Bobbsy Posted May 25, 2006 Share Posted May 25, 2006 :) Split from initial discussion on sound check music here. For me, it has always been Phantom of the Opera, but I have used iTunes to listen to some of the music suggested, and bought a couple to try next time the system is set up, and then I will get hold of the CD's later if they work for me. Pedantic old git alert: If you got the music from itunes it will be bit rate reduced and have been subjected to a brick wall filter at 15kHz. This kinda makes it useless for soundcheck purposes since there is no top end present. Bob Link to comment Share on other sites More sharing options...
mackerr Posted May 26, 2006 Share Posted May 26, 2006 If you got the music from itunes it will be bit rate reduced and have been subjected to a brick wall filter at 15kHz. This kinda makes it useless for soundcheck purposes since there is no top end present.If he bought the music at the iTunes Music Store, yes. All downloads from iTMS are 128kb AAC files. iTunes is not limited to playing content from iTMS though. iTunes can play mp3 at any bit rate, AAC at any bitrate, wav, and aif files. Music imported to iTunes from CD can be uncompressed or encoded as mp3 or AAC at whatever bitrate you choose. All my test music is in iTunes uncompressed. Most of the rest is 192kb AAC. Mac Link to comment Share on other sites More sharing options...
J Pearce Posted May 26, 2006 Share Posted May 26, 2006 If you got the music from itunes it will be bit rate reduced and have been subjected to a brick wall filter at 15kHz. This kinda makes it useless for soundcheck purposes since there is no top end present. Half the gigs I engineer, most the public present can't hear up there anyway............ And being even more pedantic the filter is at 16kHz. (Yes I was sad enough to put some iTunes bought music through a spectrum analyser) Link to comment Share on other sites More sharing options...
Bobbsy Posted May 26, 2006 Author Share Posted May 26, 2006 ** laughs out loud **...I'm just as sad. I also did a frequency analysis on a 128kb compressed file...but, never having bought anything from iTunes, I just used an MP3. I therefore bow to your analyser over mine! However, if anyone is interested in the nasties that bit rate reduction does to your recordings, have a look at http://i27.photobucket.com/albums/c188/Bobbsy/MP3freq.jpg which is the frequency analysis of a 128kb MP3 version of an original full bandwidth wave file. And this is before we talk about audible artefacts on top of the frequency truncation! Bob Link to comment Share on other sites More sharing options...
Just Some Bloke Posted May 26, 2006 Share Posted May 26, 2006 ... if anyone is interested in the nasties that bit rate reduction does to your recordings, have a look at http://i27.photobucket.com/albums/c188/Bobbsy/MP3freq.jpg which is the frequency analysis of a 128kb MP3 version of an original full bandwidth wave file. BobBob that's fascinating. Any chance you could post an analysis of the original for comparison? It would be interesting to see them side by side. Link to comment Share on other sites More sharing options...
Simon Lewis Posted May 26, 2006 Share Posted May 26, 2006 Just finished reading a dissertation where the effects of compression methods upon binaural recordings was studied. Some spectral analysis was undertaken, and a series of controlled listening tests studying the spatial information delivered by each format was carried out. This student's study suggests that Ogg Vorbis was the best, followed by AAC then MP3. There were noticeable differences in the spectral analysis. Simon Link to comment Share on other sites More sharing options...
Light Console Posted May 26, 2006 Share Posted May 26, 2006 This is potentially getting too in depth for me! Bobbsy, I read your notes about compression before downloading, but as I say, I will get original cd's if I like any of the tracks listed. Out of interest, and being a lighting boy first, what is the compression of a MD in SP mode, compared with 192kbs MP3s? I seem to remember being told that MD's had a 80% compression rate, is this so? I ask because I can rarely hear "funnies" in the sound on MD except on brassy instruments, but I can hear something odd in most MP3s.And another slightly OT, but why can I tell if a set of headphones is on back to front or L&R reversed, but not in a PA rig? Link to comment Share on other sites More sharing options...
bruce Posted May 26, 2006 Share Posted May 26, 2006 That's a fascinating graph. (Mods - can we split this part of the discussion off into a new thread?) I've just spent half an hour loading various MP3s into audacity, and running Analyse > Plot Spectrum which gives a similar picture. Virtually every MP3 (except one - see below) has a similar response, flattish, with a sharp cut-off at X KHz. However, X is variable. For some MP3s, encoded at ridiculously low rates of 56kb/s, it was at 10KHz. for others, encoded at 192, it was up around 18-19 kHz. (And one had a "hole" between 9 and 10, but then went all the way up to 18...) Link to comment Share on other sites More sharing options...
Simon Lewis Posted May 26, 2006 Share Posted May 26, 2006 Bruce, Try converting full size .WAV files to different formats and compression rates, and then look at the spectral analysis! Simon Link to comment Share on other sites More sharing options...
bruce Posted May 26, 2006 Share Posted May 26, 2006 I was thinking about trying that. But I've decided that a curry and a beer or two is a much better idea ;) Maybe later... Edit: The curry is on hold.... OK - here's some pictures. The original music was the first 23 seconds (that's all Audacity will analyse) of the first track "chez fernaud" on Martin Taylor's "Spirit of Django" First we have the track ripped from CD to a .wav file (51 Mb!) and then pulled into Audacity: http://www.zen21383.zen.co.uk/pics/wav.JPG Now, the same track, ripped with Windows Media Player at 320k http://www.zen21383.zen.co.uk/pics/mp3-320.JPG ...and at 256k:http://www.zen21383.zen.co.uk/pics/mp3-256.JPG ...and at 192k:http://www.zen21383.zen.co.uk/pics/mp3-192.JPG...and at 128k:http://www.zen21383.zen.co.uk/pics/mp3-128.JPG and finally, a 160k ogg...http://www.zen21383.zen.co.uk/pics/ogg.JPG Bruce. Link to comment Share on other sites More sharing options...
Light Console Posted May 26, 2006 Share Posted May 26, 2006 Just looked at MP3s and WAVs via SpectraLAB, and I get the 16kHz cut just like others reported on MP3s. Now then, think back to the thread I started about ITV news and the 16kHz anti yob devices, is this cut off there because people can't hear above that frequency?! Link to comment Share on other sites More sharing options...
Bobbsy Posted May 26, 2006 Author Share Posted May 26, 2006 Okay. Just for fun (and out of a sad desire to look like I was being productive and therefore be ignored by the wife!) I've done a file showing the Spectrum Analysis of the same file in the following versions: Original WaveMP3 at 128 kMP3 at 192 kOgg Vorbis at 128 kOgg Vorbis at 192 k All the files are at: http://s27.photobucket.com/albums/c188/Bob...ncy%20Analysis/, hopefully adequately labelled for you to work out what's what. The analysis was done with the appropriate tool in Audition 2.0 btw. This is not the the same file as I used last night (that was on an archive disc I put away; this is an original recording I was working on today. It's multiple instruments, a mixture of live and software synth stuff generated in Sonar using Edirol VSC, plus a solo female vocal. The original mix is 16 bit, 44.1 sampling, stereo Interestingly, in MP3, this one is filtered at 16 kHz, not the 15 kHz that last night's was. I tried several different original tracks and the brick wall filter varied between 13kHz and 16kHz, so obviously some analysis is going on to determine the filter setting. The MP3 coding was done constant bit rate using the Fraunhofer coding in Audition. Also interesting, changing from 128k to 192k in MP3 doesn't seem to make much difference to the top end frequency response but it does have slightly more definition/dynamic range. The Ogg BRR is interesting...somewhat more faithful to the original and it just stops coding at the top frequency without the heavy filtering. The upper frequency response varies with bit rate, more as I would have expected. I'm not sure what any of this proves, but enjoy the pictures! Bob Edit...just noticed that Bruce skipped his beer and curry to post while I was grabbing my pics. A quick look shows similar results to what I got. Beer and pizza for me now...curry tomorrow around here! Link to comment Share on other sites More sharing options...
Simon Lewis Posted May 30, 2006 Share Posted May 30, 2006 I'm not sure what any of this proves... Well, I guess it's useful to show what's happening in the frequency domain when these compressions alogirithms are used. What isn't as apparent is the data that's lost in perceptual coding. The dissertation that I referred to was primarily considering whether spatial information was preserved correctly (specifically in binaural recordings). The results indicate that the various compressions methods used vary greatly in their ability to retain spatial integrity. Link to comment Share on other sites More sharing options...
Light Console Posted June 3, 2006 Share Posted June 3, 2006 MP3s sounded rubbish through my EV rig, so I think it will be a case of buying some CDs over the summer! Link to comment Share on other sites More sharing options...
Bobbsy Posted June 3, 2006 Author Share Posted June 3, 2006 Dang! Light console worked it out without hours of spectrum analysis and dozens of screen shots! Bob Link to comment Share on other sites More sharing options...
Recommended Posts
Archived
This topic is now archived and is closed to further replies.