Exactly. The best way to say it is to look at what THD is...
Total Harmonic Distortion.
The first "T" tells you that THD is the sum of several things. and the the "H" says that the things we are adding up are just one kind of thing.
To get a feel for hormonics go to a piano. Play a C and another C one octave up. The higher C is a second harmonic of the first and the two sound good together. There are other compinations of piano keys that sound bad together.
THD only says how loud those notes are, not which notes and if they sound good or bad.
In some cases an amp can have fairly high THD. Say 5% (many good tube amps are this high) and the sound be be very good because all 5% happen to be the good sounding harmonics. But other amps can sound very harsh with 0.5% THD.
THD is only telling you the area under a curve on a graph, usfull but what you want to know is the shape of the graph, it's the shape that determines the "sound" of an amp.
Different amp designs produce different kinds of harmonics but THD only tells to the total, not the kind.
I understand the difference between good and bad types of distortion from guitar amps, it's similar, SS amps generally don't distort as well, SS amps usally have better "clean" channels although a little steril sounding, because they are not inflicting their own tone as much as a valve amp. Guitar amps are a little different, as the amp is part of the instrument, it produces its own sound or "colour", not necessarily reproducing the source acuratley.
What would be a good THD for SS amp, in the sence that it is not introducing much harmonic distortion, reproducing the original source without adding to many octaves?
What would be a good THD for SS amp, in the sence that it is not introducing much harmonic distortion, reproducing the original source without adding to many octaves?
Well, here is how I understand it, and how it has worked well for me.
It isn't the total, it's the content. As Bear and P10 say. So what content?
In listening tests in the '60, '70s and '80s ( and perhaps later) it has been found that a regular fall off of harmonics sounds "clean." If each higher harmonic decreases by an regular amount, the ear does not seem to hear the distortion up to fairly large amounts. There is a certain masking effect of the higher harmonics when the fall off is regular. This was determined by Wegel and Lane in the 1930s.
In contrast.
And that's just measured on steady state signals. Knowing what an amp does under dynamic conditions is another can of worms!
It isn't the total, it's the content. As Bear and P10 say. So what content?
In listening tests in the '60, '70s and '80s ( and perhaps later) it has been found that a regular fall off of harmonics sounds "clean." If each higher harmonic decreases by an regular amount, the ear does not seem to hear the distortion up to fairly large amounts. There is a certain masking effect of the higher harmonics when the fall off is regular. This was determined by Wegel and Lane in the 1930s.
In contrast.
- Dominant even order harmonics sound "warm", "soft" "round"
- Dominant odd order harmonics sound "cold" "edgy" "brittle" (but worse above 1KHz than below.)
- Suppression of the higher harmonics sounds "dull"
- A rise in higher harmonics can sound "bright" or even shrill.
And that's just measured on steady state signals. Knowing what an amp does under dynamic conditions is another can of worms!
What you want to look for is monotonically decreasing peaks of harmonics, with little (or noting above 3rd order). 4th, 5th, etc are not naturally occuring and stand-out to the ear-brain, even if 2nd & 3rd are much higher.
You also need to consider at what power levels you are actually going to be using the amplifier at... many measurements are at full power, not where the amp will be most of the time. Also each spectrum chart will be at a specific frequency.
dave
You also need to consider at what power levels you are actually going to be using the amplifier at... many measurements are at full power, not where the amp will be most of the time. Also each spectrum chart will be at a specific frequency.
dave
For years I have always searched for amplifier circuits that have the highest power and lowest THD only to have my dreams shattered by speaker specs.
If you have a speaker that produces less than or no more than 2% to 5% THD then you have really got something, as most produce well over 10% thd. jer
If you have a speaker that produces less than or no more than 2% to 5% THD then you have really got something, as most produce well over 10% thd. jer
To toss a little hand grenade into the discussion, while it needs to be shown in practice I see no reason in principle to rule out second harmonic cancellation between amp and speakers. For those unfamiliar, when two electronic devices with the proper ratios of second harmonic distortion are connected in series the resultant second harmonic distortion on the output of the chain can be less than that of either device.
If the same occurs at the amp/speaker interface, the question becomes: "what amount of distortion into Speaker X is hifi"? I'll go now.
And the best part is. After reading all of the replies *many times conflicting* you see in this thread, your going to learn the only answer. Buy an amp and speakers that sound good to you and forget what the other guys think 
And another important rule. Forget the specifications on most receivers/amps as they are fudged anyway.
And another important rule. Forget the specifications on most receivers/amps as they are fudged anyway.
Last edited:
Since this thread is in the Chip Amps forum, there could be useful reading here:
Gainclone & Sound Quality
This is LM3875, demonstrating mainly even order distortion at quite low levels. If distortion measurements say anything about how something sounds, this thing should sound very, very good.
Gainclone & Sound Quality
This is LM3875, demonstrating mainly even order distortion at quite low levels. If distortion measurements say anything about how something sounds, this thing should sound very, very good.
That's usually pretty close. It's a very rare amp where those terms don't dominate.
If an amp with 0.0004% THD sounds "rubbish," it's most likely that you're overloading it or using an inappropriate load for that design. Or that the amp actually isn't giving 0.0004% THD with a reasonable load.
SoundStage! Measurements - Bryston 4B SST Stereo Amplifier (10/2004)
If I look at this, the 5HD is just a couple of dB down from 2HD. Will this not be audible?
If I look at this, the 5HD is just a couple of dB down from 2HD. Will this not be audible?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.