A Difficult Album for Your System to Test

[honn]

I just listened to both songs and passages twice, on low volume and medium volume (but since I have no db-meter around, you'll just have to take that for what it's worth...) and could hear no distortion. (speakers: Technics SB-R1)

the distortion coincides with bands around 2 - 3 kHz. If your speakers have a FR dips around that area, that may explain why. I think the distortion is inherent in the track.

 

[Koloth]

the distortion coincides with bands around 2 - 3 kHz. If your speakers have a FR dips around that area, that may explain why. I think the distortion is inherent in the track.

Would that be reflected in these measurements?

First one bi Hifitest, september of 2015. I am listening at about 15°, where there's at most a 4db deviation from 500 hertz.

Second by Audio Test, june of 2015. I don't see any significant deviation here at all.

I just listened to the track again. Loud. There is no distortion, I can make out. His voice is loud, piercing and there is a kind of warble inherent in his voice when he bells out. Thats on the entire track though.

 

[honn]

Would that be reflected in these measurements?

First one bi Hifitest, september of 2015. I am listening at about 15°, where there's at most a 4db deviation from 500 hertz.
View attachment 102185

Second by Audio Test, june of 2015. I don't see any significant deviation here at all.
View attachment 102186

I just listened to the track again. Loud. There is no distortion, I can make out. His voice is loud, piercing and there is a kind of warble inherent in his voice when he bells out. Thats on the entire track though.

Thanks! In fact, on many speakers if we play it loud the distortion will likely be masked by other frequencies and thus reduced or inaudible, depending on the speakers' frequency response characteristics. You should try to listen on very very low volume, I think the distortion will be heard.

 

[ernestcarl]

There is no distortion, I can make out. His voice is loud, piercing and there is a kind of warble inherent in his voice when he bells out.

As honn suggested, just focus on listening within the specified time intervals. Call it warble/crackle/distortion or whatever, as he sings at his loudest most people should be able to hear something quite deviating from 'smooth' if the volume is set low. You can also try different monitors/headphones.

I cannot myself hear the distortion at loud volumes on both the Neumann KH120 and the Sceptre S8. I might be going deaf. Who knows... I haven't had my own hearing tested. At low volumes, though, I hear it clearly. I kind of suspect it's JD's microphone clipping, and as mentioned prior by others, those sections correspond with peaks around 3k coming from his own voice.

It would be interesting to know how they recorded the tracks in detail...

 

[Koloth]

As honn suggested, just focus on listening within the specified time intervals. Call it warble/crackle/distortion or whatever, as he sings at his loudest most people should be able to hear something quite deviating from 'smooth' if the volume is set low. You can also try different monitors/headphones.

I cannot myself hear the distortion at loud volumes on both the Neumann KH120 and the Sceptre S8. I might be going deaf. Who knows... I haven't had my own hearing tested. At low volumes, though, I hear it clearly. I kind of suspect it's JD's microphone clipping, and as mentioned prior by others, those sections correspond with peaks around 3k coming from his own voice.

It would be interesting to know how they recorded the tracks in detail...

That is fascinating. Following the suggestions, I do now hear the crackle of seeming distortion at 04:02-04:04 of the Granada track. It is definitively not the "warble" I described before. I can distinguish vocal peculiarities and the crackle of distortion. This is the latter. I only hear it at (really) low volumes, as you predicted.

This phenomenon, where distortion is only audible at a volume lower than x and becomes inaudible above x is very much new to me. I figured distortion would become more apparent and jarring the louder one played something and understood talk of masking to relate only to adjacent frequency bands that - if the speaker deviates sufficiently in those areas - would overpower the information in the lower amplitude frequencies. Seeing that the phenomenon discussed here is volume-dependent, I guess it is the amplitude-dependent frequency sensitivity of my own hearing that causes this 'masking' (cf. Fletcher-Munson curve)?

 

[ernestcarl]

This phenomenon, where distortion is only audible at a volume lower than x and becomes inaudible above x is very much new to me. I figured distortion would become more apparent and jarring the louder one played something and understood talk of masking to relate only to adjacent frequency bands that - if the speaker deviates sufficiently in those areas - would overpower the information in the lower amplitude frequencies. Seeing that the phenomenon discussed here is volume-dependent, I guess it is the amplitude-dependent frequency sensitivity of my own hearing that causes this 'masking' (cf. Fletcher-Munson curve)?

Don't know if you watched the video: https://www.audiosciencereview.com/...bum-for-your-system-to-test.18826/post-616965

I tried to see if I EQ'd the peak the audible distortion would go away.

Well, nope. It's still there. Clipping can produce a lot of higher order harmonic distortion so this unwanted artifact is probably baked-in the vocal track. In that case, and as @watchnerd stated, this kind of defect (likely) isn't something that can be fixed.

 

[MakeMineVinyl]

Would that be reflected in these measurements?

First one bi Hifitest, september of 2015. I am listening at about 15°, where there's at most a 4db deviation from 500 hertz.
View attachment 102185

Second by Audio Test, june of 2015. I don't see any significant deviation here at all.
View attachment 102186

I just listened to the track again. Loud. There is no distortion, I can make out. His voice is loud, piercing and there is a kind of warble inherent in his voice when he bells out. Thats on the entire track though.

On the previous page of this thread I show the waveform zoomed-in on one of the 'soft clipped' peaks. I am not aware of any measurement method which can tell when distortion occurs on dynamic material such as music if it is already hard-coded into the original recording. Even if there were a method of measuring it, there would be nothing to indicate. A screaming lead electric guitar might have '100% distortion' as an effect, but any measuring method would not know whether that distortion should be there or not - it would have no idea. The only practical method is what I've already done; if there is a suspicious point where there is 'distortion' that should not be there, the waveform must be looked at to help identify the cause.

 

[MakeMineVinyl]

It would be interesting to know how they recorded the tracks in detail...

The answer is - poorly.

 

[honn]

The answer is - poorly.

hahah spare some mercy for the Decca engineers! Recording the orchestra is already difficult, add to that some superhuman vocal power of opera singers! haha but yeah, this album wasn't really recorded as good as it can. The subsequent Decca album by the same artist, L'amour (2014) is really well-recorded.

 

[ernestcarl]

The answer is - poorly.

C'mon... It's not that bad.

Man, there are so many worse recordings out there...

For example:

L’elisir d’amore “Una Furtiva Lágrima -- Alfredo Kraus

This one released by Fonal Music is particularly egregious.

Compare that to another recording of the same artist:

Unfortunately, I can still hear some small amount of distortion but nothing I would call very poor -- I'd say it's acceptable.

 

[MakeMineVinyl]

C'mon... It's not that bad.

Man, there are so many worse recordings out there...

For example:

L’elisir d’amore “Una Furtiva Lágrima -- Alfredo Kraus

This one released by Fonal Music is particularly egregious.

Compare that to another recording of the same artist:

Unfortunately, I can still hear some small amount of distortion but nothing I would call very poor -- I'd say it's acceptable.

I rate that album as poorly recorded because of the compression and excessive brightness. As far as the clipping, the engineer just got caught with his pants down.

 

[honn]

C'mon... It's not that bad.

Man, there are so many worse recordings out there...

For example:

L’elisir d’amore “Una Furtiva Lágrima -- Alfredo Kraus

This one released by Fonal Music is particularly egregious.

Compare that to another recording of the same artist:

Unfortunately, I can still hear some small amount of distortion but nothing I would call very poor -- I'd say it's acceptable.

MAESTRO KRAUS~!!

 

[ctrl]

I tried to see if I EQ'd the peak the audible distortion would go away.
Well, nope. It's still there. Clipping can produce a lot of higher order harmonic distortion so this unwanted artifact is probably baked-in the vocal track.

In contrast to a loudspeaker, where it is not possible to remove harmonic distortion with a filter (sometimes it works by changing the crossover frequency), it should actually be possible with a recording (since the distortions are already directly included in the signal).

However, the filter must act in the frequency range of the suspected harmonic distortion, i.e. at twice the fundamental frequency for HD2 and three times for HD3,...

To do this, cut a short music passage with the location of the audible distortions and make a frequency analysis. Then find the fundamental of the distortions and find the harmonic distortions that match it.

Here is the analysis from a short music passage of the Granada piece. My guess would be that the fundamental around 2530Hz at -5dB, causes harmonic distortion. At 5060Hz there is HD2 with -40dB. That is about 1.8% related to the fundamental, which could be quite audible.
Possibly also 0.9% HD3 (I'm not sure about that, because the frequency doesn't match 100% and I don't use Audacity enough to be able to judge it)

In this case, a filter around 5060Hz would have to be set to suppress the presumed HD2.

However, at 5060Hz it could still be an overtone of the singing voice, then it would not be called harmonic distortion or the orchestra also could produce tones in this frequency range.

 

[ernestcarl]

In contrast to the loudspeaker, where it is not possible to remove harmonic distortion with a filter (sometimes it works by changing the crossover frequency), it should actually be possible with a recording (since the distortions are already directly included in the signal).

However, the filter must act in the frequency range of the suspected harmonic distortion, i.e. at twice the fundamental frequency for HD2 and three times for HD3,...

To do this, cut a short music passage with the location of the audible distortions and make a frequency analysis. Then find the fundamental of the distortions and find the harmonic distortions that match it.

Here is the analysis from a short music passage of the Granada piece. My guess would be that the fundamental around 2530Hz at -5dB, causes harmonic distortion. At 5060Hz there is HD2 with -40dB. That is about 1.8% related to the fundamental, which could be quite audible.
Possibly also 0.9% HD3 (I'm not sure about that, because the frequency doesn't match 100% and I don't use Audacity enough to be able to judge it)
View attachment 102380
In this case, a filter around 5060Hz would have to be set to suppress the presumed HD2.

However, at 5060Hz it could still be an overtone of the singing voice, then it would not be called harmonic distortion or the orchestra also could produce tones in this frequency range.

That would be interesting indeed for someone to try out... sounds doable enough in DAW since we would want to limit EQ to that section only. BUT, that is far more involved than someone like me is willing to take for a few measly seconds of audible (volume dependent) distortion. LOL

One would also have to purchase a CD or digital copy -- maybe record it -- I am only streaming the album.

It's so much easier and faster to just find another recording of the same song...

 

[MakeMineVinyl]

In contrast to a loudspeaker, where it is not possible to remove harmonic distortion with a filter (sometimes it works by changing the crossover frequency), it should actually be possible with a recording (since the distortions are already directly included in the signal).

However, the filter must act in the frequency range of the suspected harmonic distortion, i.e. at twice the fundamental frequency for HD2 and three times for HD3,...

To do this, cut a short music passage with the location of the audible distortions and make a frequency analysis. Then find the fundamental of the distortions and find the harmonic distortions that match it.

Here is the analysis from a short music passage of the Granada piece. My guess would be that the fundamental around 2530Hz at -5dB, causes harmonic distortion. At 5060Hz there is HD2 with -40dB. That is about 1.8% related to the fundamental, which could be quite audible.
Possibly also 0.9% HD3 (I'm not sure about that, because the frequency doesn't match 100% and I don't use Audacity enough to be able to judge it)
View attachment 102380
In this case, a filter around 5060Hz would have to be set to suppress the presumed HD2.

However, at 5060Hz it could still be an overtone of the singing voice, then it would not be called harmonic distortion or the orchestra also could produce tones in this frequency range.

While that sounds feasible from a technical standpoint, if you were to actually process that section you would very likely produce an area which is more distracting to the ear than the original distortion in the context of the sound before and after.

 

[Bugal1998]

Thanks! In fact, on many speakers if we play it loud the distortion will likely be masked by other frequencies and thus reduced or inaudible, depending on the speakers' frequency response characteristics. You should try to listen on very very low volume, I think the distortion will be heard.

Well what do ya know... when I set the volume not just low, but very low there is in fact crackling (distortion?) present. Heard it on the HD800s and my 305p (haven't tried any other speakers). I can hear the distortion and turn up the volume as it's playing and it 'goes away'. Not sure if it's perceptual masking or limitations of the speakers.