Tubes and Digital Audio

[Travis In Austin]

What the heck does this mean?

"The extra cool thing about this amp is that it has the ability to listen to the room via the principal of reciprocity. The loudspeaker the amp is connected to behaves like a microphone and "listens to the room." That signal is then sent back to the input to be mixed with the forward signal to help generate a more delicious sense of acoustic space by including the room in the final sound in a way other amps can't."

"Delicious sense of acoustic space" sounds like something righ out of the Bose playbook. We will tell you how it sounds, but we don't publish specs. It also sounds like a load of crap to me.

Isn't he saying that because of reciprocity, the microphone effect, that this exists regardless of whether the amp had feedback designed into it? Will not some speakers act as better microphones than others?

[Mallette]

However, new pressings usually come from a digital master so is there any advantage of a vinyl pressed from digital vs a high quality digital file from a digital master?

 

Well, that's a can of worms.  Some say yes, others no.  The sound of vinyl seems to defy science.  I don't really have any bias as it's all about the music to me and I am more interested in how well it is engineered. 

 

I'll leave it at that.  Well engineered music will sound good on any medium, and once you start compressing or otherwise losing information it will continue to sound better than the run of the mill crap.

 

Dave

[Chris A]

However, new pressings usually come from a digital master so is there any advantage of a vinyl pressed from digital vs a high quality digital file from a digital master?

 

Someone that says the phonograph record that is derived from a digital master "sounds better" than the original digital master...well, what can I say?  We are at a impasse: I don't argue, especially if someone declares "...it's my opinion...".  But I do form a subjective opinion of that person's aural judgments, and it's not complimentary.

 

 

Chris

[Mallette]

Woot! Woot!  Off and rolling... And I keep hearing that it's boring around here. 

 

No reason that if tubes take the "edge" off edgy digital that vinyl wouldn't have the same effect, even when played through SS or Class D. 

 

Dave

[Chris A]

Woot! Woot!  Off and rolling... And I keep hearing that it's boring around here.

 

I find it odd that I'm talking to a Klipsch forum moderator in this manner, on this particular subject, and in light of the recent events.

 

Chris

Edited May 13, 2015 by Chris A

[Quiet_Hollow]

I would guess that Tom's point is that there are many other topics to consider just on the subject of amplifiers (...although he does seem to run off the subject toward the end of his list ).

Pretty much what I was getting at.

 

I feel the list is composed of legitimate topics that I'd like to see papers on. The last two, especially.

 

"Holograms and clocks" would be a thorough investigation into how various recording systems preserve channel-to-channel time relationships, what and how much capacity there is to do so,  the audible effects of intereference and resulting distortion of the L/R plane of symmetry, and finally how various, current technologies perform specifically in that regard.

 

"EQ and the air we breathe" would be a discussion of treating an acoustic system as the sum of all the filters involved.....answering questions like, "if these three things sum like this, would there ever be any potential of the group to sound like that. To approach the topic of EQ as what if everything is filtered whether we like it or not, so what constitutes the ideal or "target" filter? To dispense with the notion that EQ is some box or plug-in inserted into the signal chain to somehow bastardize the source material.

Edited May 15, 2015 by Quiet_Hollow

[Chris A]

So Chris based on what you just quoted do you agree that this "microphone/reverb effect" requires that the amplifier be a design using feedback..?

 

No, I believe what Carver said (and I highlighted in brown type) is that all tube amplifiers "do it" to a greater or lesser degree.  This is the idea that if the amplifier presents any impedance to the loudspeaker, that it behaves like a mechanical resistance to any microphonic behavior coming back to the amplifier output terminals and reflecting back to the loudspeaker input - with the acoustics of the room providing the delays.  The efficiency of this transfer of room acoustic reflection energy back to the amplifier's output is a function of the ratio of the output impedance of the amplifier to the input impedance of the loudspeaker.  In my case, that is the 11-19 ohms of the TAD driver, direct coupled.  So the ratio is 1 to 10,  If you have a lower input impedance to your loudspeaker networks or drivers--like 4 ohms for a cone driver on a bass bin, that efficiency of reverb coupling is 1/4.

 

This is just like you mentioned once with a lot of loudspeakers in a demo room providing "damping"--except that it isn't really damping, but rather room-based reverberation that is reinforced by the amplifier-loudspeaker circuits themselves.

 

What Carver apparently is saying is that, if you add current feedback to a tube-based amplifier with a controlled output impedance level (1.5-1.9 ohm), this reverb effect can be "enhanced" by the gain of the amplifier, which is apparently what he is doing in his tube amplifiers. In effect, what Carver is saying is that you can design a high output Z amplifier with current feedback that resists the microphonic loudspeaker return energy and actually provides gain on its output to "hold the drivers" in place against the acoustics of the returning room reflections - especially in the corners of a room with corner horns - just like the effects of using corner bass traps enhance the effects of absorption.

 

Since the output impedance of almost all junction transistor amplifiers (BJT) is almost zero due to the use of multiple gain stages and the use of global feedback, they do not produce a reverb effect that is audible. 

 

However, Nelson Pass uses single stage (Darlington pair) FET-type amplifiers--like the First Watt F3--that have about 1 ohm of output impedance, which is very uncharacteristic of transistor amplifiers. If you look at his SIT amplifiers (SIT-1 and SIT-2), the output impedance is even higher toward 2 ohm levels, and he uses a single gain stage without feedback.

 

You already know how I feel about this effect after owning planar dipoles for 20 years and becoming fatigued by the backwave-front wall reflections,  relative to the clean/unadulterated sound of the Jubilee without an artificial depth of field being splashed around (like Bose 901s). The "listening to two rooms" effect of the recording venue plus the listening room of the sound reproduction system - this second room acoustic effect needs to be controlled to a lower reflection level in order to hear the acoustics of the original recording room/venue. 

 

For recordings that were made in separate rooms and later recombined into a single stereo image (i.e., using multitrack recorders), the loss of the acoustics of an original recording venue is great and the sense of space is lost. 

 

[For many people reading this, there is a significant percentage of those that mainly listen to popular music recorded using multitrack recorders and separate dead acoustic recording booths, so these people rarely listen to "first room" acoustic reflections, and instead add their own with "second room" reflections to make up for the loss of "first room" acoustic space/feeling.  I believe that this is why we hear so much artificial reverb used by mixing and mastering engineers to provide some semblance of first room reflections that have been otherwise stripped away by the recording process.]

 

The reason why we have 5.1 (and 7.1, 9.1, etc.) surround arrays is to reinforce the original recording venue acoustical space, which provides a much greater ambience and realism than a stereo recording played back with tubes that reinforces the "second room" acoustics.

 

One thing that occurred to me is that the concept is actually very simple to visualize: I pretty much got it from the start when I saw Carver's video on Home Theater.  I could easily visualize the electro-mechanical dynamics in my mind's eye. 

 

What's amazing to me is that the folks that are apparently having trouble seeing the effect are folks that own and use tube amplifiers.  I wonder if there is some other mental dynamics at work preventing this visualization?

 

Chris

Edited May 15, 2015 by Chris A

[Chris A]

Chris if you agree then isn't it a fact that a high output impedance Single Ended Tube amplifier without global feedback could not have this distortion by the "microphone/reverb effect"....and thus Carver's claim that all tube amplifiers exhibit this to some degree is false..?

 

Triodes themselves have internal feedback:

 

http://www.hometheaterhifi.com/volume_5_4/essaynegativefeedbackoctober98.html

 

Triode-based tube amplifiers don’t require additional feedback circuits simply because that kind of tube inherently utilizes feedback within. But, because a triode doesn’t need an external feedback circuit, you also can’t control the feedback loop.

 

In truth, every audio amplifier, from SET (Single-Ended Triode) to push/pull bipolar, relies on feedback to simply cover the bandwidth required by audio, as well as maintain linear operation. All amplifiers use feedback, period.

 

 

 

Another reference on internal feedback of triodes:  http://ken-gilbert.com/images/pdf/Inherent_FB_inTriodes.pdf

Edited May 15, 2015 by Chris A

[mikebse2a3]

 

So Chris based on what you just quoted do you agree that this "microphone/reverb effect" requires that the amplifier be a design using feedback..?

 

No, I believe what Carver said (and I highlighted in brown type) is that all tube amplifiers "do it" to a greater or lesser degree.  This is the idea that if the amplifier presents any impedance to the loudspeaker, that it behaves like a mechanical resistance to any microphonic behavior coming back to the amplifier output terminals and reflecting back to the loudspeaker input - with the acoustics of the room providing the delays.  The efficiency of this transfer of room acoustic reflection energy back to the amplifier's output is a function of the ratio of the output impedance of the amplifier to the input impedance of the loudspeaker.  In my case, that is the 11-19 ohms of the TAD driver, direct coupled.  So the ratio is 1 to 10,  If you have a lower input impedance to your loudspeaker networks or drivers--like 4 ohms for a cone driver on a bass bin, that efficiency of reverb coupling is 1/4.

 

This is just like you mentioned once with a lot of loudspeakers in a demo room providing "damping"--except that it isn't really damping, but rather room-based reverberation that is reinforced by the amplifier-loudspeaker circuits themselves.

 

Chris my question was based on Carver's comments and calculations showing the (speaker-microphone effect) signal's voltage being applied to the input of the amplifier by way of the global feedback path and thus amplified by the gain(32db) of the amplifier in his example to reach his quoted -38.6db speaker-microphone reverb level and which he describes as "just barely noticeable".    I should have made it clearer that I was referring to global feedback in my question.

 

Without the 32db of amplifier gain I don't see how the speaker-microphone level could reach an audible level and this is why I say an amplifier designed without global feedback should not exhibit the speaker-microphone effect reverb at an audible level.

 

The fact that a speaker can act as a microphone is not disputed or the fact that the internal output resistance of the amplifier will have an effect on the (speaker-microphone effect) signal's level.  What is in question for me is if or when it could reach a level that becomes audible?

 

Carver has thrown out some numbers but where is the real data to support that all tube amplifiers with high internal output impedance do it at a perceptible level?

 

----------------------------------------------------------------------------------------------------------------------

 

 

 

What Carver apparently is saying is that, if you add current feedback to a tube-based amplifier with a controlled output impedance level (1.5-1.9 ohm), this reverb effect can be "enhanced" by the gain of the amplifier, which is apparently what he is doing in his tube amplifiers. In effect, what Carver is saying is that you can design a high output Z amplifier with current feedback that resists the microphonic loudspeaker return energy and actually provides gain on its output to "hold the drivers" in place against the acoustics of the returning room reflections - especially in the corners of a room with corner horns - just like the effects of using corner bass traps enhance the effects of absorption.

 

 

I'm not clear what you are saying here Chris with "actually provides gain on it's output to "hold the drivers" in place against the acoustics of the returning room reflections". This sounds like you are saying he is trying to suppress the speaker-microphone effect in this sentence which is the opposite of how I read his comments.

 

 

---------------------------------------------------------------------------------------------------------------------------

 

 

 

 

 

 

Since the output impedance of almost all junction transistor amplifiers (BJT) is almost zero due to the use of multiple gain stages and the use of global feedback, they do not produce a reverb effect that is audible. 

 

However, Nelson Pass uses single stage (Darlington pair) FET-type amplifiers--like the First Watt F3--that have about 1 ohm of output impedance, which is very uncharacteristic of transistor amplifiers. If you look at his SIT amplifiers (SIT-1 and SIT-2), the output impedance is even higher toward 2 ohm levels, and he uses a single gain stage without feedback.

 

You already know how I feel about this effect after owning planar dipoles for 20 years and becoming fatigued by the backwave-front wall reflections,  relative to the clean/unadulterated sound of the Jubilee without an artificial depth of field being splashed around (like Bose 901s). The "listening to two rooms" effect of the recording venue plus the listening room of the sound reproduction system - this second room acoustic effect needs to be controlled to a lower reflection level in order to hear the acoustics of the original recording room/venue. 

 

For recordings that were made in separate rooms and later recombined into a single stereo image (i.e., using multitrack recorders), the loss of the acoustics of an original recording venue is great and the sense of space is lost. 

 

[For many people reading this, there is a significant percentage of those that mainly listen to popular music recorded using multitrack recorders and separate dead acoustic recording booths, so these people rarely listen to "first room" acoustic reflections, and instead add their own with "second room" reflections to make up for the loss of "first room" acoustic space/feeling.  I believe that this is why we hear so much artificial reverb used by mixing and mastering engineers to provide some semblance of first room reflections that have been otherwise stripped away by the recording process.]

 

The reason why we have 5.1 (and 7.1, 9.1, etc.) surround arrays is to reinforce the original recording venue acoustical space, which provides a much greater ambience and realism than a stereo recording played back with tubes that reinforces the "second room" acoustics.

 

One thing that occurred to me is that the concept is actually very simple to visualize: I pretty much got it from the start when I saw Carver's video on Home Theater.  I could easily visualize the electro-mechanical dynamics in my mind's eye. 

 

 

What's amazing to me is that the folks that are apparently having trouble seeing the effect are folks that own and use tube amplifiers.  I wonder if there is some other mental dynamics at work preventing this visualization?

 

I'm not having trouble visualizing the concept Chris but without some real data to support the claims of audibility I believe it is a careless assumption that this is why some people choose tube amplifiers.

 

I also want to point out that any audibly added "speaker-microphone effect" reverb should be obvious and also result in a lack of clarity in the recording's reproduction and when I compare my Cary Single Ended 2A3 Tube Amps vs my First Watt F3 the listening results do not support the claims at this point.

 

 

 

Thanks for bringing this concept to my attention Chris and it is something I intend to explore more in depth when time allows.

 

miketn

 

 

 

Chris

 

Edited May 16, 2015 by mikebse2a3

[Chris A]

Mike, I tried.  Perhaps you should investigate on your own since my responses are clearly inadequate to your intent, as you have so clearly stated here.  I'm not sure how I can go any further-since it's really not my axe to grind for a living. However, I'm okay with the information that I already have--although I will welcome any test data. 

 

In other words, I choose to believe Carver, not because of his personality or his marketing prowess, but because what he is saying actually makes sense--in an area where a lot of folks have steadfastly refused to want to quantify what is actually happening that makes the sound different.  Apparently no one other than Carver to date has run tests (and published data) to prove the effect and how to design using that effect in greater or lesser quantity, as Carver has stated he is doing.

 

Carver has a good track record of backing up what he is saying although he chooses marketing words when describing his inventions since so many people in this industry historically have been, well, audiophiles, and fail to accept common physics-based explanations as sufficient to believe and invest in his inventions.  So he's stopped trying to convince using published technical articles.

 

It's okay to ask for data and it's okay to demand proof if you're not willing to accept that there is something qualitatively different about amplifiers with high output impedance.  The problem is--you'll probably wind up having to run the tests yourself.  There are no articles on overrunning/active loads with any kind of tube amplifier that I've found other than the discussions that Carver has made.  Carver doesn't have to publish, and has no incentive to do so especially in light of his well-documented past efforts trying to prove his inventions to those that didn't want the answers he provided (i.e., the two amplifier transfer function matching challenges of the 1980s with two different sets of audiophile magazine staff golden ears, where he won the argument on both technical and subjective listening evaluation grounds).  It's called "solution aversion" and is now a well-known human psychology phenomenon that I have personally experienced and have studied the effects of in my professional career.

 

It's clear to me that all amplifiers with high output impedance will adjust their gains to compensate for shifting output voltage to match the input command (i.e., either global or local feedback on the output stage) with amplifier-gain-driven forward loop gain changes.  This really isn't rocket science.

 

Chris

Edited May 16, 2015 by Chris A

[mikebse2a3]

This really isn't rocket science.

 

 

Well one of the most brilliant minds that I've followed did work for JPL and realized the problem of (Hearing vs Measurement : published in Audio 1988 and Time Delay Spectrometry) is much more complex than he first thought.

 

Heyser:  " Surely the end product of audio technology is the listening experience. We must never lose sight of this fact. No matter how exotic our instrumentation, no matter how impressive our mathematics, it is what we hear, not what we measure or compute, that is the final arbiter of audio quality."

 

I've learned long ago that one dimensional measurements don't always correlate to our multi-dimensional listening experience and it is the "listening experience" which I hope we can agree is the end goal of our reproduction systems.

 

 

Chris your comment: " What's amazing to me is that the folks that are apparently having trouble seeing the effect are folks that own and use tube amplifiers.  I wonder if there is some other mental dynamics at work preventing this visualization? "

 

My other reply to your comment:  I have also found those who are at the extreme ends of the debate of "Hearing vs Measurement" often have biases that prevent them from advancing their understanding of what is a very complex subject.

 

 

 

 

One of Heyser's very many insightful writings and please note especially comments about distortion and also synergism (which might play a part in the thread discussion of Tubes and Digital Audio. See this unpublished paper at the link directly below for the pdf: Submissions from 1975    Untitled (subject Measuring Sound) . I strongly suggest people search out and read all they can find of Heyser's writings..!!!

 

Heyser's Unpublished Writings:

http://digitalcommons.colum.edu/cadc_heyser_unpublished/

 

 

 

Some of Heyser's Unpublished Writings edited by Douglas R Jones. Mr Jones was instrumental in acquiring

the Heyser Archives for Columbia College Chicago. :  I especially encourage reading ":Is Heyser Still Relevant?"

http://works.bepress.com/douglas_jones1/

 

 

See section Heyser Resources here for more:  http://about.colum.edu/archives/collections/manuscripts/richard-heyser.php

 

 

 

miketn

Edited May 17, 2015 by mikebse2a3

[joshnich]

While most of this discussion is way over my head, I do enjoy trying to understand it! I recently purchased a Benchmark AHB2 power amplifier and it has significantly rocked my world. I would be interested in what the experts here think about the design notes regarding this amp - and if what the designer says is indeed that unique. The specs are off the charts in terms of dynamic range and distortion and  to me it sounds incredible whether I am using a tubed preamp and phono stage or directly using my DAC as a preamp.

 

Design Notes are here

http://benchmarkmedia.com/blogs/news/14680625-the-ahb2-a-radical-approach-to-power-amplification

[Ray Garrison]

I have a suggestion for anyone curious about what effect "placing tubes in the signal chain" will have on their system.

 

I have (heavily modified) sort of KHorns [edit - Converted to 2-way, ALK ES400 crossover, B&C DE750TN 2" compression driver, Fastlane Audio Eliptrac 400 Tractrix horn, Bob Crites woofers].  I've been using a laptop with JRiver 20, a Teac UD-501 D/A, Pass XP-10 pre and Pass XA30.5 amp.  Sounds great.  I recently bought a Schiit Valhalla 2 headphone amp for my Sennheisers.  Just for the heck of it, I tried placing it between the UD-501 and the XP-10, using the line outs on the Valhalla 2.  I'm running the balanced outs from the 501 to the XP-10, and the unbalanced outs from the 501 to the Valhalla 2 and unbalanced outs from the Valhalla 2 to the XP-10, so I can flip back and forth between the direct UD-501 input and the "tubed" Valhalla 2 input.  (Yeah, have to adjust gain, and one is balanced and one is unbalanced, but read on...)

 

Now, obviously (I think?) the only thing the insertion of the Valhalla 2 into the chain can do is either ( a ) nothing, or ( b )add distortion of some sort.  There's no way (I think?) it can do anything else.  I suppose if there was some sort of massive impedance mis-match in the original chain, and the Valhalla 2 just happened to have exactly the right input and output impedances to "fix" that mis-match it could possibly, uh, "fix" something, but that seems pretty far fetched.  So I'm totally discounting that possibility.

 

I prefer the way the system sounds with the Valhalla 2 between the DAC and the preamp.

 

I suppose I could try to analyze why I like it better - more "jump factor" on transients?  More "fleshed out" voices? - but I think that's pretty pointless.  I just like it better.

 

My point is, this is a relatively inexpensive and pretty safe way to see what happens when you insert tubes into your system.  The Valhalla 2 (6N1P dual triode input, 6N6P dual triode output, OTL) is $350 and comes with a money back guarantee.

 

My guess is it's adding low order harmonics (primarily either 2nd or 3rd) and that, particularly on transients, the additional distortion adds a touch of "jump factor" by increasing the amplitude / slope / peak duration of the initial attack.  Also moderate amounts of low order distortion tend to "add body" or "flesh out" voices and, particularly, acoustic instruments.  No I can't site specific references to support that... Art Dudley has written about it in Stereophile to some extent, and I remember some Audio articles from back in the 80's that talked about it (Richard Heyser?), but God knows where or when.

 

Anyway, that's not my point.  If you have a system configuration that allows you to insert something like the Valhalla 2 into the chain someplace, and you're curious about "tube sound", it's kind of a no-lose way to see what happens.

 

[edit] - One note that might be important.  This 2-way setup has virtually NO response over about 15kHz (according to my MiniDSP UMIK-1 and Room EQ Wizard).  This doesn't impact me at all, as I can't heat anything over roughly 13 to 14kHz anyway, but if I COULD hear more, I might not like it as much.

Edited May 29, 2015 by Ray Garrison