Post by petea on Jul 29, 2022 15:08:53 GMT
I have long preferred the original vinyl / early CD mix of Tubular Bells to the 2009 remix / remaster that Mike Oldfield produced. This was made more apparent following the changes made to the file playback in the UK system.
While testing the replay of files from the Fidata server via the Uptone Audio EtherREGEN / Naim NDX2 route we listened to Tubular Bells. To say it was like hearing the album anew would be something of an understatement! However, as the album progressed the degree of distortion in some of the guitar sections was reaching levels that were enough to make one feel the speakers were on their way out (they aren't!). The Lab12 DAC has VU meters on it and in many sections it could be seen that the levels were well into areas where distortion would be apparent. After a bit of experimentation it became clear why.
As I stated earlier, I prefer the original mix to the later one. My file version of Tubular Bells was originally purchased from Qobuz as I only had the album on CD in the UK and I was in Germany and so could not rip the CD I had at the time. I mentioned this to a friend and he sent me a rip of a CD that he used and this I much preferred and copied it onto the Fidata server there, replacing the Qobuz one. The NAS backups in Germany and the UK are linked and so both versions ended up in both countries. However, I had forgotten that I had only replaced the version on the Fidata in Germany and so when I moved the album from the NAS onto the Fidata in the UK, I ended up with the Qobuz sourced version, and this is what was played during testing.
Looking at the NAS, I could see that there were two versions and that they were different in size. The metadata wasn't much help as much of it was missing from the CD rip and, as is often the case with files purchased from Qobuz, it was a complete 'dog's breakfast' on those. However, after processing them through Jaikoz by JThink Software Ltd and referencing the files with the Discogs library, it became clear that the Qobuz version was from 2009 (and 1 second longer!), and so the version that Mike Oldfield had remixed / remastered. And looking at the files using MusicScope revealed what an unholy 'mess' he'd made of some of the sections.
Now I realise that Mike Oldfield used a fair amount of distortion on certain instruments (mainly guitars) during the recording, but his later remix seems to have magnified this and an analytical playback of the files makes this very obvious. Other parts are enhanced in a more pleasant manner; for example the higher level used for Vivian Stanshall's voice and some of the gong strikes etc. Other bits get a bit lost (eg the hand bell at times).
Anyway, it was certainly an interesting and enlightening process, but the CD rip is now in place on the server, although I can still access the Qobuz file should I get the urge!
Here are the MusicScope charts of Part 1 (where the distortion is most apparent) showing the differences in the two version. Note the differences in levels in the History charts and the differences in dynamic range in both the loudness range figures (LRA) and the S-Plot: these both show that the later version is significantly more compressed than the earlier one. The stereo image is slightly wider in the later version as well.
Original Version Remixed / Qobuz
While testing the replay of files from the Fidata server via the Uptone Audio EtherREGEN / Naim NDX2 route we listened to Tubular Bells. To say it was like hearing the album anew would be something of an understatement! However, as the album progressed the degree of distortion in some of the guitar sections was reaching levels that were enough to make one feel the speakers were on their way out (they aren't!). The Lab12 DAC has VU meters on it and in many sections it could be seen that the levels were well into areas where distortion would be apparent. After a bit of experimentation it became clear why.
As I stated earlier, I prefer the original mix to the later one. My file version of Tubular Bells was originally purchased from Qobuz as I only had the album on CD in the UK and I was in Germany and so could not rip the CD I had at the time. I mentioned this to a friend and he sent me a rip of a CD that he used and this I much preferred and copied it onto the Fidata server there, replacing the Qobuz one. The NAS backups in Germany and the UK are linked and so both versions ended up in both countries. However, I had forgotten that I had only replaced the version on the Fidata in Germany and so when I moved the album from the NAS onto the Fidata in the UK, I ended up with the Qobuz sourced version, and this is what was played during testing.
Looking at the NAS, I could see that there were two versions and that they were different in size. The metadata wasn't much help as much of it was missing from the CD rip and, as is often the case with files purchased from Qobuz, it was a complete 'dog's breakfast' on those. However, after processing them through Jaikoz by JThink Software Ltd and referencing the files with the Discogs library, it became clear that the Qobuz version was from 2009 (and 1 second longer!), and so the version that Mike Oldfield had remixed / remastered. And looking at the files using MusicScope revealed what an unholy 'mess' he'd made of some of the sections.
Now I realise that Mike Oldfield used a fair amount of distortion on certain instruments (mainly guitars) during the recording, but his later remix seems to have magnified this and an analytical playback of the files makes this very obvious. Other parts are enhanced in a more pleasant manner; for example the higher level used for Vivian Stanshall's voice and some of the gong strikes etc. Other bits get a bit lost (eg the hand bell at times).
Anyway, it was certainly an interesting and enlightening process, but the CD rip is now in place on the server, although I can still access the Qobuz file should I get the urge!
Here are the MusicScope charts of Part 1 (where the distortion is most apparent) showing the differences in the two version. Note the differences in levels in the History charts and the differences in dynamic range in both the loudness range figures (LRA) and the S-Plot: these both show that the later version is significantly more compressed than the earlier one. The stereo image is slightly wider in the later version as well.
Original Version Remixed / Qobuz
