Curious. Does anyone know at what point THD numbers become worthless? Can you hear the difference between .04 and .05?
I'm not sure it's a case of 'hear it' - it's not like tracking distortion - more a sense of 'fatigue' you notice while listening a for a period to music. The absence of it is what I notice.
First Question;No and if my memory serves me correctly I recall an old article that in amplifiers distortion can't be detected below 3%. Also in the low frequencies smaller woofers distort on the order of 10% or more. I pretty sure tube amps if realistically measured across the full bandwidth are close to 1% THD and IM. Under certain conditions IM is more audible than THD. I'm sure my comments will open a can of worms so lets have the worms make a good showing.
Under controlled listening conditions the human ear can't detect anything under 1%. If people say they can they are fooling themselves.
I agree with VinylSoul. Worm free post. No worms for you yet from me.... I concur. Speakers are far more liable to distortion than even the poorest High Fidelity amp designs. That's why choosing good low distortion speakers is so important... And electronics distortion is of lesser importance, but not of NO importance. Speakers of typical design produce 3%, 5% or more harmonic distortion over much of the audio range. An order or two of magnitiude greater than most electronics driving them. Tubes and transformers designs are far higher in THD distortion than a good modern silicon type amp. But they impart other qualities to the music that pleases the ear, (to many people) more than the distortion displeases the ear. Somthing about the dynamics is different with a good tube amp... I have been trying to put my "finger" on it. Yes, humans will likely not hear the difference of .01% amp THD at all. Totally swamped by transducer THD.
Not a chance. As has been correctly pointed out, detecting anything below 1% is problematic, let alone a difference below that number that is one-hundredth of 1%.
While it's true that speakers have massive distortion compared with electronics, this does not preclude extremely subtle defects elsewhere in the chain from being heard under the right conditions.
Decades ago, I had my Dynaco Pas 3 and Stereo 70 amp tested at a local shop, who was advertising free distortion tests on any gear (probably just to get people into their store). I was aghast to find that my Stereo 70's distortion was, well, sorta below 1.0% at the lower frequencies, but climbed like the side of Mt. Everest as the frequenceis rose above 1K. By the time they were at 15K, the distortion was well over 5.0%! No misprint! I took consolation in the fact that not once had I ever heard anything like distortion while listening to those Dynas. I realized that THD was just a measurement that probably didn't have much to do with my listening pleasure. I've never paid any attention to those figures since.
I agree with the others that once you get below 1 percent, for most types of distortion anyway, it's inaudible and thus irrelevant. Another way to look at this is to treat distortions by their dB equivalents. One percent distortion means all of the distortion components together are 40 dB below the program. 40 dB is pretty soft! But not always inaudible because it depends on the masking effect. I tested this once by purposely creating "harmonics" in an audio editor program that were far removed from the fundamental frequency to minimize masking. As I recall I used 100 Hz for the fundamental, and inserted new frequency components around 3 KHz where the ear is most sensitive. At 40 dB down the added frequencies were soft but audible. By 60 dB - equal to 0.1 percent distortion - it was very difficult to hear even switching them on and off. For normal types of distortion I'd think 60 dB would be totally inaudible. --Ethan
I have read that anything under 3% THD is inaudbile. You rarely see anything over 1.5% THD mentioned in specs, and that is usually at clipping. Tactics used to reduce THD, such as use of large amounts of negative feedback, often cause the component to sound worse than the THD does! THD is not something anydoby on this forum should be worried about. It's almost a meaningless number, trumpeted by mass market brands in an attempt to give the average consumer a kind of "comparison figure" to gauge how a product sounds. Funny thing is that many products with .00001% THD sound like crud, while ones with 1% THD sound fantastic. Utterly worthless statistic, IMHO.
Do mean ever order harmonics of .04% or odd order harmonics of .04%? I have an RCA Tube manual from the 1950's that suggests that THD should be scrapped as an evaluative measurement for music reproduction.
Very interesting article that also is a good argument to put to rest the objectivist vs. subjectivist view of things.
Yeah, this article is unfortunately closely tied to the tube DIY world, so it really has not gotten out to the rest of the audio universe. It was first published about ten years ago in Glass Audio Magazine. Regards Robert
When the FCC used to require proof tests for radio stations, we would shut off all of the processing equipment and run tones to meet 5% distortion at a maximum. I always wondered what the level was with the processing turned back on. Of course they don't require that anymore that I know of. That was 23 years ago.
But - the issue isn't simply about use of negative feedback, it's how well negative feedback is implemented in such designs, in order to avoid phase shift problems. For example, the Harman-Kardon Citation II is to my ears one sweet sounding vintage tube power amp, and yet I believe it featured perhaps the most heavy use of negative feedback of any of the "classic era" vintage tube amps.
Doesn't it matter if it's primarily even ordered vs. odd ordered distortion? I have always been led to believe the human ear is much more sensitive to odd-ordered harmonic distortion and that even-ordered can even sound pleasing to the ear. And this logic leads to the reason why tube gear sounds good. Or something.
True. It's a question of designing the amplifier to produce the least Thd innately and then adding just enough Negative feedback to lower it in the region of .04% or there abouts without losing anything in the sound quality and 'timing'. Sometimes designs are done relying very much on high levels of feedback to cure Thd which just bad engineering practise. The amount of negative feedback applied isn't a statistic to be bragged about by the sales dept. thats for sure.
Another point to consider is cumulative distortion: If the amp puts out 0.1% THD, but say, 5% at low frequencies, and it goes through a speaker that is also 5%, you have more than 5% THD.
If I remember correctly, that statement was made by the late Julian Hirsch when he reviewed equipments for Stereo Review, the predecessor of Sound and Vision. He did not specifically mention odd or even order distortion for the THD measure. In his view, a 0.1% THD sounds no different from a 0.03% THD to most human ears.
I am not surprised that .1 THD will sound no different from a .03 THD as the human ear simply can't hear THD under 1%. If anyone thinks they can their fooling themselves. What they are doing is perceiving the sound differently. Under laboratory test conditions anything under 1% is just not audible to the human ear!