Test your ears: THD

[ijedouglas]

Interesting video with a cool ear test for THD

https://www.youtube.com/watch?v=sMziG4TrfXc&t=908s

[martin manning]

Ha, just watched that this morning.

[ijedouglas]

Ha, just watched that this morning.

Did you get it?

Interesting his advice on the radial cap tops too. I don't trust them

[martin manning]

Did you get it?

I can hear the harmonic easily on my MacBook speakers with the volume at normal listening level for the fundamental, and I have an idea of which of the three samples has the harmonic in it, but the answer hasn’t been posted yet. I can also hear the beating that others reported even when just the fundamental was played, so I wonder if there isn’t something else going on. I listened again with (Bluetooth noise cancelling) headphones and got the same result.

[ijedouglas]

Did you get it?

I can hear the harmonic easily on my MacBook speakers with the volume at normal listening level for the fundamental, and I have an idea of which of the three samples has the harmonic in it, but the answer hasn’t been posted yet. I can also hear the beating that others reported even when just the fundamental was played, so I wonder if there isn’t something else going on. I listened again with (Bluetooth noise cancelling) headphones and got the same result.

Me too on my laptop speakers at normal listening volume. The samples with the harmonic in it were pretty easy to spot. BTW, you can validate your answers in his follow up video

[martin manning]

I had it right :^)

But this comment by frednerks should give pause:

"Hey Tony. Due to what I was hearing with the last video, I stuck my scope onto my soundcard out and for the initial pure tone labeled 1kHz, when zoomed out it showed the volume changed significantly in a repetitive fashion over time for that tone. So it was not a pure tone. This is the warble/beat tone I heard in the pure 1kHz, as well as the first and third sounds in the test. The first and third sounds in the test had this volume change in a repetitive fashion too. The second tone in the test had a smoother more steady volume with only slight variation in volume. I then put the scope in fft mode to look at the frequencies and it showed that the original 1kHz pure sound as well as all three sounds in the test had second and third harmonics present, though the first and third sounds in the test has the 2kHz harmonic slightly higher. I suggest the only reason people picked one and three is because they had the rhythmic volume change - rather than actually hearing the 2kHz tone. Can you try playing this from the youtube video (rather than your original sample) into your scope? Can you explain why the original sound marked pure 1kHz has the varying volume, while the second sound in the test looks different to the one marked 1kHz? I think this is just a bad test and needs to be redone without the varying volume in the pure tone. Then I expect selecting the one or ones with distortion would be random due to inaudibility."

[ijedouglas]

Interesting. A few of the comments mentioned the "warble". I wonder if YouTubes compression algorithm has something to do with it? Let's see if he updates the test

[pompeiisneaks]

I'd bet a ton of money that the audio compression is what does it. The MP4 video format container allows a TON of different audio formats, and even if he used a high quality one during recording (if he even had the ability to choose it, his camera audio recording system may have locked that) then Youtube does a big auto conversion from the format you upload to multiple different sized video, and may not even host the original version as playable if they find it too large or some other random criteria. Thinking youtube audio is as good as high quality wav or lossless formats like FLAC or OGG is not worth the effort, I'd liken it to bluetooth quality honestly BUT I'm not sure if there's a way to determine it.

I could hear the 2kHz sample for sure, but it was barely audible with no other noise, I do have hearing damage, but I believe it's in the higher frequencies. I thought I knew which one it was in, but I need to find the follow up video to be sure now.

~Phil

[pompeiisneaks]

I was wrong, or at least only partially right... I got two of the three right, but was wrong on one... so as to not give away the answer

I did find this and indeed they do standardize audio:

https://www.h3xed.com/web-and-internet/ ... 720p-1080p

And it seems like it's AAC which is still a lossy audio format.

~Phil

[roberto]

1st and 3rd samples have "beats" on them, so there are different frequencies in them, whilst 2nd is pure.
It is interesting to perform same tests with different harmonics, and how you can hear them more the higher their order is.

[TUBEDUDE]

I got all three with my phone speakers.
I have old damaged ears too.
This was easy but it was constant level with two tones. Not a reasonable sample of what you'd hear in dynamic music. Also, it was 2nd harmonic, which most find pleasing. Higher order odd harmonics are the ones we want to avoid, that and transient intermodulation distortion.

[roberto]

I would not generalize.

We are talking on an instrument amplification forum, and in some cases high order odd harmonics are part of the sound.
I would also say that the hype some amps with low PSU filtration got, is transient intermodulation distortion.

Going into hi-fi the situation changes, but even there the lowest distortion is not the only approach.
Again, it is something personal and cannot be generalized (you can prefer lowest THD and highest DF or better way to overdrive and medium DF, or... etc...).

[TUBEDUDE]

True Roberto,
Tonequest is a personal journey. I prefer cleaner tones generally. Many others do not.
Personal preferences and tone goals aside, my main point is, detecting distortion in a 2 tone harmonic distortion test is not the same as detecting it in dynamic music. Nor is it the only distortion mechanism present in music.

[Roe]

even on a crappy laptop at a noisy office clip 1 and 3 stod out.

[martin manning]

even on a crappy laptop at a noisy office clip 1 and 3 stod out.

But see the post above where someone measured the sound card output with a scope and FFT. Another fatally flawed attempt, Im afraid.