Compare original and distorted audio from any .WAV file.
The latest version of the program matches RMS power for comparison pairs. This makes it easier to discern small differences in sound quality. You can now try fourth-harmonic distortion as well as second or third. Fourth is particularly nasty. Finally, you can specify distortion in either percent or dB.
How Much Distortion Can You Tolerate?
Brian
The latest version of the program matches RMS power for comparison pairs. This makes it easier to discern small differences in sound quality. You can now try fourth-harmonic distortion as well as second or third. Fourth is particularly nasty. Finally, you can specify distortion in either percent or dB.
How Much Distortion Can You Tolerate?
Brian
It's generally accepted that distortion levels under 0.5% are not noticeable, and with amplifiers having 0.25%, (vintageMcIntosh) and down as low as 0.008% (my Technics amp) are of little to no issue or worry.
However, you have multitudes of "nitpickers" that obsess over such inaudible, trivial values, and clutter up the internet with endless threads of nothings.
However, you have multitudes of "nitpickers" that obsess over such inaudible, trivial values, and clutter up the internet with endless threads of nothings.
Depends on the type of distortion.
Marshall made a fortune from the right type of distortion which was many %.
On the other hand very little distortion of the wrong type can sound bad.
Marshall made a fortune from the right type of distortion which was many %.
On the other hand very little distortion of the wrong type can sound bad.
People in such threads are not necessarily claiming that they can hear a difference between the two. Designing an amplifier that can achieve .002% THD across the spectrum is far from trivial and there is much to be learned in the process.
Furthermore, THD can highlight PCB layout issues.
That said... low THD does not mean that an amplifier will sound good. I've found more than a few Crown amplifiers that measure reasonably well with relatively low THD (for that type of amplifier) but are rather unpleasant to listen to.
I always suggest measuring as much as you can. That said, don't forget to do that A/B comparison with an amplifier that is known to be a good-sounding reference design and note the difference, if there is one.
Crown amps aren't particularly good for "hifi" setups at home, never were.
However, for some reasons, people did just that, (because of coaxing/hype by others?) and as you mentioned, learned something.
Crown stuff was better suited for clubs and PA use.
Yeah, I've never been a big fan of Crown amplifiers- but they were good for illustrating my point.
In this case I had a Macro-Tech 3600VZ in for a while. I got it cheap because it had a bad circuit breaker on one channel. Before I gave it to someone for an experiment, I decided to do a comparison between it and a few other amps for a near-field monitoring application. The point I'm making is that the measured THD+N of that amplifier was very close to those I tested it against... but it definitely didn't sound the same.
It's not just that it's a pro amp though. MC2 amps sound amazing in a home environment, and some of the better QSC and Crest amps, though not quite on par with the MC2, sound quite good in that application too.
In this case I had a Macro-Tech 3600VZ in for a while. I got it cheap because it had a bad circuit breaker on one channel. Before I gave it to someone for an experiment, I decided to do a comparison between it and a few other amps for a near-field monitoring application. The point I'm making is that the measured THD+N of that amplifier was very close to those I tested it against... but it definitely didn't sound the same.
It's not just that it's a pro amp though. MC2 amps sound amazing in a home environment, and some of the better QSC and Crest amps, though not quite on par with the MC2, sound quite good in that application too.
If I may go back to the original subject of this thread ;-)
Bill Waslo has a similar app, with extensive user interface, Audio DiffMaker.
His fav test is playing a piece of music you select, and then you can add in a complete symphony orchestra at -60dB.
Hard to impossible to discern. Talk about an eye-opener.
Jan
Bill Waslo has a similar app, with extensive user interface, Audio DiffMaker.
His fav test is playing a piece of music you select, and then you can add in a complete symphony orchestra at -60dB.
Hard to impossible to discern. Talk about an eye-opener.
Jan
Last edited:
Would've been interesting to use that when we tested all those power supplies!
The differences were very notable, Waslo's would have left us with an actual physical record.
I measured that I can detect -65dB of 7th harmonic of a 100Hz signal, -63dB of 5th harmonic of 200Hz (in a comparitive test) which suggests a limit around 0.05% for higher harmonics. This suggests < 0.1% is a good level to aim for (for me) in a class B amplifier so that the individual high harmonics are well below that.
The lower order harmonics aren't themselves easy to hear due to perceptual masking, but the IM products from lower order distortion can be.
So what you're saying is, that oh-so-adored and coveted by purists fabulous McIntosh amp, with its miserable 0.25% distortion level would drive you bonkers.
That translates to those purists and audiofools accepting such distortion levels just to own a "name" to be proud of and flaunt.
And that's so pathetic, yet common, isn't it?
A lot of it depends on exactly *what* that 0.25% THD is comprised of. THD (which often gets confused with THD+N) expressed as a percentage is really not all that helpful, because we don't know what the harmonic profile looks like. If that 0.25% is mostly second order, it will likely sound very different than if it is mostly higher order.
0.25% THD from a really badly designed solid-state amp (or a solid-state amp that's way outside of its normal operating range) is likely to be far more offensive than the 0.25% residual from a McIntosh tube amp.
It's also worth remembering that tube amps tend to be rather noisy in comparison with their solid-state counterparts, and also comparatively low-power, so if the measurement is just THD+N, that can, in some cases, affect the number as well.
Interesting issue. My habit when hearing distortion is to turn down the volume to a level that's "acceptable" and then, for some material even 10% is acceptable (for a while). However it also depends on the source material. When limited to for instance 300...3000 Hz, distortion generated by the playback system will be outside that limited spectrum (no longer masked) and can become annoying at just a few % THD.