|Venue||Japan, Shinjuku Kosei Nenkin Hall, Tokyo|
|Notes||Source: dimeadozen.org #638907
Original folder name: Miles Davis 1973-06-19 Tokyo, Japan acetboy master
Miles Davis 1973-06-19 acetboy.txt
Miles Davis Septet
Shinjuku Kosei Nenkin Hall
lineage: pre-fm or fm
more likely lineage per plaz:
FM > Reel > ? > dat(x) > analog > bootleg CD then
DIME > flacs > WAV > WavLab+SoundForge > flac(lvl8) and now:
DIME > flacs > xACT > WAV > ProTools LE > 24 bit > ocenaudio > ProTools LE >
16 bit > xACT > flac > back into the wild
02 Tune in 5
03 Right Off
05 Unknown in F
02 Agharta Prelude
This show has been shared numerous times over the years.
From a bootleg: Unreachable Station 2 CD Legendary Collection.
That's either the original bootleg that was shared previously or one just like the
boots that were shared before.
This and the earlier versions all had quite a few problems that were taken care of
and then shared in this version from plaz_restore:
In the included original text from plaz you can read about the many issues that
he fixed from the earlier shares. Many gaps and glitches and scratches and a speed/pitch fix.
As you can read in those notes plaz worked on a version that had some other fixes done.
One thing that was done to that earlier version was that one channel had been inverted
to fix a phase problem. The first set was pretty much 180 degrees out of phase between the two
channels. That fix though was also applied to the second set. The second set did not need it.
So that left the second set almost 180 degrees out of phase. Oops.
What I have discovered is that going back to the earliest shares the two sets actually are not
exactly the same in many ways. So at some point they were derived from different sources.
Or maybe the same source but between the first and second set there were changes made in the
audio being fed to the recorder. Kind of hard to say.
So the first set inverting the phase of one channel was better than nothing but it still left
some phase issues intact. All of those microphones on the stage....
And inverting the second set of course through it off. But not inverting it still left it with
some phase issues. All of those microphones.....
So here I have all phase issues sorted.
Then there were eq problems.
In the low end the first set is lacking somewhat. The second set is better, more low end than the first.
With the high end the first set is good in both channels but in the second set the right channel
is similar to the first set but the left channel is way low. Lots of highs missing/way down level-wise
in the left channel.
Then through out the mids there are various level discrepancies. In the lower mids the first set
is still down a bit compared to the second set. And in the upper mids the second set left channel
is lacking compared to the first set.
When I work on audio like this I will check out what's going on eq wise by using low pass and
high pass filters together to zero in on frequency bands to see what is lacking and where.
So if you lowpass everything at 120 Hz you can see nothing but the low end below that point.
I will use a high pass filter and cut everything below 8000 and then I'm seeing in the waveform
nothing but the high end. That's when you will see a problem like the right channel is fine
but the left is seriously missing what should be there.
So this way I discover what needs eq and where it's needed. Then I fix it.
No blindly turning knobs or pushing sliders. Of course in the end it has to sound good.
You can look all day but at some point you have to listen.
So with this version all phase issues are taken care of and any missing frequencies are brought back.
My techniques in effect, un-master audio. And I create a new master.
The levels were border line blown out in the original share, here and there anyway.
My final levels are equal and consistent through out. And the final level of the share
is not stepped on with compression, no brick-walling. Turn it up!
length expanded size cdr WAVE problems fmt ratio filename
12:49.08 135670460 B --- -- ---xx flac 0.6622 Miles Davis 1973-06-19 set 1 01.flac
9:28.06 100209356 B --- -- ---xx flac 0.6023 Miles Davis 1973-06-19 set 1 02.flac
1:19.56 14067356 B --- -- ---xx flac 0.6845 Miles Davis 1973-06-19 set 1 03.flac
10:47.48 114243740 B --- -- ---xx flac 0.6233 Miles Davis 1973-06-19 set 1 04.flac
10:43.32 113500508 B --- -- ---xx flac 0.5999 Miles Davis 1973-06-19 set 1 05.flac
22:23.30 236975804 B --- -- ---xx flac 0.5633 Miles Davis 1973-06-19 set 2 01.flac
9:51.36 104337116 B --- -- ---xx flac 0.6115 Miles Davis 1973-06-19 set 2 02.flac
13:35.09 143787212 B --- -- ---xx flac 0.5726 Miles Davis 1973-06-19 set 2 03.flac
90:58.00 962791552 B 0.6011 (8 files)
Miles Davis Tokyo 1973-06-19 (fm) Info original.txt
June 19, 1973 (a)
Shinjuku Kosei Nenkin Hall
prefm or fm, 90:58, A
Set 1 [45:08]
1-1 Turnaroundphrase 12:49
1-2 Tune in 5 9:28
1-3 Right Off 1:19
1-4 Funk 10:47
1-5 Unknown F 10:43
Set 2 [45:50]
2-1 Ife 22:23
2-2 Aghartha Prelude 9:51
2-3 Zimbabwe 13:35
Miles Davis - trumpet, organ
Dave Liebman - tenor and soprano saxophones
Pete Cosey - guitar, percussion
Reggie Lucas - guitar
Michael Henderson - electric bass guitar
James "Mtume" Heath - congas, rhythm box, table percussion
Al Foster - drums
Original editor's/seeder's notes:
Lineage: pre-fm station Reel > ? > bootleg CD > CDR > EAC > FLAC
First, much thanks to the original taper and to the folks who distributed the recording
on EZT in April 2005. There were torrent comments indicating that some folks believed
the phase of the left channel was inverted. I've made this kind of repair on torrents
before, listened to the original, came to the conclusion it had the same type of problem,
and made the same type of repair to this one. There are sound men who believe that
inverting the phase of one channel will increase the perception of stereo separation.
There are also sound men that will knowingly invert one channel when taping as an
anti-bootlegging strategy. It's also quite common to run into ALD recordings and mono
recordings with one channel inverted. I have no proof that the left channel was inverted
on the original material, it is just a hunch. Listen, compare, and see for yourself.
I tested the original FLAC files with "flac -t", visually verified the original FFP against
"shntool md5", converted them to WAV with "flac-d", then wrote a short script to run "sox"
on each track, inverting the phase of the left channel. The sox command line for each file
sox -V $infile $outfile avg -1,0,0,1
I created WAV md5 checksums for the new wav files with "md5sum --binary *.wav" and created new
FLAC files with "flac -e --verify --best *.wav". I used "metaflac --show-md5sum" to create a
new FFP file. I did a test decode of the FLAC files and verified that the WAV files from the
new FLAC files matched the WAV md5 checksums. I was feeling sassy and so I used easytag to set
some basic FLAC tags from the contents of the info file and then used "metaflac --add-replay-gain"
to add ReplayGain tags as well.
All work was done on White Box Enterprise Linux 3 (RHEL3 clone).
EasyTAG 1.0 (Nov 1 2004)
Thanks to Bocci April 2005
add.lineage: DIME > flacs > WAV > WavLab+SoundForge > flac(lvl8)
Comments by editor 2 (plaz):
Many thanks to the previous editor and uploader. I have compared the 2 versions and believe this DIME
remastered version has better balance and a more realistic stereo image. The bootleg CD's left-channel
phase inversion issue is complicated by an over saturated left channel. This deficiency is less apparent
in the DIME remaster as well. It is the same if not worse in the other version (description of
source B will follow).
While enjoying this recording I noted some drop outs and a few clicks and pops and some crackle here and
there. So I eventually set about and, after extensive analysis, removed the drop outs and repaired the
clicks, pops and crackle. See the 'Editing' manifest below for details. As I repaired these minutely
audible glitches, I found that the dat-type drop outs were not entirely silent, but had analog noise
superimposed. No music seems to have been lost to these drop outs, so that removing them corrects the
sound of the whole band skipping a beat, or dropping the time simultaneously. As these drop outs were
not completely digitally silent, I suspect that the source went through an analog generation after being
digitized (either that, or some noise reduction or other analog signal processing was applied to
the digital source).
Based on what I hear, I believe this is a really good FM > Reel recording and not from the station reel
as stated by the bootleg source. Also it is unlikely that the station reel was faded before the end of
the concert. (Likely lineage: FM > Reel > ? > dat(x) > analog > bootleg CD.) Furthermore, on comparison
with my FM>R>CDR alternate version (source B) of this broadcast, this source has at least 1 additional
analog generation as well as dat in its lineage.
I am not sure if the original broadcast was faded put at the end or not, and the other source I have
for this concert ends more abruptly. The other source... I have a cdr transfer from a master off-air reel
that dues not have the above-noted digital dropouts and other glitches (source B), but there is something
definitely weird with the radio broadcast mix or the source off-air reel recording. I suspect that the phase
is inverted in the left channel of this recording as well, so it was either broadcast that way, or the CD
bootleg remastered here, is a descendant of the source B off-air source reel. The sound is a little harsher
(running fast) in Version B and has slightly more noticeable distortion of the left channel bass than is
apparent in the bootleg CD and in its subsequent phase-inversion remaster. The B version also has an
overdriven left channel which leads me to suspect the master for this deficiency. Source B needs the
left channel phase inversion correction, and amplitude restoration (de-clipping). Then the 2 channels
could be balanced and mixed. In the meantime, I finished editing the phase-corrected descendant of the
CD bootleg that was posted on DIME.
This is audio restoration: I have not altered the music in any way: I have done no frequency equalization
no balancing, nor adjusted the mix. The changes in file sizes is the result of removing null blocks
(digitally-silent drop outs commonly caused by dat>dat cloning and AD/DA conversion processes) and the
retracking that entailed. I removed the null blocks, repaired clicks, pops, crackle, and small-scale
gain fluctuations (partial, analog drop outs), and clipped some opening and closing applause segments
by 5 to 6 dB. All this was done in WavLab. Finally, flambay recommended speed correction of -10 cents
and this was preformed using SoundForge.
Since this is a well-circulated recording, I provide a detailed account of my editing activities.
Editing Manifest: source A
SET 1 [44:52]
Track 1: Turnaroundphrase (12:44)
- removed 15.8ms dropout at 6:12 (no music missing; file length changed)
Track 2: Tune in 5 (9:24)
- removed crackle at 5:58 and 6:35
Track 3: Right Off (1:19)
- clipped applause at end of track
Track 4: Funk (10:43)
- repaired click at 0:15
- repaired crackle at 0:21
- repaired gain fluctuations at 1:44-5
- repaired 9.9ms right-channel gain drop at 4:03
- repaired click and crackle at 6:51
- repaired click at 7:16
Track 5: Unknown F (10:39)
- repaired 2 click at 10:05
- repaired gain fluctuation at 10:20
- clipped closing applause
SET 2 [45:34]
Track 1: Ife (22:15)
- clipped opening applause
- repaired small pop in right channel at 0:23.497-9
- removed 15.4ms drop out at 3:26 (no music missing; file length changed)
- repaired pop at 5:52 (stubborn pop remains but is less audible)
- removed 15.4ms drop out at 3:26 (no music missing; file length changed)
- repaired 4 clicks at 4:01-2
- repaired 2 clicks at 4:22
x thud at 5:52 (not repaired)
- removed 11.4ms drop out at 6:58 (no music missing; file length changed)
- bumped left-channel gain up by 0.5dB from 11:48-15:32
- removed 11.2ms drop out at 13:02 (no music missing; file length changed)
- repaired pop and clicks at 15:58
- repaired clicks at 16:28
- removed 15.2ms drop out at 17:15 (no music missing; file length changed)
x repaired crackle and pops from 18:58-19:09 (heavily damaged; some crackle and thuds remain)
- removed 12.5ms drop out at 19:16 (no music missing; file length changed)
- repaired pop at 19:20
- repaired crackle and pops from 19:20-19:26
Track 2: Aghartha Prelude (9:48)
- removed 11.4ms drop out at 3:11 (no music missing; file length changed)
- repaired pop at 8:19
Track 3: Zimbabwe (13:30)
- removed 16.3ms drop out at 0:43 (no music missing; file length changed)
- increased gain of left channel by 1.2dB from 1:19-2:28
- repaired click at 1:57
- repaired pop at 3:39
- repaired pop and crackle at 4:24-5
- repaired pops at 4:34
- repaired pops at 4:41
- repaired pops at 4:52
- clipped closing applause
- speed correction: -10 cents (all tracks)
- retracked and packed to flac(lvl8)