Marantz AV7705 Home Theater Processor Review

[Matthew J Poes]

I was looking forward to the release of AV7706 but after reading the review I can only hope that it fixes at least the worst problems AV7705 seemed to have. It would be great to have review for the AV7706

I have one coming for review. However, the only changes are the addition of 8K. No changes to the digital or analog section. That doesn't mean these issues weren't resolved, if any are software, but it does mean that its unlikely to be different.

 

[Sal1950]

I'm hoping Amir's measurement of the Marantz and Denon gear will soon have a positive influance on the Marantz designs. If I were working for Marantz I'd feel as though I had some egg on my face after comparing the results of the two's latest AVR products. It was too soon I'm afaid for the 7706 but we might have some hope for the next 8806 which probably won't be around for another year..

 

[Sal1950]

I have one coming for review.

BTW, Thanks Matt for your continued participation here.

 

[Taddpole]

I'm hoping Amir's measurement of the Marantz and Denon gear will soon have a positive influance on the Marantz designs. If I were working for Marantz I'd feel as though I had some egg on my face after comparing the results of the two's latest AVR products. It was too soon I'm afaid for the 7706 but we might have some hope for the next 8806 which probably won't be around for another year..

Suspect if it sells more they won't care.

 

[digicidal]

I'm hoping Amir's measurement of the Marantz and Denon gear will soon have a positive influance on the Marantz designs. If I were working for Marantz I'd feel as though I had some egg on my face after comparing the results of the two's latest AVR products. It was too soon I'm afaid for the 7706 but we might have some hope for the next 8806 which probably won't be around for another year..

Unfortunately, I think it's clear from their correspondence with Amir (i.e. "that agrees with our internal testing...") that the deficiencies that many of us on this site perceive in their gear is a conscious design decision rather than an oversight or omission. For whatever reasons the management or marketing branches cling to - it appears the desire for a "signature sound" profile is paramount.

On the other hand, as long as I ignore the objective performance metrics... my 8801's sound perfectly fine - even in comparison to my DX7 Pro... which is objectively far more transparent to the source. Brass ears are a definite advantage here, since if I can't hear the difference - I don't care about it as much. Now if they would only make a processor with fewer channels and analog inputs in a smaller form factor...

 

[peng]

Unfortunately, I think it's clear from their correspondence with Amir (i.e. "that agrees with our internal testing...") that the deficiencies that many of us on this site perceive in their gear is a conscious design decision rather than an oversight or omission. For whatever reasons the management or marketing branches cling to - it appears the desire for a "signature sound" profile is paramount.

On the other hand, as long as I ignore the objective performance metrics... my 8801's sound perfectly fine - even in comparison to my DX7 Pro... which is objectively far more transparent to the source. Brass ears are a definite advantage here, since if I can't hear the difference - I don't care about it as much. Now if they would only make a processor with fewer channels and analog inputs in a smaller form factor...

I don't think the AV8801 has the >10 kHz roll off based on bench test results. The second one below has a typical roll off, like Denon's down only 0.33 dB at 20 kHz, when the sampling frequency was 44.1 kHz. I guess Marantz started using the slow filter since the AV8802, 7701 or 7702, and SR7008 or 7009.

https://hometheaterhifi.com/reviews...ntz-av8801-11-2-surround-sound-processor-ssp/

https://www.soundandvision.com/cont...ocessor-amp-mm8077-amplifier-ht-labs-measures

 

[bigguyca]

The AV8801 used the two channel TI PCM1795 current output DAC IC. See page 10 or 41 of the datasheet below for the two reconstruction filter choices.

https://www.ti.com/lit/ds/symlink/p...=https%3A%2F%2Fwww.ti.com%2Fproduct%2FPCM1795

The PCM1795 is a good DAC IC, but implementing the I/V conversion added opamps and complexity to the design. Use of the AKM voltage output DAC IC likely cut costs.

 

[Martin_320]

By the way, I noted when reading the Hometheatrehifi review (which you linked to above) that Paul Belanger, Technical Product Manager for Marantz says:
"...The HDAM SA2 is the circuitry sitting between the OP-Amps and is mainly defining the characteristic of the sound. The OP-Amps might not look best on the data sheet, however in this configuration with the HDAM and current feedback topology it was our choice for a good balance in sound."

 

[bigguyca]

By the way, I noted when reading the Hometheatrehifi review (which you linked to above) that Paul Belanger, Technical Product Manager for Marantz says:
"...The HDAM SA2 is the circuitry sitting between the OP-Amps and is mainly defining the characteristic of the sound. The OP-Amps might not look best on the data sheet, however in this configuration with the HDAM and current feedback topology it was our choice for a good balance in sound."

The AV8801 has a typical two opamp output buffer (voltage follower, shunt-mode inverter) for each XLR output using mediocre NJM4565 dual opamp IC's. No further circuitry would be required to provide the output buffer function. Certainly better opamps would nice. Marantz could make a really nice product using just this simple buffer with higher quality opamps. It is sad that this isn't done

The HDAM, four transistor SA2 circuit is used in a stage before the output buffer. This stage can only add noise and distortion. It adds no functionality to the design. It could be replace with a wire.

Overall this stage is an NJM4565 opamp to SA2 HDAM, to common emitter transistor stage for voltage gain, to another NJM4565 opamp. Another definition of "Good balance," much be, "adds noise and distortion."

 

[Sal1950]

Unfortunately, I think it's clear from their correspondence with Amir (i.e. "that agrees with our internal testing...") that the deficiencies that many of us on this site perceive in their gear is a conscious design decision rather than an oversight or omission. For whatever reasons the management or marketing branches cling to - it appears the desire for a "signature sound" profile is paramount.

True but when that decision becomes a sales liability in the public's perception of the product they'll change their minds. All that's needed are more honest negative reviews from us and others, they can only take that beating for so long. If they see the scales numbers sliding to Denon they'll change their minds in a hurry.

On the other hand, as long as I ignore the objective performance metrics... my 8801's sound perfectly fine - even in comparison to my DX7 Pro... which is objectively far more transparent to the source. Brass ears are a definite advantage here, since if I can't hear the difference - I don't care about it as much.

Same here, at 70 yo, my 7703 sounds fine to me. LOL

 

[peng]

By the way, I noted when reading the Hometheatrehifi review (which you linked to above) that Paul Belanger, Technical Product Manager for Marantz says:
"...The HDAM SA2 is the circuitry sitting between the OP-Amps and is mainly defining the characteristic of the sound. The OP-Amps might not look best on the data sheet, however in this configuration with the HDAM and current feedback topology it was our choice for a good balance in sound."

I have to agree with him in one sense, because there are only two differences between Marantz and comparable Denon models are the HDAM and the slow roll of DAC filter used in the Marantz (except the AV8801 and older models). I never believe the dac filter would explain the claimed difference heard because it only affects digital inputs used on 44.1 sampling freq contents. On the other hand, the HDAM thing cannot explain it either because it is an opamp based buffer (discrete and/or IC/discrete combination) that shouldn't affect freq response in meaningful ways though if it adds enough distortions, it could in fact define the characteristic of the sound, but in what way and by how much? Evident is that the overall SINAD measured on ASR showed very little difference at pre out level up to 1.4 V. In the end, it boils down to marketing strategy, and I would bet at least 2 to 1 that those who could tell the two apart would no longer be able to do the same in a DBT. Easily for me to say because we are not going to know, but it is also easy for people from both camp to make their claims.

 

[Martin_320]

...I never believe the dac filter would explain the claimed difference heard because it only affects digital inputs used on 44.1 sampling freq contents. On the other hand, the HDAM thing cannot explain it either because it is an opamp based buffer (discrete and/or IC/discrete combination) that shouldn't affect freq response in meaningful ways though if it adds enough distortions, it could in fact define the characteristic of the sound, but in what way and by how much? ...

I have noted that Marantz likes to talk about fast slew rates and fast transient response performance in its products. That could explain their decision for setting the DAC filter to slow -- since a slow filter produces a faster impulse response and less ringing etc.. (that's even noted in AKM's own data sheet). Fast transient response could also be the reason they use HDAMs.

There may be something to this. Indeed, I can certainly feel that with music there is so much more attack/punch/drive/slam (whatever you want to call it) with AV7705 than I ever got with my previous Pre-pro -- the Onkyo PR-SC886 (and the difference is so noticeable for it to have been due to any 'placebo' effect.)
And the Onkyo's treble, whilst it was 'ice-pick' accurate, nevertheless would get very fatiguing when listening for extended periods, and would sometimes leave my ears ringing (no such problem with the Marantz). And as for the bass, while the Onkyo's went very deep, it never really had any punch to it and was actually quite boring with music. Again, these things was rectified in my system once I had removed the Onkyo and replaced it with the Marantz -- HDAMs and all

 

[peng]

I have noted that Marantz likes to talk about fast slew rates and fast transient response performance in its products. That could explain their decision for setting the DAC filter to slow -- since a slow filter produces a faster impulse response and less ringing etc.. (that's even noted in AKM's own data sheet). Fast transient response could also be the reason they use HDAMs.

There may be something to this. Indeed, I can certainly feel that with music there is so much more attack/punch/drive/slam (whatever you want to call it) with AV7705 than I ever got with my previous Pre-pro -- the Onkyo PR-SC886 (and the difference is so noticeable for it to have been due to any 'placebo' effect.)
And the Onkyo's treble, whilst it was 'ice-pick' accurate, nevertheless would get very fatiguing when listening for extended periods, and would sometimes leave my ears ringing (no such problem with the Marantz). And as for the bass, while the Onkyo's went very deep, it never really had any punch to it and was actually quite boring with music. Again, these things was rectified in my system once I had removed the Onkyo and replaced it with the Marantz -- HDAMs and all

Problem is, there could be another one of those "Marantz......by design..." as noted in the Audioholics review (see Gene's comments below the graph).

About the slew rate BS, if that's what they claimed, it would make no sense at all. The slew rate of the HDAM cannot not a factor because even if it has an infinite slew rate, it would be limited by the upstream volume control IC that is the same one use for all D&M AVRs and AV processors!! It is the last stage (buffer) of the preamp, so how the heck can it improve on slew rate, especially when its gain is "1". As an EE I felt like I have to call them out on this, may be, and I hope @amirm could chime in and tell me I am wrong on the logic but I bet he would agree, on condition that the HDAM is in fact the last stage, immediately ahead of the power amp input (Amir may not know that but I have the SM, that's what the diagrams show)...

Gene's comments:
https://www.audioholics.com/av-receiver-reviews/marantz-sr8015/conclusion

"I've been asked by several audioholics to start including square wave response curves in our measurements. Ideally there'd be no overshoot or ringing like we are seeing in the curve above. BUT Marantz confirmed this was a deliberate design attribute due to the short delay, slow roll-off DAC filter curve employed on ALL Marantz products to help better preserve high frequency phase response."

 

[Martin_320]

...It is the last stage (buffer) of the preamp, . . . . the HDAM is in fact the last stage, immediately ahead of the power amp input ...

I don't see a problem with a feedback mechanism starting at the output of the last stage of the system which you are applying feedback to. (I would have thought that to be standard practice.)
Here's a detailed discussion about HDAM & feedback here (esp. see posts by user "Bonsai") and they don't seem to be ridiculing it...
https://www.diyaudio.com/forums/sol...ck-amplifier-topologies-considerations-4.html

Regarding the square wave, the one depicted in the Audioholics review looks pretty good indeed! -- I see: (a) no pre-ringing; (b) some small initial overshoot; (c) after that virtually no post-ringing ripples

NB. If they had used a 'fast' filter on the same square wave, then you would have seen a lot more ringing/ripples, rather like the plot shown in the image here...

http://nihtila.com/wp-content/uploads/2019/10/W-DAC_v13A_Filt_SDSharp_Square.png
Above we see: long and high post-ringing (if there is pre-ringing it may even be masked by post-ringing) and high over/undershoot.

 

[peng]

I don't see a problem with a feedback mechanism starting at the output of the last stage of the system which you are applying feedback to. (I would have thought that to be standard practice.)
Here's a detailed discussion about HDAM & feedback here (esp. see posts by user "Bonsai") and they don't seem to be ridiculing it...
https://www.diyaudio.com/forums/sol...ck-amplifier-topologies-considerations-4.html

I don't think what Bonsai discussed would be relevant though, as my comments was based on logic only. The signal flow from the preamp section input (for simplicity, forget the digital, just go with analog for now) to the vol control IC, then from the vol control IC output to the HDAM and then from HDAM output to the power amp section input. So slew rate is already set by the vol control IC, the HDAM is an extra stage added there, Denon's do not have this stage. Have you seen the block diagrams that have been posted a few times? I can find it yet..

Again, I am assuming the HDAM's slew rate is infinite (that would be for an ideal op amp but obviously impossible in the real world), it still cannot output at slew rate faster than the vol control IC's output can provide. I understand the advantage of current feedback, but in this case the logic just won't work. It would be like using the a very good DAC with an average quality LSI chip ahead of it for volume control, like the AV8801 and AV8802, the LSI chip would be bottleneck as Dr. Rich of HTHF once said, in his 5 part article on AVR build quality.

https://hometheaterhifi.com/technical/technical-reviews/options-by-supplier-and-price/

"A key takeaway: circuit quality in the direct mode (stereo or 7.1) is almost always invariant to AVR prices in the range of $400 to $2,000. As examples, the $250 Yamaha RX-V367 and Marantz AV8801 ($3000) use the same Renesas LSI chip (R2A15220FP). With the LSI analog chip in these products, the sound of the direct mode is relatively constant, although a more robust power supplies, addition a quality output buffer and enhanced DC blocking capacitor quality can make small differences."

Bonsai seems to be talking about his Nx power amp that uses current feedback. I am only talking about the HDAMs used in the Marantz prepros and AVR's pramp section, and the fact is, it is an extra stage buffering the vol control IC output and power amp input, the current feedback technique affects only that stage. Also, to be clear, I am only talking about the claimed benefit in terms of slew rate that you mentioned earlier, obviously the HDAM might offer other benefits, such as improving output impedance at higher output level etc., though that needs to be supported by bench tests.

https://training.ti.com/system/files/docs/1221 - Slew Rate 1 - slides.pdf

So I take it that you have the AV7705? Don't get me wrong, I am not knocking the AV7705 at all. I do like my AV8801 though I like the X4400H that replaced it just the same. The AV7705 did not measure well on Amir's bench, but I am sure in a controlled test, even single blind test I wouldn't be able to tell a difference. I am only commenting on what I considered market hypes of Marantz, and Denon as well, I think they all do that so it is normal really.

 

[Martin_320]

So I take it that you have the AV7705? Don't get me wrong, I am not knocking the AV7705 at all.

Yes -- I assumed you knew I had this processor -- which has replaced my Onkyo pre-pro. And I wasn't suggesting you were 'knocking' it. We're simply talking about slow vs fast DAC filters, and HDAMs etc. -- something that we both have first-hand experience of.

Regarding the filter question, I'm confident that adults cannot hear the 2dB roll-off between 16kHz and 20kHz, ie. when a sample rate of 44.1kHz is being used. (I've done repeated tests between the same CD played into the Marantz using (a) SPDIF and (b) pure analog inputs respectively -- and I cannot hear any difference whatsoever.)

So given that we can now safely discount any audibility of the freq roll-off for a 'slow' filter, then what we are left with is the benefit of the slow filter -- ie. the superior transient & impulse response -- see one of my earlier posts. You have to admit that's a deliberate and tangible advantage of slow filters?

I do like my AV8801 though I like the X4400H that replaced it just the same.

A couple of questions:
1. Are you using the Denon receiver as a pre-amp to a separate power amp?
2. If you could do an apples-to-apples comparison, would you honestly notice any significant difference in sound quality between your Denon and your Marantz? (apart from Atmos codec-related differences etc.)
3. If the answer to (2) is "no"then I presume you got the Denon because you needed the latest Dolby Surround / Atmos codecs?
(same reason that I retired the Onkyo )

 

[peng]

So if can safely discount any audible freq roll-off for a 'slow' filter, then what we are left with is the benefit of the slow filter -- ie. the superior transient & impulse response -- see one of my earlier posts. You have to admit that's a deliberate and tangible advantage of slow filters?

I know it was by design. In fact I was the one who posted the response from Marantz Japan earlier this year about their rationale for choosing the slow roll-off filter. They never cited superior transient and impulse response as the reason. And regarding the pre-ringing issue that apparently is inherent from the use of fast roll off filter, of I understood it right, it would only apply when lower sampling rate such as 44.1/48 kHz are used, and even then it is not proven that people could hear the effect. At higher sampling freq, the ringing effect would be well outside the audible range. One of the site I read up on this was probably from the one linked below, time permitting I would read up on the topic again as I am now going by distant memory so I could be wrong and you may be right.

http://archimago.blogspot.com/2013/06/measurements-digital-filters-and.html

A couple of questions:
1. Are you using the Denon receiver as a pre-amp to a separate power amp?
2. If you could do an apples-to-apples comparison, would you honestly notice any significant difference in sound quality between your Denon and your Marantz? (apart from Atmos codec-related differences etc.)
3. If the answer to (2) is "no"then I presume you got the Denon because you needed the latest Dolby Surround / Atmos codecs?
(same reason that I retired the Onkyo )

1. As prepro.
2. Not compared in HT, but in two channel stereo. Also compared it with a X3400H and one of my separate pair, the Cambridge Audio preamp and Halo A21. All sounded close enough to be indistinguishable to me, and based on REW FR curves, they were virtually identical from 20-20000 Hz. I know FR is just one factor but it is an important one. I can tell a difference if I use PEQ on the very audible range such as between 200 and 8 kHz, even if adjustments are limited to just a couple dB. It was just my subjective experience that there weren't the expected/or claimed, rather, Marantz, Denon, or Cambridge Audio sound. Being subjective, my findings would be totally meaningless to others, or anecdotal at best.
3. Yes, Atmos was the triggered, but it was also a wake up call for me that I would likely have to update/upgrade the prepro/AVR every 5 years or sooner. I can afford spending 4-5K once in a while but not every 5 years, so it was time to switch to AVR that typically could be had at deep discounts a few times a year for a variety of reasons.

By the way, I found the Marantz response for you in case you haven't read it before:

"The cause of distortion deterioration is quantization aliasing noise and the characteristics of the digital filter are intended and in accordance with the specifications of the DAC. (AK4490).

It does not apply to the point that there is a bug in the filter characteristics of the DAC. This characteristic was selected by the filter setting according to Marantz's sound policy. Marantz has selected slow roll-off filters for DAC output for many years. If it attenuates before 22.05 kHz, then the audio band will not be flat and the sound quality will be sacrificed.
We consider that the reviewer's perception and expectations for the slow roll-off filter are different from what we intended. The AV8805 is a product that has sufficiently considered the sound quality of Marantz."

When I asked them the question, I requested that they get the answer from their engineering, and if I remember right they did forward it to Japan. It took quite some to get the response.

 

[Martin_320]

I know it was by design. In fact I was the one who posted the response from Marantz Japan earlier this year ....
"The cause of distortion deterioration is quantization aliasing noise and the characteristics of the digital filter are intended and in accordance with the specifications of the DAC....

Ah good one -- I didn't see your answer from Marantz in Japan before.

However, when I read the statement, he hasn't really been very clear, or these are the words of a comms spokesperson and something got 'lost in translation'.
For example, I don't follow how he can be talking "aliasing" and "quantisation" noise -- as I've always understood that anti-aliasing is process you do in the PCM sampling (A-to-D) stage, rather than in the PCM decoding (D-to-A) process.
eg. You'd get aliasing in the encoded waveform if you hadn't previously low-passed the source analog signal (eg from a vocal microphone or guitar interface etc. ) at just below half the sample rate. (Nyquist/Shannon) .
Here I don't see how that can apply to a reconstruction filter situation, which is there to take out DAC noise (and there is no re-sampling going on -- as the fs rate stays exactly the same as what was used during the recording/sampling phase -- so no further aliasing is would happen anyway).

 

[peng]

Ah good one -- I didn't see your answer from Marantz in Japan before.

However, when I read the statement, he hasn't really been very clear, or these are the words of a comms spokesperson and something got 'lost in translation'.
For example, I don't follow how he can be talking "aliasing" and "quantisation" noise -- as I've always understood that anti-aliasing is process you do in the PCM sampling (A-to-D) stage, rather than in the PCM decoding (D-to-A) process.
eg. You'd get aliasing in the encoded waveform if you hadn't previously low-passed the source analog signal (eg from a vocal microphone or guitar interface etc. ) at just below half the sample rate. (Nyquist/Shannon) .
Here I don't see how that can apply to a reconstruction filter situation, which is there to take out DAC noise (and there is no re-sampling going on -- as the fs rate stays exactly the same as what was used during the recording/sampling phase -- so no further aliasing is would happen anyway).

Agreed, it did look like something got lost in translation. My main concern at the time was whether it was something by design, or an over sight, and on that count they answered my questions. I have been dealt with D+M customer support before so I know for technical questions, it is better to insist they get a response from level two, or from Japan, and take a chance of the "translation" issue that may or may not be an issue.

 

[Blumlein 88]

Ah good one -- I didn't see your answer from Marantz in Japan before.

However, when I read the statement, he hasn't really been very clear, or these are the words of a comms spokesperson and something got 'lost in translation'.
For example, I don't follow how he can be talking "aliasing" and "quantisation" noise -- as I've always understood that anti-aliasing is process you do in the PCM sampling (A-to-D) stage, rather than in the PCM decoding (D-to-A) process.
eg. You'd get aliasing in the encoded waveform if you hadn't previously low-passed the source analog signal (eg from a vocal microphone or guitar interface etc. ) at just below half the sample rate. (Nyquist/Shannon) .
Here I don't see how that can apply to a reconstruction filter situation, which is there to take out DAC noise (and there is no re-sampling going on -- as the fs rate stays exactly the same as what was used during the recording/sampling phase -- so no further aliasing is would happen anyway).

With a slow roll off filter you can get imaging above nyquist. A mirror imaged reflection of the below 22 khz stuff. If you had high level in the upper frequencies it can image at fairly high level just above 22 khz. Whether that can cause IM distortion back into the audible band in tweeters etc is dependent upon the particulars and wouldn't do so very often. I'm not sure that is what they meant prior to translation however.