I got into a discussion a while ago about 'euphonic distortions' where I just couldn't pin the other chap down on whether the intention was to deliberately build them into the gear, or whether they were just a serendipitous by-product of certain types of equipment. Nor could I pin him down on whether he would prefer perfectly linear equipment, or equipment featuring these distortions.
I think the discussion also mentioned that there's no such thing as harmonic distortion on a complex music signal* - it's almost completely IMD.
*Unless using DSP you identify and separate the contributing waveforms of the individual sources, apply harmonic distortion to these separately, then add them up again.
Groucho,
that is an interesting point.
Because I know Paul Miller has commented several times about some of the best engineered products to have an engineered measurement/performance (one example was his analysis of Pass Labs pure class A model).
Regarding the Pass Labs he states the measurements are incredibly good and the structure of the THD/snr across the spectrum and power showed Nelson Pass has exceptional engineering knowledge/skill to create the subtle curve it provides.
But one could say the context goes beyond this if considering the more recent research I was hinting at because measurements should use complex tone that also have a complex envelope - IMO and I appreciate others will disagree.
Cheers
Orb
Regarding the Pass Labs he states the measurements are incredibly good and the structure of the THD/snr across the spectrum and power showed Nelson Pass has exceptional engineering knowledge/skill to create the subtle curve it provides.
So the amplifier has deliberately-engineered distortion characteristics that behave in a certain way with power and spectrum? And that there are set criteria that make it possible to classify these as better than those of other amps? Would the amp be considered as better than a hypothetical zero distortion amp? Or is it the idea that a zero-distortion amp is impossible and therefore it is better to tailor the distortion deliberately? Is it better to actually raise the distortion level in a certain way deliberately in order to mask lower level, but undesirable, distortion characteristics?
I like those book thingies. Did this GE study exclusively link loudness perception to odd-order harmonics, or find that they were noticably stronger in creating that perception? Because I don't think there's any argument that they contribute. The question is are they any more significant than even order harmonics, or simply content in the range where human hearing is most sensitive?
According to the GE study, the odds were more significant. The same study showed the even orders were tolerated by listeners up to 30%(!) before objections were raised.
In the context of this conversation, I think it's about a theory that odd-order harmonics create more dramatic loudness perception, are more prevalent in SS gear, and create harshness at high volume in that gear that is not present in tube equipment. I've asked for clarity around that, but I think that's the point.
Mostly. I like to think of it as being on the cutting edge- if you can build a transistor amp that lacks the trace odd orders that seem to be the issue, you may well be also building the amp that might make people give up on tubes.
WRT the tube/transistor thing, you don't really have to know anything technical to understand that there is an issue. Tubes have been obsolete now longer than they were the only game in town, and somehow the market still wants them around. Its the market that is speaking with the louder voice, not techno-geeks like myself.
I got into a discussion a while ago about 'euphonic distortions' where I just couldn't pin the other chap down on whether the intention was to deliberately build them into the gear, or whether they were just a serendipitous by-product of certain types of equipment. Nor could I pin him down on whether he would prefer perfectly linear equipment, or equipment featuring these distortions.
I think the discussion also mentioned that there's no such thing as harmonic distortion on a complex music signal* - it's almost completely IMD.
*Unless using DSP you identify and separate the contributing waveforms of the individual sources, apply harmonic distortion to these separately, then add them up again.
I also see HD as being a form of IMD. I am also of the opinion that certain amplifier topologies are indeed exploited for their distortion signatures. I think SETs *can* be a good example, depending on how careful the user was to set up the SET with a speaker that is efficient enough to really take advantage of the SET's better qualities (IMO most people don't do this...).
Sunn made a more or less legendary solid state guitar amp back in the early 1970s that took advantage of topology such that it had a reputation of being fairly rich sounding. It thus is one of the few solid state guitar amplifiers that are sought-after and have respect with guitar players. The amplifier had a single-ended preamp section that used FETs operating zero feedback, and the power amp section was entirely single-ended until the driver transformer converted to push pull to drive the output devices. Such a circuit would have made more lower ordered harmonics.
Where do active crossovers fit into this debate? If amp distortion is a real issue, then the active crossover differs on several fronts does it not?
- restricted signal bandwidth to each amplifier
- much easier load on each amp, much less power required
- the already-much reduced IMD is restricted to each driver's bandwidth.
But if the 'euphonic distortion' idea is valid, it is completely scrambled by the active crossover. Different signals will cause differing amounts of distortion in each driver with no clear link between overall signal volume and the type and level of distortion. Is this considered a good thing or the opposite?
(Nailing my colours to the mast: my preference is definitely for active speakers).
(...) I think the discussion also mentioned that there's no such thing as harmonic distortion on a complex music signal* - it's almost completely IMD.
*Unless using DSP you identify and separate the contributing waveforms of the individual sources, apply harmonic distortion to these separately, then add them up again.
Deliberately introducing very small distortions in digital recordings using DSP tecnhiques and listening to them using SOTA linear systems would be a way of studying all these aspects - unhappily the people who could do it properly will tell you that all of them sound exactly the same.
Deliberately introducing very small distortions in digital recordings using DSP tecnhiques and listening to them using SOTA linear systems would be a way of studying all these aspects - unhappily the people who could do it properly will tell you that all of them sound exactly the same.
So just enquiring a little further: you are saying that there are linear systems out there, but that they are not your personal preference? So by what criteria do you judge a system if not by its linearity? Are you saying that there are systems that you prefer, and you also know them not to be linear? If so, what is it about their sound that you like better than a linear system? (I'm not trying to catch you out or anything!)
So just enquiring a little further: you are saying that there are linear systems out there, but that they are not your personal preference? So by what criteria do you judge a system if not by its linearity? Are you saying that there are systems that you prefer, and you also know them not to be linear? If so, what is it about their sound that you like better than a linear system? (I'm not trying to catch you out or anything!)
By "linear" I just mean the typical audio equipment having 0.001% distortion that it is said by some people to sound all the same way. And yes, I do not prefer such systems. If high-end is just adding euphonic distortions - something I do not believe - if someone adds them on the files they should be able to sound like high-end when played in "linear".
By "linear" I just mean the typical audio equipment having 0.001% distortion that it is said by some people to sound all the same way. And yes, I do not prefer such systems. If high-end is just adding euphonic distortions - something I do not believe - if someone adds them on the files they should be able to sound like high-end when played in "linear".
So basically like the legendary (mythical?) Carver challenge? Try to emulate 'high-end' equipment using cheapo equipment. Sounds like fun, except that I would find it hard to be motivated to do it. If I thought that high end equipment sounded special, I would already be trying to do it. As it is, I really haven't heard anything better than my own system - and I built it to be as linear as possible using stuff that people around here would call mid fi (those mass produced amps with 0.001% distortion etc.). But I am open to the idea that I'm missing something, although mainly sceptical.
I have a simple wisheople who love to talk about how great their system is and poke fun at other people who have invested more in their systems should share with us the details of said system. I think it's a fair request.
So basically like the legendary (mythical?) Carver challenge? Try to emulate 'high-end' equipment using cheapo equipment. Sounds like fun, except that I would find it hard to be motivated to do it. If I thought that high end equipment sounded special, I would already be trying to do it. As it is, I really haven't heard anything better than my own system - and I built it to be as linear as possible using stuff that people around here would call mid fi (those mass produced amps with 0.001% distortion etc.). But I am open to the idea that I'm missing something, although mainly sceptical.
You are really a happy man - I have heard to several systems significantly better than my own, although I am very pleased with it. And we, living at this side of the Atlantic, are relatively free from the risk of a visit to Steve's room and his tapes!
I have a simple wisheople who love to talk about how great their system is and poke fun at other people who have invested more in their systems should share with us the details of said system. I think it's a fair request.
I don't understand why anyone (and everyone) interested enough to read and post here frequently doesn't put their system (and their real name) in their Profile. I'm amazed at how many members' here are empty of gear.
I have a simple wisheople who love to talk about how great their system is and poke fun at other people who have invested more in their systems should share with us the details of said system. I think it's a fair request.
My system doesn't need each component to be extraordinary or 'high end', because life is very easy for each of them. A desktop PC is the source, with software to do the crossover filtering and driver correction. Three stereo amps drive the woofers, mids and tweeters. Enclosures are sealed. Woofers are big. It's a recipe for a lack of stress on every element, with direct connection from amp to driver and different from what most people here have ever heard. There's no point discussing exactly what make of mass-produced 0.001% distortion amp I use because it isn't being used in the same way as an amp normally is. No point discussing what brand of speakers they are, because they don't have a brand, being made of individually-chosen drivers that, needless to say, aren't made of diamond and whose type isn't critical. Speaker cables are very thick zip wire and the interconnects cost about $5 each. The most high end bit is the software, but I wrote it, and it's got tools in it to allow me to set the system up very accurately - and I hear no sign whatsoever that straightforward objective accuracy isn't the best policy.
As I say, I've never heard anything as good, but hopefully I'm going to get to hear some Meridians at some point, and I'm expecting great things of them.
What do you reckon? Am I the same as every self-deluding amateur speaker builder who believes their latest chipboard monstrosity is the best thing ever, or might it just be possible that I've built something good? The individual components are nothing special, but the configuration of them is a 'paradigm shift' compared to most systems. Can you be totally sure it'll sound crap compared to your favourite turntable - valve amp - passive speakers?
Deliberately introducing very small distortions in digital recordings using DSP tecnhiques and listening to them using SOTA linear systems would be a way of studying all these aspects - unhappily the people who could do it properly will tell you that all of them sound exactly the same.
Depends. If one set of speakers (for example) is really pretty and the other set ugly as **** that's rather an important preference to some folks.
And that is as it should be. And, yes, I could accomplish that, with two sets of speakers that are identical to rather close measurements, with one set painted flat black and safety orange, with flashing green LED's on the sides powered by a battery, and the other set having Bobinga veneer, lovingly polished with 22 layers of shellac.
I like those book thingies. Did this GE study exclusively link loudness perception to odd-order harmonics, or find that they were noticably stronger in creating that perception? Because I don't think there's any argument that they contribute. The question is are they any more significant than even order harmonics, or simply content in the range where human hearing is most sensitive?
Why? Just look at vinyl distortions. L+R is symmetric except for skating error. L-R is not symmetric. How will an active crossover do anything beyond send the appropriate signals to the appropriate places?
