Shop OBEX P1 Docs P2 Docs Learn Events
128kbps vbr holiday music — Parallax Forums

128kbps vbr holiday music

xanaduxanadu Posts: 3,347
edited 2013-12-23 14:19 in General Discussion
Public places like malls hire sound engineers to design speakers systems. It is something I have witnessed first hand and mind blowing the amount of stuff they take into account when designing them.

So why am I listening to satellite radio on a bad day quality radio? It's worse than AM quality but its spatial. DSPd nonsense. The mall needs a DJ, with a Dolby cert.

Who is inspired by lossy cymbals that sound like they're underwater?

I was about to call mall security but i opted for the suggestions box.

Comments

  • Peter KG6LSEPeter KG6LSE Posts: 1,383
    edited 2013-12-22 07:49
    i never was in to MP3 ........ its AIFF or WAV for me ...HDDs are soo cheap that I cant find any reason I may want to butcher my mozart with compression .......
  • potatoheadpotatohead Posts: 10,261
    edited 2013-12-22 13:46
    The problem with compressed audio isn't that it's compressed. If you encode with LAME at 256 - 320kbps, you are going to get a very good reproduction, though that all is 16Khz bandwidth. For many people, that's enough, but that is one problem right there, I never see anything encoded at greater than 16Khz, and frankly I wouldn't mind 18 - 22Khz, though my own hearing rolls off about 17Khz and ends somewhere around 19.

    Of course 16hz is plenty, but the problem is the slippery slope regarding those decisions. If we can shave this, we can shave that, and these people can't hear this and those people don't care about that, and eventually we get to an ugly mess that really isn't all that accurate above 5Khz.

    In fact, the HD Radio digital system in the US only encodes up to about 8Khz, and that 8Khz is reasonably accurate. Above that is synthesized on the receiver end. !!!

    Annoying to say the least.

    Good encodes used to take a little bit more time and they took a little bit more space and the whole thing hit when it was just possible to do and make sense, so we are stuck with decisions made back then, and expectations set back then.

    Garbage today.

    I am very sensitive to vocals. Poor encoders mangle voices just as badly as they mangle higher frequency things. The watery cymbals stand out just as badly as the warbles you will hear in the secondary+ formants and noise content found in human vocalizations. I hate it. I much prefer low bandwidth, but accurate signals. Give me 10Khz that's spot on rather than 16 that is a wet mess anyday.

    For the mall, 10 to 12Khz with an emphasis curve closer to what we used for AM radio would actually sound absolutely great! The overall noise level of the mall would fill in the lack of higher frequency content, and the AM emphasis curve would punch vocals and primary instruments right through the bustle without fatigue. It would just "stand out" in the overall profile nicely. Sadly, the people who design these things make that possible. The people who program for them don't get it. At. All.

    Ever hear carolers in the mall? Notice how they just stand out, right there, present? That is that sound profile at work, and it's beautiful and completely unappreciated today. Again, nobody gets it.

    So we are left with highly compressed garbage, no attention to the environment, and very little appreciation for passive EQ and or environmental effects beyond the basic engineering done to put the systems in. Maybe those people should do the sound and pass the info along. We all would seriously appreciate it.

    Beyond all of that, I have found it very interesting. Older people, who grew up with analog sound have a basic noise tolerance that younger people simply do not. Younger people will very regularly trade accuracy for lack of noise and lack of dynamic range too. (I don't understand the latter)

    The other dynamic we've seen play out is choice vs accuracy. When you've got a bitstream, you can maximize it, or offer choices. Choice wins every single time, so it all gets diluted down to offer the maximum on both axis.

    Of course, there is this little experiment:

    You have two choices! One is very high quality, but boring. The programming is simply not compelling at all. The other is low quality, but it's compelling and interesting to you.

    Which do you listen to?

    The vast majority of people will choose entertaining over quality, and that is precisely why we don't have quality. With broadcast media, I get it. I mostly understand this dynamic on streaming media too, but I don't understand it for malls and other places where there isn't choice. Where there isn't choice, it needs to be quality, but that gets botched and here we are.

    BTW: I thought I would add there is the initial encoding. For some reason, there are a ton of sloppy encoders out there. One problem. The other is the playback engine. Lots of sloppy messes there too.

    I had a program called "amp" on the SGI that would play nearly *anything* however poorly encoded with very few of the warbly, messy sounds. On my droid phone, the Google music thing is similar, generally teasing out very reasonable sound from even low bit rate files. Having downloaded a few apps with better interfaces, I have found most of them just play garbage. IMHO, might be worth a look one of these days to understand just why that is. I wonder what the better players do in order to improve the temporal accuracy of the playback stream like they do.
  • Beau SchwabeBeau Schwabe Posts: 6,566
    edited 2013-12-23 00:42
    Not that I am an audio buff, but I do understand a potential problem that I hear (no pun) very few people mention. Although we (humans) may have a cutoff near 20kHz, when we digitize and truncate near or below that level we also truncate a critical part of the music that can only truly be reproduced by heterodyning the very frequencies we are cutting off making any reproduced musical experience sounding flat, lacking the original depth of the music.
  • Heater.Heater. Posts: 21,230
    edited 2013-12-23 02:39
    Potatohead,
    I much prefer low bandwidth, but accurate signals.

    You said it. Back in the day we had analog radio that we listened to via tube receivers and crumby speakers.

    No idea what the resulting bandwidth might have been but I'm sure it was far less than you can get with digital systems, amplifiers and speakers today.

    But then we compress the **** out of the data and instead of cutting off those cymbals we end up desperately trying to preserve them and ending up with a warbly, garbled mess.

    Good grief, at my age I'm not even supposed to be able to hear such things.

    How do the youngsters put up with it? Well I guess if that is what you have lived with all your life and most of the music you listen to is of electronic origin anyway you just don't hear anything wrong with it.
  • Heater.Heater. Posts: 21,230
    edited 2013-12-23 02:43
    Beau,
    ...when we digitize and truncate near or below that level we also truncate a critical part of the music that can only truly be reproduced by heterodyning the very frequencies we are cutting off...

    It's not clear to me how this "lack of things to heterodyne" is different when signals are digitized as when the are analog.

    Presumably the cut of is done in the analog domain anyway prior to digitizing else we are into the world of aliasing problems.
  • Heater.Heater. Posts: 21,230
    edited 2013-12-23 02:57
    A strange observation:

    A DJ used to come to a small bar in Helsinki once a month or so. His equipment consisted of one 1960's vintage portable record player and a pile of 45's of pop music ranging from the late 1950's to late 1970's. He used to make trips to New York to scour the record dealers for collectable disks.

    Now this DJ was not into being a big personality, he did not even have a mike, he did not even introduce the records. No, he just played them. Occasionally people would sift through the pile of 45's and make requests.

    By any modern standards the sound was horrible. No bandwidth, scratchy, noisy. But there we have a bar full of happy people dancing and partying for a couple of hours.

    Now I always thought this was very strange because pretty much all the same songs were played on the CD jukebox or over the bar's stereo system everyday anyway and normally the place was dead, no dancing and no partying.

    What's that all about?
  • potatoheadpotatohead Posts: 10,261
    edited 2013-12-23 06:34
    Atmosphere.

    That DJ knows how to cultivate one. Old 45's have a great, quirky sound. Many audio programs have digital effects that do a fairly reasonable job of replicating it today. I DJ'ed quite a bit in my 20's, introducing people to imports from around the world --that I had to pay nicely for, while weaving in the pop staples with just a sprinkle of "can't miss" 70's gold. :) BTW: Doing that is a LOT of fun. If you can read the people, you play to them. You take them on an emotional story ride, over hear, down, up, sideways, slow, fast, sensual, quiet, etc... and that all provides the foundation for whatever experiences they are having and when it's done right, everybody feels absolutely great, wanting more. That's when you turn it all off. Don't go too far, end it solid so they remember who you are and come back. :)

    I've kept one 45 for just that reason. It's a copy of "Video Killed The Radio Star" by the Buggles. It's not the best copy, but it is my favorite one, and that sound comes through pretty nicely after being digitized. Little bits of stereo imaging shift, back and forth due to quick pressing, little clicks, pops, and the odd, very light rattle and clip here and there due to the cartridge and you've got some great sounding music!

    That one represents a transition. Prior to that tune, it was one era, 50's, 60's, 70's, and then came the 80's. Digital was a thing, MTV began to air it's formerly ecclectic videos and cool music news, and that kicked us off into the 90's where we got connected, and P2P decimated FM and the major labels control on culture. The 00's were the first generation to not have a set genre, instead choosing music for pure style reasons, plucking whatever they wanted from any era, country, genre in the world, through today.

    You can still pick up that old vibe by following the top charts, but it's not the same, more like a distant signal in the noise now. We all hear it, but we all don't care so much either.

    A vintage portable player will output a very warm 70's era sound that is full in the mid and low end, but not dominant, and likely soft below 100hz, while rolling off the highs about 14Khz or so. Think AM and FM table radio type sound and you've got it about right. For a bar, assuming it's not too large, he can be off in the corner nicely placing the sound in the overall scene without that sound dominating the place. People can gather, talk, and during an idle muse, perhaps collecting their thoughts, or seeking inspiration from the music, can do so without the nice ambiance getting in the way of things.

    Think, "just another voice in the room" and you've got the intent just about right. Sometimes music can fill a room, and put the people "in" it, this is what a club does. It's all about the party, and the intense vibe of the music can make emotions run high. Mix in the booze, people spiffed up and seeking, and there you go. Good times. What this guy is doing is much different. The music is there, like an interesting regular would be, but not obtrusive, if that makes any sense at all. So that is "with" the music, not "in" the music, and the difference is atmosphere.

    They aren't there to jam or party. They are there to feel good with one another, or persue thoughts, reminince, etc...

    That guy can tickle your mind with the 45's, invoking memories, thoughts, moods, times, places, while leaving you free to do what you do in a bar and it's your choice on what you fixate on. Ever notice how people will "go outside" a club, or enter a quiet place sometimes? This guy inverts it, and he's the quiet place, still entertaining, but not obtrusive so people can be people on a "let's talk about stuff" or "let me sit and think about the thoughts I need to" kind of level.

    When I play that 45, or my digitized recording of it, I get taken back to a pretty great time. The people in the bar will too, and that's the atmosphere.

    This is the same reason why dives stay in business. It's like a trip back to time, laced with nostalga, quiet, feeling just right, comfy like old, worn shoes do.
  • potatoheadpotatohead Posts: 10,261
    edited 2013-12-23 06:55
    Beau,
    ...when we digitize and truncate near or below that level we also truncate a critical part of the music that can only truly be reproduced by heterodyning the very frequencies we are cutting off...


    It's not clear to me how this "lack of things to heterodyne" is different when signals are digitized as when the are analog.

    A few things:

    1. Beau is right. Truth is, reproducing great live music feel needs about 30Khz. We are phase blind in that we hear only frequency and ampthitude over time. But, we also hear the product of phases and the room all adding up. Many will refer to this as "imaging", and all those subtle effects place music and the things in the room to us in ways that are missing from a lower bandwidth recording. This is true for most people, even when their hearing has rolled off to 12Khz or so.

    (all the really great stuff is under 10Khz, maybe even 8Khz, and that was the bandwidth Heater refers to in the tube days. 5 - 8Khz AM on a tube radio, wood frame speakler, great receiver coupling to take the bass down to 20Hz, and you've got wonderful, warm sound...)

    2. The phase differences get mashed away when using the CD sampling rates. This is the biggest argument vinyl advocates make when comparing the two forms. My own personal experience is a well mastered CD performs very well. Under age 30, my hearing was 22Khz. I could hear those differences and they were minor at best, but they are there. Now I've rolled off to somewhere around 18? I don't want to check... (but I will eventually) I really don't hear them to the same degree. It's a feel, and it needs to be loud for it to happen.

    As a little kid it was even higher. I could hear the people sensors at the grocery store easily, and the old "clang the metal bar" type acoustic remote was clear as a bell, each high sound a specific note. I demonstrated this by clanging small objects around the house, until they contained mostly that note, successfully turning the TV OFF for my "demo" of said things. :)

    I don't have perfect pitch, in that the various notes don't map to a scale or numbers for me. But I do remember specific ones and can recall them and compare very well over time. And I'm thankful for that, because nothing was perfect in the 70's. Nothing. When we got crystal controlled audio, it was so darn nice! You can see an artifact of that with higher end "DJ" or "Pro" style CD players with a pitch control on them. When it's loud, most people experience a pitch shift, and a good DJ will compensate that to put the sound closer to that "reference" we all expect now because our audio is accurate enough for many people to experience discordant perception of music played at the wrong pitch... Why they pitch up on FM is beyond me, but they still do. Well, I know why, and it's to gain a couple few extra commercial spots per day, and convey "excitement" by raising the pitch some. Mostly it conveys suckage, but there you go. That's why they won't ever let me on the radio :p

    3. Compressed audio completely eliminates these things. They are simply not there anymore. Gone. Those of us exposed to analog music in any sense can easily hear this on even the best compressed audio. 320Kbps and up to where you might as well be using AIFF or WAV straight up CD style.
  • xanaduxanadu Posts: 3,347
    edited 2013-12-23 10:33
    I listened to some nice ol analog music on Sunday. I had to replace a record player needle to do it, but I'm glad I did.
  • Peter KG6LSEPeter KG6LSE Posts: 1,383
    edited 2013-12-23 11:43
    potatohead wrote: »
    I've kept one 45 for just that reason. It's a copy of "Video Killed The Radio Star" by the Buggles.

    WOW just wow ........ * and no tape flutter *

    was that not the first music video on MTV ? Gahh Iam showing my age !

    My digital copy of staying alive ( BeeGees ) is right off my dads LP . sounds WAY better then the ones you can buy online .
  • potatoheadpotatohead Posts: 10,261
    edited 2013-12-23 11:46
    I will drop box a pretty great MP3 later today, for those who might want to hear that 45. :)
  • Heater.Heater. Posts: 21,230
    edited 2013-12-23 11:50
    It's Christmas right?

    I want you to all listen to (and watch) this: http://www.youtube.com/watch?v=7HItFqKBAQE
  • Heater.Heater. Posts: 21,230
    edited 2013-12-23 11:51
    potatohead,

    That's wicked:)
  • potatoheadpotatohead Posts: 10,261
    edited 2013-12-23 12:05
    Up for a day. :)

    BTW: That is what 128Kbps can do, done right. :)

    I'm not where I have the original sound file, or I would put that up, but this one actually does a reasonable job conveying the feel of the 45.

    Encoded with LAME, high quality setting on SGI IRIX.
  • Heater.Heater. Posts: 21,230
    edited 2013-12-23 12:17
    Are we supposed to be pirating copyrighted material here?
  • potatoheadpotatohead Posts: 10,261
    edited 2013-12-23 12:31
    Oh yeah, that's fair. I'll pull it. Done.

    Admittedly, I'm not always on top of that stuff given the availability of streaming, etc... In my peer group for music related things, we don't worry so much about a track. Get it, listen for whatever critique, delete it, move on, etc... I'll make a "fair use" edit and close the loop that way. Lots of times, I just stuff it up on SoundCloud for discussion.

    Frankly, there isn't any way to discuss this stuff *without* sharing sound.

    https://dl.dropboxusercontent.com/u/20952046/Buggles-encode-sample.wav

    Snippet. :) We get to do that. And it's a WAV to avoid another trip through the LAME codec and the artifacts that will come with it.

    Up for a day. smile.gif

    BTW: That is what 128Kbps can do, done right.
  • Heater.Heater. Posts: 21,230
    edited 2013-12-23 12:42
    Not that it worries me. I firmly believe it is every humans duty to copy everything they find valuable so as to preserve our cultural and scientific heritage from rotting away in corporate basements.

    But the management here might have a different view..
  • potatoheadpotatohead Posts: 10,261
    edited 2013-12-23 13:02
    Totally.

    Glad you mentioned it No worries at all.

    That's a good encode too. So many of these 128Kbps ones are run through the Faunhoffer codec which is regularly licensed commercially, and it's absolutely terrible! One would think they would have tuned it up over the years, but I still hear absolutely morbid 128Kbps files all the time. And that's highly likely to be what people are complaining about more than they are compressed audio, which has it's issues, but not as many as people think overall.

    Which is why I wanted to link the sound. So there we go.

    A snippet is just fine for this discussion anyway. Anyone who wants the full tune can just go here:http://grooveshark.com/#!/s/Video+Killed+The+Radio+Star+The+Buggles/3svDzG?src=5 (link updated) It's a nice comparison to the snippet I put up.
  • xanaduxanadu Posts: 3,347
    edited 2013-12-23 13:55
    Next problem, I need a pink noise LP so I can tune my EQ.

    Hey, whadya know - http://www.discogs.com/No-Artist-Test-Record-Pink-Noise/release/859821

    I agree that 128kbps isn't as much of an issue as the sample rate. I used to listen to some shoutcast streams on Winamp at 128kbps and it wasn't that bad.
  • Heater.Heater. Posts: 21,230
    edited 2013-12-23 14:19
    I prefer White Noise myself : https://www.youtube.com/watch?v=MH9atqqkUuY&list=PL2lzdz1r6T4eIMwmPJqoIlrFSWMWtRyMr

    Tune your EQ with your ears.
Sign In or Register to comment.