Distortion Listening Test

[anmpr1]

I think if you are under 1% (that's one percent) and are listening to real music, and not test tones, it's all going to be about the same--at least from a "Can I hear this in my living room?" standpoint. Note: this is not an argument that electronics ought to have that much distortion.

My question is: how additive is the distortion. Let's say your gear is producing 0.1%, but your loudspeakers are at 5%. Does this equate to 5.1% at the listening position?

Totally unrelated, but reminds me of the time I bought the Mobile Fidelity half-speed original mastered LP of a Rolling Stones record. Pressed on pristine heavyweight JVC virgin vinyl encased in a a special warp-defying non-shrink wrapped biodegradable save-the-planet sleeve, blessed by the Salvador Dali Lama.

Anyhow, on the front of the record jacket was a small sticker that said something like: When you play this record you will hear distortion in certain passages. That is not a fault of the record or our mastering, It is there on purpose. It is a trademark of the band.

 

[Tks]

I ask that question a while back. Basically it seems to depend on the type of distortion. Cat remember what I was told precisely, but for harmonics that are exactly the same (like if there is a distortion on both products at 3kHz, then the distortion of the worse devices swamps over the one with lesser, but if one devices has no distortion at 3kHz, and instead is only at 4kHz, then you’ll have the THD measurement rise in numerical fashion, though I’m not sure at what point is it multiplicative, and which point it’s additive).

 

[Alexanderc]

I have a couple of questions about the listening test. First, what options should I choose? I don’t understand why I would choose one over the others except for the playback method (headphones vs. laptop speakers vs. full range speakers...). Second, I took the test twice, once on my main stereo system and once on my MacBook Pro (which was sitting on my lap), and got the same score. All parameters were the same except I chose the corresponding playback source. Is that supposed to happen? Does that mean my stereo system isn’t lower distortion than my laptop?

 

[Blumlein 88]

I think if you are under 1% (that's one percent) and are listening to real music, and not test tones, it's all going to be about the same--at least from a "Can I hear this in my living room?" standpoint. Note: this is not an argument that electronics ought to have that much distortion.

My question is: how additive is the distortion. Let's say your gear is producing 0.1%, but your loudspeakers are at 5%. Does this equate to 5.1% at the listening position?

Two reasons you can't add distortion quite like that. One is harmonic content. If your gear is .1% 3rd harmonic only, and your speaker is some mixture of 2nd and 3rd, then obviously they don't just add up. Another reason is various circuits have distortion at the harmonics at different phase relative to the signal. You can see this if you look at distortion using REW test signals. It is possible if both your gear and your speaker were 3rd harmonic only the two harmonics could be out of phase and combined you get 4.9% as a result.

 

[GelbeMusik]

Let's break the ice here - first with music, …

Had a change from AKG702 to Sennheiser58xx and scored -42 with T.C BL only. I still don't like this kind of bling-bling musical style. That's not the program I chose my gear for, if that is relevant.

- I love Tracy Chapman's music and found her recording to be decent back then, but ...

No, I found this hype immediately terrible. Hifi test music, the good woman really does not deserve this label. I despised the boys who had seriously taken this record as relevant, maybe even representative for their music listening. Oh, ah, yes, mhhh, I can remember only too well, and I still feel cold rushes running down my back.

Or Yello, or Dire Straits, kind of cruel too.

Anyway, today, I think the sound just sucks. The mix is just dead, unimaginative, artificial, cramped. Doesn't anyone hear the compression? The spectacularly applied cherries on top? The same goes for the Mrs. Jones. No friggin' way!

Well, I only listened to the first few seconds, and still managed to improve my score from -22dB to -33dB (BJ, all in) in a second pass (or so). I first had to teach myself which version would be the "real" one. So strange to me is this kind of entertainment with hi-fi fake music.
Not to mention the fact that the playback device at home, in my case an EQ'ed AKG702 and a Sennheiser 58xx, on themselves can produce -40dB distortion at higher volume, which then starts from the deep bass. Either that, or the compressor mentioned above.

With this test, with all good intentions, Prof Klippel does not give himself a good reference.

On the other hand, it is all too clearly shown, that HD/IM matters! There is no use in trying to do the HiFi with ever smaller speakers. The real thing starts at about an 8" bass/mid. Maybe plus sub, no less.

 

[outerspace]

Feels like I'm some kind of unique. I just hear the subtle hiss on the first second of distorted track. Or maybe test was changed during the year?

 

[JohnYang1997]

Feels like I'm some kind of unique or something. I just hear the subtle hiss on the first second of distorted track. Or maybe test was changed during the year?
View attachment 82049View attachment 82048

That's not what it's for. Yeah, it's a way to hack the test.

 

[outerspace]

That's not what it's for. Yeah, it's a way to hack the test.

But this hiss is part of distortion, no? And why nobody spot this in this thread? All listened first seconds of audio.

 

[j_j]

Distortion is distortion. If you can hear a difference in ABX or ABC/hr, it's different.

 

[JohnYang1997]

But this hiss is part of distortion, no? And why nobody spot this in this thread? All listened first seconds of audio.

Of course not. I remember someone mentioned before. Not sure if it's this thread.

 

[MediumRare]

But this hiss is part of distortion, no? And why nobody spot this in this thread? All listened first seconds of audio.

No. Hiss is added sound of frequencies related to the electronics. Distortion is generally added tones related to the signal. For example, harmonic distortion is made of of new tones at specific multiples 92x, 3x, 4x, etc.) of the signal. They are both modifications of the output different from the input, but their causes and effects are different. (EEs, feel free to correct my layman’s explanation.)

 

[j_j]

No. Hiss is added sound of frequencies related to the electronics. Distortion is generally added tones related to the signal. For example, harmonic distortion is made of of new tones at specific multiples 92x, 3x, 4x, etc.) of the signal. They are both modifications of the output different from the input, but their causes and effects are different. (EEs, feel free to correct my layman’s explanation.)

If something added a hiss, something added a hiss, and it's audible. And no, it doesn't have to be "related to the electronics". I can add hiss to anything in Matlab, no "hardware" need exist. You can get hiss from limit cycles of a recursive codec. There are many ways to get 'hiss', and if it's audible, it's audible.

Is it "distortion"? That's a semantic argument. It's an audible artifact, and that's the real question.

 

[MediumRare]

If something added a hiss, something added a hiss, and it's audible. And no, it doesn't have to be "related to the electronics". I can add hiss to anything in Matlab, no "hardware" need exist. You can get hiss from limit cycles of a recursive codec. There are many ways to get 'hiss', and if it's audible, it's audible.

Is it "distortion"? That's a semantic argument. It's an audible artifact, and that's the real question.

I don’t dispute your comment at all. I only point out that using Matlab or recursive codecs are synthetic cases; I (and I believe the poster) were referring to a more typical audio systems set up for listening to music.

 

[j_j]

All analog audio is "synthetic" in some sense, isn't it? It's using electronics and storage to capture a teeny tiny bit of a soundfield for future playback. Yes? No?

 

[bobbooo]

Feels like I'm some kind of unique. I just hear the subtle hiss on the first second of distorted track. Or maybe test was changed during the year?
View attachment 82049View attachment 82048

So we have a 'tell' at both the beginning and end of the track. Even if they're not consciously listening for the tell, the listener may subconsciously register it and so cause them to identify the distorted track. This seems to void all results of this test and render invalid any individual or collective claims/conclusions from it of distortion audibility thresholds when listening to actual music.

 

[outerspace]

Of course not. I remember someone mentioned before. Not sure if it's this thread.

@draftcombo mentioned noise at the end. Nobody said about start of the track in this thread.

To be clear, what I called "hiss" maybe not an actual hiss but part of IMD or other distortions, who knows? It's too subtle to decide. You all can now pay attention to this and tell.

 

[JohnYang1997]

@draftcombo mentioned noise at the end. Nobody said about start of the track in this thread.

To be clear, what I called "hiss" maybe not an actual hiss but part of IMD or other distortions, who knows? It's too subtle to decide. You all can now pay attention to this and tell.

It's not 'who knows'... If it's hiss it's obviously test artifacts. The IMD is the overtone appearing in the ears. It should be very clear that in low IMD range -40db you can still hear the IMD but on both of them. That's because you will be hearing the IMD in the ears not the source. (It could be headphones/earphones but usually earphones have lower distortion).

 

[outerspace]

The IMD is the overtone appearing in the ears.

IMD can be added to the source, why not. It can appear in DACs, amplifiers and even at speakers. Why you associate them only with the ears? The question is what kind of distortions was added to the test. Also I want to note this "hiss" seems became more subtle with lowering of overall distortions. So it's seems like part of them.

Do you hear this thing at the start of distorted track? How would you describe it?

 

[JohnYang1997]

IMD can be added to the source, why not. It can appear in DACs, amplifiers and even at speakers. Why you associate them only with the ears? The question is what kind of distortions was added to the test. Also I want to note this "hiss" seems became more subtle with lowering of overall distortions. So it's seems like part of them.

Do you hear this thing at the start of distorted track? How would you describe it?

Didn't you start with the most distorted one? That's not hiss right? As the level decreases, the only thing that decreases is the distortion. New thing that appears is not part of the test. Or did you mean it's not actual hiss but some noises that's generated that sounds close to hiss or buzz?

 

[outerspace]

Didn't you start with the most distorted one? That's not hiss right? As the level decreases, the only thing that decreases is the distortion. New thing that appears is not part of the test. Or did you mean it's not actual hiss but some noises that's generated that sounds close to hiss or buzz?

It's actually too subtle to decide is it correlated with signal magnitude or not. But it's almost for sure correlated with overall distortions level. So at high distortions it's seems easier to hear than on low. Maybe someone can listen to it and correct me.

Didn't you start with the most distorted one?

Yes.

That's not hiss right?

Distortions of lower frequency region which dominate at the start of the test are not like hiss for sure.