Take this 2-minute test of the audibility of EQ errors and distortion when driven by clicks and pops.
The attached file contains four 30-second segments of an impulse with a 10Hz repetition rate. To that, I have added at a 0.5Hz repetition rate:
(1) Time 0-30s: EQ +3dB at 20KHz
(2) Time 30-60s: EQ +3dB at 200KHz
(3) Time 60-90s: 2KHz sine at 1%
(4) Time 90-120s: 2KHz sine at 0.01%
(1) and (3) calibrate the listener's expectations.
(2) is representative of the +1 error in phono pre-amps with non-inverting active EQ and no correction.
(4) is a proxy for distortion.
In all cases, the small difference is switched on and off at a 0.5Hz rate. You are listening for a small change every second.
The impulse is a challenging waveform due to its flat frequency spectrum and high peak-to-average. The software-generated waveforms are perfectly bandwidth-limited to 20KHz.
Ed
The attached file contains four 30-second segments of an impulse with a 10Hz repetition rate. To that, I have added at a 0.5Hz repetition rate:
(1) Time 0-30s: EQ +3dB at 20KHz
(2) Time 30-60s: EQ +3dB at 200KHz
(3) Time 60-90s: 2KHz sine at 1%
(4) Time 90-120s: 2KHz sine at 0.01%
(1) and (3) calibrate the listener's expectations.
(2) is representative of the +1 error in phono pre-amps with non-inverting active EQ and no correction.
(4) is a proxy for distortion.
In all cases, the small difference is switched on and off at a 0.5Hz rate. You are listening for a small change every second.
The impulse is a challenging waveform due to its flat frequency spectrum and high peak-to-average. The software-generated waveforms are perfectly bandwidth-limited to 20KHz.
Ed
Last edited:
Ed,
"Du bist ein Teufelskerl"
You are a son of a gun 👍🙂.
greetings,
HBt.
"Du bist ein Teufelskerl"
You are a son of a gun 👍🙂.
greetings,
HBt.
And the error is?
ETA - I see it. I will fix. Unfortunately, the board allows only choices to be added.
Ed
ETA - I see it. I will fix. Unfortunately, the board allows only choices to be added.
Ed
I designed the test to make the EQ and distortion apparent. It is testing one's ability to hear small effects.
Ed
Ed
The second and third choices are the same. I will appreciate if you can delete the second choice.
Ed
Ed
@EdGr , if you wanted to use a real life vinyl click instead of artificial impulses that do not exist as a resulting voltage from any phono cartridge, then you are welcome to do do. The click is recorded without RIAA EQ, useful signal is 315Hz sine. I am probably missing something, but I do not understand the purpose of your test.
@PMA - Thanks for the sample!
I am using an impulse because a bandwidth-limited version can be computed in the frequency domain. I expect that the impulse makes the EQ error more apparent than a real click.
The purpose is to test whether people can hear small effects (i.e. the +1 error and 0.01% distortion).
Ed
I am using an impulse because a bandwidth-limited version can be computed in the frequency domain. I expect that the impulse makes the EQ error more apparent than a real click.
The purpose is to test whether people can hear small effects (i.e. the +1 error and 0.01% distortion).
Ed
I should have mentioned that computer monitors will not play the waveform loudly enough without distortion.
So far, we have 23 downloads and 3 votes. Please vote.
Ed
So far, we have 23 downloads and 3 votes. Please vote.
Ed
EdGr, I'm not sure what your presentations 3 & 4 are.
Is '3' a 2kHz sine at -40dB full scale (FS) ? It doesn't sound distorted to 1% to me.
If so, is '4' a 2kHz sine at -80dB FS?
Is '3' a 2kHz sine at -40dB full scale (FS) ? It doesn't sound distorted to 1% to me.
If so, is '4' a 2kHz sine at -80dB FS?
We have enough votes to start drawing conclusions.
That 2/6 people could hear the "+1 error" is more than I expected. This error is only +0.04dB and -6 degrees at 20KHz. The results support the belief that phono pre-amps that do not have the +1 error seem quieter.
That no-one (so far) could hear the 2KHz sine at 0.01% is showing that the impulse masks low-level signals due to its broad spectrum. In my previous "THD challenge", one-quarter of people could hear a seventh harmonic at 0.01%, but in the presence of an impulse, distortion needs to higher to be audible.
Thanks to everyone for listening to clicks!: 🙂
Ed
That 2/6 people could hear the "+1 error" is more than I expected. This error is only +0.04dB and -6 degrees at 20KHz. The results support the belief that phono pre-amps that do not have the +1 error seem quieter.
That no-one (so far) could hear the 2KHz sine at 0.01% is showing that the impulse masks low-level signals due to its broad spectrum. In my previous "THD challenge", one-quarter of people could hear a seventh harmonic at 0.01%, but in the presence of an impulse, distortion needs to higher to be audible.
Thanks to everyone for listening to clicks!: 🙂
Ed
The result also surprises me a lot, because I know which group I belong to now, I know where my cross is.
I also thank you for both testing options. Great.
HBt.
I also thank you for both testing options. Great.
HBt.