• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Are we looking at Distortion the wrong way?

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,425
Likes
7,941
Location
Brussels, Belgium
Reading about all the things KEF are doing to optimize their drivers made me think, how come they have mediocre distortion measurements?

But looking at the KEF LS50 meta there is nothing mediocre about it at all, if we look past the the second harmonic it has 3rd harmonic levels competetive enough with the Purifi woofers.

1680193059510.png


1680193127632.png


A look at the KEF Reference Meta shows incredible distortion performance as well (-60 on this graph is comparable to 26 dB on the previous ones, third harmonic hovers around -66, pay attention to the different colors used). It's compareable to the LS50 Metas.

Kef%20Reference%201%20Meta%20Harmonic%20Distortion%20%2886dB%20%40%201m%29.png


And like this got me thinking, there is sufficent reasearch on harmonic distortion and masking to make a relatively accurate ''weighting'' system for harmonic distortion, Would you welcome such a thing? do you think it's a good idea to make a more relevant THD ''score''.
 
Last edited:

thewas

Master Contributor
Forum Donor
Joined
Jan 15, 2020
Messages
6,755
Likes
16,199
I fully agree that 2nd cannot be compared to 3rd HD, although in my experience HD measurements of full loudspeakers are quite less correlating to audibility than multitone ones like S&R, Nuyes or Erin do so I would rather have these instead (for driver and loudspeaker design HD measurements are very helpful).
 

NTK

Major Contributor
Forum Donor
Joined
Aug 11, 2019
Messages
2,657
Likes
5,820
Location
US East
Here is the harmonic distortions criteria ANSI/CTA-2034 (section 8.2) uses to determine maximum loudspeaker peak SPL. The input signals for the measurements are 6.5 cycles raised cosine window (Hann window) shaped tone bursts. Center frequencies of the bursts are in 1/3 octave increments from 20 Hz to 5 kHz.

Frequency bands are band 1: 20 Hz to 50 Hz; band 2: 63 Hz to 200 Hz; and band 3: 250 Hz to 5 kHz, to account for our higher sensitivity to HD at higher frequencies.
cta_2034.png

[Edit] This is how the tone burst signal looks like (center frequency normalized to 1).

tone_burst.png
 
Last edited:

fineMen

Major Contributor
Joined
Oct 31, 2021
Messages
1,504
Likes
679
... a more relevant THD ''score''.
No score. Test the speakers individually for yourself and decide on personal preference. As a DIYer I've got a greater throughput of speaker boxes, drivers or alignments than the regular Joe. I prefer low distortion speakers for esthetical reasons, because I know, from measuring it, about the fact. Preocupation aside my peak sensitivity for distortion lies about 1kHz @0,3% HD3 with pure tones. HD4 etc comes below ambient noises and such quite quickly.

Seriously, why speculating about general limits and so called 'scientific' accessments while a decent headphone, a PC with soundcard and signal generator(s) are readily at hand. Doesn't make any sense to me except you hold a stereo as a show piece rather than as a source of enjoyment. Test yourself, your ears, your brain, as they put it, and, foremost, don't worry to miss!

Intermodulation IM, and especially Doppler (FM) transformed, by room interaction, to amplitude modulation (AM) must not be measured. It is just a not avoidable physical fact that can be calculated from the cone excursion and bandwith of the used individual drivers. Score from worst to best:

wideband single driver / two-way / no-way / three-way .. .. four way w/ waveguide ...


Nearly forgot: musical harmonics often greatly mask HD, but not IM. An upright bass has like 1000% of 'distortion' of the base tone. But not so with the electric guitar. If overdriven, which is commonplace, the rectangular waveform won't comprise that much of even harmonics.
 
Top Bottom