The SEPARATES vs. AVR Thread

[PENG]

Exactly what I’m thinking since AVR has FR that are like +/- 0.1dB 20Hz-20kHz.

Yeah, and I assume both you and @RichB have seen my REW plots comparing my AVR-3805 with my other preamp/amp separates combos including the Halo A21/Cambridge Audio 840E. Using smoothing as low as 1/48 the graphs practically overlaps each other almost perfectly. Even without smoothing they look practically the same. Yet if you compare the same plots of different speakers, or the same speakers but at a slightly different toe in angle or distance, the difference between the plots would be obvious to naked eyes.

So if flat FR is most critical (Dr. Floyd seemed to be saying just that..), keep in mind on/off axis makes no difference to most electronics, and that THD+N for the DUTs involved are not an issue, how the heck could one pass a SBT, let alone SBT using my DUTs in my room? That is, regardless of Dr. F.Toole's assertion without elaborating the details that those little Omni mic, Umik-1 mics etc., aren't really measuring what we we are hearing. I am sure he's right in the related theories and even in practice, but if the mic tell us FR is flat under a certain condition, even though our ears won't hear the same, but so what, as long as our goal is flat FR right? Never mind mics, we all hear differently too, and have different taste. That's why I think Dr. Toole is most likely right in theory, but I do find him contradicting himself on occasions (not theories, but only in his talks on topics related to REQs, psychoacoustics such as human preference to different speakers etc.) in his various video I have seen so far.

 

[AcuDefTechGuy]

Yeah, and I assume both you and @RichB have seen my REW plots comparing my AVR-3805 with my other preamp/amp separates combos including the Halo A21/Cambridge Audio 840E. Using smoothing as low as 1/48 the graphs practically overlaps each other almost perfectly. Even without smoothing they look practically the same. Yet if you compare the same plots of different speakers, or the same speakers but at a slightly different toe in angle or distance, the difference between the plots would be obvious to naked eyes.

So if flat FR is most critical (Dr. Floyd seemed to be saying just that..), keep in mind on/off axis makes no difference to most electronics, and that THD+N for the DUTs involved are not an issue, how the heck could one pass a SBT, let alone SBT using my DUTs in my room? That is, regardless of Dr. F.Toole's assertion without elaborating the details that those little Omni mic, Umik-1 mics etc., aren't really measuring what we we are hearing. I am sure he's right in the related theories and even in practice, but if the mic tell us FR is flat under a certain condition, even though our ears won't hear the same, but so what, as long as our goal is flat FR right? Never mind mics, we all hear differently too, and have different taste. That's why I think Dr. Toole is most likely right in theory, but I do find him contradicting himself on occasions (not theories, but only in his talks on topics related to REQs, psychoacoustics such as human preference to different speakers etc.) in his various video I have seen so far.

We could sure use him chiming in.

Freqency Response seems to be the salient factor according to his audio/acoustic research.

That’s why I would disagree w/ M Code on what Floyd Toole would say regarding the sound quality of AVR vs Separates since both AVR and Separates have ruler-flat FR among other things like extremely low THD and great SNR.

There are clear technical differences between separates vs AVR. But theoretically, they should all sound extremely similar in Direct or Bypass mode, especially when the AVR is being used only as a preamp, not amp.

 

[RichB]

We could sure use him chiming in.

Freqency Response seems to be the salient factor according to his audio/acoustic research.

That’s why I would disagree w/ M Code on what Floyd Toole would say regarding the sound quality of AVR vs Separates since both AVR and Separates have ruler-flat FR among other things like extremely low THD and great SNR.

There are clear technical differences between separates vs AVR. But theoretically, they should all sound extremely similar in Direct or Bypass mode.

They sound the same, except perhaps driving speakers.

There can be current limiting and other protection circuits which impede their performance driving real loads. That does not mean all AVRs or amps have these feature(s).

- Rich

 

[PENG]

We could sure use him chiming in.

Freqency Response seems to be the salient factor according to his audio/acoustic research.

That’s why I would disagree w/ M Code on what Floyd Toole would say regarding the sound quality of AVR vs Separates since both AVR and Separates have ruler-flat FR among other things like extremely low THD and great SNR.

There are clear technical differences between separates vs AVR. But theoretically, they should all sound extremely similar in Direct or Bypass mode.

I have not bought his book only because I have been too busy to read anything that serious, but I will eventually. However, in all the free videos and presentations of his that I have read/viewed, I do not recall him commenting on the significance of AVRs and separates on sound quality, but he certainly had commented on the importance of the quality of the loudspeakers and that if they perform well (Free from nasty resonances, flat FR) on both on and off axis)in anechoic chamber, they would sound good.

 

[RichB]

Yeah, and I assume both you and @RichB have seen my REW plots comparing my AVR-3805 with my other preamp/amp separates combos including the Halo A21/Cambridge Audio 840E. Using smoothing as low as 1/48 the graphs practically overlaps each other almost perfectly. Even without smoothing they look practically the same. Yet if you compare the same plots of different speakers, or the same speakers but at a slightly different toe in angle or distance, the difference between the plots would be obvious to naked eyes.

So if flat FR is most critical (Dr. Floyd seemed to be saying just that..), keep in mind on/off axis makes no difference to most electronics, and that THD+N for the DUTs involved are not an issue, how the heck could one pass a SBT, let alone SBT using my DUTs in my room? That is, regardless of Dr. F.Toole's assertion without elaborating the details that those little Omni mic, Umik-1 mics etc., aren't really measuring what we we are hearing. I am sure he's right in the related theories and even in practice, but if the mic tell us FR is flat under a certain condition, even though our ears won't hear the same, but so what, as long as our goal is flat FR right? Never mind mics, we all hear differently too, and have different taste. That's why I think Dr. Toole is most likely right in theory, but I do find him contradicting himself on occasions (not theories, but only in his talks on topics related to REQs, psychoacoustics such as human preference to different speakers etc.) in his various video I have seen so far.

Using Dirac and a curtain set to not process above 30Hz, I can produce measurements that mirror yours. Little, if any, difference between pure direct and Dirac engaged. However, playing 2-channel music and sitting 8 feet from each speaker, the change in the soundstage is obvious.

I don't know what to make of it. Perhaps, there is some other processing, timing related involved. I have no idea what processing is still engaged when the curtain is set this low, but something is happening. Until, I find a processor that does not behave this way, I will consider REQ a cost/benefit.

With PEQ, I discern little or no difference in the sound-stage when limited to below 60Hz. At this point, I prefer using PEQ to REQ since it seems to have less side-effects and is backed up by after measurements.

- Rich

 

[PENG]

There can be current limiting and other protection circuits which impede their performance driving read loads. That does not mean all AVRs or amps have these feature(s).

- Rich

Of course, but in my comparisons, effects of such factor would show up in the REW plots. I have also measured the current output capability of my amps including the Bryston, Adcom, Anthem, Marantz and the A21, the AVR-3805 and 4308 absolutely matched the power amps up to spl I could not withstand. My measurements would sort of, and in general, collaborate with bench tests of some AVRs and separates by different magazine labs, and AH's. Based on my documented records, 100 WPC will go a long way in medium sized room even when 20-23 dB peaks found (rarely) in some music are factored in.

 

[PENG]

Using Dirac and a curtain set to not process above 30Hz, I can produce measurements that mirror yours. Little, if any, difference between pure direct and Dirac engaged. However, playing 2-channel music and sitting 8 feet from each speaker, the change in the soundstage is obvious.

I don't know what to make of it. Perhaps, there is some other processing, timing related involved. I have no idea what processing is still engaged when the curtain is set this low, but something is happening. Until, I find a processor that does not behave this way, I will consider REQ a cost/benefit.

With PEQ, I discern little or no difference in the sound-stage when limited to below 60Hz. At this point, I prefer using PEQ to REQ since it seems to has less side-effects and is backed up by after measurements.

- Rich

My REW plots were done to compare AVRs, preamps, power amps, and DACs, all done in pure direct with no processing that I was aware of. Have you tried that too, again, no Dirac or any other REQ, no PEQ, tone control, nothing, just pure direct using analog connections from DAC out to AVR/Preamp/amp analog inputs only?

 

[RichB]

Of course, but in my comparisons, effects of such factor would show up in the REW plots. I have also measured the current output capability of my amps including the Bryston, Adcom, Anthem, Marantz and the A21, the AVR-3805 and 4308 absolutely matched the power amps up to spl I could not withstand. My measurements would sort of, and in general, collaborate with bench tests of some AVRs and separates by different magazine labs, and AH's. Based on my documented records, 100 WPC will go a long way in medium sized room even when 20-23 dB peaks found (rarely) in some music are factored in.

100 WPC that remained linear and at low distortion driving any speaker load would sound the same and perhaps the Denon’s do. You have conducted enough experiments at your home to conclude that in this environment these components are sufficient.

I can say that the comparing the Yamaha RX-A820 100WPC to the ATI AT522NC (level matched) was easily identifiable in SBT's. The RX-A820 was also adding another pre-amplification so there are more variables and my speakers are not the same.

I hauled the AT6002 down for a quick compare and they sounded different enough, that I have decided to compare two 200 WPC amps from ATI: the AT522NC (NCore) and the AT4002 (fully balanced class A/B).
I should be able to get a good volume match using the fully balanced Oppo HA-1 as a preamp.

- Rich

 

[RichB]

My REW plots were done to compare AVRs, preamps, power amps, and DACs, all done in pure direct with no processing that I was aware of. Have you tried that too, again, no Dirac or any other REQ, no PEQ, tone control, nothing, just pure direct using analog connections from DAC out to AVR/Preamp/amp analog inputs only?

I only do comparissons with Pure Direct and all processing disabled. For the AT522NC/RX-A820, the A820 was in Pure Direct and amps set to 8 ohms (after the factory reset).

- Rich

 

[AcuDefTechGuy]

They sound the same, except perhaps driving speakers.

There can be current limiting and other protection circuits which impede their performance driving real loads. That does not mean all AVRs or amps have these feature(s).

- Rich

I modified to say if the AVR is being used as a Preamp, not amp. So preamp-only comparison between AVR and Pre-Pro using the exact same amp and speakers and everything else.

 

[RichB]

I modified to say if the AVR is being used as a Preamp, not amp. So preamp-only comparison between AVR and Pre-Pro using the exact same amp and speakers and everything else.

I still use an Onkyo 505 (I think) AVR driving 4 Revel M20 and C20's in a vacation home. The Sunfire 200 WPC amp is approaching 20 years so it may be time to get ready for a replacement.

The Onkyo is fine and was not expensive but I would rather not have the amps becuase the unit gets hot in the cabinet. AVR's with 7.1 analog outs are priced higher and I would still consider them for this application if they would allow me to turn off the amps to reduce heat. I already have fans but this house does not have AC so heat is not easily controlled. That is also why class-D is of interest here.

I'd be more inclined to buy one of these low cost products:

• Emotiva MC700: https://emotiva.com/collections/processors/products/mc-700
  
  
• Outlaw Audio 976: https://www.outlawaudio.com/products/976.html

The Outlaw has good PEQ options.
The only feature that is missing is Bluetooth/Airplay.
I know you like DEQ so those are not options for you.

- Rich

 

[M Code]

We could sure use him chiming in.

Freqency Response seems to be the salient factor according to his audio/acoustic research.

That’s why I would disagree w/ M Code on what Floyd Toole would say regarding the sound quality of AVR vs Separates since both AVR and Separates have ruler-flat FR among other things like extremely low THD and great SNR.

There are clear technical differences between separates vs AVR. But theoretically, they should all sound extremely similar in Direct or Bypass mode, especially when the AVR is being used only as a preamp, not amp.

The audible differences between separate component amplifiers and AVR amplifier sections can generally be attributed to how well the respective output stage interfaces with the loudspeaker. Component amplifiers typically have more over-design and are overbuilt compared to than an AVR, since the AVR is built on tighter budget to meet market price. Since the AVR's amplifier output stage is usually lower power and less robust, the design engineer sets the protection and thermal circuits tighter to maintain product reliability. And the audible byproducts of having the protection circuit activated is what causes its sonic performance to hampered. Note that Dr.Toole's expertise is more focused on the listening room and interaction of a given loudspeaker especially for positioning and room furnishings/layout...

Just my $0.02...

 

[M Code]

We could sure use him chiming in.

Freqency Response seems to be the salient factor according to his audio/acoustic research.

Correct...
Dr.Toole clearly states in his books, papers and lectures a very crucial factor is a loudspeaker's frequency response both on-axis and off-axis. Thats why his research studies often relate to loudspeaker x-over design, driver materials and cabinet design. For example, an often overlooked spec is a loudspeaker's frequency response on-axis & off-axis... Every loudspeaker driver element be it a woofer, mid, tweeter has an optimum frequency response capability, and if a loudspeaker driver's x-over frequency point is too high it begins to beam reducing frequency response while simultaneously decreasing its dispersion pattern. The optimum goal is for the loudspeaker's system frequency response to be flat on a power basis, here the physics for loudspeakers follow similar to light as the wave length increases its dispersion narrows.

Thats one reason why the later Harman developed Room EQ software system included (2) measurements 1 for near-field and 1 for far-field. Note that I don't want to speak for Dr.Toole and his superb expertise for audio/acoustics but working closely with him for many years on various design & research projects some of his great experience/knowledge has rubbed off to me...

Though electronics including amplifiers are certainly a contributing factor, its ranks below the listening room and loudspeaker in importance.


Just my $0.02...

 

[KEW]

I can say that the comparing the Yamaha RX-A820 100WPC to the ATI AT522NC (level matched) was easily identifiable in SBT's. The RX-A820 was also adding another pre-amplification so there are more variables and my speakers are not the same.

I don't know, but I might think the A820 and the A860 may have similar amp sections.

When Gene reviewed the A860, he found the amp section underwhelming!
(so poor that he opted not to even measure CFP-BW into 4 Ohms!)
He did measure 1kHz powersweep for 5 channels at 8 Ohms and got 37WPC at 1% distortion... and he commented:

While the RX-A860 is fine driving small bass-managed 8 ohm speakers, this is the first time I'd actually caution people against using 4 ohm speakers or even running 8 ohm tower speakers on the “large” setting. I’ve never said this about a Yamaha before. It's sad that a $400 predecessor from the very same company offered a more robust amp and power section than this product, which has AVENTAGE moniker stamped on it.

While the $900 A860 is (on paper) a 100WPC avr, it is clearly a very poor one!
For comparison purposes, the $1000 Denon AVR-X3300W is a 105WPC avr that provides 88.6 WPC 1kHz into 5 channels at 8 Ohms at 1% distortion.

https://www.audioholics.com/av-receiver-reviews/yamaha-rx-a860/conclusion

https://www.audioholics.com/av-receiver-reviews/denon-avr-x3300w-1/measurements

I don't know what speakers you were driving, but could this reasonably account for the difference you heard?

 

[RichB]

I don't know, but I might think the A820 and the A860 may have similar amp sections.

When Gene reviewed the A860, he found the amp section underwhelming!
(so poor that he opted not to even measure CFP-BW into 4 Ohms!)
He did measure 1kHz powersweep for 5 channels at 8 Ohms and got 37WPC at 1% distortion... and he commented:


While the $900 A860 is (on paper) a 100WPC avr, it is clearly a very poor one!
For comparison purposes, the $1000 Denon AVR-X3300W is a 105WPC avr that provides 88.6 WPC 1kHz into 5 channels at 8 Ohms at 1% distortion.

https://www.audioholics.com/av-receiver-reviews/yamaha-rx-a860/conclusion

https://www.audioholics.com/av-receiver-reviews/denon-avr-x3300w-1/measurements

I don't know what speakers you were driving, but could this reasonably account for the difference you heard?

Yes, it could very well be the A820's amp section and I have been careful to point out that this is not an indictment of all AVRs. The Revel M20 bookshelf are not particularly efficient nor easy to drive:

My comparissons were at relatively low voltage which should not have exceeded a couple of watts driving the M20's.

Here are the Specs from Yamaha:

RX-A820 AV Receiver
Amplifier Section Channel 7.2
Rated Output Power (1kHz, 1ch driven) 160W (4 ohms, 0.9 % THD [European Model]), 130W (8 ohms, 0.9% THD)
Rated Output Power (20Hz-20kHz, 2ch driven) 100W (8ohms, 0.09% THD)
Maximum Effective Output Power (1kHz, 1ch driven) (JEITA) 160W (8ohms, 10% THD)
Dynamic Power per Channel (8/6/4/2 ohms) 140/180/210/250W

Nothing in the specs indicate that there should be any problem driving the M20s at low to moderate listening levels. I suspect nothing in the A860 specs indicated that AH would measure poor performance.

If SV did the review, they would have measured 1K sweeps into 4 and 8 ohm loads and I suspect there would be no problem. The student is tought the tests. Some implementations are created to squeak by.

AH detected issues with the A860 but others did not. That is the problem with specs. After a while, manufacturers become known for over-designing and therefore their spe

I had a similar experience when my brother-in-law was driving his B&W CM10's with a Pioneer Elite SC-07 Class-D (ICE) AVR. I did not like it and we replaced it with Outlaw M2200 mono-blocks directly driven by an Oppo BDP-105. Later I found that AH/Gene found issues when driving 4 ohm loads.

The good news in these cases is that if the A820 is like the A860, then it is possible to measure issues with products that have real-world performance issues. There have been amps that AH measures well did not fare as well in their listening sessions. This has also occurred

- Rich

 

[MrBoat]

I have quite a few different amps. Pass DIY mono-blocks, Adcom GFA-5500/preamp, old integrated, and two Denon 3805's and all are for music. There is a reason I bought two of the Denons. Between Pure Direct, coupled with bass management, I have been pretty much able to emulate any type system, including speakers that use wide band drivers. If they told me I had to trade in my old faithful 3805's in exchange for new separates, I'd tell them to pound sand.

Now in spite of the most apparent and audible difference being the different speakers, what all the amps have in common is cause for being anal with the source material above all else. The audible (subtle) differences between amps ends up becoming moot, comparatively. Everything else resides in a time frame of about 20 minutes with user break-in, otherwise.

Oddly enough, my audio nemesis, who insists on using my paltry efforts as a benchmark to his system, which has about $30k more invested, has to constantly be reminded that he is up against a dated, 2 channel system with mostly DIY components and analog at that.

 

[RichB]

^ It's a good thing your focusing on audio because when you get into new 4K UHD displays, the industry makes life very difficult.

Emotiva is targeting May 1'st for the XMC-1 user installable HDMI 2.2b board ($399). I know they have time-table issues but this is really a boon.

- Rich

 

[lovinthehd]

I'd also like to point out that RichB's comparo with the Yamaha avr was using the avr as a power amp, set to +8 with the volume dial, especially with the comments about protective circuits kicking in....

 

[PENG]

Yes, it could very well be the A820's amp section and I have been careful to point out that this is not an indictment of all AVRs. The Revel M20 bookshelf are not particularly efficient nor easy to drive:

View attachment 24056

My comparissons were at relatively low voltage which should not have exceeded a couple of watts driving the M20's.

Here are the Specs from Yamaha:

RX-A820 AV Receiver
Amplifier Section Channel 7.2
Rated Output Power (1kHz, 1ch driven) 160W (4 ohms, 0.9 % THD [European Model]), 130W (8 ohms, 0.9% THD)
Rated Output Power (20Hz-20kHz, 2ch driven) 100W (8ohms, 0.09% THD)
Maximum Effective Output Power (1kHz, 1ch driven) (JEITA) 160W (8ohms, 10% THD)
Dynamic Power per Channel (8/6/4/2 ohms) 140/180/210/250W

Nothing in the specs indicate that there should be any problem driving the M20s at low to moderate listening levels. I suspect nothing in the A860 specs indicated that AH would measure poor performance.

If SV did the review, they would have measured 1K sweeps into 4 and 8 ohm loads and I suspect there would be no problem. The student is tought the tests. Some implementations are created to squeak by.

AH detected issues with the A860 but others did not. That is the problem with specs. After a while, manufacturers become known for over-designing and therefore their spe

I had a similar experience when my brother-in-law was driving his B&W CM10's with a Pioneer Elite SC-07 Class-D (ICE) AVR. I did not like it and we replaced it with Outlaw M2200 mono-blocks directly driven by an Oppo BDP-105. Later I found that AH/Gene found issues when driving 4 ohm loads.

The good news in these cases is that if the A820 is like the A860, then it is possible to measure issues with products that have real-world performance issues. There have been amps that AH measures well did not fare as well in their listening sessions. This has also occurred

- Rich

It has been a while so I no longer remember the details, other than we have had some back and forth on whether that little RX-A820 was clipping driving the M20s, even if the average power output was calculated to be only a few watts. I seemed to recall the way you did it, I figured the A820 could possibly be clipping if there were high enough peaks in the music contents used for the tests. If power wasn't the issue then that AVR should probably be avoided based on your comparison test results. If you do the same comparison but substitute the RX-A820 with a Denon AVR-X3400H or Marantz SR5012, I doubt the difference would be audibly significant, even for your discerning ears. I am not saying you couldn't hear any difference, but in a DBT I think it would be more subtle than using the A820.

 

[MrBoat]

^ It's a good thing your focusing on audio because when you get into new 4K UHD displays, the industry makes life very difficult.

Emotiva is targeting May 1'st for the XMC-1 user installable HDMI 2.2b board ($399). I know they have time-table issues but this is really a boon.

- Rich

I would never get into any of that. I've been impressed by very little since CDs were invented. Video, and it's effects for some reason, has become so utterly predictable. I'd be over it before I ever managed to feel I got my $'s worth from it.