The often recommended AVR-X3500H got measured, finally..

[PENG]

It finally got measured over at Audiosciencereview.com.

I thought the results clearly raised doubts about the effects of the lack of preamp mode. Amir found similar shutdown issues Gene found when he (actually may as well say "they") did the 20-20,000 Hz sweep with no speakers connected.

His results also showed pre out up to 1.5 V yield decent results but got ugly past that point, up to 0.02 % neighborhood THD at 2 V. Even 0.05% wouldn't look bad to us but Amir is used to much lower values when he tested real ext. DACs so needless to say he wasn't impressed.

If you ignore the DAC performance (not exactly a dac but it's Amir you know..) part, as an amp using analog inputs the thing left the NAD T777 Vs and 758 V3 in dust, that's his words not mine. It was better than the RX-A1080 too, but only for analog inputs, otherwise the Yamaha won, thought all on paper only; and imo the differences wouldn't be audible to most users.

Anyway, his measurements sort of explained why I thought the X3400H sounded as good as my A21/840W amp/preamp, as I was using my external DACs connected to the X3400H's analog inputs.

 

[AcuDefTechGuy]

It finally got measured over at Audiosciencereview.com.

I thought the results clearly raised doubts about the effects of the lack of preamp mode. Amir found similar shutdown issues Gene found when he (actually may as well say "they") did the 20-20,000 Hz sweep with no speakers connected.

His results also showed pre out up to 1.5 V yield decent results but got ugly past that point, up to 0.02 % neighborhood THD at 2 V. Even 0.05% wouldn't look bad to us but Amir is used to much lower values when he tested real ext. DACs so needless to say he wasn't impressed.

If you ignore the DAC performance (not exactly a dac but it's Amir you know..) part, as an amp using analog inputs the thing left the NAD T777 Vs and 758 V3 in dust, that's his words not mine. It was better than the RX-A1080 too, but only for analog inputs, otherwise the Yamaha won, thought all on paper only; and imo the differences wouldn't be audible to most users.

Anyway, his measurements sort of explained why I thought the X3400H sounded as good as my A21/840W amp/preamp, as I was using my external DACs connected to the X3400H's analog inputs.

Besides the “unforgiven” 0.02% THD @ 2V, did you ”translate“ his numbers to the usual SNR, XT?

 

[PENG]

Besides the “unforgiven” 0.02% THD @ 2V, did you ”translate“ his numbers to the usual SNR, XT?

XT wasn't great, SNR good.. translation is for tomorrow

 

[AcuDefTechGuy]

XT wasn't great, SNR good.. translation is for tomorrow

I gave up on that site. So I’ll have to wait for your extrapolations.

 

[PENG]

For those, like ADTG who don't want to search the jungle for treasures, my own quick take, or as ADTG called it, "translations/extraplations":

A) Pre out (Amir called in DAC, using HDMI and coax input) performance:

- No issue with shutting down at 2 V output but THD (the worse channel), 0.021585% are higher than the RX-A1080's 0.005805% but the Yamaha fell short of 2 V before it would trip, and that's just the usually 1 kHz test that others (S&V) typically did.

- THD at 1 and 1.2 V was about 0.0017%, Amir said "it goes crazy after 1.5 V (vol = 80), that's volume "0"., he didn't show any results for below 2 V, so it must be between 0.0017% and 0.02%, he called it crazy but we (like @Irvrobinson would likely know what he meant right....
Note: The RX-A1080 had much higher THD if Toslink input is used, only a few dB better than Denon's, Amir couldn't think of any reasons.

Well, at least that's good news for those who are using this $599 machine ext. amp such as the Emo gen3, Monolith, Outlaw, ATI amps that have gains >28 dB as the THD should be excellent at below 1.5 V and still very good for most of us at 2 V or higher, just don't compare it to Anthem, Bryston, Audio Research class of preamps, may be okay to compare with our favorite bashing target the P5. Sorry @ematthews..for the unintended aggravation..

Dynamic range (I pasted the Yamaha's so we can compare):

AVR-X3500H...........................RX-A1080
Ch1.....109.903........................98.679
Ch2.....110.191........................98.716

B) Amplifier tests:

THD+N at 5 W, 4 ohm load, analog/digital coax input):
Ch1.........................0.006007%/0.007163%
Ch2.........................0.007391%/0.008626%

He commented the higher numbers for the coax input was due to increased noise floor.

Note: He did not engaged pure direct, I guess that's because he found no difference when he did so when measuring the NAD, Yamaha, or Anthem? (don't remember) it made no difference. I have seen the schematics of some Denon's and I believe if he engaged pure direct, noise would have dropped a little.

SNR (Analog in, near max power output)
Ch1.........................102.657 dB
Ch2.........................103.050 dB

Note that Amir never used weighting so those figures were very good.

Now the interesting part, remember how often we got told don't look at specs, NAD's power outputs were very conservative and better for 4 ohm loads blablabla...

Also remember I said before, any amps can drive 4 ohm loads depending on your power requirements (like how loud you listen, sitting distance, spkr sens and how aggressive the protective system were set etc...?

Using the coax input, he measured 159 Watts at 0.007% THD+N (knee point) vs NAD T777 V3's 0.02% at approx the same point.

Below are Amir's comments on the Power vs distortions @ 4 Ohm Both Channels Driven tests:

"Again, the NAD T777 V3 is left in the dust with respect to noise and distortion.

Note that the amplifier is not rated for 4 ohm but works anyway."

Ranking chart that includes only AVRs he recently measured:

To see the original please visit:

https://www.audiosciencereview.com/forum/index.php?threads/denon-avr-x3500h-avr-review.10053/;

)

 

[ryanosaur]

@PENG
Thanks for posting!

 

[Irvrobinson]

For those, like ADTG who don't want to search the jungle for treasures, my own quick take, or as ADTG called it, "translations/extraplations":

A) Pre out (Amir called in DAC, using HDMI and coax input) performance:

- No issue with shutting down at 2 V output but THD (the worse channel), 0.021585% are higher than the RX-A1080's 0.005805% but the Yamaha fell short of 2 V before it would trip, and that's just the usually 1 kHz test that others (S&V) typically did.

- THD at 1 and 1.2 V was about 0.0017%, Amir said "it goes crazy after 1.5 V (vol = 80), that's volume "0"., he didn't show any results for below 2 V, so it must be between 0.0017% and 0.02%, he called it crazy but we (like @Irvrobinson would likely know what he meant right....
Note: The RX-A1080 had much higher THD if Toslink input is used, only a few dB better than Denon's, Amir couldn't think of any reasons.

I don't understand your comment regarding my mention. Why would I know what he meant?

 

[PENG]

@PENG
Thanks for posting!

You're welcome!

Based on those measurements, I would continue to recommend the X3400/3500/3600/4400/4500H but would be cautious about people with golden ears using them as preamp for serious two channel use and expect near "separates" performance levels. I highly doubt 0.02% THD+N preout voltage >=2 V will be problematic at all, but if you know it, can see it, you may just hear it. I hope Amir will get to measure a Marantz SR6013/14, though Gene has already measured the SR8012 so we can reasonably extrapolate the numbers for the lower models, and conclude that the results between Denon and Marantz's should be very similar.

Technically speaking, the Marantz models do have an extra buffer stage, namely the HDAM, so in theory they should have cleaner output at above 1.5 V though Dr. Rich found that the much improved preamp vol control IC and HDAM vs the AV8805 over the AV8802A did not improve the results overall because while noise improved (due the vol control upgrade), distortions has not. Iirc he seemed suspicious of the HDAM adding distortions, so if that's the case/logic, I suspect Marantz may do better with less sensitive power amps but the overall distortions might not be better. In any case it is all academic as imo unless you really have fantastic hearing with trained ears/brains, AVRs, integrated, separates are not the bottleneck. Recording quality, room acoustics, speakers, noise floor etc. are likely the real bottlenecks relatively speaking.

 

[PENG]

I don't understand your comment regarding my mention. Why would I know what he meant?

I just thought because you might have visited his site way earlier than many of us here at AH, so you would know when he referred to 0.02% THD or 74 SINAD at 2 V output as crazy, he did not necessarily mean "crazy" in terms of audibility but in terms of it being very poor relative to many of the DACs he measured, like those in the blue/green buckets on his SINAD ranking charts.

As example, in post#49 in that review thread he said:

"We would slash our wrists if we had a DAC with a SINAD of 95. But yes, it is much better at lower output level. "

Denon AVR-X3500H AVR Review

Although it would be hard to arrange, I think it would be interesting to do a blind A/B/X study to see if many could reliably hear the difference between something like this mediocre unit and a home theater setup with SOTA DAC quality on a mix of Blu-Rays & DVDs- music+video discs (concerts) and...

www.audiosciencereview.com

By the way, 95 dB is about 0.0018%, I highly doubt people here would think about slashing wrists, but he seemed to think we might..... Seriously though, like him, I also tend to get fixated on numbers, the lower the better while I am also a believer in the inaudibility of THD < 0.01 or even 0.05%.

 

[Irvrobinson]

I just thought because you might have visited his site way earlier than many of us here at AH, so you would know when he referred to 0.02% THD or 74 SINAD at 2 V output as crazy, he did not necessarily mean "crazy" in terms of audibility but in terms of it being very poor relative to many of the DACs he measured, like those in the blue/green buckets on his SINAD ranking charts.

As example, in post#49 in that review thread he said:

"We would slash our wrists if we had a DAC with a SINAD of 95. But yes, it is much better at lower output level. "

Denon AVR-X3500H AVR Review

Although it would be hard to arrange, I think it would be interesting to do a blind A/B/X study to see if many could reliably hear the difference between something like this mediocre unit and a home theater setup with SOTA DAC quality on a mix of Blu-Rays & DVDs- music+video discs (concerts) and...

www.audiosciencereview.com

By the way, 95 dB is about 0.0018%, I highly doubt people here would think about slashing wrists, but he seemed to think we might..... Seriously though, like him, I also tend to get fixated on numbers, the lower the better while I am also a believer in the inaudibility of THD < 0.01 or even 0.05%.

I think I mentioned a while back that, IMO, Amir is a numbers guy. He's using measurements like SINAD as a figure of merit, and he has never correlated specific SINAD levels of results to audibility. He does compare them to, for example, Redbook CD limits, which themselves are typically better than audibility. I admit that, as with John Atkinson's measurements in Stereophile, I do tend to take Amir's opinions as just that, opinions, and I come to my own conclusions. To Amir, better measurements are just better. All other things being equal, I prefer better measurements for their own sake too, as it is a sign of superior engineering execution, but all other things are seldom equal. It should also be noted that Amir takes the time to train himself as a listener, so he does claim to be able to hear differences (with evidence of blind test software) that untrained listeners using music probably couldn't.

I don't know where my Marantz AVR (acting as a pre-pro) would be positioned in his graph, but even if it came out last I wouldn't replace it. For my purposes it works pretty much flawlessly.

 

[Swerd]

I think I mentioned a while back that, IMO, Amir is a numbers guy. He's using measurements like SINAD as a figure of merit, and he has never correlated specific SINAD levels of results to audibility.

I have to agree with Irv on this question. Until Amir can correlate any of his measurements with audibly improved sound, they mean nothing in terms of improved home audio.

Frankly, I got tired of reading Amir's posts a long time ago, when he used to argue incessantly on AVR with the late Arnie (I forget his last name). Amir's posts were elaborate and complex. I wasn't impressed, as I thought he over-complexificated things. As long as he refused to correlate his electronic lab bench measurements with audibly improved sound, they resembled the pseudo-science arguments that too often take the approach of "if you can't dazzle them with brilliance, baffle them with BS". Amir may not be full of BS, but his arguments and logic failed to convince me that I should learn to wade through his complexity.

 

[ryanosaur]

over-complexificated

That’s a technical term?

 

[Irvrobinson]

I have to agree with Irv on this question. Until Amir can correlate any of his measurements with audibly improved sound, they mean nothing in terms of improved home audio.

That's not a compelling argument, since the audibility of many measurements, especially for those of electronics, but sometimes even for speakers, may not have a direct relationship to the listening experience. Unless the measurements reveal substantial problems audibility may be questionable. The reason for measurements is to validate that a product meets or exceeds its specifications, or doesn't exhibit anomalous behavior. Some products do exhibit anomalous behavior.

Frankly, I got tired of reading Amir's posts a long time ago, when he used to argue incessantly on AVR with the late Arnie (I forget his last name). Amir's posts were elaborate and complex. I wasn't impressed, as I thought he over-complexificated things. As long as he refused to correlate his electronic lab bench measurements with audibly improved sound, they resembled the pseudo-science arguments that too often take the approach of "if you can't dazzle them with brilliance, baffle them with BS". Amir may not be full of BS, but his arguments and logic failed to convince me that I should learn to wade through his complexity.

I think you're holding Amir to a higher standard. Gene doesn't correlate electronics measurements to audible behavior either. Measurements, especially raw measurements, are rarely published these days. I appreciate his measurements; the rest is just opinion.

 

[PENG]

I think I mentioned a while back that, IMO, Amir is a numbers guy. He's using measurements like SINAD as a figure of merit, and he has never correlated specific SINAD levels of results to audibility. He does compare them to, for example, Redbook CD limits, which themselves are typically better than audibility. I admit that, as with John Atkinson's measurements in Stereophile, I do tend to take Amir's opinions as just that, opinions, and I come to my own conclusions. To Amir, better measurements are just better. All other things being equal, I prefer better measurements for their own sake too, as it is a sign of superior engineering execution, but all other things are seldom equal. It should also be noted that Amir takes the time to train himself as a listener, so he does claim to be able to hear differences (with evidence of blind test software) that untrained listeners using music probably couldn't.

I don't know where my Marantz AVR (acting as a pre-pro) would be positioned in his graph, but even if it came out last I wouldn't replace it. For my purposes it works pretty much flawlessly.

From what I can see in the schematics, your 1609 would measure extremely similar to the x3500h's on the preamp side but won't be as good on the power amp side. I will very sure about this.

 

[PENG]

I have to agree with Irv on this question. Until Amir can correlate any of his measurements with audibly improved sound, they mean nothing in terms of improved home audio.

Frankly, I got tired of reading Amir's posts a long time ago, when he used to argue incessantly on AVR with the late Arnie (I forget his last name). Amir's posts were elaborate and complex. I wasn't impressed, as I thought he over-complexificated things. As long as he refused to correlate his electronic lab bench measurements with audibly improved sound, they resembled the pseudo-science arguments that too often take the approach of "if you can't dazzle them with brilliance, baffle them with BS". Amir may not be full of BS, but his arguments and logic failed to convince me that I should learn to wade through his complexity.

I have no issue with his site or Amir because I always read the subjective parts if any sites for fun only and only if I have free time. His measurements seem credible because 1) the numbers typically corroborate with other site's where applicable, and 2) he is an EE and did practice albeit may be long time ago (not quite sure how long).

 

[Swerd]

I have no issue with his site or Amir because I always read the subjective parts if any sites for fun only and only if I have free time. His measurements seem credible because 1) the numbers typically corroborate with other site's where applicable, and 2) he is an EE and did practice albeit may be long time ago (not quite sure how long).

I do understand what you say. But Amir has never been able to say, in plain English, why his measurements should matter. Do any of his measurements lead to better sounding home audio?

At present, his work seems to appeal only to other EEs. And it strongly suggests, to myself and others, that his findings fall below the level of audibly improved sound quality. Why would anyone want to pay extra for electronic gear that might have superior design or build qualities, but fail to deliver on improved sound?

 

[Irvrobinson]

Frankly, I got tired of reading Amir's posts a long time ago, when he used to argue incessantly on AVR with the late Arnie (I forget his last name).

I think you're talking about Arnie Krueger. Some commentary from Amir no less:

Arny Krueger Has Passed Away

I woke up with very somber mood to hear that our member ArnyK (Arney Krueger) passed away a few days ago: https://www.legacy.com/obituaries/name/Arnold-Krueger-obituary?pid=188925019&view=guestbook He is best known for popularizing ABX testing and championing blind audio listening tests in...

www.audiosciencereview.com

 

[Swerd]

I think you're talking about Arnie Krueger.

Yes, it was Arny Krueger. Thanks.

That note on his passing, from Amirm no less, was very decent.

 

[Irvrobinson]

Yes, it was Arny Krueger. Thanks.

That note on his passing, from Amirm no less, was very decent.

Many times even adversaries respect each other.

 

[PENG]

I do understand what you say. But Amir has never been able to say, in plain English, why his measurements should matter. Do any of his measurements lead to better sounding home audio?

At present, his work seems to appeal only to other EEs. And it strongly suggests, to myself and others, that his findings fall below the level of audibly improved sound quality. Why would anyone want to pay extra for electronic gear that might have superior design or build qualities, but fail to deliver on improved sound?

Also, he knows a lot about the bits and byes, and testing protocols but I have the feeling that he knows a lot more about desktop and portable DACs/headphone amps than av receivers. It is encouraging that he seems receptive to start testing more non dac/headphone amps lol.. It is a shame that Stereophile won't bother with AVRs, perhaps they are just being practical minded?