Never read such balanced, relaxed post and attitude from an Axe-Fx user :)

    Quote from ampjunkie

    I would agree -- tube amps have their place and are nice to have. Not just tube amps though. The Pearce G2r combo SS amp is killer.


    Regarding turntables and vinyl -- I'd be interested in the demographics of who's buying those. If I had to guess, it would be the baby boomer crowd with a huge dose of nostalgia -- not the millenials. :)

    The 1 million figure comes from a recent annual report of the German phono industry association, the 2-figure percentage also (as well as magazine articles). I am following this subject with interest for quite some time.
    Even large record stores stock up their vinyl offer every year, it is not only the small specialist record shop anymore that supports vinyl records ... and this is not predominantly electronic music anymore either.
    Almost every "important" album (sic) release in these days is also being released as a vinyl record. This would have been unimaginable 5,6 years ago.


    Actually a large percentage of vinyl buyers are people in their twenties and thirties, so not really the baby boomers (who actually show a declining interest in vinyl).


    The Audio-CD is dead.
    Try to get a decent CD-player in your electronic chain store these days. It is almost impossible. You can rather go to a thoroughbred HiFi shop to buy one.
    Philips has stopped developing CD-drives (which also affects audiophile manufacturers who relied on their drives for their own products).


    So it is either audio files or vinyl in the near future.
    The 192 kHz technology however is an interesting trend to keep an eye on in regards of digital music.

    There are basically two reasons why I disagree with the linked article:


    1. It has been showed that
    a. many musical instruments carry information above 20 kHz;
    b. people (us) perceive ultra-high frequencies, using other "organs" than ears.


    2. When a signal is very low in level, having a higher resolution array (24 or 32 Vs. 16-bit) leaves much more resolution available. Think of a person in a theater coughing in the last row of chairs, an actor whispering his hate while dying, or the fading echo of a strings chord resonating in a church. With 16 bit there are no more than 7-8 possible different sound levels, which of course become much more when the word lenght is higher, rendering a much more realistic sound scene.

    Quote from viabcroce

    There are basically two reasons why I disagree with the linked article:


    1. It has been showed that
    a. many musical instruments carry information above 20 kHz;
    b. people (us) perceive ultra-high frequencies, using other "organs" than ears.


    2. When a signal is very low in level, having a higher resolution array (24 or 32 Vs. 16-bit) leaves much more resolution available. Think of a person in a theater coughing in the last row of chairs, an actor whispering his hate while dying, or the fading echo of a strings chord resonating in a church. With 16 bit there are no more than 7-8 possible different sound levels, which of course become much more when the word lenght is higher, rendering a much more realistic sound scene.

    Here's my opinion on this subject ...


    1a. Yes, music instruments carry info above 20 kHz. But the power in that spectrum is very low
    1b. I would like to see any scientific study that shows people can perceive (double-blind testing) frequencies significantly above 20 kHz at the power levels that normal musical instruments demonstrate. People can definitely feel low frequencies, but I doubt they can really perceive > 20 kHz.


    2. 16-bit means 2^16 different levels. That is 65536 different levels. 24-bit is even better, and is only available in consumer audio today due to much better A/D and D/A techniques (mostly oversampling and delta-sigma) which enabled using lower tolerance, cheaper components. 32b converters are rare -- because at that level, the least significant bit (LSB) is much less than the thermal noise of the passive components at room temp. See http://en.wikipedia.org/wiki/Analog-to-digital_converter.


    In general, I am not sure high-bit, high sample-rate audio will take off. Most people can't perceive the difference. But then again, if it is perceived to be better and doesn't cost that much more, the placebo effect will certainly take hold and it could be popular....

    I perfectly know where you're coming from :)
    Usually, when I state something I'm pretty sure of my sources. I'd never write something on the "this has been demonstrated" line if that was not the case. I've a scientific background, so I'm careful in telling my opinions from facts.


    I'm currently on holidays, and have no access to my database. I'll send you details about what i wrote in September, if you're really interestes. It would be of help if you sent me a PM just as a reminder tho :)


    Meanwhile, some notes about what you wrote.


    The energy of the musical content above 20 kHz is not in discussion here. Once agreed it is perceivable and makes for a different sound experience, high sampling rates suddenly become more interesting...


    I respect your position, but I'm always surprised at how people can so passionately deny something they have no real prove against. It's like saying "I doubt there's any other intellingent life in the Universe": I can take it as an emotion, a feeling, a sensation, but it is worth nothing in terms of constructive discussions. Hope you do not take this bad tho :)


    I'm not sure you agree with me that a very low-level - and fading - signal (such as the ones I named as examples) is left with very few discreete voltage levels to be described if you're using 16-bit words. If there are just 3-4 bits left, you easily see that there are just 8-16 different level, which is quite *squared*, and not certainly realistic.


    A different matter is whether a specific person can hear a difference in a specific application. I'd agree that in a thrash metal tune the difference between 16 bit and 24 bit is not as meaningful as in a live, full of ambience recording of events rich in low-level signals.


    I'm also aware that the ability of hearing and discriminating sonic events is not only a matter of education but also of personal qualities. There are very skilled people out there, capable of extraordinary sensitivity. The mistake we should avoid is believing that "I (we) could no hear a difference, so there's none" is a meaningful statement.


    Try and run 100 m as fast as you can. If you're an average, sedentary individual around your 40, it would probably take between 15 and 20 seconds. You try and compete with your sedentary friends, and you all record around that range.
    Now, imagine you have no access to media, sport events, news. Suddently someone comes from Jamaica and tells you that there is a guy there who runs in 9". Would you believe that? I highly doubt it, but this would not make it false, while I also agree that you might declare yourself unconvinced as long as someone can prove it to you.
    What I mean is that something can be true even tho we don't have access to a prove yet, or we find it to be unbelievable.


    :)

    The person who is happy to stroll along a street listing to music from a squeaking mobile phone speaker can never be won for 192 kHz-technology. The same applies to the ordinary music consumer.
    This technology is for knowledgable people (or with a certain "awareness") and with a particular interest.


    The same applies to audiophile HiFi, vinyl records, the KPA, a handbuilt boutique amp or instrument.


    However, don´t get me wrong "ampjunkie" it is still an interesting article to read, although I have another point of view on this.
    I have listened to 192kHz recordings as well as music through a 192kHz converter and to me (perhaps only placebo effect ?) it sounded like a master.


    There are people for who a 4x12 box with Celestion Vintage 30 chassis is the ultimate electric guitar sound reproduction, but there are also people who feel that a FRFR loudspeaker with a frequency range up to 18 or 20 kHz offers the best sound projection, just because they know (or believe to know) that there are overtones in the electric guitar sound that are above the 5 kHz that a Celestion V30 can deliver.


    For those who worship a 4x12" box with V30s, a good FRFR speaker system will sound "too HiFi" and some of them would ever never think about buying such a "ridiculous digital thing" like the KPA.


    Of course this is yet another subject that can be endlessly discussed, as there are as many opinions as there are thread contributors.

    Why?
    Let's those who enjoy this discussion go on, whIle all the others can play as much as they like!
    A discussion, when respectful and civil, is always good, specially if there are new cards on the table :)


    REVILO:

    Quote

    The person who is happy to stroll along a street listing to music from a squeaking mobile phone speaker can never be won for 192 kHz-technology. The same applies to the ordinary music consumer.
    This technology is for knowledgable people (or with a certain "awareness") and with a particular interest.


    This might be true. This discussion is much more critical tho, since it's related to whether there are any differences at all among differently sampled signals.
    Also, the notion that there's a meaningful and perceivable quote of signal above the 20 kHz should make everyone jump on their chair, while it seems no-one cares...


    :huh:


    :D

    I may add my impressions to this topic.


    - The article Timo has linked is probably the most accurate and most complete description about the quality aspects of digital audio.
    - Every single music instrument produces audio frequencies well above 20 kHz. A guitar amp with cabinet does it too!
    - There is no single human being known that can hear above 20 kHz, even though it is damn easy to show and prove that he/she could. Is there anybody out there?
    - Very soft signals do not get lost in 16 bit signals. If that was the case, it could be shown easily by providing a representative 16 bit signal and the respective 24 bit signal. Did I miss something?

    It has been proved that humans don't *hear*, but actually *perceive* ultrasonic sounds. We use other organs for the purpose.
    The difference is meaningful indeed, but inplications are such as well.


    As for 16-bit Vs. longer words rendering, what I was saying is that, when rendering a low-level sound, the converter is left with less bits a ailable the shorter the word is. If a sound is rendered with the least meaningful 4 bits and it fades down, the next possible rendering will have 3 bits available. The fading will have a jump in level, where nothing can be said about *how* the sound moved from the 4-bit rendering to the 3-bit rendering.
    With a longer word available, the same absolute signal can use more bits, so what makes a difference is the perceived resolution of the sound change (fading).
    Hope my English translates properly...


    :)

    I could not find a reference or link on the internet to find a proof, study, test or human being, that can do what you say.
    Can you?


    What you say about the Bits does indeed follow a natural intuition, but it is simply not correct.
    If a signal approaches the "few bit" loudness, it does not fade out in steps, it does not jump at all.
    What happens between the 4 bits to 3 bits, is well known. It is continuous.


    A digital signal mastered with dithering is pretty much equivalent to an analog signal.

    Thanks for your contribution :)


    Provided you're referring to the high-frequencies perception, as I wrote I'm currently away from my database (holidays!). I'll be happy to share the sources once back home, if you're interested (a PM would help me to not forget about it till September ;D)


    As regards the sampling resolution issue, I'm not reporting an intuition of mine but what I've read from Eng. Paolo Nuti's works (http://www.key4biz.it/Players/…/N/Nuti_Paolo_184267.html).
    The page is in Italian but google works quite well to English


    I'm not sure I fully understand the discrepancies between common sense and the sampling theory. Let me just invent numbers for discussion's sake, ok? :)


    If a sampler with a 16-bit resolution senses a signal for which its uses 4 bits, let's say 0.0200 V, the next lower level would be 0.0100 V, corresponding to 3 bits. If the signal progressively fades from 0.0200 to ground noise, where does the sampled file stores the information about how the signal changed from one value to the other?
    It is called "resolution" for a reason, I suspect.


    I'm aware that the sample rate has to be at least twice as big as the highest sampled frequency, but what about the maths about resolution?


    Thanks again :)

    I could not find an article from Paolo Nuti about sampling.
    But I am interested in that.


    If you have a signal that is only 4 bits (4 steps) high, it will go through all 4 steps. If you make it even softer, it will still go through 4 steps, but will hit the 4th step shorter.
    If even softer the 4th step will not be reached eventually. But that is a smooth transition.
    The steps will be noticed as "quantization noise".


    Quantization noise is easily noticable at a signal of 6 or 7 Bits of resolution, that is about 100 value steps.
    And when the signal gets lower than half a bit, it will indeed disappear!
    But the level difference between noticing quantization noise in the signal and dissappearance of the signal is about 40 dB.
    Thus, if anyone tells you that soft signals disappear in 16 bit digital audio, without mentioning quantization noise first, this person has never experienced the disappearance of sound, and is just talking blabla.


    However, quantization noise, once you hear it, is not nice. But it can be completely killed by adding analog noise to it, which is a nice thing. That process is called dithering.
    The nice side effect of dithering is that - once applied - no signal disappears anymore. Instead it will drown in the analog noise when attenuated, just as it would when working with analog tape.


    Did anyone know that you can easily produce quantization noise in purpose with the Profiler?

    Yeah thanks redrum, the guy is cool! I'll watch it carefully :)


    PS: I must have already told you, but...


    ... Great nick! :D


    @ ckemper: I've never researched for Nuti 's works in English, not even sure there's any available online. In case you're interested to get in touch with him, you might want to contact the magazine he founded (http://www.audioreviewmagazine.it). Just click on Menu and then Contatti.


    Please let me know if I can bef of further help


    :)

    Quote from viabcroce

    It has been proved that humans don't *hear*, but actually *perceive* ultrasonic sounds. We use other organs for the purpose.
    The difference is meaningful indeed, but inplications are such as well.
    :)


    Hi Viabroce,
    you were certainly referring to this matter:
    http://en.wikipedia.org/wiki/Psychoacoustics


    I was reading a book about tube amplifier technology the other day, when I came across a chapter about Intermodulation Distortion (IDM) with these two quotes that I would like to add:


    1).
    "The brain electrical activity mapping experiments show that high frequencies above 26 kHz in Gamelan music are able to induce activation of a EEG that persists in the absence of continued high frequency stimulations".
    In other words, if you expose someone to high frequencies, certain brain patterns will be activated which continue to persist when the reproduction of these high frequencies has stopped.


    2).
    "The results from the subjective sound quality evaluation experiment by Scheffe´s method indicated that the music containing high frequency components was perceived as more pleasant and rich in nuance than music from which high frequency components were eliminated".