1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

What makes piano sound so hard to reproduce?

Discussion in 'Sound Science' started by matti620, Dec 20, 2012.
4 5 6 7 8 9 10 11 12 13
  1. Doctor Fuse
    Naxos - Not Any Xcellent & Outstanding Sounds
  2. Dinos Gonatas
    above 20kHz??  You can really hear that? Are you a cat?
    The other week I listened to rising wave forms with my 17 yr-old daughter. She could hear up to 19 kHz... my ears quit at 13.
  3. Doctor Fuse
    When I listen to an LP (especially well-recorded ones with mics that have a frequency response above 20kHz) with a Grado cartridge (50kHz), NAD 3020/Nikko bi-amps (50-100kHz frequency response) through NHT tweeters (25kHz), the sound is much more real, with a sensation of being in the actual room (I am a professional symphony musician). "Air" is a common descriptor. I am a firm believer that we do indeed sense these supra-aural frequencies.
  4. Doctor Fuse
    Further, when I play different violas (my instrument - yes, poor me, I know!) a big cause of differences in sound is how strong the overtones are, and which ones are emphasized. This makes me suspect that very high overtones are perceived by the ear and brain, and they do affect the sound we hear and how we perceive it.
  5. gregorio
    Your observation and your conclusion based on your observation are logical and not at all uncommon amongst classical musicians. Unfortunately, the reliable and accepted science demonstrates that your observation and therefore your conclusion based on that observation are flawed. Just to be clear, I'm not disputing what you're hearing, just your perception of what you're hearing. For example, CD and other lossless digital audio formats are almost perfectly accurate in the very high/ultra sonic frequencies, whereas vinyl is very inaccurate. So, one question we can ask is: Is an almost perfectly accurate audio recording of what is a relatively inaccurate output from a microphone really going to sound better than an inaccurate recording on vinyl which inaccurately reduces those very high frequencies and/or masks them with relatively benign, low level, high frequency noise (hiss)? The answer is "not necessarily". In fact, the addition of this type of noise very commonly results in exactly what you're describing, a better sense of "air" and "room". Nowhere is this fact exemplified more than in the film audio world, where one of the main Editors in the audio post-production team spends much of their time doing nothing else but manufacturing/recreating very specific "hisses" (called "room tones"). Each of which is specific to a single location in the film and sometimes there needs to be several different room tones for even a single location, depending on perspective. Audience members are generally completely unaware of the existence/inclusion of these room tones (or of the hard work and fairly sophisticated tools/techniques used to create them), unless of course we don't do a good job, in which case a public audience wouldn't specifically recognise a room tone weakness/fault, just that the scene/shot feels wrong or unconvincing.
    Your conclusion about the importance of ultrasonic frequencies is not at all unreasonable based on your observation but is if we look at the actual demonstrated/proven facts. For example, just 4% of the energy produced by a violin is above 20kHz. So even if you could hear beyond 20kHz (which you can't), that 4% would be totally drowned out by the other 96% anyway. I'm certainly not saying that overtones (harmonics) are not important, they're vitally important, but only up to around two octaves above the highest note a violin is likely to play (about 16kHz).
    BTW, I'm speaking as a conservertoire trained and then pro symphony orch musician myself for a number of years. So firstly, no viola jokes I promise :) and secondly, I was shocked at my transition from classic musician to audio engineer. It took much longer than I expected and much of what I thought I knew about sound had to be thrown out and started again. I'd learnt/been taught a great deal about how music in general and I in particular could manipulate perception but very little about how that perception actually works. Not a problem for a musician who has little or no control over many of those aspects of audio perception anyway but a serious handicap for an audio engineer who does.
  6. NeoG
    I always imagined there was someone in charge of location ambience, but I never thought it was that specific. That is awesome.
  7. gregorio
    Talking film sound is rather off topic but it is related to much of what's been discussed so: Don't confuse room tone with ambience. Room tone is just one component of ambience (more commonly called "background atmospheres" or just "atmos"). Typically, on a well budgeted film, there is an entire team dedicated purely to manufacturing/creating atmos. Room tone is recorded on set/location during filming by the production sound team, in theory 30 secs of it is recorded either immediately prior or after the "take". Only about 50% of the time can this room tone be used, instead, the Dialogue Editor has to scrounge up and stitch together bits of clean room tone from the gaps between words/sentences actually in the "take" but usually this is not enough so more has to be manufactured using sampling/regeneration techniques, convolution techniques or DX extraction. There are some interesting factoids about room tone, such as the simplest way to get more of it is to copy/duplicate what you've already stitched together, reverse the copy and butt it up against the original, thereby doubling the length. It sounds completely authentic/continuous, the ear cannot identify/differentiate the reversed room tone from the original, even though it can easily identify even relatively minor differences/changes between different room tones. And, this is the problem/difficulty with room tone, in any one scene we're likely to have two or more angles, plus potentially edits from different takes, each of which has slightly (or significantly) different room tone and all of which have to be reconciled because a change in room tone implies (creates the perception) of a change in time/location, not at all what we want in what is supposed to appear to be a continuous scene in the same location. Room tone is most closely related to white noise but unfortunately, not closely enough that we can ever just use white noise.
    Atomos is another whole field, a sub-division of the SFX team, which typically also comprises Hard FX and Sound FX Design teams, plus there's the Foley team, ADR team, music team and finally the Mix (re-recording) team. In total, the entire audio post team for medium and high budget films usually comprises about 35-70 audio post personnel, including the Sound Designer who has overall artistic control of the sound and the Supervising Sound Editor who manages the entire process. Audio post is split into two phases, editorial and mixing and the whole process typically takes 3-6 months, although there are many exceptions, Interstellar for example took 9 months just for the mix phase!
    The reason I mention all this is because it goes to the heart of hearing perception, of what we think we're hearing and of the ears being fooled. For example, typically somewhere around 50% (and sometimes virtually 100%) of the dialogue you hear in a film is not the dialogue which was spoken during filming, it's dialogue which has been re-performed weeks or months after the filming has "wrapped" and is recorded not "on location" but in a studio (ADR Stage) which commonly isn't even in the same country as the shooting location! Of all the sound present in a finished film, typically only a few percent is sound which was recorded during filming, the rest is all manufactured in audio post. And finally, the Sound Designer's role is an artistic role, just recreating reality would be a mainly technical task rather than an artistic one and therefore the boundary of sound design is believability, NOT reality! This provides considerable wiggle room for manipulating the audience; what they're feeling, how they're interpreting what they're seeing/hearing, even what part of the frame/picture they're focusing on. With the exception of the big flashy sound design statements (such as explosions, for example), the vast majority of sound design is subliminal, the audience are never consciously aware that they're being manipulated, let alone have any idea of how and this is why sound design can be such a powerful filmmaking tool, it directly accesses the sub-conscious and bypasses the thinking/analysing/reasoning centres of the brain. One of my favourite lectures (to give) was where I dissected a single scene from a film, explaining what was done in audio post and why. It was one of my favourites because of the invariably stunned/amazed reaction of the students, who, despite they're obvious interest in audio, had no idea of the depths of perception manipulation going on. Common descriptions of this revelation were; "that's unbelievable!","it's like a dance", "you're just constantly playing with (or teasing) them [the audience]" and "wow, that's a lot of work!". Quite a satisfying reaction from usually apathetic 18-20 year olds, trying to appear cool/unimpressed.
    Audiophiles asserting that they know what they're hearing and trust their ears, is just so ridiculous it's laughable. If that were actually true, pretty much every film with sound (from about 1927 onwards) would be un-watchable, let alone enjoyable! Do these audiophiles never watch films or TV? Are they also immune to the illusion of stereophonic reproduction, the expectation and other biases generated by harmonic progressions (the basis of nearly all western music, of any genre, composed in the last 500 years or so), the human perception that notes and chords even exist, does all music just sound like unrelated collections of frequencies, nothing more than non/semi-random noise?
  8. spruce music

    Excellent comment as is the norm for you G.
    I would like to point out, that despite the 9 months spent on Interstellar (darn good movie) they left something like a continuous 15 khz tone running several minutes in a few places of the film.  I wondered what it was and how that happened.  Or was it intentional like the overblown organ in places?  15 khz or so is the limit of my hearing these days and I barely heard it, but it was there and must have been pretty loud.
  9. Pham Nguyen
    wow, that's very interesting, I've never think about this before. Thx guys:wink:
  10. gregorio
    I was not involved in making Interstellar so I cannot say whether it was a fault or deliberate and I haven't seen the film since it's theatrical release and don't recall the parts you're referring to. However, if it did take 9 months to mix (as Christopher Nolan stated), the re-recording mixers, of which there would have been at least 3, would have extremely carefully listened to every scene dozens of times, maybe even hundreds of times, so it's inconceivable something like this could have just been missed. A blockbuster like Interstellar is going to have many different print-masters/versions, possibly as many as 70 or so and the probably of such a mistake is higher on some of those versions than on others, although still fairly slim.
    The organ was a tad over the top in places IMHO too, although exactly how much over the top it's difficult to judge. The particular cinema/listening environment where you heard Interstellar could easily have exacerbated the problem. "Translation", how a film mixed on a reference Dub Stage translates to individual cinemas, is a bit of a "thing" in the industry at the moment.
  11. spruce music

    I barely perceived a fleeting tone at times.  It also seemed to 'wear my ears out' when it was present.  I later saw others having the same complaint.  Some think it a CRT leakage though where someone would have that these days seems an odd question in itself.  People with a copy of the movie pinned down the frequency and confirmed it presence.  The movie where I saw it seemed to be overblown at both the low and high end.  This isn't a general concern with this theater so I do believe it was the choices made for Interstellar.  It could have been some mis-adjustment by the theater as well however.
    There were a few other discussions of it shortly after the movie was released, but this one is the only one I see currently. 
  12. gregorio
    Mmm, interesting, thanks for the link. A few things don't ring quite true or logical. Like for example the use of CRTs. It's been standard practice for a couple of decades to work with DIs (digital intermediates) during audio post. If using a modern digital TV, there is latency as the TV renders the DI, a CRT has very little latency in comparison but then you can't feed a CRT a DI in the first place, the DI would have to go through a video card which converts the DI to an analogue signal (for the CRT), which introduces the same sort of latency as a digital TV, although the amount is variable depending on the make/model of TV. I'm not disputing reports that CRTs may have been used but I don't understand why they would have been used.
    Also, while I haven't been able to hear 15.7kHz for a few years now, that still doesn't mean I'd have missed it. There are a number of points in the editorial and mixing phases where a spectral output is likely to occur and a 15kHz tone noticed even if no one involved can actually hear 15kHz. And, I personally wouldn't rule out the possibility that this 15kHz tone was deliberate! Nolan has a history of using sound in some of his films against standard convention. By this I mean that Nolan has deliberately done certain things which would normally be considered mistakes, because he's after a specific, not immediately obvious, audience response. It doesn't always work in practice but I admire him for experimenting and pushing the boundaries of the art.
    To get slightly more back on topic though: Probably somewhere around 100 million people have seen Interstellar, including most of the world's professional film reviewers and I'd assume a reasonable percentage of the world's audiophiles but it looks like only a minute fraction of people heard and commented on this annoying tone. How is this possible? Audiophiles apparently need/can sense audio freqs far beyond the limits of CD, all the way up to the limits of 192kHz sample rate, IE. Audio frequencies from 22kHz to 96kHz. An annoying tone at 15.7kHz is a very high frequency but still significantly below the conventionally accepted limits of human hearing and even further below the ~22kHz limit of "standard definition" (44.1kHz CD). So where were all the posts from annoyed audiophiles who can hear a "night and day" difference with Hi-res? Have audiophiles avoided seeing Interstellar for some reason or are they in effect saying that they can hear frequencies way above 22kHz but have some sort of mental block specifically/only at 15.7kHz?
  13. Don Hills
    Playing Devil's Advocate for a moment, they might have a physical block. Most audiophiles grew up with CRT TVs. It's possible they have a notch in their hearing at TV line frequency.
  14. pofdstudios
    After slogging through all this scientific argument I must say I feel pretty good about my 48 year old hearing because I cant hear half the crap you folks are arguing about and therefore ignorance is bliss and these days Im very blistful! :)
  15. warbles
    Love that 'audiophile lunatic fringe"
4 5 6 7 8 9 10 11 12 13

Share This Page