4Fs presumably refers to a sample rate of 4 times the redbook CD rate. 4 x 44.1 kHz = 176.4 kHz)

Best,
-- Al

05-11-15: Unsound
Al, I agree that the advantages the DEQX bring to the table might (probably) outweigh the down sampling issue, but in that sub $100 DVD players can decode (yes I realize that's all) the native higher sampling rates, it is still somewhat disconcerting that an item that can cost close to $5,000 can't do it all in the native higher sampling rates.

Hi Unsound,

Your reaction is natural and understandable, and as you indicated you recognize that the DEQX processing is much more extensive than what a DVD player has to do. But I think that my use of "much more extensive" understates it considerably. What I envision is that the mathematical computations that are involved in the digital signal processing the DEQX has to perform on the fly, fast enough to keep up with the music data, are HUMONGOUS. My understanding of it is that in real time it has to divide the spectrum into thousands of frequency segments, mathematically determine the contents of each of those segments by converting the series of data samples from the time domain to the frequency domain (that conversion involving a huge amount of mathematics), mathematically adjust the delays and amplitudes of the contents of each of those frequency segments in accordance with the speaker and room calibrations that have been established, as well as in accordance with any additional equalizations that have been programmed, and then put everything back together and convert it back to the time domain (again, a huge amount of mathematics), while digitally adjusting the volume and then converting the data to analog. I think it would not be unfair to characterize the processing a DVD player has to do as not much more than a drop in the ocean compared to that.

I'd imagine that if 24/192 processing could have been implemented in the PreMate which JA measured in a reasonable manner without drastically complicating the design, and/or delaying release of the design considerably (perhaps by lessening the extent to which that design could draw upon their previous design work), and/or making the unit much more expensive, DEQX would have done so.

I don't know, btw, if the design of the more recent HDP-5 which I am getting processes 24/192 as 24/192 or downsamples it similarly. But I'm not concerned either way.

Best regards,
-- Al

Update: I received my new DEQX HDP-5 from Acoustic Frontiers a few days ago. See my post dated today in the ”Is DEQX A Game-Changer?” thread for a detailed update, which is all good at this early stage (I have not yet attempted any measurements or calibrations). Further updates will follow in that thread in the next few weeks.

Regards,
-- Al