Fas42’s Stereo ‘Magic’

[Archimago]

16 hours ago, opus101 said:

 

Guilt by association? Actually what he's saying is true based on subjective experience - small FR deviations make less subjective impression than additive noise - sibilance for example. Sibilance gives an uncomfortable experience in listening longer term but an HF boost/dip of a dB or two can be borne, subjectively.

 

Harping on about sibilance for a while longer - for me its the first and most obvious sign a system is under-performing subjectively but to date I've not seen any objectivist attempt to make a measurement of it or even talk about it much. Must be a lacuna.

 

Wondering @opus101 if you can give me an example of a musical track that you use to evaluate for sibilance?

 

Good topic and I think it'd be much more useful to have some specifics to test out the assertion that there's any kind of lacuna. Sibilance is something I've generally found to be a function of the recording itself rather than DAC or separate from the frequency response of one's other hardware like speakers...

 

[Blackmorec]

 

2 hours ago, Archimago said:

 

Wondering @opus101 if you can give me an example of a musical track that you use to evaluate for sibilance?

 

Good topic and I think it'd be much more useful to have some specifics to test out the assertion that there's any kind of lacuna. Sibilance is something I've generally found to be a function of the recording itself rather than DAC or separate from the frequency response of one's other hardware like speakers...

 

Try the announcers’ voices on Swiss Radio Classics. Whenever I’m waiting for a new piece of kit to run in, sibilance will make an appearance at one stage or another during the process. A few days after the sibilance appears the treble will make a leap in SQ and the sibilance will disappear. 

 

[fas42]

3 hours ago, Archimago said:

 

Wondering @opus101 if you can give me an example of a musical track that you use to evaluate for sibilance?

 

Good topic and I think it'd be much more useful to have some specifics to test out the assertion that there's any kind of lacuna. Sibilance is something I've generally found to be a function of the recording itself rather than DAC or separate from the frequency response of one's other hardware like speakers...

 

 

Sibilance is THE easy test for playback distortion ... I don't have a single CD with sibilance "issues", 😁. Why? Because if the playback has any irritating distortion, then the sibilance sound in the human voice immediately strikes one as being incorrect, when heard over speakers. It is easy to hear that a certain recording has been made in manner such that the ess sound is more prominent, than on others - just as one can go close to a person speaking loudly, live, and be very aware of the strength of that part of his speech, or singing. But it sounds completely normal - because, well, that's how the human voice box is made ... if it's irritating to hear, on a recording, that's because it's serving as an obvious marker for playback distortion.

 

Plenty of other sounds also sound wrong with faulty replay - annoying sibilance just makes it really, really obvious ...

[Archimago]

2 hours ago, Blackmorec said:

 

Try the announcers’ voices on Swiss Radio Classics. Whenever I’m waiting for a new piece of kit to run in, sibilance will make an appearance at one stage or another during the process. A few days after the sibilance appears the treble will make a leap in SQ and the sibilance will disappear. 

 

Interesting. Will have a listen to Swiss Radio Classics. Would still be good to hear about a specific audio track that folks use as some kind of "standard" of what should or should not sound sibilant.

 

[I had a listen to Swiss Radio Classic - yup, female announcer sound harsh. To me it's also the lossy artifacts that I hear causing the issue/distortion.]

 

The other day, I was listening to Toni Braxton's self-titled album (1993) and I found the track "Seven Whole Days" as an example of a song that was sibilant at points with Toni's voice or the background singers through my headphones. To me, this is an example of a track where the sibilance is just "burned in". A bit of "de-essing" in the studio would be useful. Sure, I can EQ the highs on playback to reduce the general harshness of the song but it's the recording itself so there's no point in fooling with the hardware as this would affect fidelity for other material.

 

2 hours ago, fas42 said:

Sibilance is THE easy test for playback distortion ... I don't have a single CD with sibilance "issues", 😁. Why? Because if the playback has any irritating distortion, then the sibilance sound in the human voice immediately strikes one as being incorrect, when heard over speakers. It is easy to hear that a certain recording has been made in manner such that the ess sound is more prominent, than on others - just as one can go close to a person speaking loudly, live, and be very aware of the strength of that part of his speech, or singing. But it sounds completely normal - because, well, that's how the human voice box is made ... if it's irritating to hear, on a recording, that's because it's serving as an obvious marker for playback distortion.

 

Plenty of other sounds also sound wrong with faulty replay - annoying sibilance just makes it really, really obvious ...

 

Hang on a second Frank...

 

I don't know what kind of music you listen to, but if a music lover has an eclectic collection of albums including vocals, rock, pop, it would be rather unusual to not come across some tracks here and there where sibilance sounds a bit excessive I think.

 

Obviously if a system sounds sibilant frequently with known good recordings, then we have a problem! But a good system would be able to reproduce a sibilant track as recorded and this would have nothing to do with whatever "irritating distortion" you're talking about.

 

This is why I think talking in vague generalities like not having a "single CD with sibilance issues" suggests to me that this is more of a selection bias than any kind of tweak I would like to fool around with! More useful to have specific examples of music to listen for sibilance and judge if there is a problem or not.

 

[fas42]

22 minutes ago, Archimago said:

Hang on a second Frank...

 

I don't know what kind of music you listen to, but if a music lover has an eclectic collection of albums including vocals, rock, pop, it would be rather unusual to not come across some tracks here and there where sibilance sounds a bit excessive I think.

 

There will always be a variation, depending upon how the microphones, etc, were set up - the point is whether the natural sibilance of that voice is disturbing or not.

 

Quote

Obviously if a system sounds sibilant frequently with known good recordings, then we have a problem! But a good system would be able to reproduce a sibilant track as recorded and this would have nothing to do with whatever "irritating distortion" you're talking about.

 

Again, the naturalness of the sibilance is key - on YouTube it is quite easy to find a track being played on an ambitious rig, where the sibilance is out of control - this is distortion, pure and simple.

 

 

Quote

 

This is why I think talking in vague generalities like not having a "single CD with sibilance issues" suggests to me that this is more of a selection bias than any kind of tweak I would like to fool around with! More useful to have specific examples of music to listen for sibilance and judge if there is a problem or not.

 

 

Nothing like a bit of bias to explain everything, eh? 😉 You see, how it works is that you listen to a setup, and amongst other indicators, the level of sibilance distortion is quite offputting - yes, on a specific track, if you want that 🙂. So, you know the system is below par; and then go through your various diagnostic testing, etc, procedures to work out where the problems are - fix the issues ... voila, sibilance is now back to a normal, realistic standard for that track. Which is purely incidental - everything you play is now of a much higher, subjective standard - it's part of the deal when you sort out issues, 😉.

[fas42]

Where those of an objectivist stance do everyone else a deep disservice, is by poo-pooing all the various methods people use as workarounds for the lack of integrity of playback chains - I'm thinking here of the thread about the junilabs player. Because the workaround seems silly, is not easily measurable, it therefore is nonsense, has absolutely zero merit - is the objectivist's POV. But those who try them often find they are effective, at least to some degree - therefore, there is, "something going on". And zero progress in understanding is achieved, by the dumping of ridicule on those people's findings, by others who just don't like those sort of results.

[opus101]

7 hours ago, Archimago said:

 

Wondering @opus101 if you can give me an example of a musical track that you use to evaluate for sibilance?

 

Good topic and I think it'd be much more useful to have some specifics to test out the assertion that there's any kind of lacuna. Sibilance is something I've generally found to be a function of the recording itself rather than DAC or separate from the frequency response of one's other hardware like speakers...

 

 

Generally sibilance issues are an artifact of the playback system, IME. But I do have one or two recordings that exhibit a sibilant flavour and one (that I have no idea of the whereabouts right now) that does demonstrate it on female voice, not just subjectively but in Audacity FFT too. I will look out these examples.

[fas42]

Just looking up a classic example where poor standards specified, by the electronics industry, regularly create a headache for people - yep, those miserable Ethernet plugs, where the tiny sliver of piece of junky plastic that is intended to stop the plug coming adrift, snaps off at the slightest provocation. Meaning, a dodgy connector which is always likely to pull out, giving zero or poor contact - simple, throw the whole cable in the rubbish bin, and buy a whole new one, to properly fix it. Smart thinking there, by some individual, or committee 🙄 🙄 ...

 

Which has led to a whole industry of "snake oil" - things to fit onto the defective plug, to make it work a bit better - Blu Tack, rubber bands, you name it, tweaks to save throwing a perfectly good cable away ... all because the industry couldn't bother doing it properly, and hasn't done anything about it, since .

 

Why just now? Smart TV no longer smart - of course, because the cable pulled out ...

[Archimago]

1 hour ago, fas42 said:

Where those of an objectivist stance do everyone else a deep disservice, is by poo-pooing all the various methods people use as workarounds for the lack of integrity of playback chains - I'm thinking here of the thread about the junilabs player. Because the workaround seems silly, is not easily measurable, it therefore is nonsense, has absolutely zero merit - is the objectivist's POV. But those who try them often find they are effective, at least to some degree - therefore, there is, "something going on". And zero progress in understanding is achieved, by the dumping of ridicule on those people's findings, by others who just don't like those sort of results.

Oh yeah... This is good 😉.

 

What we're talking about here is neither "subjectivist" nor "objectivist". The poo-pooing of the Junilabs Player - ie. bit-perfect "optimization" and exact copies sounding different - is a fallacy of logic and goes against the body of knowledge we know as computer science. Especially these days with energy-efficient computers and SSDs without moving parts.

 

Just because some person or company claims that their "player" makes a difference doesn't mean it does, does it?

 

Arguing about these things is akin to arguing with a Flat-Earther. Whether one has a more scientific worldview, or one is more religious, I trust both camps could still recognize that the model of "truth" conveyed by a Flat-Earther is false, perhaps even delusional. So too I believe both "more objective" and "more subjective" folks can look at that Junilabs Player, read some of the others' experiences, and recognize that the balance of evidence goes against their claims and it would simply be wise to ask questions and expect good answers before being "committed" that there's anything there. This is not ridicule, it's being assertive with knowledge and experience that one has accumulated over time. If the OP of that post is truly certain that there's something, then he should have stuck around and provided the evidence including listening tests for example. That his "hard drive failed" is simply suspicious - even with HD failure, there's no reason not to come back at a later date and continue an honest discussion with evidence.

 

As I mentioned in one of my earlier posts here, if there is "something going on", it's more likely in the domain of psychology than computing, or engineering, or physics.

 

Anyhow, it's been fun! Best of luck with the magic. I don't think I have anything more to add.

 

[Archimago]

21 minutes ago, opus101 said:

 

Generally sibilance issues are an artifact of the playback system, IME. But I do have one or two recordings that exhibit a sibilant flavour and one (that I have no idea of the whereabouts right now) that does demonstrate it on female voice, not just subjectively but in Audacity FFT too. I will look out these examples.

 

Cool, yeah, would love to have a listen to those examples!

 

[opus101]

Here you go, I chopped the track to under 30s hoping its within 'fair use'. This is the female voice example which I managed to dig out off an old SDcard.

 

Listening to this on my current system its not objectionable in the way a sub-optimal replay system generates sibilance, just easily noticeable.

ALS-Rebecca Pidgeon_sample.flac

[fas42]

32 minutes ago, opus101 said:

Here you go, I chopped the track to under 30s hoping its within 'fair use'. This is the female voice example which I managed to dig out off an old SDcard.

 

Listening to this on my current system its not objectionable in the way a sub-optimal replay system generates sibilance, just easily noticeable.

ALS-Rebecca Pidgeon_sample.flac 2.29 MB · 1 download

 

Yes, here the reverb could create issues; the listening ear/brain needing to separate the direct from the reflected sound, compounds any distortion factors ... the laptop has no such issues, 🙂.

 

Which reminds me ... there are some killer drum sounds on Paul Simon's Gracelands album, where 'excessive' reverb is used; these moments are another easy giveaway, of system competence.

[Archimago]

28 minutes ago, opus101 said:

Here you go, I chopped the track to under 30s hoping its within 'fair use'. This is the female voice example which I managed to dig out off an old SDcard.

 

Listening to this on my current system its not objectionable in the way a sub-optimal replay system generates sibilance, just easily noticeable.

ALS-Rebecca Pidgeon_sample.flac 2.29 MB · 0 downloads

 

Awesome, thanks man! Yeah, that segment definitely sounds sibilant (not horrible) and could have used a little de-ess'ing.

 

We can easily see the "heat" in the spectral frequency display between 5-14kHz or so during the most obvious of these portions (typically sibilance frequencies live in the 5-10kHz range). Would certainly be exacerbated if speakers or headphones have accentuation in those frequencies.

 

 

Will keep a copy of this segment to try out with various devices!

 

[fas42]

1 hour ago, Archimago said:

Oh yeah... This is good 😉.

 

What we're talking about here is neither "subjectivist" nor "objectivist". The poo-pooing of the Junilabs Player - ie. bit-perfect "optimization" and exact copies sounding different - is a fallacy of logic and goes against the body of knowledge we know as computer science. Especially these days with energy-efficient computers and SSDs without moving parts.

 

Right, fell straight into the trap, 😄 ... there is no such thing as an "exact copy" - at some level of the physical world, any copy is always different - people used to argue that a CD-R version sounded better than the original CD; why could this be, you ask? Well, maybe, just maybe, the read electronics had an easier time accessing the copy, used less current to drive the parts, and so generated less noise which worked its irritating way into analogue areas ...

 

Yes, this is a physical mechanism, as in there are clunky mechanical bits moving about - but if the electronics alter the pattern of their behaviour, just a little bit, to access data stored differently, that could be enough - in any storage mechanism.

 

Quote

 

Just because some person or company claims that their "player" makes a difference doesn't mean it does, does it?

 

Arguing about these things is akin to arguing with a Flat-Earther. Whether one has a more scientific worldview, or one is more religious, I trust both camps could still recognize that the model of "truth" conveyed by a Flat-Earther is false, perhaps even delusional. So too I believe both "more objective" and "more subjective" folks can look at that Junilabs Player, read some of the others' experiences, and recognize that the balance of evidence goes against their claims and it would simply be wise to ask questions and expect good answers before being "committed" that there's anything there. This is not ridicule, it's being assertive with knowledge and experience that one has accumulated over time. If the OP of that post is truly certain that there's something, then he should have stuck around and provided the evidence including listening tests for example. That his "hard drive failed" is simply suspicious - even with HD failure, there's no reason not to come back at a later date and continue an honest discussion with evidence.

 

As I mentioned in one of my earlier posts here, if there is "something going on", it's more likely in the domain of psychology than computing, or engineering, or physics.

 

The audio friend down the road had a good session with me, many years ago - he had a decent media player, the palm sized thing, hooked up to the big rig - and precisely how he selected a track to play made a distinct difference. The same song, each time, but the exact sequence of keystrokes to initiate actual play "did the damage".

 

A cute parlor trick - which says what? That the precise electrical activity in even a cheap device, before the playback, is enough to alter the SQ - anyone who takes SQ really seriously knows how these wacky things are in the picture ... the real goal is to not fool around with them, but to work out how to make playback consistent, no matter what the circumstances are.

 

Quote

 

Anyhow, it's been fun! Best of luck with the magic. I don't think I have anything more to add.

 

 

Bye ...

[fas42]

Of course, where the subjectivists muck it up for themselves is that they regularly invoke magic workings of the universe as causing all this variations in SQ. Absolutely none of this is necessary; it's all borin' old physics, the same stuff you learnt to whatever level you needed - it's just that it is very difficult to completely isolate any electronics from all sorts of nasty parasitic behaviours; that is, some electrical activity will occur where and when you least want it, from just about anything - heat, vibration, radio frequency mutterings, age, corrosion - all will be doing their rotten bit. The Art, ahem, of getting audio Right is to deal with all these things effectively; so that it doesn't get in the way of best SQ - this can be done with enough focus on the job ... the Good News is that these days it's getting easier and easier to jump high enough, to get over these hurdles ...

[March Audio]

2 hours ago, fas42 said:

 

Right, fell straight into the trap, 😄 ... there is no such thing as an "exact copy" - at some level of the physical world, any copy is always different -

What trap?

 

What absolute nonsense Frank.  Of course digital data can be an exact copy.

 

 

[March Audio]

1 hour ago, fas42 said:

Of course, where the subjectivists muck it up for themselves is that they regularly invoke magic workings of the universe as causing all this variations in SQ. Absolutely none of this is necessary; it's all borin' old physics, the same stuff you learnt to whatever level you needed

 

- it's just that it is very difficult to completely isolate any electronics from all sorts of nasty parasitic behaviours; 

Now the first bit I do agree with, second not so much.  You just need to understand and quantify the problem at hand and implement appropriate solutions.

 

 

[fas42]

2 minutes ago, March Audio said:

Now the first bit I do agree with, second not so much.  You just need to understand and quantify the problem at hand and implement appropriate solutions.

 

 

 

Yes. The arguments are, how far does one need to go, and what are the most effective methods to achieve the necessary levels of robustness.

[Confused]

5 hours ago, fas42 said:

Where those of an objectivist stance do everyone else a deep disservice, is by poo-pooing all the various methods people use as workarounds for the lack of integrity of playback chains - I'm thinking here of the thread about the junilabs player. Because the workaround seems silly, is not easily measurable, it therefore is nonsense, has absolutely zero merit - is the objectivist's POV. But those who try them often find they are effective, at least to some degree - therefore, there is, "something going on". And zero progress in understanding is achieved, by the dumping of ridicule on those people's findings, by others who just don't like those sort of results.

Could you hear any difference using the Junilabs player?  

 

I kept an open mind.  Without knowing what it is actually doing, who knows, and maybe there is something going on here?  I tried it, and could hear no difference whatsoever.  

[March Audio]

6 hours ago, fas42 said:

Where those of an objectivist stance do everyone else a deep disservice, is by poo-pooing all the various methods people use as workarounds for the lack of integrity of playback chains - I'm thinking here of the thread about the junilabs player. Because the workaround seems silly, is not easily measurable, it therefore is nonsense, has absolutely zero merit - is the objectivist's POV. 

No Frank, what objectivists do is look at the evidence. If you have any explanation as to how the Junilabs software can possibly do anything to affect sound quality, then let's hear it.

 

This has got nothing to do with objectivist V subjectivist, it's more like choosing to believe, or not believe, in fairies.

[March Audio]

5 hours ago, Archimago said:

Oh yeah... This is good 😉.

 

What we're talking about here is neither "subjectivist" nor "objectivist". The poo-pooing of the Junilabs Player - ie. bit-perfect "optimization" and exact copies sounding different - is a fallacy of logic and goes against the body of knowledge we know as computer science. Especially these days with energy-efficient computers and SSDs without moving parts.

 

Just because some person or company claims that their "player" makes a difference doesn't mean it does, does it?

 

Arguing about these things is akin to arguing with a Flat-Earther. Whether one has a more scientific worldview, or one is more religious, I trust both camps could still recognize that the model of "truth" conveyed by a Flat-Earther is false, perhaps even delusional. So too I believe both "more objective" and "more subjective" folks can look at that Junilabs Player, read some of the others' experiences, and recognize that the balance of evidence goes against their claims and it would simply be wise to ask questions and expect good answers before being "committed" that there's anything there. This is not ridicule, it's being assertive with knowledge and experience that one has accumulated over time. If the OP of that post is truly certain that there's something, then he should have stuck around and provided the evidence including listening tests for example. That his "hard drive failed" is simply suspicious - even with HD failure, there's no reason not to come back at a later date and continue an honest discussion with evidence.

 

As I mentioned in one of my earlier posts here, if there is "something going on", it's more likely in the domain of psychology than computing, or engineering, or physics.

 

Anyhow, it's been fun! Best of luck with the magic. I don't think I have anything more to add.

 

This covers it all.

 

I agree there is nothing further to add.

[fas42]

2 hours ago, Confused said:

Could you hear any difference using the Junilabs player?  

 

I kept an open mind.  Without knowing what it is actually doing, who knows, and maybe there is something going on here?  I tried it, and could hear no difference whatsoever.  

 

Haven't downloaded or considered anything with that player; a thought I had before was that he could be trying to compensate for, or balance vibration  - he mentions that it's best if no-one is with the computer while it does the copying. Most likely he heard an effect with his machine; explored it, and it evolved into something that worked best for him. Which doesn't mean that it will be of any benefit to someone else - it will be highly hardware configuration dependent.

[fas42]

2 hours ago, March Audio said:

 

This has got nothing to do with objectivist V subjectivist, it's more like choosing to believe, or not believe, in fairies.

 

Unfortunately, fairies exist ... that is, seemingly bizarre variations in what one does, or how a playback is configured, "matter" - my number one interest is getting it under control, rather than understanding why the weaknesses in a specific system make the SQ change when various things are tried - the latter is like trying to hold two things together with a mess of rubber bands that are on hand, which is unstable as hell; as compared to welding, or using an industrial adhesive to do it properly.

[Blackmorec]

7 hours ago, Archimago said:

Interesting. Will have a listen to Swiss Radio Classics. Would still be good to hear about a specific audio track that folks use as some kind of "standard" of what should or should not sound sibilant.

 

[I had a listen to Swiss Radio Classic - yup, female announcer sound harsh. To me it's also the lossy artifacts that I hear causing the issue/distortion.]

Interesting reply, which goes right to the heart of this thread.  In fact, the announcers voices were recorded in different studios, by different people at different times over a period of about 20 years so by definition, they all sound different.  BUT, they can all sound pretty much perfect. No sibilance, certainly no harshness and a spooky focus and presence that puts a speaking human being right there in your room. When artefacts are present, its not the bit rate or lossy artefacts that’re to blame. Its the system that’s playing it and more specifically, the network stream. Remove every shortcoming and fault in the stream’s components and the announcers sound pretty much perfect, ditto the music. Sure you can hear shortcomings vs CD and hi-res, it certainly doesn’t produce the grandeur, power and depth that full and hi res audio can achieve but nor does it have any discernible artefacts, which makes it an ideal tool to use when trying to extract the maximum from a system. The announcers voices lack the last nth of detail that would  make them sound entirely human, but you’ll can still hear the intake of breath before the announcement  however if they sound coloured,  sibilant or harsh, that’s distortion and its down to the replay system, one way or another. Lossy audio is lacking in detail but it isn’t distorted. The system processing the lossy material is where the distortion comes from. Lossy audio’s sin is one of omission rather than distortion, which makes them ideal for troubleshooting streaming systems. 

[March Audio]

11 minutes ago, fas42 said:

 

Unfortunately, fairies exist ... that is, seemingly bizarre variations in what one does, or how a playback is configured, "matter" - my number one interest is getting it under control, rather than understanding why the weaknesses in a specific system make the SQ change when various things are tried - the latter is like trying to hold two things together with a mess of rubber bands that are on hand, which is unstable as hell; as compared to welding, or using an industrial adhesive to do it properly.

Fairies arent bizare, they just don't exist.

 

If you are referring to real effects, can you explain what matters? Please describe it specifically, followed by how you get, what ever it is, under control.