DSD comparison to PCM.

Thanks for the info. The Marantz SA113S uses the TI/Burr-Brown DSD1792 delta-sigma converter, with specs here:
http://www.ti.com/lit/ds/symlink/dsd1792a.pdf

Going by the spec sheet, the Burr-Brown DSD1792 does not accept 128fs DSD, and is restricted to SACD and 64fs DSD-download playback. The distortion is 0.0004% at 44.1kHz, 0.0008% at 96kHz, and 0.0016% at 192kHz. The Resonessence Invicta, based on the ESS 9018, is slightly lower at less than 0.00032% (with no mention of frequency).

http://resonessencelabs.com/tech-specifications/
http://resonessencelabs.com/wp-content/uploads/2012/05/InvictaMeasNotes.pdf

Page 20 of the PDF has an interesting section mentioning how noise level in delta-sigma converters are related to the DC level of the signal. DC-dependent noise can increase as much as 20 db (!) with some delta-sigma converters; the relationship between DC level and noise depends on the selection of noise-shaping algorithm by the chip vendor.

The steady-state distortion specs don't tell us much about the sound. The Burr-Brown 1792 and ESS 9018 are both delta-sigma architectures, but with different noise-shaping algorithms, so we can expect them to sound different.

The noise-shaping algorithm (which is designed by the chip vendor) is responsible for a 40 to 60 dB improvement in converter performance. 40 to 60 dB of multiple-loop digital feedback around a high-speed 5 or 6-bit converter is a big deal; without noise shaping, a 5-to-6 bit converter would not be suitable for high-quality audio at all.

This is a purely subjective opinion, but I feel that the noise-shaping algorithm dominates the sound of delta-sigma converters, although low-speed analog electronics can mask the differences between converters and overlay another whole layer of coloration. In particular, low-speed electronics will react quite differently to the ultrasonic comb spectra of PCM and the ultrasonic spread spectra of DSD.

Here's a link to a seemingly unrelated topic:
http://arstechnica.com/science/2013...tric-grid-by-keeping-generators-out-of-synch/

This is what happens when you have multiple feedback loops that interact with each other; small, difficult-to-predict transient conditions can make the whole system go unstable. Noise-shaping, as used in delta-sigma and DSD converters, are high-order feedback systems wrapped around an array of switches, and as mentioned in the ESS and Resonessence literature, can behave unpredictably under dynamic conditions. Steady-state sinewave testing will not reveal the instabilities; the sliding-DC test exposes some of the instabilities, but not all of them.

Part of what I like about ladder/R2R converters is they are flash converters; no noise-shaping, no feedback, no stability issues, just a switch array. Yes, there are issues with monotonicity (due to physical limits of resistor-matching), but Philips dynamic-matching and Burr-Brown's CoLinear are effective ways around this. The monotonicity errors are also steady-state; that is, they are always there, instead of a dynamic distortion that comes and goes with the signal. Loudspeakers, after all, have pretty substantial amounts of distortion compared to amplifiers and digital systems, but the distortion is steady-state.

At the risk of going all meta on you guys, I'm attracted to non-feedback audio technology, with the signal only traveling in the forward direction, and backward shunt paths through the power supplies kept to a minimum.

The monotonicity errors are also steady-state; that is, they are always there, instead of a dynamic distortion that comes and goes with the signal.

Click to expand...

That depends how you look at it;
Since the resistors are not equally calibrated in relation to each other (so, say they don't have the exact same resistance), the voltage (current) coming from any set (combination) of bits will generate another distortion level. So, for easy thinking, supposed that one bit needs one resistor and an 8 bit word (with 256 levels) thus needs 8 resistors, this is what will happen when the 3rd bit/resistor (from the right) is off(-spec) :

01100011
01100100

Here the current should have increased with 1/256 of this 256 bit domain but because the 3rd resistor is off, it may be close to nothing or close to 1/128.
Now, since this bit (resistor) is always part of the signal, whether on or off, you can well say that in 50% of audio words it flips and thus implies a non-linearity in the signal itself. This is seen in THD and this is how lowering the digital volume such that one bit drops off (from say -5.9dBFS to -6.0dBFS) can show a better THD figure. If only that one poor resistor isn't used too often for that level.

This is not so easy to grasp, because when looking at a sine, the level changes all the time, as if all the time we would be attenuating and increasing the level again. It's just what the level of the sine-slope needs (for on/off bits) at the time. Only with DC this would show very good, but I don't think analysers will show THD on a DC signal (if sustaining a DC offset would be possible at all).

Also think of this :
Datasheets often show the THD for -0dBFS and for -20dBFS. Something like : when both are good, all in the middle of that will be better than -20dBFS. No way this is the case ... I have seen differences of 9dB on THD going *better* when attenuating more. Remember, this is ladder chips.

Of course, relative to the SDM and its "DC anomalies" the R2R could be called steady state.

Peter

Just came across this article on PCM DACs...originally written in '99/'00 and updated in '10. It was written by Thorsten Loesch, who is Director of Technology at Abbingdon Music Research (AMR).

http://www.scribd.com/doc/105561243...rt-Overview-of-the-Subject-by-Thorsten-Loesch

It is just over my head, but i can understand parts of it...any thoughts on the merits of this article? thanks for any opinions/guidance.

PeterSt said:

That depends how you look at it;
Since the resistors are not equally calibrated in relation to each other (so, say they don't have the exact same resistance), the voltage (current) coming from any set (combination) of bits will generate another distortion level. So, for easy thinking, supposed that one bit needs one resistor and an 8 bit word (with 256 levels) thus needs 8 resistors, this is what will happen when the 3rd bit/resistor (from the right) is off(-spec) :

01100011
01100100

Here the current should have increased with 1/256 of this 256 bit domain but because the 3rd resistor is off, it may be close to nothing or close to 1/128.
Now, since this bit (resistor) is always part of the signal, whether on or off, you can well say that in 50% of audio words it flips and thus implies a non-linearity in the signal itself. This is seen in THD and this is how lowering the digital volume such that one bit drops off (from say -5.9dBFS to -6.0dBFS) can show a better THD figure. If only that one poor resistor isn't used too often for that level.

This is not so easy to grasp, because when looking at a sine, the level changes all the time, as if all the time we would be attenuating and increasing the level again. It's just what the level of the sine-slope needs (for on/off bits) at the time. Only with DC this would show very good, but I don't think analysers will show THD on a DC signal (if sustaining a DC offset would be possible at all).

Also think of this :
Datasheets often show the THD for -0dBFS and for -20dBFS. Something like : when both are good, all in the middle of that will be better than -20dBFS. No way this is the case ... I have seen differences of 9dB on THD going *better* when attenuating more. Remember, this is ladder chips.

Of course, relative to the SDM and its "DC anomalies" the R2R could be called steady state.

Peter

Click to expand...

Peter, you know the monotonicity issues with ladder converters. The TotalDAC uses an array of discrete 0.01% accuracy Vishay resistors. Since each resistor is in free air, instead of sharing a monolithic die with a common substrate, temperature-tracking issues may arise, depending on the temperature coefficient of the discrete Vishay resistors. On the other hand, massive power can be routed through the resistors (compared to an integrated circuit), making an amplified I/V converter unnecessary.

The tradeoff, if I understand correctly, is monotonicity accuracy (and temperature drift) versus the potential coloration of an active I/V converter. As we say in America, half a dozen of one, or six of the other (two ways of solving the same problem).

How much improvement does the dynamic matching of the TDA 1541 buy compared to the Colinear system of Burr-Brown? Going further, what are the relative merits of paralleling converters versus balanced (push-pull) operation? I can see the utility of balanced operation for deglitching, but the measurements I made of the raw current output of the PCM-63 showed very little overshoot.

lloydelee21 said:

Just came across this article on PCM DACs...originally written in '99/'00 and updated in '10. It was written by Thorsten Loesch, who is Director of Technology at Abbingdon Music Research (AMR).

http://www.scribd.com/doc/105561243...rt-Overview-of-the-Subject-by-Thorsten-Loesch

It is just over my head, but i can understand parts of it...any thoughts on the merits of this article? thanks for any opinions/guidance.

Click to expand...

I'm largely in agreement with Thorsten, although I wouldn't expect analog designers who use transistors to agree with him. What I haven't seen - and would like to - is an intelligent hybrid approach, using a moderate-ratio microphone transformer to fully electrically isolate the converter from the analog circuitry.

This would have many benefits; a well-designed transformer has a 2nd-order lowpass function in the 30 to 70 kHz range, depending on what is requested, it can very accurately convert from balanced-current output to single-ended, and also isolates the grounds of the digital and analog sections. In other words, it does an excellent job of signal-conditioning for the analog stage.

Some folks are advocating eliminating the amplification stage entirely with a high-ratio MC-type transformer, but I think this is a mistake. High-ratio transformers do not perform very well, as moving-coil phono-cartridge enthusiasts know. Bandwidth and impulse response are badly compromised once you start using voltage/turns ratios much beyond 1:4; talk to a transformer designer and ask them what the optimum ratios for good impulse response are. The ones I've spoken to recommend anything between 1:1 and 1:4, and a 80% nickel or amorphous core, along with a screened secondary and a mu-metal case.

So the proposed architecture is a TDA1541, BB1704, or TotalDAC resistor array, an isolation/balancing transformer as mentioned above, and a vacuum-tube or solid-state analog stage with 200~500 kHz full-power bandwidth. This gives a generous margin between full-power bandwidth and the passband of the transformer, along with extremely high isolation between digital and analog sections. I haven't seen any commercial DACs with this topology; maybe I can persuade John Atwood or my other audio-friends to build a DAC like this.

LynnOlson said:

How much improvement does the dynamic matching of the TDA 1541 buy compared to the Colinear system of Burr-Brown?

Click to expand...

They're solving different issues - DEM is for addressing weighting errors, the CoLinear system is primarily about moving the glitch away from the zero crossing. Though Colinear also claims to share one array of resistors (or part thereof, I'm not recalling very clearly) between positive and negative halves so could help minimize 2nd harmonic distortion in relation to INL.

lloydelee21 said:

It is just over my head, but i can understand parts of it...any thoughts on the merits of this article? thanks for any opinions/guidance.

Click to expand...

I broadly agree with Thorsten on DACs too - just for my money he doesn't go far enough with the passive filtering, preferring to leave the imaging frequencies directly above 22kHz largely unfiltered.

More questions than answers

Hi Lynn,

LynnOlson said:

Peter, you know the monotonicity issues with ladder converters. The TotalDAC uses an array of discrete 0.01% accuracy Vishay resistors. Since each resistor is in free air, instead of sharing a monolithic die with a common substrate, temperature-tracking issues may arise, depending on the temperature coefficient of the discrete Vishay resistors. On the other hand, massive power can be routed through the resistors (compared to an integrated circuit), making an amplified I/V converter unnecessary.

Click to expand...

If it were only about that, there are more ways, like leaving it all passive with more gain in the main amps (no preamp), transformers, etc.
But all is about noise. I once worked on a "Total DAC" but I couldn't see it going to work with whatevel mil spec resistors because of the noise. And 0.01% ? that could be quite worse than any trimmed die'd resistor.

The tradeoff, if I understand correctly, is monotonicity accuracy (and temperature drift) versus the potential coloration of an active I/V converter.

Click to expand...

Sadly no any two solutions lead to comparable results. And this already in the measurement department, let alone the listening. But I think these kind of deliberations can be killed quite quickly by means of noticing that so far we are all listening to resonstruction filters. So, *these* determine the sound first, and next comes everything else. Well, sort of and you will get the grasp. For example, when I tried to implement my in-DAC volume control and listen through that to any of this killing reconstruction filters, what could I ever complain more because it was sh*t to begin with ? So yes, I still could see the in my view wrong noise lines and spikes which shouldn't be there, but that's all really overwhelmed by the filter. Notice that the ringing also smoothens those spikes or other anomalies.
What am I actually saying ? well, that we shouldn't be too theoretical, which it will be as long as we all *ought* to listen through these reconstruction filters. Only when that does not happen - and which for 100% sure also does not mean that no filter at all is allowed - only then these considerations become practice. And for practice : only then any volume control means kills the sound suddenly (so yes, with these reconstruction filters the sound was killed in advance).

And so, no way you are wrong anywhere, but you'd have to make it practice for yourself first. Otherwise it's a rather moot thing and it even could be better to have more noise to flatten anomalies (what about adding random jitter to mask deterministic).

How much improvement does the dynamic matching of the TDA 1541 buy compared to the Colinear system of Burr-Brown? Going further, what are the relative merits of paralleling converters versus balanced (push-pull) operation? I can see the utility of balanced operation for deglitching, but the measurements I made of the raw current output of the PCM-63 showed very little overshoot.

Click to expand...

I wanted to have some sort of smart answer here, but I'm in a answer questions with (sort of) questions mood as it seems;
I don't think you can't have any close clue to what software impeded jitter can do *first*. So, now I am talking about stupid playback software with some dials and all they do (all they can do !!) is change jitter signatures. My invention ? fine. But with this knowledge at hand, the current NOS1 is the third DAC (including a proto which was totally different) which completely failed on not letting this influence pass through but which was the explicit target.
Important ? well, as long as I, say twice per year, can let your amps sound like 20K more value only by means of new playback software versions, what to say. But what it should say is that we're focusing on the wrong targets, first.
I'm now into my 4th DAC design and when that can still be influenced by my own software I'll drop dead. Promise (reminds me from two angles of a lyric Killing yourself to Live (Black Sabbath) ).
So, the Colinear system of Burr Brown ? well, I tackled that one all right so it's not even necessary. In playback software. I know, now I'm really talking riddles. But some days I feel like that ...

Peter


PS: Lynn, a 40GHz nice device is underway to me, and will finally allow me to see 200fs of whatever jitter plus all the sh*t-beyond I could never see. Thank you for you help on this ...

Few random comments from a (high-speed, not audio) DAC designer:

1. Monotonicity is a static and dynamic problem. Calibrate the DC errors and dynamic errors are still killer...
2. Thermal drift and offsets are also static and dynamic. I have a program showing the impact of signal induced power causing thermal bow in a data converter. Fixing that is a pain...
3. Even if the reference ladder is static, switches couple charge in when they change state, adding dynamic offsets.
4. Stability is still an issue in the output buffer and filter stages (if active). DACs put out very wideband glitches that can cause the output stage to slew-limit, leading to all sorts of nasty things.
5. Dynamic matching can add correlated LF noise as elements rotate through the bits, and HF noise from the switches if their charge is not carefully filtered, a challenge in an IC and with lots of matched elements.

A fully-unary DAC might help but thats a lot of cells! Low bandwidth, big chip...

NorthStar said:

---- Hi Lynn; are those Marantz components still in your possession? ...And playing?
Are they kind of a Reference to you?

* The AV8003 uses indeed a delta-sigma Cirrus Logic CS4382A 192-kilohertz/24-bit DAC, which is an eight-channel DAC on a single chip. ...And the Cirrus Logic CS5361 ADCs.
It also has HDCD decoding; the 'Microsoft' Pacific Microsonics PMD-200 decoder/digital filter. ...And XLR balanced inputs and outputs.

As for Audyssey; only Audyssey MultEQ, and with some restrictions, because of limited DSP horsepower. ...Totally inadequate IMO.

Click to expand...

The Audyssey auto-EQ is horrible; it reminds of me of the 10-band graphic EQ's that were so popular in mid-fi Japanese equipment in the Seventies ... except maybe worse. It's seriously misdesigned, and seems unaware of the difference between direct-arrival sound and overall room energy. As a result, it boosts HF by a ridiculous amount, and is very shrill sounding. Maybe if the speakers were really, really bad, it would be an improvement, but even my college system back in the late Sixties had better speakers than that.

With the Audyssey switched out, the sound is ... OK. Faint praise, I know, but far better than the gruesome Denon, Onkyo, and Pioneer receivers. Man, those things are bad. Worse than the mid-fi receivers from Pioneer, Kenwood, and Sansui in the Seventies, which I wouldn't think possible. Based on glowing reviews and high prices, I thought Anthem would be good. Nope. Just more expensive.

Granted, I'm not a fan of transistor electronics in general, but I'm pretty happy with the sound of the iPod, as long it's connected to my Sennheiser HD580 headphones and playing uncompressed AIFF files. Actually, the Marantz sounds pretty much like that, just bigger and more expansive, thanks to 5 speakers and a REL Strata II subwoofer. The Dolby Pro-Logic II and DTS Surround is pretty good at 2 -> 5 synthesis, as these things go.

But the net result of the Marantz combo is that I didn't listen to much music with it; most CD's sound flat and uninvolving. Blu-Ray movie soundtracks, though, sound great, much better than in the movie theater, and spatial localization is a lot better than most HT receivers. So it's good at it's primary task, so I can't complain too much.

-- For untreated rooms, like mine, Audyssey MultEQ XT32 (Integra DHC-80.3 pre/pro),
in combination with DSD from multichannel music SACDs sounds pretty good to my older yEARS. :b

LynnOlson said:

The Audyssey auto-EQ is horrible; it reminds of me of the 10-band graphic EQ's that were so popular in mid-fi Japanese equipment in the Seventies ... except maybe worse. It's seriously misdesigned, and seems unaware of the difference between direct-arrival sound and overall room energy. As a result, it boosts HF by a ridiculous amount, and is very shrill sounding. Maybe if the speakers were really, really bad, it would be an improvement, but even my college system back in the late Sixties had better speakers than that.

With the Audyssey switched out, the sound is ... OK. Faint praise, I know, but far better than the gruesome Denon, Onkyo, and Pioneer receivers. Man, those things are bad. Worse than the mid-fi receivers from Pioneer, Kenwood, and Sansui in the Seventies, which I wouldn't think possible. Based on glowing reviews and high prices, I thought Anthem would be good. Nope. Just more expensive.

Granted, I'm not a fan of transistor electronics in general, but I'm pretty happy with the sound of the iPod, as long it's connected to my Sennheiser HD580 headphones and playing uncompressed AIFF files. Actually, the Marantz sounds pretty much like that, just bigger and more expansive, thanks to 5 speakers and a REL Strata II subwoofer. The Dolby Pro-Logic II and DTS Surround is pretty good at 2 -> 5 synthesis, as these things go.

But the net result of the Marantz combo is that I didn't listen to much music with it; most CD's sound flat and uninvolving. Blu-Ray movie soundtracks, though, sound great, much better than in the movie theater, and spatial localization is a lot better than most HT receivers. So it's good at it's primary task, so I can't complain too much.

Click to expand...

I have the the 7005/7055 combo. I use them strictly for watching TV and Movies. As you said, it does a very good job with that.

Kal Rubinson, who many consider a MC guru, thinks highly of the Audyssey EQ. Maybe he can comment.

-- Lynn's Marantz duo has Audyssey MultEQ.
=> http://www.hometheater.com/content/marantz-av8003-processor-and-mm8003-amplifier

- Your Marantz duo, Andre, has Audyssey MultEQ XT (one notch higher) . ...The replacement of Lynn's Marantz system.
=> http://www.hometheater.com/content/marantz-av7005-surround-processor-and-mm7055-amplifier

- And the latest Marantz duo (Andre's replacement) has Audyssey MultEQ XT32 (highest notch). ...& 32-bit DACs.
=> http://www.hometheater.com/content/marantz-av8801-surround-processor-amp-mm8077-amplifier

_________

* Me (XT32): => http://www.hometheater.com/content/integra-dhc-803-surround-processor-and-dta-701-amplifier

NorthStar said:

-- Lynn's Marantz duo has Audyssey MultEQ.
=> http://www.hometheater.com/content/marantz-av8003-processor-and-mm8003-amplifier

- Your Marantz duo, Andre, has Audyssey MultEQ XT (one notch higher) . ...The replacement of Lynn's Marantz system.
=> http://www.hometheater.com/content/marantz-av7005-surround-processor-and-mm7055-amplifier

- And the latest Marantz duo (Andre's replacement) has Audyssey MultEQ XT32 (highest notch). ...& 32-bit DACs.
=> http://www.hometheater.com/content/marantz-av8801-surround-processor-amp-mm8077-amplifier

_________

* Me (XT32): => http://www.hometheater.com/content/integra-dhc-803-surround-processor-and-dta-701-amplifier

Click to expand...

Thanks for info!!

-- I believe that these four pre/pros from above can accept DSD stream, straight from SACDs.
Or you have the choice to select PCM.

The last two pre/pros have the TI Burrr-Brown PCM-1795 32-bit/192kHz DACs. One per each channel, for all eleven channels plus sub (six stereo DACs).

LynnOlson said:

I'm largely in agreement with Thorsten, although I wouldn't expect analog designers who use transistors to agree with him. What I haven't seen - and would like to - is an intelligent hybrid approach, using a moderate-ratio microphone transformer to fully electrically isolate the converter from the analog circuitry.

This would have many benefits; a well-designed transformer has a 2nd-order lowpass function in the 30 to 70 kHz range, depending on what is requested, it can very accurately convert from balanced-current output to single-ended, and also isolates the grounds of the digital and analog sections. In other words, it does an excellent job of signal-conditioning for the analog stage.

Some folks are advocating eliminating the amplification stage entirely with a high-ratio MC-type transformer, but I think this is a mistake. High-ratio transformers do not perform very well, as moving-coil phono-cartridge enthusiasts know. Bandwidth and impulse response are badly compromised once you start using voltage/turns ratios much beyond 1:4; talk to a transformer designer and ask them what the optimum ratios for good impulse response are. The ones I've spoken to recommend anything between 1:1 and 1:4, and a 80% nickel or amorphous core, along with a screened secondary and a mu-metal case.

So the proposed architecture is a TDA1541, BB1704, or TotalDAC resistor array, an isolation/balancing transformer as mentioned above, and a vacuum-tube or solid-state analog stage with 200~500 kHz full-power bandwidth. This gives a generous margin between full-power bandwidth and the passband of the transformer, along with extremely high isolation between digital and analog sections. I haven't seen any commercial DACs with this topology; maybe I can persuade John Atwood or my other audio-friends to build a DAC like this.

Click to expand...

opus111 said:

They're solving different issues - DEM is for addressing weighting errors, the CoLinear system is primarily about moving the glitch away from the zero crossing. Though Colinear also claims to share one array of resistors (or part thereof, I'm not recalling very clearly) between positive and negative halves so could help minimize 2nd harmonic distortion in relation to INL.

Click to expand...

Thank you both for taking the time to read it and comment back. I will have to re-read these later this weekend to try to catch up with the discussion. In any event, this is a positive for me as i found the article interesting and instinctively felt it credible. The designer's conclusions are also reflected in some design elements in my own current favorite digital, Zanden. Thanks again.

Andre Marc said:

I have the the 7005/7055 combo. I use them strictly for watching TV and Movies. As you said, it does a very good job with that.

Kal Rubinson, who many consider a MC guru, thinks highly of the Audyssey EQ. Maybe he can comment.

Click to expand...

I do not want to get into a point-by-point analysis or argument because (1) I am a fan of solid state equipment, (2) I detest headphone listening and (3) most of my music listening is in discrete multichannel and that already puts Lynn and me in different spheres. That said, there is no question that my NYC system which generally runs without any EQ (except for the subwoofer) is vastly superior to my CT system which runs with Audyssey. However, the two rooms are quite different and my equipment investment in NYC is much greater. Not really comparable situations.

What I will say is that I have not found that Audyssey boosts the HF except in the cases where I have positioned the microphone improperly.

Kal Rubinson said:

I do not want to get into a point-by-point analysis or argument because (1) I am a fan of solid state equipment, (2) I detest headphone listening and (3) most of my music listening is in discrete multichannel and that already puts Lynn and me in different spheres. That said, there is no question that my NYC system which generally runs without any EQ (except for the subwoofer) is vastly superior to my CT system which runs with Audyssey. However, the two rooms are quite different and my equipment investment in NYC is much greater. Not really comparable situations.

What I will say is that I have not found that Audyssey boosts the HF except in the cases where I have positioned the microphone improperly.

Click to expand...

Hi Kal:

Thanks very much for the input. I too have no issues with solid state, even though I lean toward tubes.

I use Audsyssey in my living room with the Marantz combo because I think it does make a positive difference
due to the shortcomings of the room.

I use a full suite of Paradigm Monitor series 5 speakers and a PSB sub.

Agree that in your situation it is tough to compare since the set ups are quite different.

Part of my problem with Audyssey might be the Ariels are flatter and have tighter impulse response than most HT-friendly loudspeakers, which are more optimized for dynamics and vocal "pop-out" quality (mostly the result of tightening of directivity pattern in the vocal range).

This isn't a knock per se on HT-optimized speakers; there was a phase in the early development of the Ariel when they sounded *exactly* like HT speakers, while still measuring pretty flat. What got rid of the HT sound was reduction in diffraction (by using very large radii on the cabinet edges), matching the phase angles between tweeter and midbass to 10 degrees or better, and chasing out small resonances and standing-wave patterns in the 1~5 kHz region. Kind of tedious work, and took 13 different crossovers and another 6 months with MLSSA at my side.

The version of Audyssey that I have is reasonably good in the difficult 100~500 Hz region, where room colorations dominate. The version I have makes a complete mess of the 1kHz and above region, though, which is why I suspect it is trying to EQ the room instead of the first-arrival response from the loudspeaker. If that's what it's attempting, that's wrong; perceptually, first-arrival response dominates in the 1 kHz and above region.

The quality of the converters in the Marantz AV-8003 seem good enough; the opamps are probably what let it down. I sincerely hope they are not JRC4558's, which are dual versions of the notorious 741, which is unsuited for any kind of audio application. More likely, they're 5532/5534's, which are pretty ancient, dating back to 1979, and are mysteriously popular with the Japanese. Typical Japanese practice (this includes the Sony SACD-1 player) is plenty of 5532/5534's, all coupled in cascade with "selected" 100uF electrolytic caps for the DC-blocking function.

I'm not a fan of $1 opamps and 30-cent electrolytics in a product that sells for more than $1500; the Marantz AV-8003 (and MM8003) were built in China, so it's not like the labor was expensive. On the other hand, I bet all the Audyssey, DTS, Dolby, HDMI, etc. etc. licensing fees added up to a pretty high per-unit cost, particularly for the (relatively) small production run of Marantz flagship products. The licensing-fee issue alone is probably what keeps most audiophile manufacturers out of home theater.

P.S. If you are listening to DSD through Audyssey, it ain't DSD no more. Audyssey is a PCM signal processing system. Whether it runs at 24/88.2, 24/96, or 24/192 depends on how much processing power is available in the HT pre-pro. Conversion of DSD to non-integer speeds like 96 or 192 kHz, although simplifying the signal path for movies (which are based on 48 and 96 kHz sample rates), does no favors for DSD, which is based on multiples of 44.1 kHz.

Same story for bass management and user-selectable loudspeaker delay. These are PCM processes, and in a HT product, quite probably running at 96 or 192 kHz speeds. You'd have to look for HF artifacts on a spectrum analyzer to find out the effective speeds of the internal PCM signal processing system.

-- Hi Lynn,

Interesting. I give you only two examples here:

1. Onkyo TX-NR818 A/V receiver ($1,200 MSRP) or $649 street. ...Has Audyssey MultEQ XT32! Wow!
2. Sherwood Newcastle R-972 A/V receiver ($1,800 MSRP) or $599 street. ...Has TRINNOV! Wow!

Licencing fees?

Nowadays you have receivers that are Audyssey MultEQ Pro Ready for less than a thou! Or roughly $600 street!
Denon for example. Even some of their lower price receivers have Audyssey MultEQ XT and are Audyssey MultEQ Pro Ready!

_______________

I agree with what you just said from your above post; Audyssey is not necessarily for all type of loudspeakers, and is not much of a benefit at all (rather a detriment) for well acoustically treated rooms.

In my own personal case, with an untreated room, and 'inferior' speakers; if I engage Audyssey I got clarity in the highs, tightness in the lows, and an overall well fully balanced sound (for Movies, and Music also; Stereo, and Multichannel).
If I disable Audyssey (XT32), I got a sound that is dull, muffled, unclear, and just dead boring.
And the bass coming from my two subs is overpowering, booming, blotted, and just plain simply and totally undefined. ...Bland!

Now, I don't have a decent stereo sound system ($50,000 or so, including the speakers).
And I don't have a well treated room.
But if I would; screw Audyssey! ...Why? Because simply is is digitally 'pumping' the audio signals, the end sound.

For home theater, alright. For serious music listening, no way Jose!

At the end it's all very relative to each one's own systems and applications in their own environments.
...Just like some people prefer more DSD, and others PCM.