Effect of DAC performance on Audyssey?

[Lattiboy]

Hello all,

I currently run a Marantz 6012 (which, from what I understand, is quite similar to the Denon 3500H). I was reading the ASR review of the 3500H where Amir claims it has a terribly noisy DAC with overall sub par performance.

This didn’t concern me much as I use an external DAC and the rca ins for music and he said the analog was uneffected by the digitization, but now that I’ve come to rely on Audyssey, I wondered if this poor performing DAC would be used in the Audyssey system.

I only ask because I think I can get a good deal on the 3600H, which is apparently totally re worked with a DAC section that is one of the better performing AVRs anywhere near the price. I’d prefer not to redo all my calibration and go through selling the 6012.

Thanks!

 

[Pogre]

Have you ever really done a comparison between the DAC in your Marantz and the external one? I'll bet you a dollar you wouldn't be able to consistently pick which was which in a DBT. I'd just use my Marantz and put my money into stuff that will yield audible improvements.

Amir does some great work, but for all of his hyperbole 99.9999% of the stuff he rails at is inaudible to the point of being a complete non-issue. There are so many other things that have far greater impact and can yield much better results than chasing after inaudible specs. If folks put half of the energy into worrying about speaker performance as they do DACs and amps they'd be much further ahead.

 

[lovinthehd]

Might read the various threads on the subject of dac audibility at asr. Amir also gets worked up over jitter which is not much of an issue either, and when he says a dac is terribly noisy he isn't talking about audibility. While it's nice to have better engineering, changing your avr is not likely to make any real audible difference, altho you might convince yourself it did.

 

[PENG]

Hello all,

I currently run a Marantz 6012 (which, from what I understand, is quite similar to the Denon 3500H). I was reading the ASR review of the 3500H where Amir claims it has a terribly noisy DAC with overall sub par performance.

This didn’t concern me much as I use an external DAC and the rca ins for music and he said the analog was uneffected by the digitization, but now that I’ve come to rely on Audyssey, I wondered if this poor performing DAC would be used in the Audyssey system.

I only ask because I think I can get a good deal on the 3600H, which is apparently totally re worked with a DAC section that is one of the better performing AVRs anywhere near the price. I’d prefer not to redo all my calibration and go through selling the 6012.

Thanks!

You have to read Amir's commentary part carefully in order to decipher what would/could actually mean to you, practically speaking in your own application.

For example, at the 1.2 V level (that's how Denon and Marantz rate their preout), your AVR-X3500H actually measured better, that is 95 dB SINAD for the X3500H, vs only 92 dB for the AV7705.

I am using the AV7705 as an example because it is a prepro, so inherently not affected by the power amp section that isn't there. In practical term, I personally between even SINAD as low as 70 dB is not going to make an audible difference for probably 99% of the home HT users.

Think logically, let's say you have a very quiet room with noise floor between 25 to 30 dB. And let's assume your SR6012 would have SINAD 72 dB, that is 3 dB worse than the AV7705, at 1.2 V.

72+25 = 97 dB, that means under the worst case scenario you won't be able to hear distortions unless you listen to SPL above 97 dB from you listening position. ASR is going to measure a SR6014 soon so stay tune, but again based on the measured AV7705, I doubt it would do better than the X3500H, just to prepare you for the worse so that you may be nicely surprised.

Regarding the DAC, as others mentioned, it is not the bottleneck at all!! So don't worry about it.

 

[AcuDefTechGuy]

Are people saying the same thing over on the ASR forum as we're saying here?

Has anyone brought up these excellent points over there?

 

[Pogre]

Are people saying the same thing over on the ASR forum as we're saying here?

Has anyone brought up these excellent points over there?

Well I never made an account over there.

I think they would argue that audible or not, the comments are based on equipment claiming to be sota and it doesn't measure like sota. It's kind of a valid argument, imo.

Tho I am really starting to see the effect this is having on Joe regular when he visits and reads about sad and noisy performance without understanding that noise is still well below the threshold of audibility.

 

[PENG]

Well I never made an account over there.

I think they would argue that audible or not, the comments are based on equipment claiming to be sota and it doesn't measure like sota. It's kind of a valid argument, imo.

Tho I am really starting to see the effect this is having on Joe regular when he visits and reads about sad and noisy performance without understanding that noise is still well below the threshold of audibility.

I made similar arguments over there, but the emotion is so high I doubt I changed any minds. It is just so sad that so many don't have enough understanding of the technical merits and practical meanings of such measurements.
The reviewer sets a high bar because he meant to use those measurements as indicators of engineering excellence,or lack of, but when the high bar is not reached, the resulting comments that followed, and they always followed, tend to mislead the seemingly vast majority, who, again, simply don't understand the numbers well enough, yet they easily, and naturally understood the resulting negative comments well enough, the wrong way..........

The reviewers cautionary notes against some of those same measurements, they included verbiage such as "not that audible, fortunately not that audible, fortunately not very audible, fortunately not audible..." seemed to have always been overlooked, or fallen on deaf ears.

 

[RickyT]

I made similar arguments over there, but the emotion is so high I doubt I changed any minds. It is just so sad that so many don't have enough understanding of the technical merits and practical meanings of such measurements.
The reviewer sets a high bar because he meant to use those measurements as indicators of engineering excellence,or lack of, but when the high bar is not reached, the resulting comments that followed, and they always followed, tend to mislead the seemingly vast majority, who, again, simply don't understand the numbers well enough, yet they easily, and naturally understood the resulting negative comments well enough, the wrong way..........

The reviewers cautionary notes against some of those same measurements, they included verbiage such as "not that audible, fortunately not that audible, fortunately not very audible, fortunately not audible..." seemed to have always been overlooked, or fallen on deaf ears.

I think you have a Denon 4400H: what frequency range are you using XT32?

 

[PENG]

I think you have a Denon 4400H: what frequency range are you using XT32?

I have been using for as low as 200 hz, and as high as 600 hz, but I started playing with the App using the Ratbuddyssey UI to tweak the higher freq. So I think I have it up the 15K now.

 

[AcuDefTechGuy]

I made similar arguments over there, but the emotion is so high I doubt I changed any minds. It is just so sad that so many don't have enough understanding of the technical merits and practical meanings of such measurements.
The reviewer sets a high bar because he meant to use those measurements as indicators of engineering excellence,or lack of, but when the high bar is not reached, the resulting comments that followed, and they always followed, tend to mislead the seemingly vast majority, who, again, simply don't understand the numbers well enough, yet they easily, and naturally understood the resulting negative comments well enough, the wrong way..........

The reviewers cautionary notes against some of those same measurements, they included verbiage such as "not that audible, fortunately not that audible, fortunately not very audible, fortunately not audible..." seemed to have always been overlooked, or fallen on deaf ears.

Speaking of ENGINEERING EXCELLENCE, what do you think about how he got "poor" measurements from the new Emotiva pre-pro at first, which initially meant "poor engineering"?

Then almost overnight, Emotiva tweaked something in the pre-pro and it became "excellent", which means excellent engineering?

How can a machine go from poor engineering to excellent engineering overnight with some tweak?

It just seems so "superficial" or fake to me. Engineering a machine isn't overnight.

So how true are his numbers when predicting "excellent" engineering? Can people like Emotiva just tweak their machines to become excellent engineering overnight?

Is excellent engineering nothing more than some numbers?

And as I've said before, none of these numbers have proven anything in terms of RELIABILITY or Sound quality, which to me are the 2 salient measures of excellent engineering.

 

[AcuDefTechGuy]

I think they would argue that audible or not, the comments are based on equipment claiming to be sota and it doesn't measure like sota. It's kind of a valid argument, imo.

Tho I am really starting to see the effect this is having on Joe regular when he visits and reads about sad and noisy performance without understanding that noise is still well below the threshold of audibility.

SOTA measurements - but does he claim to be the only true standard for SOTA measurements? He said that Gene's 90dB standard is "too low" for him.

Is SOTA measurements solely based on his numbers? What about numbers from Audioholics or S&V or Stereophile?

 

[RickyT]

I have been using for as low as 200 hz, and as high as 600 hz, but I started playing with the App using the Ratbuddyssey UI to tweak the higher freq. So I think I have it up the 15K now.

And XT32 is doing your subwoofer down to ~ 20hz?

On my Marantz 7704, I currently have the non-subwoofer cutoff at 1k hz. The rest of my gikacoustics bass traps / acoustic treatments arrived yesterday. Over the next week or so, I will install those and learn how to use the Ratbuddyssey UI.

 

[Lattiboy]

You have to read Amir's commentary part carefully in order to decipher what would/could actually mean to you, practically speaking in your own application.

For example, at the 1.2 V level (that's how Denon and Marantz rate their preout), your AVR-X3500H actually measured better, that is 95 dB SINAD for the X3500H, vs only 92 dB for the AV7705.

I am using the AV7705 as an example because it is a prepro, so inherently not affected by the power amp section that isn't there. In practical term, I personally between even SINAD as low as 70 dB is not going to make an audible difference for probably 99% of the home HT users.

Think logically, let's say you have a very quiet room with noise floor between 25 to 30 dB. And let's assume your SR6012 would have SINAD 72 dB, that is 3 dB worse than the AV7705, at 1.2 V.

72+25 = 97 dB, that means under the worst case scenario you won't be able to hear distortions unless you listen to SPL above 97 dB from you listening position. ASR is going to measure a SR6014 soon so stay tune, but again based on the measured AV7705, I doubt it would do better than the X3500H, just to prepare you for the worse so that you may be nicely surprised.

Regarding the DAC, as others mentioned, it is not the bottleneck at all!! So don't worry about it.

thanks Peng!!


That 1.2v bit is really critical and not highlighted nearly enough over there! I mean, if you’re looking at dangerous listening levels being where distortion happens, who cares? What actual human buying a consumer AVR is worried about listening at levels that would undoubtedly damage you’re hearing?

I kind of appreciate what Amir does, But that place has the feeling of a cult meet up a lot of the time. It’s incredible the pull that dude has online. I clearly fell into it a bit mysel!

 

[PENG]

Speaking of ENGINEERING EXCELLENCE, what do you think about how he got "poor" measurements from the new Emotiva pre-pro at first, which initially meant "poor engineering"?

Then almost overnight, Emotiva tweaked something in the pre-pro and it became "excellent", which means excellent engineering?

How can a machine go from poor engineering to excellent engineering overnight with some tweak?

It just seems so "superficial" or fake to me. Engineering a machine isn't overnight.

So how true are his numbers when predicting "excellent" engineering? Can people like Emotiva just tweak their machines to become excellent engineering overnight?

Is excellent engineering nothing more than some numbers?

And as I've said before, none of these numbers have proven anything in terms of RELIABILITY or Sound quality, which to me are the 2 salient measures of excellent engineering.

I might have missed that Emo incident, will look into it. He seems to have been testing a lot of devices lately, since I joined anyway. As such, probability is naturally higher that something got missed. In regard to measurements as indicators, that's just indicators, and may not be 100% reliable and/or consistent.

As you know, I prefer more balanced view so I should try to hold a more balanced/neutral approach myself in most things. I don't want to jump to conclusions too quickly, that's just me, can expect others to do the same..

 

[PENG]

And XT32 is doing your subwoofer down to ~ 20hz?

On my Marantz 7704, I currently have the non-subwoofer cutoff at 1k hz. The rest of my gikacoustics bass traps / acoustic treatments arrived yesterday. Over the next week or so, I will install those and learn how to use the Ratbuddyssey UI.

About 12 Hz for the subwoofer, from the look of the curves. For the mains I set XO to 80 Hz. I forgot the Audyssey's low limit, but I think I have it included in the 1st post of my App users thread.

 

[PENG]

thanks Peng!!


That 1.2v bit is really critical and not highlighted nearly enough over there! I mean, if you’re looking at dangerous listening levels being where distortion happens, who cares? What actual human buying a consumer AVR is worried about listening at levels that would undoubtedly damage you’re hearing?

I kind of appreciate what Amir does, But that place has the feeling of a cult meet up a lot of the time. It’s incredible the pull that dude has online. I clearly fell into it a bit mysel!

I tried to wake them up once in a while lol..

 

[lovinthehd]

thanks Peng!!


That 1.2v bit is really critical and not highlighted nearly enough over there! I mean, if you’re looking at dangerous listening levels being where distortion happens, who cares? What actual human buying a consumer AVR is worried about listening at levels that would undoubtedly damage you’re hearing?

I kind of appreciate what Amir does, But that place has the feeling of a cult meet up a lot of the time. It’s incredible the pull that dude has online. I clearly fell into it a bit mysel!

There's a recent thread there about a guy who only wants preouts measured at 1.2V as he feels that's a more typical use. There are a few threads about the inaudibility of the differences in dac spec/testing results, but avrs tend to get dumped on in general because I think you're mostly dealing with a two ch crowd who've recently added digital with external dacs when not using vinyl

ps Anyone remember the Arny K/Amir arguments about jitter?

 

[Pogre]

SOTA measurements - but does he claim to be the only true standard for SOTA measurements? He said that Gene's 90dB standard is "too low" for him.

Is SOTA measurements solely based on his numbers? What about numbers from Audioholics or S&V or Stereophile?

I'm not trying to argue for him. I just know what they'd say and I lack the technical chops to be calling folks out. I'm lying if I act like I don't get what he's saying tho. He does give some context, but it gets lost in all of the negative language. I've always said I'm not a big fan of his hyperbole.

Like I alluded to earlier, I'm not so sure if he's having an overall positive effect. I really am beginning to wonder if a lot of people are starting to get the wrong message. Based on some of the questions I've been seeing lately it seems folks go there, read up and leave convinced that they now own garbage.

 

[RickyT]

Speaking of ENGINEERING EXCELLENCE, what do you think about how he got "poor" measurements from the new Emotiva pre-pro at first, which initially meant "poor engineering"?

Then almost overnight, Emotiva tweaked something in the pre-pro and it became "excellent", which means excellent engineering?

How can a machine go from poor engineering to excellent engineering overnight with some tweak?

It just seems so "superficial" or fake to me. Engineering a machine isn't overnight.

So how true are his numbers when predicting "excellent" engineering? Can people like Emotiva just tweak their machines to become excellent engineering overnight?

Is excellent engineering nothing more than some numbers?

And as I've said before, none of these numbers have proven anything in terms of RELIABILITY or Sound quality, which to me are the 2 salient measures of excellent engineering.

Emotiva collaborated with Amir and ended up fixing the issue with a firmware upgrade....a great outcome for current and future Emotiva owners! In other words, the Emotiva engineering was fine...it seemed be a bug that Amir/ASR unveiled. Who knows how long, or if ever, this would have been fixed if Amir didn't discover it.

Denon engineers have been in recent discussions with Amir regarding the 4700H. In the next couple of weeks, Denon will post something on their Denon blog. IF the issue can fixed via firmware, that'd be another great outcome. If it can't be fixed via firmware, perhaps future models will have the fix.

Amir will also get a 3700H to test soon...that will provide a good comparison of 3700H vs 3600H, and 4700H vs 3700H.

I'm not an EE, I don't know or quote things like dac model numbers, transistors, etc. I leave that to people like Peng or bigguyca on AVS LOL!

 

[PENG]

In regards to the AVS, the resident D&M expert jdsmoothie,


Emotiva collaborated with Amir and ended up fixing the issue with a firmware upgrade....a great outcome for current and future Emotiva owners! In other words, the Emotiva engineering was fine...it seemed be a bug that Amir/ASR unveiled. Who knows how long, or if ever, this would have been fixed if Amir didn't discover it.

Denon engineers have been in recent discussions with Amir regarding the 4700H. In the next couple of weeks, Denon will post something on their Denon blog. IF the issue can fixed via firmware, that'd be another great outcome. If it can't be fixed via firmware, perhaps future models will have the fix.

Amir will also get a 3700H to test soon...that will provide a good comparison of 3700H vs 3600H, and 4700H vs 3700H.

I'm not an EE, I don't know or quote things like dac model numbers, transistors, etc. I leave that to people like Peng or bigguyca on AVS LOL!

Thank you, so in that case it was a happy ending anyway. Hopefully the very bad 32 tone test results for the X4700H is also due to something of the same sort and therefore FW fixable too.