24th May 1972
AUD: mono crappy Pye cassette recorder with hand-held mike > Ampex R2R > CD-R > EAC > CEP > CDWAV > .shn & CD-R.
Original recording by Chris Jones. Transferred (untracked) to CDR by Graham Harper. EAC extraction and all subsequent modification by Alasdair Macdonald.
Most songs have minor cuts at beginning and/or end.
Disc #1: Total Time 77:10
1. Cold Rain & Snow 4:26
2. Beat It On Down The Line 3:17
3. Mr Charlie 3:49
4. Deal 4:19
5. Me & My Uncle 3:15
6. It Hurts Me Too 7:31
7. Dire//Wolf // 3:24
8. Black // Throated Wind 5:48
9. Chinatown//Shuffle 2:38
10. Chinacat Sunflower> 5:56
11. I Know You Rider 5:30
12. Playing In//The Band 12:11
13. You Win Again 4:05
14. Jack // Straw 4:48
15. Casey Jones 6:14
Disc #2: Total Time 68:43
- 2nd set -
1. Rockin' Pneumonia 5:28
2. Mexicali Blues 3:31
3. Black Peter 8:56
4. Mexican Hat Dance// 0:43
5. //Truck//in'> 10:47
6. Drums> 1:53
7. The Other One 29:44
8. Sing Me Back Home (fades) 7:40
9. Sugar Magnolia> (missing)
10. Turn On Your Lovelight (missing)
- encore -
11. One More Saturday Night (missing)
All editing carried out in Cool Edit Pro.
The supplied "raw" discs are transfers from a 1st gen Reel copy of the master (see lineage above). These "raw" discs (or rather retracked copies of them) are the source of the following CEP editing.
Note: Very many recording interruptions were present (and removed, leaving short splices) in Black-Throated Wind.
1. Removal of between-track extraneous noises (by judicious cuts). A duplicated section right at the very start of Dire Wolf was also removed.
2. Resulting .wav file then converted to 32 bit.
3. Entire .wav file then de-hissed. The amount of de-hissing removes the most obvious and loudest hiss, and although some remains I have tried to make the optimum balance (erring on the side of caution) between reducing hiss and adding artifacts. The settings used were the "low" hiss noise floor in CEP (-90db) and fine-tuning it to -3db. Other values were left at their defaults.
4. Fixes: At 15:43 there is a loud microphone handling noise: Click removal - fill single click). Same noise at 26:37 (right at the end of It Hurts Me Too) - Click removal/fill single click. Click at 51:36 (during PITB) tidied (Click removal - fill single click). Click at 40:26 - ditto. A small dropout in the dying note of BTW (at 35:44) was deleted.
5. Identification of different segments, ie with differing levels, tonal balance.
The following timings relate to the .wav file after removal of between-track clicks (see (1) above). (OK, so I ended up not quite following these segment breakdowns, but this is how it looked to me at the beginning)...
(a). d1 00:00.000 - 07:43.828 (CR&S, biodtl) - very low level, bassy.
(b). d1 07:43.828 - 11:32.630 (Mr Charlie) - low level, bassy.
(c). d1 11:32.630 - 12:34 (First part of Deal) - louder than (b)
(d). d1 12:34 - 26:37.734 (remainder of Deal, M&MU, It Hurts Me Too) - almost loudest part of d1.
(e). d1 26:37.734 - 30:01.857 (Dire Wolf) - lower level (about the same as (b).
(f). d1 30:01.857 - 49:53.755 (BTW, Chinatown, China>Rider) - a little louder than (b)
(g). d1 49:53.755 - 66:05.834 (PITB, You Win Again) - Like (d), almost loudest part of d1.
(h). d1 66:05.834 - 70:58 (Jack Straw) - Loudest part of d1, VERY bassy.
Each of these sections requires different FFT filter / normalisation.
(a): Normalisation: sample values x 4 (+12db)
(b): Normalisation: sample values x 3 (+9.5db)
(c) + (d): Normalisation: sample values x 2 (+6db)
(e): Normalisation: sample values x 3 (+9.5db)
(f) + (g): Normalisation: sample values x 2 (+6db)
(g): FFT points: -5db @ 21.39Hz, -5db @ 129Hz, -5db @ 187Hz, +5db @ 571Hz, +15db @ 2726Hz, +10db @ 18614Hz, +3db @ 22050Hz, with Spline curves option.
(h): FFT points: -10.1db @ 21.39Hz, +11.95db @ 22050Hz, then Normalisation: sample values x 2.5 (+8db)
I then applied the following FFT filter to d1 sections (a) to (e):
FFT points: -5db @ 21.39Hz, -5db @ 129Hz, -5db @ 187Hz, -1db @ 571Hz, +7db @ 2726Hz, +3db @ 18614Hz, +2.6db @ 22050Hz, with Spline curves option.
I then applied the following FFT filter to Chinatown Shuffle:
FFT points: -2.5db @ 21.39Hz, -2.5db @ 129Hz, -2.5db @ 187Hz, +1db @ 571Hz, +8db @ 2726Hz, +4.5db @ 18614Hz, +-0.4db @ 22050Hz, with Spline curves option.
I then applied the following FFT filter to China>Rider (ie d1 section (f) excluding BTW & Chinatown Shuffle):
FFT points: 0db @ 21.39Hz, 0db @ 129Hz, 0db @ 187Hz, +3db @ 571Hz, +9db @ 2726Hz, +6db @ 18614Hz, +1.8db @ 22050Hz, with Spline curves option. (A couple of peaks had to be softened first, reduced one at 46:48 to 90%, also at 49:51 (to 80%)).
I then applied the following FFT filter to d1 section (h):
FFT points: 0db @ 21.39Hz, 0db @ 129Hz, 0db @ 187Hz, +3db @ 571Hz, +9db @ 2726Hz, +6db @ 18614Hz, +1.8db @ 22050Hz, with Spline curves option.
6. A small segment of silence (about 1 second) was trimmed from the start, and the end of the large .wav for d1. d1 was then converted back to 16 bit, saved and tracked using CDWAV.
FFT filter applied over Casey Jones, with following settings:
FFT points: +1.5db @ 21.39Hz, +1.5db @ 129Hz, +1.5db @ 187Hz, +6.2db @ 571Hz, +15.6db @ 2726Hz, +10.9db @ 18614Hz, +4.32db @ 22050Hz, with Spline curves option.
Note that on the ORIGINAL disc layout Casey Jones was present as the first track on disc #2.
1. Removal of between-track extraneous noises (by judicious cuts).
2. Resulting .wav file then converted to 32 bit.
3. I did NOT perform the de-hissing on d2 that I did on d1. There is hiss, but it is either quieter or less noticeable or the recording levels on this disc successfully mask it.
4. Equalisation & normalisation. Casey Jones (final track of 1st set) has the same boomy bass sound of the last 3 tracks on disc #1 (PITB, You Win Again, Jack Straw). The settings used to improve Casey Jones are different from those used on the 2nd set tunes.
FFT filter applied over Rockin' Pneumonia to Sing Me Back Home, with following settings:
FFT points: +3db @ 21.39Hz, +3db @ 129Hz, +3db @ 187Hz, +7.32db @ 797Hz, +15.96db @ 2708Hz, +13.26db @ 18614Hz, +6.24db @ 22050Hz, with Spline curves option.
5. Converted to 16 bit & tracked using CDWAV. Then conversion to .shn and burning of CDs.