Marantz AV7705 Review (AV Processor)

[LTig]

Had a private request to measure the RCA output at full 2 volts and reduced 1.4 volt. Here they are:

View attachment 148864

View attachment 148865

I have another processor to test so please don't ask me for more measurements of this one.

At 1.4V it's in its specs.

 

[ah-ra]

The PM800X have different versions of HDAM. Also, those HDAMs in the PM series are actually doing something as they are not just an extra unity gain buffer added behind the volume control IC right at the end of the signal path. /QUOTE]

That's what I was thinking/hoping. I own a PM8006 and have to say I'm really quite happy with it. I would send it in for review, but I'm living in Germany... Would also be interesting to see measurements of a NR1710/11, as those slim line AVR should have a different architecture.....

 

[Koeitje]

First, I said more than a dozen brands tested here, not seven, nor what is available in the market. Do please read before you disagree. Finally, are you really saying that there are 700 Hi-Fi brands on the market competing with each other?

A dozen or seven makes no difference. In terms of affordable options there are around 7 anyway if you take into account parent companies. And yes I am saying hundreds of brands are competing for other types of components. You can design a dac today and sell some. Same for amplifiers or speakers.

Want to design and sell your own home cinema surround processor? Good luck getting even started. First thing you need to do it get licensed for Dolby and DTS. That is a very high barrier to entry. Then you need to have the technical know how to program the actual firmware on the machine. Which goes beyond even DSP controlled stereo amplifiers because you have much more to deal with including video support.

 

[don'ttrustauthority]

I used to think that. But that isn't it. Pro gear can have much higher density and channels than AVR and the best of it is very good measurement-wise. So it could be possible to make an AVR that has great numbers, and my guess is other than bothering to make it a goal the cost vs what they have would be somewhere from minimally higher to perhaps not at all higher.

Take Marantz's insistence on the HDAM. It isn't particularly great, they make it with quite a few discrete transistor circuits. My guess is they could make an all IC version 20 db better in every metric and far cheaper. Why don't they do it? I don't know.

I think you'll find buyers make decisions more on features than sound. If that's their preference, why do you think sound should be more prioritized?

 

[Tks]

Curious, what source(s) do you use that only has composite video output?

SNES being one off the top of by head.

 

[polmuaddib]

Why does this unit exhibit worse linarity then the first unit measured? Is the way it's measured different? Or this sample is worse then the first unit measured? There's almost 20db difference which seems significant. Also, how would we percieve bad linearity?

 

[peng]

Why does this unit exhibit worse linarity then the first unit measured? Is the way it's measured different? Or this sample is worse then the first unit measured? There's almost 20db difference which seems significant. Also, how would we percieve bad linearity?

I didn't see any major difference between the two tested except in linearity as you pointed out but that's a difference of 2 bit, not a huge deal.
The difference in preamp/DAC SINAD is negligible if you compare them at the same output voltage level, both managed about 74 to 75 dB.

 

[polmuaddib]

I didn't see any major difference between the two tested except in linearity as you pointed out but that's a difference of 2 bit, not a huge deal.
The difference in preamp/DAC SINAD is negligible if you compare them at the same output voltage level, both managed about 74 to 75 dB.

 

[Martin_320]

From the above charts, one can conclude that the SP-DIF coax input signal (from Amir's first benchtest in 2020) is superior to the Toslink (optical) connection from this most recent August 2021 benchtest.

 

[peng]

From the above charts, one can conclude that the SP-DIF coax input signal (from Amir's first benchtest in 2020) is superior to the Toslink (optical) connection from this most recent August 2021 benchtest.

He said almost 20 dB difference so that confused me a little. Now I see what he's saying, it looks like the it's a case of 16 bit vs 18 bits. I don't know how significant that is in practical sense. I noticed at least a few other AVP such as Monolith HTP-1 and Emotiva XMC-1 did do that well either. Though one would think between the two samples of the AV7705, the difference shouldn't be that much, but again, practically speaking it most likely won't make a difference. At least the preamp/DAC SINADs are consistent for the two.

 

[MasterApex]

I hope the AV7706 measured better. Just purchased one a few months ago to replace my old Anthem AVM50v for multichannel purpose.
I did stereo listening comparison between:
Oppo 105D (HDMI out) --- -------AV7706 ---- --------------------------Levinson---- 802D3
vs
Oppo 105D (XLR Analog out) ----GFP750 pre-amp (passive mode) ----Levinson----802D3

They sounds very similar for pop and classical songs, playing up to 98dBC peak at 11ft listening distance.

 

[bigguyca]

I hope the AV7706 measured better. Just purchased one a few months ago to replace my old Anthem AVM50v for multichannel purpose.
I did stereo listening comparison between:
Oppo 105D (HDMI out) --- -------AV7706 ---- --------------------------Levinson---- 802D3
vs
Oppo 105D (XLR Analog out) ----GFP750 pre-amp (passive mode) ----Levinson----802D3

They sounds very similar for pop and classical songs, playing up to 98dBC peak at 11ft listening distance.

The AV7706 measures about the same as the AV7705. This is to be expected as little has changed between the two models. Unfortunately the HDAM's haven't been upgraded with currents sources as is seen in the SR8015. The current sources are still resistor-based. This restricts "reasonably" good performance to a maximum from the XLR outputs of about 2.4V RMS.

The AV7706 offers better distortion performance if the digital input is restricted to about -4dBFS and below, that is -5dBFS, etc. The X6700H for example, doesn't work this way.

This effect can be seen using three scenarios:

#1: 0dBFS input and using the volume control to set output at say 2V XLR.

Internally the AV7706 outputs 2V from the DAC IC/filter circuit. The volume control is set at -6dB internally. This outputs 1V to the + output, the - output is thus -1V and the overall outputs is 2V.

#2: -6dBFS input and setting the volume control 6dB higher to again reach 2V from the XLR.

Internally the AV7706 outputs 1V from the DAC IC/filter circuit at -6dBFS. The volume control is set at 0dB internally. This outputs 1V to the + output, the - output is thus -1V and the overall outputs is 2V.

#3: 0dBFS input to say the left channel with that channel set at -6dBFS in the channel offsets. Volume control set as in #2 above, that is +6dB vs. #1.

Internally in the AV7706, the left channel input is 0dBFS. This is offset -6dB to -6dBFS. The output is thus 1V from the DAC IC/filter circuit. The volume control is set at 0dB internally. This outputs 1V to the + output, the - output is thus -1V and the overall outputs is 2V.

Scenarios #2 and #3 have lower distortion with the input to the DAC IC at -6dBFS vs. 0dBFS in #1.

Channel offsets in the AV7706, and other Denon/Marantz AVR's and AVP's using Audyssey are often -4dB and less. These lower digital input levels in practice may be improving the THD performance in practice in the AV7706.


On possible reason for this worse performance at digital input levels approaching 0dBFS is that the distortion of the AKM AK4458 DAC IC increases levels close to 0dBFS. Of course the X6700H also uses the AK4458 (or did, another story), but perhaps there is a difference in how the X6700H adjusts digital input level behind the scenes.

The attached file is the distortion performance of the AK4458 vs. input level from the AK4458 Evaluation Board manual


An alternative explanation to the above is that the DAC IC/filter circuit has higher distortion in the AV7706 than in the X6700H. The X6700H and Marantz SR8015 AVR's, and some other Denon AVR's, have current sources that feed the NJM8080 opamps in the DAC filter circuit such that the opamp is operated in Class A. The AV7706, and AV770X series in general, do not have these current sources.

The current sources are new with the X6700H. The X6500H did not have these current sources. The current sources must have been added for performance reasons since they are not mentioned in English language advertising.

It may be possible that these current sources cause the opamps in the DAC filter circuit in the X6700H and other recent Denon units to operate with significantly lower distortion in the 1V to 2V output range than in the AV7706.

 

[Qbd]

Comparing this review to the first one, why is the SNR so much worse here than there? In this review it’s 90 dB (Toslink), which is pretty bad

While in the first one it’s decent at ~107 dB (coax in)

Now I only really care about HDMI input, but why would there be such a huge difference?

 

[martijn86]

Comparing this review to the first one, why is the SNR so much worse here than there? In this review it’s 90 dB (Toslink), which is pretty bad
View attachment 164238
While in the first one it’s decent at ~107 dB (coax in)
View attachment 164236
Now I only really care about HDMI input, but why would there be such a huge difference?

To go from anoptical signal to an electrical signal, a conversion step needs to take place. That's being done by an active component and they're not all created equal. It could be that the conversion step is being done internally at 16 bit which would result in 15 bits worth of signal after it is processed and outputted over XLR/

 

[Martin_320]

Comparing this review to the first one, why is the SNR so much worse here than there? In this review it’s 90 dB (Toslink), which is pretty bad
View attachment 164238
While in the first one it’s decent at ~107 dB (coax in)
View attachment 164236
Now I only really care about HDMI input, but why would there be such a huge difference?

I remember 20 years ago when Paul Miller (tech editor / tester) at Hi-Fi-News magazine made the point that Optical/Toslink was inferior to coax sp-dif.
I also had an Audio-Alchemy DAC for my CD player with both sp-dif and Toslink inputs. With coax it sounded way better and much clearer vs optical.

 

[JJB70]

I think most of the AVR reviews here miss the point. It's a segment where feature set matters a great deal, and when you look at what is included within some of these AVRs (channels and software) it's a completely different ball game to 2 channel. As long as the SQ is good enough then I don't think it will matter to those buying these units. How many people do we really think would reliably discern a difference between this and a top scoring amplifier in a level matched double blind test?

 

[respice finem]

The perfectionist will say it's a lame excuse, but you're essentially right, even more so because it's the room that matters (and spoils many a home cinema). I ended up with two systems, one multichannel "for the fun" and one for "serious" stereo, with active monitors.

 

[bigguyca]

I think most of the AVR reviews here miss the point. It's a segment where feature set matters a great deal, and when you look at what is included within some of these AVRs (channels and software) it's a completely different ball game to 2 channel. As long as the SQ is good enough then I don't think it will matter to those buying these units. How many people do we really think would reliably discern a difference between this and a top scoring amplifier in a level matched double blind test?

How many valid double-blind tests are run? Essentially none.

Bringing up double-blind tests is frankly nonsense. Mentioning double-blind tests, as is done hundreds of times a year on this forum and others, is apparently an attempt to seem to contribute something when the poster is likely bankrupt of valid, technical points or information.

 

[JJB70]

How many valid double-blind tests are run? Essentially none.

Bringing up double-blind tests is frankly nonsense. Mentioning double-blind tests, as is done hundreds of times a year on this forum and others, is apparently an attempt to seem to contribute something when the poster is likely bankrupt of valid, technical points or information.

Or maybe it's an affront to some that all those alleged differences seem to mysteriously disappear when people aren't aware of what they are listening to.

 

[Echolane]

This is a review and detailed measurements of the Marantz AV7705 Audio/Video Processor (AVP). It was kindly purchased as a refurbished unit by a member and drop shipped to me. Normal price on Amazon is US $2,399.

I continue to not be a fan of Marantz design with the port hole requiring opening the panel below to see real information:

View attachment 148691

Back panel has the silly inputs no one uses these days such as composite and component video inputs, in gold plated no less:
View attachment 148692

I powered the unit and let it go through a firmware update which took an eternity. But all was well when it restarted.

I focused my testing on the front left and right balanced outputs. For inputs I tested HDMI and Toslink.

Marantz AV7705 Measurements
As usual we start with our dashboard using HDMI although Toslink performance was identical. I adjusted the volume until I got 4 volts which is the nominal output of most stand-alone DACs:

View attachment 148693

Well, this is not good! There is a ton of distortion and it is 3rd harmonic dominated so can't hide behind "it is good distortion" line. I played with various "pure" modes and it didn't make a difference. The SINAD of 75 dB is one of the worst in any AV receiver or processor I have tested:

View attachment 148694

You might say that 4 volt output is high so let's test different outputs:

View attachment 148695

Best performance is at 2.5 volt but even then, SINAD doesn't improve beyond the red ("poor") range of AV products we have tested.

I was hoping dynamic range was better but it is not:

View attachment 148696

Jitter performance was poor indicating a number of interference sources:

View attachment 148697

Could linearity be better? Of course not:

View attachment 148698

Intermodulation vs level shows the high noise floor, not being better than a freebie phone headphone dongle:

View attachment 148700

The DAC reconstruction filter is the typical very slow and wrong type:

View attachment 148699

This exerts a penalty on our THD+N vs frequency test due its wide bandwidth:

View attachment 148701

But there is more that is wrong here as changing the sample rate to 192 kHz would eliminate that factor. We see a bit of reduction at the top end but performance is still quite poor. We can see why if we look at the spectrum at 1 kHz:

View attachment 148702

The spray of distortion seems to go on forever so adds up to high THD+N with 90 kHz bandwidth (dashboard is at 22.4 kHz).

Multitone produces subpar performance as well:

View attachment 148703

Conclusion
It is simple: Marantz group takes perfectly good platform uses in its sister Denon division and screws it up across the board. Everything is degraded no matter what we test. It really is time for Marantz to go back to producing well engineered products rather than chasing subjective aspects they can't demonstrate. And poor notions like the slow reconstruction filter. It is depressing to test an AV product in 2021 with such poor performance.

Needless to say, I can't recommend the Marantz AV7705. Get a Denon AVR, shut off its amps and use it as a processor instead.

----------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

Those silly back panel connections are of crucial importance to me as I still have video sources that use either S-video or component video. I have a Pioneer Elite CLD-99 laser disc player and a Panasonic VHS player. I love opera and I have a lot of operas recorded on VHS as well as many operas on laser disc. So thank goodness for these silly inputs.

Of course there are work around in small cheap devices that convert these older signals to HDMI but they are clumsy and no doubt degrade the sound.

I am a music lover and LOVE GOOD SOUND. AVRs and Pre/Pros are major disappointments in audio quality and I would love to find something that measures at least decent. I bought the expensive Arcam SR-250 and was so disappointed in its audio that I bypassed it and connected all my video sources by RCA composite cables to my good stereo receiver, a refurbished Pioneer SX-1050. The expensive Arcam was reduced to just the job of video switching. I got a major improvement in sound but it’s hugely disappointing to spend so much money on such a poor sounding product.

BTW, I came to this review because I was about to buy this Marantz product. I’m sure glad I read this first. Thanks for doing this! Unfortunately, I have no idea how to disconnect the amps inside the Denon. I will just have to continue my present practice. Luckily, it works.