IPB

Welcome Guest ( Log In | Register )

Limitations of a synthetic tests (THD, IMD etc.)
IgorC
post Jan 13 2013, 21:50
Post #1





Group: Members
Posts: 1578
Joined: 3-January 05
From: ARG/RUS
Member No.: 18803



THD is measured on one single frequency (tone) at once, IMD with two tones.
So these measurements are made with synthetic signals while music and speech aren’t just pure tones but rather complex sounds.

For example IMD is calculated on two tones 19 kHz and 20 kHz like here. http://nwavguy.blogspot.com.ar/2012/04/odac-released.html
Just one example of possible many. IMD between 20 kHz and 19 kHz tones ends with 1 kHz tone (20-19 kHz) in audible band. But at the same time there can be 20.1, 20.11 … and 19.1, 19.11… kHz tones those will also cause the same 1 kHz tone in audible band. A classic IMD measurement is calculated only for one pair of tones and not for all other. Wouldn’t be it a huge oversimplification? I’m afraid a testing of audio equipment will be very hard on a real life signals (music, speech).

This post has been edited by IgorC: Jan 13 2013, 21:55
Go to the top of the page
+Quote Post
 
Start new topic
Replies (1 - 8)
saratoga
post Jan 13 2013, 22:01
Post #2





Group: Members
Posts: 5116
Joined: 2-September 02
Member No.: 3264



QUOTE (IgorC @ Jan 13 2013, 15:50) *
THD is measured on one single frequency (tone) at once, IMD with two tones.


Thats one way you can do it, but not the only way. RMAA for instance will do an IMD sweep. The link you posted measures THD as a function of frequency. So you can look at just a few frequencies, or you can look at lots.

QUOTE (IgorC @ Jan 13 2013, 15:50) *
A classic IMD measurement is calculated only for one pair of tones and not for all other. Wouldn’t be it a huge oversimplification? I’m afraid a testing of audio equipment will be very hard on a real life signals (music, speech).


Its only an oversimplification if you expect non-linearity to be strongly frequency dependent. Generally if a device is linear at a couple frequencies, its linear at all the ones in between as well, so just testing a subset of combinations gives you a very good idea of the overall performance.
Go to the top of the page
+Quote Post
IgorC
post Jan 13 2013, 22:19
Post #3





Group: Members
Posts: 1578
Joined: 3-January 05
From: ARG/RUS
Member No.: 18803



QUOTE (saratoga @ Jan 13 2013, 18:01) *
Its only an oversimplification if you expect non-linearity to be strongly frequency dependent.

Let's suppose if IMD for 19-20 kHz is the same as for 19.1-20.1 kHz. However the measurement of IMD with 4 tones 19, 19.1, 20 and 20.1 will be as twice as worse comparing to IMD with 2 tones (?).

IMD 19-20 kHz, spurious tone at 1 kHz -100 dB.
IMD 19.1-20.1 kHz, spurious tone at 1 kHz too - 100dB.
IMD with 4 mentioned tones would be twice. -97 dB (?)

Edit: math

This post has been edited by IgorC: Jan 13 2013, 22:34
Go to the top of the page
+Quote Post
greynol
post Jan 13 2013, 22:29
Post #4





Group: Super Moderator
Posts: 10253
Joined: 1-April 04
From: San Francisco
Member No.: 13167



Have you considered the math?


--------------------
Your eyes cannot hear.
Go to the top of the page
+Quote Post
IgorC
post Jan 13 2013, 22:33
Post #5





Group: Members
Posts: 1578
Joined: 3-January 05
From: ARG/RUS
Member No.: 18803



-97 dB. smile.gif

This post has been edited by IgorC: Jan 13 2013, 22:33
Go to the top of the page
+Quote Post
saratoga
post Jan 13 2013, 22:34
Post #6





Group: Members
Posts: 5116
Joined: 2-September 02
Member No.: 3264



QUOTE (IgorC @ Jan 13 2013, 16:19) *
IMD 19-20 kHz, spurious tone at 1 kHz -100 dB.
IMD 19.1-20.1 kHz, spurious tone at 1 kHz too - 100dB.
IMD with 4 mentioned tones would be twice. -94 dB (?)


Usually tests like this are done at full scale amplitude, so if you want to do 2x as many tones, each has to have half the intensity. Nonlinear terms are obviously not linear, so the IMD intensity does not necessarily scale in proportion, but probably using 4 tones will result in lower peak intensity since you will now have cross terms (eg. f1+f2-f3) that will land out of the audible spectrum.
Go to the top of the page
+Quote Post
IgorC
post Jan 13 2013, 22:58
Post #7





Group: Members
Posts: 1578
Joined: 3-January 05
From: ARG/RUS
Member No.: 18803



I see, will need to read about measurements and their standards more.

Thank You for your answers.
Go to the top of the page
+Quote Post
Mach-X
post Jan 14 2013, 07:08
Post #8





Group: Members
Posts: 288
Joined: 29-July 12
From: Windsor, On, Ca
Member No.: 101859



Unfortunately the "standards" are far too loose and allow for inferior equipment makers to make dubious claims. For example I feel that amplifier makers should be forced to run full 20-20k white noise before publishing power handling/thd levels rather than "100 watts x 6......" (but only if you run one channel at once at 1khz)
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Jan 14 2013, 14:39
Post #9





Group: Members
Posts: 4253
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Mach-X @ Jan 14 2013, 01:08) *
Unfortunately the "standards" are far too loose and allow for inferior equipment makers to make dubious claims. For example I feel that amplifier makers should be forced to run full 20-20k white noise before publishing power handling/thd levels rather than "100 watts x 6......" (but only if you run one channel at once at 1khz)


Despite all that, most audio gear with high fidelity pretensions is sonically transparent as used. It really doesn't take a lot of technology by 21st century standards to fool the ear.

Compared to electronics, speakers are major POS with orders of magnitude more distortion of just about every kind.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 24th November 2014 - 08:07