Ethan Winer's definition of Audio transparency

[amirm]

I would say that in the analog world, we have all the measurment capability we need, yes, manufacturers dont post it all, but audio designers of long standing, such as Nelson Pass, who holds numerous patents on eliminating all distortions in analog signals, have got us amplifiers that are at 0.00001% and below, and you cant get there without measuring in the first place..........digital stuff, OK, not my strong suit of expertise ATT.

Tom

I just got done saying THD is blind to psychoacoustics and you give me a THD number Tom . One of the best tests of amplifiers I have seen is the one that backs out its gain and compares it to its input in a blind AB test. Such tests has shown that hardly any amp is transparent to its input. It is this kind of innovation I like to see to advance our understanding.

Going back to audio compression, through full understanding of how they work, we have developed a set of music clips that are highly revealing. We use them to advance the art there. Yet if I asked people what is the best piece music is for testing amps, CD players and such, I would not get any consensus or even a defensible list.

 

[MylesBAstor]

All I said was ss units have the better measurments. And I said that the audio manufacturers, do not provide and will never provide adequate measurments, because frankly, they all look pretty bad, especially using spectrum analysis, two tones or many tones.

Tom

Obviously not the ones that count. You still have not addressed issues specific to SS circuit design. This is the same old tired argument presented at the dawn of the SS era. Yes SS had lower THD distortion, yet was unlistenable. Briitle. Hard. Metallic. These were all adjectives used to describe the sound of the earliest ss amps. If you really want to hear how bad this gear was, listen to recordings that were recorded with the first gens of ss equipment. (well maybe you liked it.). So why was that? You darn well know that there were other more important distortions in ss gear that wasn't being accounted for. (not to mention which order distortions were the worst.)

This subject has been flogged to death on WBF in other threads-- to lower THD to the levels you desire creates far worse problems and distortions eg. The cure was worse than the disease. Also show me some data that shows how far down you need to reduce discretion levels so it's not perceptible. Finally, you have yet to address the issue of switching distortion in transistors.

Show me ONE original, first or second generation SS amplifier that's not in a landfill. Not to mention they were far from reliable.

 

[opus111]

I have had the opposite experience with LM4562 - I bought a pair of active speakers which used them in the crossovers. But I found the sound harsh at the top end. Rolling out the LM4562 in favour of a (generally considered inferior) NE5532 improved the sound - the harshness was reduced but not eliminated. I have a hypothesis for what's happening but I dont have any measurement data.

 

[amirm]

LOL...but , but this IS the transparency thread..!!
I tried two lm4562 op amps in series, compared with a piece of wire. Output was about 3Vrms. I could not tell when the two amps were in the circuit or the wire, they are audibly transparent to me. And IMD is at 0.00005% at under 4Vrms output. THD is just a tad higher, i think 0.00003%.

At some point, we have to cave in and admit that something is audibly transparent. OK, I have to, the rest of you don't

Tom

I am there too but I just can't set it as low as some do. In my first post, I said I shoot for honest 16 bits of fidelity. In this day and age, there is no excuse to do worse. Yet folks constantly defend much worse performance. Here is a simple graph on my audio jitter article:

That level of jitter at -80 db exists in a ton of mass market gear over HDMI. Many people argue that jtiter is inaudible and hence the bar is -80 db. I think that gives a free pass to manufacturers to build sub-par equipment for no good reason. The DACs inside those AVRs have specs better near 100 db yet we reward them with an in-circuit performance that is 20 dB lower. Why?

 

[Phelonious Ponk]

Miles, I have no problem with your preference for tube gear, however, really, all this stuff you are saying...it applies to Krell, bryston, mark levinson, burmeister, dartezeel blah blah, etc nowadays?

It doesn't even apply to a Pioneer AV receiver these days.

Yes SS had lower THD distortion, yet was unlistenable. Briitle. Hard. Metallic. These were all adjectives used to describe the sound of the earliest ss amps.

These are all adjectives used to describe clarity and precision by Audiophiles now. Add cold and analytical and were about done. I don't know if we have the measurements to to determine amplifier transparency; if we do, I'm pretty sure we aren't using them. Why? Because it's down to such hair-splitting minutiae that no one but us, and studio pros, cares anymore.

Tim

 

[jkeny]

Anecdotally interesting. Glad you did it. But like my own listening, it's evidence a personal experience only.

Yes, Tim, but can I point out that you are so convinced by your personal experiences of blind testing that you are an advocate for it. This doesn't seem consistent with what you are advising me to do?

It was published in 1984, which doesn't have anything to do with the quality of the research, but could have everything to do with the quality of the digital gear and the audibility of it's distortions.

Tim

Yes, I'm open to being persuaded by this if you can show that the digital equipment used is in some way audibly deficient.

 

[Phelonious Ponk]

Yes, Tim, but can I point out that you are so convinced by your personal experiences of blind testing that you are an advocate for it.

Sure. What I learned from my little round of blind listening sessions, more than anything else, is that these differences between electronics that Audiophiles describe as dramatic, night and day, the difference between music and crap...fill in your favorite hyperbole...they're not. They're not even close. Even if what I cannot hear is audible to someone else, we are splitting the finest of hairs here. And that's OK, this is "high-end" this is "what's best." We're going for the last couple of percentage points. But the secret I uncovered is that in that last couple of percentage points it really is subjective. Even when you can hear it, and consistently tell whether X is A or B, if you're honest with yourself, you'll have a hell of a time telling which one is better. It's just too close.

So what I learned is that "the Zeuss DAC opened up a world of inner detail I hadn't heard before, dramatically deepened the micro dynamics, broadened the soundstage 25% and was so...musical" is illusion. It's a delicate balance of expectation bias and BS. So yes, unless you want to pay $500 for a DAC and $4500 for an illusion, I advocate blind listening. You might reach the same conclusions.

This doesn't seem consistent with what you are advising me to do?

I'm advising you not to do a public victory dance, singing nanny nanny booboo in our little cyber end zone, because Ethan's transparent noise floor requires 17 bits instead of 16, when you're own "evidence" is statistically valueless, John.

Yes, I'm open to being persuaded by this if you can show that the digital equipment used is in some way audibly deficient.

Well, all I have is a quote from a study. I don't know what equipment was used. But I know digital audio has made some progress since 1984. I'll bet you've noticed that yourself.

Tim

 

[MylesBAstor]

Miles, I have no problem with your preference for tube gear, however, really, all this stuff you are saying...it applies to Krell, bryston, mark levinson, burmeister, dartezeel blah blah, etc nowadays?

We have measurments (what manufacturers provide us and testers provide us) and we have people ears.......you can measure an amps replication by itself, but you can't hear an amps output by itself, there must be a transducer involved...measurements trounce ears, prefernces trounce measurments...

the point here is, your ears give out, when the distortions are below some level, Ethan says 80db or may be 100db, in any case, at some point the distortions go off the radar screen, can we not agree?

Tom

Tom:

Your original post had nothing to do with preference; you simply stated that because of timing errors, tube amplifiers could never be SOTA, ignoring any issues that ss might have.
Clearly these issues still exist as people continually try to eke as much Class A out of their A/B amps as possible--or like Nelson, build humongous Class A amplifier (might be add that aren't cheap!).

To each their own; there are probably even a few ss amplifiers that I could live with today.

 

[Phelonious Ponk]

That level of jitter at -80 db exists in a ton of mass market gear over HDMI.

Yes, but is anyone using HDMI for audio? I'll be showing my prejudices here, but frankly, while I'd like to see them hit a modern standard, even in AV receivers through HDMI, when listening to a home theater system playing hyper realistic fantasy shoot-outs, explosions and car chases, jitter at 80 dB is not what's going to drive me out of the room. I know I'm not typical here, but I wish they'd give me back the dynamic range in my music and compress the hell out of the movies. I'm tired of riding the volume on the remote control, and I don't have a surround sound system.

Tim

 

[Robh3606]

Yes, but is anyone using HDMI for audio?

Hello Tim

Of course they are. There is plenty of stuff out on Blue ray and if you have a universal player hooked up through HDMI then your SACD's and DVD-A are coming through the same pipe. I have a OPPO BDP-93 and have an Integra 9.8 Pre Pro. No idea what the jitter specs are through the HDMI port. It sounds fine to me.

Rob

 

[jkeny]

..........
I'm advising you not to do a public victory dance, singing nanny nanny booboo in our little cyber end zone, because Ethan's transparent noise floor requires 17 bits instead of 16, when you're own "evidence" is statistically valueless, John.

No victory dance - I posted files which were a test of pre-echo audibility - nothing to do with noise floor. I reported listener results for these files. I happened to mention that the files contradicted Ethan's criteria for audio transparency (Frequency Response: 20 hz to 20 Khz +/- 0.1 dB). Ethan introduced "artefacts" as being down @ -39dB as a counter argument to the validity of these results. These are not "artefacts" so his argument is mute.

The question of the SNR is another issue altogether & one that Amir raised with Ethan. I suggest it might be best to address your points to Amir & not me as they seem to be irrelevant to what I'm talking about - pre-echo audibility. The issue of Frequency Response: 20 hz to 20 Khz +/- 0.1 dB has still not been clarified, I believe.

 

[Phelonious Ponk]

Hello Tim

Of course they are. There is plenty of stuff out on Blue ray and if you have a universal player hooked up through HDMI then your SACD's and DVD-A are coming through the same pipe. I have a OPPO BDP-93 and have an Integra 9.8 Pre Pro. No idea what the jitter specs are through the HDMI port. It sounds fine to me.

Rob

And I've listened to plenty of audio through HDMI when I was in the business. It sounded fine to me too. But if it is an audible problem, the simple solution seems to be to do your critical listening through another cable, leave HDMI for the noisy action movies.

Tim

 

[Phelonious Ponk]

No victory dance - I posted files which were a test of pre-echo audibility - nothing to do with noise floor. I reported listener results for these files. I happened to mention that the files contradicted Ethan's criteria for audio transparency (Frequency Response: 20 hz to 20 Khz +/- 0.1 dB). Ethan introduced "artefacts" as being down @ -39dB as a counter argument to the validity of these results. These are not "artefacts" so his argument is mute.

The question of the SNR is another issue altogether & one that Amir raised with Ethan. I suggest it might be best to address your points to Amir & not me as they seem to be irrelevant to what I'm talking about - pre-echo audibility. The issue of Frequency Response: 20 hz to 20 Khz +/- 0.1 dB has still not been clarified, I believe.

The victory dance I'm referring to is this whole thread, John. You caught him in an error and launched a thread, with his name in the title, to call attention to it and question his entire philosophy. Somebody should close this damn thing. It serves no purpose.

Tim

 

[jkeny]

The victory dance I'm referring to is this whole thread, John. You caught him in an error and launched a thread, with his name in the title, to call attention to it and question his entire philosophy. Somebody should close this damn thing. It serves no purpose.

Tim

Again I have to correct this. I introduced the concept & files on the 24/192 thread as a possible reason for high-res files sounding better - pre-echo in linear phase, equiripple, interpolation filters. I later mentioned that the audibility of the results contravened one of Ethan's criteria for audio transparency (not the noise floor, SNR one). As the ensuing discussion became a pollution to that thread I branched it to here. You are free tp put your own interpretation on these facts.

I didn't catch him in an error & I don't know if Ethan even admits to the error that you are suggesting he made? It was Amir's questioning that uncovered the movable nature of the stated -100dB SNR, not me.

It is your interpretation that I question Ethan's whole philosophy - I don't. I am very favourable towards measurements in audio. As I mentioned already, Amir's link to Harmon's speaker measurements correlation with listening is very compelling & I hav espent some time investigating where I could purchase the Infinity Primus speakers but unfortunately they are not available in Europe, it would seem

My interest is in having accurate measurements that clue us into the quality of audio reproduction - I'm sure you are striving for the same goal! Simply just having measurements is not satisfactory to me although it might appear that it is acceptable to some?

 

[Steve williams]

I'm a very big fan of WBF. I read a lot and learn a lot. I don't post as often as I should so perhaps I have little or no influence here, but here goes anyway.

One of the things I love best about this forum is that people are respectful and that divergent views are discussed and tolerated. This sets WBF apart from pretty much every other "audiophile" forum out there. Can we please keep it that way?

Great thought for the day. I would like to keep it that way

 

[amirm]

Yes, but is anyone using HDMI for audio?

Almost everyone does these days if you count the entire market. Folks use the same transport for audio and video and the latter requires HDMI. I don't know of hardly anyone who uses two interconnects even though they should.

I'll be showing my prejudices here, but frankly, while I'd like to see them hit a modern standard, even in AV receivers through HDMI, when listening to a home theater system playing hyper realistic fantasy shoot-outs, explosions and car chases, jitter at 80 dB is not what's going to drive me out of the room. I know I'm not typical here, but I wish they'd give me back the dynamic range in my music and compress the hell out of the movies. I'm tired of riding the volume on the remote control, and I don't have a surround sound system.

Tim

I am not picky about movie sound either. But per above, HDMI has all but replaced other forms of digital interconnect in people's systems. Every other day I see someone asking what connection they should use from their PC to their AVR and even that advice is to use HDMI. Just read a thread on this a few days ago on AVS. People absolutely believe digital is digital and don't think twice about any differences here.

 

[Robh3606]

People absolutely believe digital is digital and don't think twice about any differences here.

Hello Amir

Maybe this isn't the right thread but what causes the jitter to be higher in HDMI vs Coax or Optical??

Rob

 

[Ethan Winer]

This rules out 16 bit CD format as being transparent which ironically, agrees with Bob Stuart assertion:

I missed the 120 dB SPL reference in the caption the first time around. That is ridiculously loud, unless it refers to peak level rather than average. But if it's peak, the crest factor is crucial. So I'll assume average, since that's how noise is usually assessed. To put this in context, 100 dB SPL is ear-splitting loud. That's the loudest I can stand to listen to normal music in my living room, and I like loud music! So once you scale everything down by 20 dB to make this graph practical, even 16 bits is enough to be transparent.

Just curious, where is this graph from? A book? A web article? I'd like to read more about the context.

--Ethan

 

[Phelonious Ponk]

Almost everyone does these days if you count the entire market. Folks use the same transport for audio and video and the latter requires HDMI. I don't know of hardly anyone who uses two interconnects even though they should.


I am not picky about movie sound either. But per above, HDMI has all but replaced other forms of digital interconnect in people's systems. Every other day I see someone asking what connection they should use from their PC to their AVR and even that advice is to use HDMI. Just read a thread on this a few days ago on AVS. People absolutely believe digital is digital and don't think twice about any differences here.

I know you're right about HDMI, Amir. Personally, I use it to connect my son's Playstation 3 to his Panasonic plasma. That's all the HDMI in my house. I know you're right that most people absolutley believe digital is digital. The funny thing is, for most people, that's right. They never listen critically to their sound. Most of the time music plays in the background as the soundtrack to their lives. Maybe once a week when the wife and kids are out of the house for a couple of hours dad might crank it up and really enjoy the music...but that's what he's listening to critically -- the music. He is not going to hear the difference between the jitter in an HDMI cable and coax. We are not normal.

Tim

 

[Robh3606]

He is not going to hear the difference between the jitter in an HDMI cable and coax. We are not normal.

Well what exactly do you listen for?? How do you hear jitter?? What kind of audible artifacts does it introduce where you would could reliably say, Ah Jitter there it is.

Rob