Great article on "Analogue Warmth"

[esldude]

Mr. Ponk, I'M not complaining. Blizz has been telling me all day how he's going to get this magic box to me. I just CANNOT wait. 300-600W of Class D, all that power but also as intimate as you could want w.3W/ch SET beauty, all the analog warmth that every Class D amp I've heard can only dream of providing. Sounds like win-win to me.

Why don't you do the experiment I suggested by feeding a class D amp your SET output? You are correct that listening to another speaker will not sound the same as speakers aren't transparent or even close. You would however get to hear what a 500 watt SET would sound like on some big speakers. I am pretty sure most of the SET sound would be intact.

The other approach would be to connect the class D amp between your SET and the speakers it uses. You will probably need to actually lower the gain on the class D amp considerably.

Or you can simply deny the possibility of something you appear not to like. The idea you like your SET due to its colorations, and those colorations can be copied.

 

[Blizzard]

You guys are going at speed of light here . I will make two comments:

1. The original article talks about a lot of sources of distortion in analog audio chain. Have not seen people say those distortions are not real. If so, how many of you hear these analog distortions and can identify them? Do you always hear them as a collection of goodness???

2. Emulation of analog distortions as described in SOS article collectively and with high fidelity is very hard to impossible. Imagine just modelling scrape flutter of tape. How that tape jumps from one speed to another every few milliseconds depend on the tape, temperature, machine and heaven knows dozens of other causes. Correct emulation would involve physical simulation of the entire unit with measurements of every component. You cannot profile this as scrape flutter will change as the tape spools fill, machine warms up, etc. Ditto for channel crosstalk, self-erasure, etc. These are time variable things that cannot be modeled in single sitting/measurements.

That said, I think a much better approximation can be done than current plug-ins in pro world. It is something I like to build one day (in software) but it is a lot of work.

So emulation of analog distortions while feasible in theory, as a practical matter, is not in front of us as a matter of simple profiling.

We have to take into consideration that this article was written in 2010 and based on 2010 technology. Even today what I'm talking about isn't available. And I do agree that exact emulation of an R2R machine will be very difficult. Especially since even the tape played back on the same machine will never sound exactly the same twice.

This technology is more for emulation of DAC's, preamps and amps. It's not for gear with mechanical moving parts.

 

[TBone]

So then you have the strange idea that LP can be better than the mic feed at creating a musical event which necessarily means it changed the signal. Yet many will poo poo the idea some digital manipulation of the signal can be beneficial in the same manner or that digital manipulation can do the same 'beneficial' thing that LP can do.

So true ... remember when certain CD players were advertised as sounded more "analog". The original Rega Planet, case in point; it included an obvious warmer darker tonality compared to it's competitors which came across as brighter and more "digital" in direct comparison. They sold a ton of those ...

 

[esldude]

You guys are going at speed of light here . I will make two comments:

1. The original article talks about a lot of sources of distortion in analog audio chain. Have not seen people say those distortions are not real. If so, how many of you hear these analog distortions and can identify them? Do you always hear them as a collection of goodness???

2. Emulation of analog distortions as described in SOS article collectively and with high fidelity is very hard to impossible. Imagine just modelling scrape flutter of tape. How that tape jumps from one speed to another every few milliseconds depend on the tape, temperature, machine and heaven knows dozens of other causes. Correct emulation would involve physical simulation of the entire unit with measurements of every component. You cannot profile this as scrape flutter will change as the tape spools fill, machine warms up, etc. Ditto for channel crosstalk, self-erasure, etc. These are time variable things that cannot be modeled in single sitting/measurements.

That said, I think a much better approximation can be done than current plug-ins in pro world. It is something I like to build one day (in software) but it is a lot of work.

So emulation of analog distortions while feasible in theory, as a practical matter, is not in front of us as a matter of simple profiling.

Look into what Universal audio is doing for emulation. Scrape flutter is too variable yes. But they are going into detail even about physical layout. For gear with transformers they model the leakage due to a pair of them being within a few inches of each other. It isn't a simply black box input to output change sort of modeling. I see no reason this is impossible to the point of being audibly identical.

 

[Audiophile Bill]

well the more relivent thing is how you have condisioned your brain sound wise and a lot of that is done as children.

you go to more live then me then plus my first unamplified jazz gig was last year but i listen to my system for 2-6 hours a day at times so i went to my first jazz gig with my home sound as a ref and was very happy when i came back. i hate most other live music sound wise amplified that is and only hear problems but dont feel bad for me i drink throught it lol.
i prefer my lounge to a lot of big gigs i go to but when i go to the little jazz venue i come back and stick on a album i love what i hear. just the beer is limited at my house in choice and volume.. thats my only problem .

if your comming from a musical backround into hifi then i think your screwed lol and might never be happy.

Hi Blizz,

I come from a musical background into hifi and am happy with what I have

 

[spiritofmusic]

Esl, when someone wants to hoist some over, I'll give it a go. Unfortunately I don't know anyone who has any.
I have other priorities.
Yr idea does sound intriguing. Esp since all the Class D I've ever heard I wouldn't want here for long powering anything other than my subs.
And I can't see how doubling the number of components/boxes/cabling can do anything except add further losses.

 

[microstrip]

Micro tut tut, you will give digital a bad reputation, the waveform is reconstructed perfectly in a band limited system, dynamics are completely unaffected.
Keith.

No, we are trying to save it. Perfect sound forever was responsible for the bad reputation, since 1982.

 

[Blizzard]

Hi Blizz,

I come from a musical background into hifi and am happy with what I have

Excellent. Millions were also very happy with the Ford model T as well as the Volkswagen beetle.

 

[Ron Resnick]

Digital perfectly reconstitutes the analog waveform . . .

This is the second time this view has been asserted. I will await the proof of the assertion.

 

[Ron Resnick]

Ron, I look forward to your visit! I would recommend that, while you are in the area, you also listen to Madfloyd's (Ian's) system, it's pretty incredible to my ears.

Peter A will be in charge of all listening session arrangements, but I would like to meet in person you and Ian as well!

 

[esldude]

This is the second time this view has been asserted. I will await the proof of the assertion.

What would you accept as proof? The Monty video offers some good evidence to the idea. Have you watched it? If so, what part of it is unconvincing?

 

[microstrip]

When people don't understand how things work, you'll always have the same guys with the same mindset trying to bash the idea. It doesn't matter what the topic is. Ask anyone who's ever accomplished anything noteworthy, and they will tell you criticism is the story of their lives. You would think people would just say "Wow great idea, but since I have no idea how stuff like this works, I'll just sit back and let the pros work on it" but instead you get a wave of revolt from naysayers who cringe at innovative ideas if they are outside the bubble of their understanding and comfort. Well can't do nothing about those guys. But one thing for sure, once you prove them wrong, and the majority hops on board with your system, the same guys will be the ones defending your product, if someone else comes along threatening to make a better mouse trap through innovation.

All you need is persuade Vladimir Lamm to give you his mathematical models and implement them with success. Then we will applause you!

A straight question - how can we emulate the sound of different types of high quality film capacitors? Do you think their distortions are linear or non linear?

 

[Blizzard]

All you need is persuade Vladimir Lamm to give you his mathematical models and implement them with success. Then we will applause you!

A straight question - how can we emulate the sound of different types of high quality film capacitors? Do you think their distortions are linear or non linear?

I won't need any mathematical model's, just a simple measurement from the analog outputs. 100% of everything will be captured, what we understand, as well as what we don't understand. That's the problem with past emulation algorithms. They are based on what we understand, what we can measure, and plot out with today's measurement gear.

Same applies with different caps. Swap the caps out in the component, and take another measurement sweep. The change of sound the cap swap made, will be captured with 100% accuracy, even if it can't be measured with today's technology.

 

[Ron Resnick]

Well stereo is incapable of recreating the original musical event. Fidelity on playback can only be to the source material.

Which means your camp #1 becomes a variant of camp #3. If the master tape accurately reproduced doesn't sound like music, and LP or any other medium moves closer to the sound of an original musical event it can only be because the original source signal has been altered. Colored, improved, processed, but in any case it has been altered.

So then you have the strange idea that LP can be better than the mic feed at creating a musical event which necessarily means it changed the signal. Yet many will poo poo the idea some digital manipulation of the signal can be beneficial in the same manner or that digital manipulation can do the same 'beneficial' thing that LP can do.

This reply is an example of why members so often talk past each other.

I was very careful to explain the three alternative philosophical objectives of high-end audio. Each objective is valid, to me. To write that 'fidelity on playback can be only to the source material" is logically wrong and ignores the other philosophical objectives. There is no analytically principled basis for declaring one of the objectives valid and dogmatically and blithely ignoring the other objectives.

This member misquotes me twice. I wrote "recreate the sound of an original musical event." I did not write: "recreating the original music event" which, obviously, we cannot do." I also did not write: "creating a musical event."

 

[Ron Resnick]

What would you accept as proof? The Monty video offers some good evidence to the idea. Have you watched it? If so, what part of it is unconvincing?

I thank Keith for posting the video. I promise I will watch it the first week in January.

 

[microstrip]

Not quite. Sampled audio on its own does what you say, but "digital" means there's quantisation error.

Touché - you are right, perfection would imply no quantization error up to the electron level.

 

[Al M.]

I guess my question to you is if your tube amps sound like transistors why not use transistors? I am not saying it isn't possible to make clean tube amps that sound like good solid state because it is. Just why put up with tubes if they sound un-tube-like?

Because I have no reason to switch to SS on my current speakers. Why spend a boatload of money for essentially the same result from an SS amp? Yet if I ever would upgrade to Magico Q series speakers (the speaker type that would interest me in the more expensive price range), for example Magico Q1 minimonitors, then these would be much harder to drive. My amps couldn't do it, and instead of looking for another tube amp I would go straight to a Spectral amp in that case.

But at this point I am not interested in that upgrade path. I'd rather put my money into an even better source than my Berkeley DAC, for example.

 

[esldude]

Okay, sense you feel I misquoted you let me try this again.

I think this proves, once again, that much honest disagreement here originates in the differing philosophies of high-end audio.

I believe there are three primary alternative objectives of high-end audio:

1) recreate the sound of an original musical event,

By this do you mean a perhaps new musical event or a musical event occurred at a given time and place which we intend to recreate?

2) reproduce exactly what is on the master tape, and

This is normally all you have to work with at best. Without additional information it is the only thing playback can have fidelity to.

3) create a sound subjectively pleasing to the audiophile.

I subscribe to the first philosophy. I want my audio system to recreate as realistically and as believably as possible the sound of an original musical event. See http://www.whatsbestforum.com/showthread.php?19261-Introduction-and-Listening-Biases

See this is where the issue is. You cannot reach your goal with merely stereo.

Since I subscribe to philosophy 1), I am saying that the LP can be better than the mic feed (with "better," as defined by philosophy 1), as better able to recreate the sound of an original musical event).

This comment is why I placed your #1 camp as a subset of #3 of creating a subjectively pleasing sound to the audiophile. The LP cannot be better than the mic feed unless you alter it in some way. Since stereo is limited in how close it could accomplish your goal, and you said LP can be closer you are inherently making choices that convince you the sound is closer to what live musical events occurred. Choices that are pleasing to that goal. The nature of LP and stereo prevent achieving that goal and your choices are just that choices based upon preference and pleasure. That is fine, but a variant of camp #3 all the same.

 

[microstrip]

This is the second time this view has been asserted. I will await the proof of the assertion.

What part are you requesting proof - the quantification or the sampling? Are you addressing theory or implementation?

BTW, I am not sending letters to the "Misquoting officer", I am sure it was not intentional, but you are distorting my original statement when you replace part of the sentence by ... .

 

[spiritofmusic]

Excellent. Millions were also very happy with the Ford model T as well as the Volkswagen beetle.

Well I run a 1968 Citroen DS21, so maybe you need to put me down as incurably romantic. And I have an analog 1996 Barco 1209S 3-gun CRT projector for movies, an analog Naim AV1 for surround duties, an analog Nikon SLR. Still watch movies on a cutting edge Pioneer HLD X0 laserdisc player, yes an analog picture stage. You get my drift. I'm sitting on my log being totally anal about things LOL.