digital eq/room correction trade-offs


I am very confused about digital room correction.

For many years, it seemed the common wisdom was to have as clean a signal path as possible, with as little processing and as few conversions as possible: use a high quality DAC to get the signal to analog and then a pure pre-amp/amp to speakers.

But it now seems that many would argue that the benefits of digital eq are such that even an extra analog-digital-analog step is worth it for the benefits of digital room eq.

So, for example, I enjoy listening to CDs and SACDs using my Bel Canto PL-1A. I go analog out to my pre-amp. Is it worth it to contemplate the extra step of analog to digital for room EQ and then back to analog to the pre? I find it hard to believe that any benefits of the room EQ won't be substantially offset by the additional conversions.

Your thoughts most appreciated. Let's assume for the sake of this discussion that my room is imperfect but not horribly so (which I think is accurate).
dgaylin
re:
LYNGDORF: Goal is to preserve the speaker's frequency response & remove only room effects. You measure the drivers up close and then at the listening position. The software filters the difference between predicted (anechoic?) response at the listening position and actual measured response.

This explanation came from the folks at Lyngdorf.
Martykl (Reviews | Threads | Answers)

comment:
This LYNGDORF plan makes no sense to me. So please pardon this rave.
1. Speakers, especially multi driver units, have complex output, phase, etc. patterns. How can these be separated from room effects?
2. Designers assume that their products will be used in rooms (not anechoic) and design accordingly.
3. Where exactly does one measure a multi driver system to get an "accurate" measurement?
4. The LYNGDORF amps appear to be analog input only. Why ad an extra set of AD/DA plus interconnect?
5. Finally, why would anyone want to preserve a speaker's frequency response? No passive analog crossover and driver system can get within shouting distance of digital correction. Look at the specs claimed (if not always achieved) and they are +- 2DB. Anybody should be able to hear a bunch of four DB variations.
Here are some opinions from people that have ACTUAL EXPERIENCE with the TacT RCS system. I've only included the RCS 2.0, as you've not mentioned any desire to actively bi-amp. Note the dates, and that many improvements have been made by Boz since these were published.(http://www.stereophile.com/amplificationreviews/437/index2.html) (http://www.audioreview.com/mfr/tact/others/rcs-room-correction-system/PRD_118079_1590crx.aspx) (http://stereotimes.com/acc110299.shtml)
As a corollary to what I said above - part of the reason people change gear so much and like re-mixes/re-masters by clever musically trained sound engineers is that they get a different accent on the musical experience - it changes where emphasis was made. The room does this too. In a sense, the new equipment or the same equipment in a new room creates a new version of what is heard with sometimes hugely wrong (but pleasing to some listeners) differences in emphasis.

Nothing can completely control all this - but measurements and careful attention to gear selection can get you, on average, much closer to what was intended or heard by the persons producing your recorded music. Whether this is worthwhile to you or not is debatable as there can be pleasure in creating new interesting sounds or changing emphasis to make old sound new (even if it isn;t close to what was originally produced)
I'd suggest to

Step 1: find a good mains positioning and listening position so that you can run mains full range and untouched - the straight wire and gain approach.

Step 2: find a good spot for your subwoofer that gives you again the least suckouts in room response when combined with your main speakers - run the sub up to 80 or 90 Hz.

It is important to worry about the suckouts because ultimately you don't mind the bumps so much as you can EQ these down. Whilst the suckouts are gonners and may create "holes" in what you hear - a bass note nearly disappears for example.

Step 3: Apply precise narrowband notch filtering (with any device of your choice) to the signal going to the subwoofer.

If you have paid careful attention to the 100 to 300 Hz range when setting up speakers and sub position then your notch filters on the sub (20 to 90 Hz)should get you pretty close to flat.

Although this process can take days it is worth doing.

=> Be psychologically prepared to be slightly disappointed when your previoulsy favorite demo track with that absolutely awesome devastating pounding earthmoving bass note - that jumps out at you and your neighbours every second or third bar - suddenly sounds tamed and controlled and musical (instead of a gong show).

==> be psychologically preapred to hear more ambience cues and more details in the lower midrange that you did not hear before.

====> be psychologically prepared to discover that you can easily follow bass player notes cleanly, evenly and clearly

FWIW: A precise sound is NOT as impressive in terms of "sound" unless you retrain yourself to focus on the musical details (a musician's careful accenting on particular beats in the bar) rather than the odd musically unrelated kaboom emphasis as a the note happens to hit a room mode.
I'll offer a slightly different take.

Room effects are almost always most pernicious below app 150hz. Room correction in this range - IMHO - will far outweigh any benefits you get from maintaining a purist signal path. As you go upward in frequency, different rooms will make the cost/benefit equation of DRC vary enormously.

My solution is a Velodyne SMS-1 sub controller (room analysis + PEq below 200hz - coupled with a versatile x-over) and a pair of subs. You can keep the SMS out of the signal path above the x-over point, if you so choose. In my book, this as close to a "have your cake and eat it, too" approach as you are likely to find.

Caveat: In some rooms full range DRC may well be worthwhile.

Also, my understanding of the TACT/Lyngdorf approaches is.

TACT: Goal is flat response at the mic position (i.e. corrects for anomalies in the speaker frequency response and room induced frequency effects.

LYNGDORF: Goal is to preserve the speaker's frequency response & remove only room effects. You measure the drivers up close and then at the listening position. The software filters the difference between predicted (anechoic?) response at the listening position and actual measured response.

This explanation came from the folks at Lyngdorf.
On the Tact, is the volume control analog or digital? Either way I guess it's good as your feedback from it as a preamp is very good.

If it's on the analog side (which is usually recommended) and I wanted to use an external DAC, then it would need to be one with volume control so to drive an amp directly. Correct?

Thank you!
TacT versus Lyngdorf. Good question, especially as they were recently partners in the same company (Tact).
My tech says that Tact (Baz) tries to accomplish more correction parameters than Lyngdorf with the result that the Tact system is more difficult to understand and apply. I would love to hear the specifics from someone more knowledgeable than myself.
My living room is far from ideal for good sound reinforcement, as there is lots of glass, stone chimney, wood floors, etc., and I will not 'decorate' the room with lots of audio panels. I decided about a year go to try the Tact 2.2XP; it completely transformed the sound. No longer is there boomy bass, shrill treble and depressed midrange. I should mention that it takes a lot of time and patience to understand the unit, it's software, and much tweaking until it's 'right'; it certainly was not plug-n-play for me! However, it was certainly worth it in the end. I also found that using a good external DAC (I use a Lector Tube DAC) makes a substantial difference over Tact's internal one (for the main speakers; it's fine for subwoofers).
Lewinskih01
To answer your question, I have my Tact 2.2x preamp (w/ Maui Mods) connected to two Tact 2150 amps (one with full Maui Mods). The 2150 is actually a DAC that swings enough voltage to power a speaker -- hence, it acts as an amp.

I'm no longer using the internal DAC in the 2.2x. It's pretty good, but the 2150 is better.. One of the 2150s drives my main speakers, and the other drives the corner subs using the digital crossover in the 2.2x. So, no analog IC cables at all.
Thanks again everyone! any thoughts on TacT versus Lyngdorf versus some of the RCS systems built in on the newer pre/pros (Anthem D2, Integra 9.9, Classe)?
"I'm not convinced that I wouldn't just be winging it with these new systems. Not to mention the problems introduced with additional jitter as the signal gets switched back and forth from digital to analog to digital to analog."
I wish I could say that the automatic equalization (Tact) was the end of it, but it is rather the beginning of several rounds of trial and error, probably over several months. However, what's new! It's a hobby after all, and we are in it because we can hear and appreciate changes that others ignore. The good news is that one can hear the improvements (and errors) immediately and at no additional cost.
As to the jitter. No additional jitter is added in the digital realm. I do have the balanced digital signal running to an internal DA in the Tact, but I had a Wadia DA originally and both work well.
Finally, the analog source AD/DA conversions. I use a SOTA table with a Souther arm, an AT magnetic cart and a CJ EF1 preamp. I run the TT to the CJ and the to the Tact. I had a Quicksilver tube full function preamp with Mullards etc. and compared the pure analog with the AD/DA. To my ear the gains of the Tact system far outweighed the "loss" of the AD/DA conversion.
My experience and evaluation is much like Richard's.
My BAT VK-D5 has a tubed output stage(fully differential) that I've stuffed with six pre '68, NOS Siemens CCa's(wonderful, lifelike reproduction). It's digital out is a BNC connector. I've not tried the BNC to my TacT for lack of a suitable interconnect. I'm not disposed to spend the kind of money for a test cable to equal the Kimber KS-1130's I'm using(for a valid comparison). Yes- I am using the TacT as a preamp and active bi-amp crossover, feeding analog signal to my amps. That's the intended purpose of the RCS 2.2X with DACs at all the outputs. TacT does offer a unit without bi-amp capability(http://www.tactlab.com/Products/RCS20/index.html) That will still do time domain correction, parametric EQ, etc. As mentioned by Richards: Experiencing it in your own system is the only way to gauge your benefit. I do believe you'd find it enlightening. If you were able to audition one with a MauiMod power supply, you might find it a revelation. One might call me a VERY satisfied customer of both companies. =8^)
I'm also interested in this.

Richards/Rodman: how do you have your TacTs connected? Are you taking a digital signal straight into the TacT and using its DAC (effectively using it as a preamp too), and from the TacT in analog to the amp? Or using an external DAC after the Tact?

I measured my room's response and seems to be rather good, and this has me wondering how big a gain I might get from RCS.

Thanks!
Well, I asked, I should be prepared to suffer the consequences! Thanks to all for the info and suggestions. Now I guess I get to go on one of these escapades that we all do in this hobby to try out the new thing!
I was a straight-wire-with-gain guy(using a modded Dahlquist DQLP-1 to bi-amp my system) for 25 years. My last attenuator was a Placette Passive Linestage. I eschewed any signal manipulation, analog or digital, until reading some reviews about the TacT RCS products/algorhithms. If the(pickier than I) live-music-listeners at TAS thought it was transparent, I figured I'd try one(the 2.2X). Now I couldn't live without it. Replacing it's power supply with the MauiMods FRED/Sanyo OSCON unit took everything(system-wide) to an even more dynamic, controlled and transparent level. A TacT RCS would replace your present pre, with fully differential and single-ended inputs/outputs. If your CDP has digital output: you could feed the pre directly from it's transport, eliminating the CDP's DAC. Unless you rode the "short bus" to school: set up and operation are a piece of cake. If you go TacT, there's no going back!
If you can get a home audition--hearing for yourself in your system is always the best way to decide. Every room and system (and listener) responds differently and priorities differ.

My experience with digital room correction (RCS) has been very positive. A friend brought over a Tact preamp and we set it up and it was as if a blanket had been removed from the speakers. I also though I had a good room, but measurements (especially in the bass) showed otherwise. My speakers have changed twice since then, but the RCS remains (Tact 2.2x and two 2150s).

Main benefits are in frequency response. The bass humps and suckouts that most of us have learned to live with are gone, which creates a subjective experience of opening up the mids. And time alignment helps add clarity, depth and realism to the soundstage. The only trade-off is a slight lack of warmth and bloom. These are very slight IMO, and insignificant compared to the advantages.

A year ago I thought maybe I was missing something with the likely recent advances in digital conversion, so I bought a $5K tube DAC and $4K amp to substitute for the Tact amp. Yes it was slightly more liquid, but less open and detailed so they were quickly sold (actually, I kept the amp for another system). There is a wonderful advantage to keeping the signal digital all the way to the speaker.

If analog or SACD is your main thing, the trade-offs with RCS might not be worth it -- hard to say. For me it wasn't and I sold my SACD, as my CD with RCS sounded about as good overall as SACD without. Depends on your room and listening preferences.

The presets are quite good by themselves, but heck, we're audiophiles and we like to tweak. I've tweaked my correction curves to fit my listening priorities, and also to have alternate options for different recordings (ie, some have too much bass for my taste, etc.) I've also got power supply, etc. mods to the preamp and amp which further increase the fidelity and minimize the digital conversion artifacts, but I could have happily lived without these.

In the end, different strokes for different folks -- but try to experience it in your system. You may be surprised. Products from Tact and Lyngdorf and Behold, etc. have really upped the ante.
Thanks Kirkus and Samujohn. I agree that it's all signal processing and it's all how well you can use the tool. I guess my thinking is that these systems are supposed to self-configure, but my understanding is that they don't do that very well. So you end up tweaking them. And even with lots of experience running mixing boards in live sound situations, I'm not convinced that I wouldn't just be winging it with these new systems. Not to mention the problems introduced with additional jitter as the signal gets switched back and forth from digital to analog to digital to analog.

To Samujohn's point -- I get the idea of elimminating an analog stage and going with a digital pre, but that's going pretty far out on a limb before I even know what the technology can do, and given how many more options there are for nice analog pre's versus the handful of digital ones. The answer is probably a home audition of one of these systems to see how it works for me...
I think you're hearing so much about it right now mainly because there are a TON of new products out in the past few years, and this is a result of some new directions in the OEM (chip-level) DSP products. It used to be that fast DSPs in general were all Harvard-model processors with fairly limited memory, so the code written for them had to be clean and concise, and thus it wasn't easy to quickly configure them for different processing arrangements. Nowadays, there are many Von Neumann-model DSPs available, with much more memory, which makes it MUCH easier to write the code . . . and many of the IC manufacturers have developed "modular" programming environments with huge chunks of common applications pre-written.

So for end-user applications, especially in professional sound, there are likewise tons of multi-configurable products that are replacing dedicated boxes. For i.e. a typical simple church sound-reinforcement application, instead of an analog mixer, analog equalizer, analog (active) crossover, and analog peak-limiter(s), you can buy a product like a dbx "Driverack", Ashly "Protea", Rane "RPM-88", etc. that do it all in the digital domain in a single box, all configured however you want it, via a PC. And this is (in general) better, cheaper, and more flexible.

The consequent of this is that if you don't know how to properly tune a system with all the analog stuff . . . suddenly finding the digital equivalents with far more virtual "knobs" to tweek isn't going to make things any better. Some of the consumer products try to get around this by automating the process . . . but IMO very, very few of these approaches hold any promise.

In general, if you get past the whole "it's better because it's new because it's digital!" thing . . . it's all just signal processing. Digital and analog products are somewhat different tools, but the thing that makes it good or bad ALWAYS comes down to the quality of the tool, the skill of the person using it, and its suitability for the intended application.
If you listen to CD, digital is a given, as is the processing of same prior to conversion to analog. Digital equalization degrades nothing and is of considerable benefit in adapting speakers to a specific environment. The straight wire with gain folks had the upper hand as long as equalization was done in analog, but this doctrine no longer makes any sense for CD. For a real improvement, get a digital preamp and skip an analog stage instead.