In Ambient measurement: THD% and IMD%

This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello how much should be a good THD% (@1W @1Khz) and a good IMD%...

I have no intuition about IMD in speakers.

But my impression is that for ordinary listening levels, THD below 1% is pretty acceptable for mid and treble speakers playing music. (My electrostatic panels run about -45dB, which includes fire trucks in the distance, etc.)

For woofers playing music, "acceptable" is pretty hard to nail down because bass music is full of harmonics and so distortion is less of an annoyance. Below 40 Hz, a woofer can rise above 10% and be considered OK.
Here's my 1956 AR-1, tested at ordinary listening levels.

BTW, tricky to do these measurements. As levels go up to a point, the measurements start looking better because the noise and background sound diminish in impact. But when getting loud, distortion rises, as you'd expect.


  • AR-1 distortion.jpg
    AR-1 distortion.jpg
    46.2 KB · Views: 138
Last edited:
At this point it looks that IMD using 60Hz + 7kHz doesnt make sense as 60 Hz comes already too much distorted even without summing the 7kHz tone.

How should be the two tones for Imd speaker measurement?

I thought long ago it was concluded that IMD didn't add information to HD sweep results. So IMD evaluation has gone away. (I lived through the period when IMD was heralded as a new great discovery thought to reveal bad sound quality not revealed by HD.)

There is no agreed benchmark or standard by which to judge IMD on an absolute or calibrated basis (as there is at least intuitively for HD), esp if you are changing the standard test set-up used for amps. On the other hand, if you are developing speakers, you can use any measurement you please iteratively as you improve the speaker.

I don't think there is anything less relevant in the whole world than loudspeaker THD @ 20 kHz. It even starts loosing relevance below 10 khz already.
I guess I know where your motivation to do so comes from: Amplifiers will usually have more THD at higher frequencies due to the dropping feedback factor.

To the OP: Is that a small fullrange driver that you are testing ?


Last edited:
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.