What are the advantages/disadvantages of different amp topologies?
- Thread starter christoph
- Start date
I use LPs that I recorded myself as reference. Knowing how the recording is supposed to sound because you were there is an enormous benefit. If you want to know if something is neutral, a good reference is a handy tool
Neutral does not mean its boring unless the music itself is boring. If an amp sounds boring, there's a very concrete reason for it doing so, likely quite measurable too.
Neutral does not mean its boring unless the music itself is boring. If an amp sounds boring, there's a very concrete reason for it doing so, likely quite measurable too.
Not sure! I bought an amp from a brand more associated with pro audio, so it measures very well - but turned out drearily dull. Give me an amp with a bit of character any day.
If it was dull it wasn't neutral. The measurements that were published might look pretty good. But they likely were not all the measurements that needed to be made.
No one disagrees that the quantitative magnitude of speaker errors is generally larger than source and electronics components. They happen to be less pernicious in their deleterious effects on sound believability.
And how can you judge this? You can’t hear a recording without a system. Where is your reference? Unless you made the recordings yourself you have no idea what it should sound like. You can maximise variability between recordings…this doesn’t necessarily tell you your accuracy to a given recording but at least it tells you that you don’t have major coloration that obscures the distinctness of each recording.
Frequent live experience and a good ability conceptualise how close a system gets to that live sound/experience (needs appropriate recordings for judgement…good live recordings of classical or Jazz, for example) another approach and with the right recordings might well converge with the maximal contrast approach.
Fidelity to the source is unobtainable due to unknowable information…like how a recording SHOULD sound.
It has nothing to do with minimizing the need for compensation…compensation is inevitable because you are trying to maximize your conception of what sounds more “real” to you. The degree of compensation is actually unknown because you will never know exactly what a recording should sound like. Clearly some recordings sound more real than others almost regardless of the system, but all systems…at least those from sophisticated listeners, have been deliberately voiced in a way that the owner thinks is better representative of their concept of “real”.
You pick gear (mostly) that you think is low in coloration but in relation to what? Recording fidelity? How do you judge that? Only through the same system you are trying to optimize…it’s recursive. Benchmarking to what you hear live helps but what is on the recordings may not be at all like what you hear live. Assuming you want the recordings to sound like live , or maybe just the really good ones, you will pick gear with compromises that you think move you in the “right” direction. But it is all compensation between elements that get you to a particular sonic presentation that to you sounds “neutral” or “real” or whatever.
Last edited:
The degree of compensation is directly related to whether a voice sounds like a voice, or a piano like a piano. Needless to say, there are records that no one knows what they're supposed to sound like, so it becomes a mere matter of getting the playback to where one likes it. To my mind, as explained further above, it seems perfectly valid to me if the latter is all you're expecting playback equipment to deliver. Not to me: I expect improvement, not in terms of what I like, but in terms of making a voice sound like a voice (believably human), and a piano like a piano (such as the Steinway on which I heard my sister practice on a regular basis). You can claim all day long that people can't agree on what sounds real and are only imagining things, but then I wonder, what and why we're discussing here at all? If better is only better according to your personal taste, logic would have it, your opinion must be meaningless to others. If the benchmark for "better" and "worse" were a moving target for each individual, why even bother makings claims on the superiority of the harmonic spectrum of one type of amplification over another etc. All this while, you were not saying one contributes to sound that is more natural whilst another more synthetic etc.? Weren't those your words? Note that is what I agree with: that we humans are most capable judging whether playback sounds more or less real. That we may not agree on matters of personal taste has nothing to do with it.
Greetings from Switzerland, David.
Last edited:
That patently doesn't make sense: if it sounds bad but measures well, this only means what's being measured is irrelevant - in other words, it does not measure well.
Greetings from Switzerland, David.
Until there are clear psychoacoustic measurement guidelines, “measures well” means low THD, low IMD , usually high power and high load tolerance. The amp Hear Here talks about probably meets those criteria and so by common objective terms measures well. I would wager the amp in question measures objectively at least as good as the Spectral gear you like…can you look at their measurements and tell which one will sound better?
Now we are only talking about the degree of compensation…good! Of course I try to minimize the synthetic in my playback and make things sound as realistic as possible to my mind. But to think that putting in a new piece of gear that sounds “better” isn’t doing some kind of compensation for other parts of the system is IMO a simplistic view of what’s really happening.
You’re bending your argumentation to your agenda: I remember you’re the one who’s pointing to e.g. spectral analysis of the harmonic behavior of amplification types as relevant to sound that human beings perceive as natural. All I just said was I doubt the relevance of measurements that can’t do so much as explain bad sound.
To the other point: yes, over the years I have come to a tentative conclusion that sound quality, in particular impulse response (along with it, phase linearity and settling time) is related to bandwidth. Feel free to argue with that, too.
Greetings from Switzerland
Last edited:
Other way round, I was always referring to the degree of compensation, picking you up further above on “truisms” (nothing’s perfect - what else is new?) and absolutes (your disregard for the spectrum in-between, which directly translates to a lesser degree of compensation being the goal, the point @DasguteOhr tried to make from the get-go).
Greetings from Switzerland, David.
No, you said if it sounds bad and measures good then it actually measured bad. I pointed out that by objective standards it almost certainly measured good…probably as good or better than Spectral, which I know you like (or you did in the past).
Been a while since I occupied my window seat in physics, but back then “megahertz” referred to some greater than “kilohertz”…
Greetings from Switzerland, David.
Measurements of something that sounds bad can logically only be deemed irrelevant is my point, yes. Besides, if you want to talk to people who buy gear based on measurements, you might want to join the nerds over at ASR. I buy mine based on listening. That doesn’t mean that something that has good sounds needs to measure badly. That correlation has never made sense to me.
Greetings from Switzerland, David.
I think of audio and measurements as work in progress: hopefully one day we’ll be able to correlate measurements to (at least) our perception of bad sound. It can’t be that music that sounded great e.g. in the orchestra hall sounds bad at home due to the absence of side products caused by speakers and gear. It’s only logical that the reproduction is not neutral, and hence the measurements flawed.
Greetings from Switzerland, David.
Last edited:
Your understanding of physics is ok, your understanding of what amps I mean is not. Your SETs go to MHz? I think not…
No, you said must measure bad…without qualification about what you mean by bad. If you mean psychacoustically irrelevant, then that holds for all measurements unless put through an algorithm that correlates the numbers with listening results.
Similar threads
