venerdì 18 ottobre 2019

Black Sabbath - 1983-08-21 - Helsinki, Finland (AUD/FLAC)

(Audience FLAC)

Black Sabbath 
Helsinki, Finland 
Audience recording 
Master tape transfer* 

SHN fileset (16/44.1) labelled "Master" (running time 90:18 min) > unknown software (converted to FLAC) > FLAC > Foobar (decoded to one big WAV file) > WAV > iZotope RX 4 (speed-correction by switching the sample rate tag to 46379 and subsequent resampling)** > WAV (running time 85:53 min) > Foobar (split, tagged and encoded to FLAC tracks) > FLAC > YOU 

Ian Gillan: vocals 
Tony Iommi: guitar 
Geezer Butler: bass 
Bev Bevan: drums 
Geoff Nicholls: keyboards 

101. Stonehenge (intro tape) (02:52 min)
102. Children of the Grave (05:11 min)
103. Hot Line (05:25 min)
104. War Pigs (07:38 min)
105. Born Again (06:30 min)
106. Supernaut (01:46 min)
107. Drum solo (03:12 min)
108. Rock 'n' Roll Doctor (03:41 min)
109. Disturbing the Priest (06:23 min)
110. Keep it Warm (05:54 min)
111. Black Sabbath (07:05 min)
112. The Dark (00:47 min)
113. Zero the Hero (06:20 min)
114. Guitar solo (02:45 min)
115. Digital Bitch (03:54 min)

201. Iron Man (07:51 min) 
202. Smoke on the Water (04:53 min) 
203. Paranoid (03:45 min)

Running time: 85:53 min

(Target speed determined by Rnranimal by playing "War Pigs" from this audio master tape transfer and from the incomplete AUD video at the same time, adjusting the sample rate of the master tape transfer until it was in sync with the audio track of the video.) 
In iZotope RX 4: 
- Sample rate tag switched from 44100 Hz to 46370 Hz (without resampling) => running time accordingly changed from 90:18 min to 85:53 min (without affecting the audio data). 
- At this point, the software automatically changed bit-depth from 16-bit to 32-bit float. (A few people criticised this as unnecessary and potentially downgrading. However, the bit-depth increase as such does NOT alter the audio data - it just adds 16 bits which could - at this point - still be removed to return to the 100% 16-bit original version. Only after some sound-processing - in this case the resampling to 44100 Hz, see next step below - will the added bits be used. I tested this by exporting the audio from iZotope after adjusting the sample rate to 46370, with 16-bit and NO dither applied, and the fingerprint of the exported WAV file did match the original file, even though the sample rate tag was differently set - 46370 vs 44100. However, since an audio file with an unusual sample rate like 46370, when played, is very likely to be resampled "on the fly" by our DAC, it's safer to do the resampling now with a good audio software. And contrary to what many people still believe, sound-processing with 32-bit float rather than 16-bit does allow for a better quality of the process - although, especially with bootlegs, the difference may sometimes or often not be notable. Thus, the Audacity Wiki <> mentions that "32-bit float is chosen to give an extremely low noise floor and to provide good headroom to avoid sound distortion even when performing heavy editing and manipulation of the audio", and that, "with floating point, rounding errors during intermediate processing are negligible". For an online conversation along the same lines, see here: <>. I am a laywoman in audio issues; my (very simplified) understanding of the matter assumes that we have much more numbers and more space at our disposal if we process in 32-bit float, so we can minimise unfavourable side effects. Accordingly, as far as I know, any current, proficient digital audio workstation will process in 32-bit float anyway, EVEN IF IT DOES NOT TELL YOU SO. Your software may give you the impression you're doing everything in 16-bit mode, but usually that isn't so - for good reasons. iZotope does show the bit-depth in which it is operating all the time, so it is simply more transparent in this matter than some other software - which is definitely not a bad thing. See my further comments follow below.) 
- Resampled to 44100 Hz. (See my notes above and below in order to get an idea why it makes sense that most or all proficient audio-editing software does such processing in 32-bit float rather than 16-bit mode, whether they tell you or not.) 
- Bit-depth changed back from 32-bit float to 16-bit with dither preset "CD dither - safe". (Again, see my comments above. Yes, reducing bit-depth does add truncation or dither noise - the latter being preferable in our case. But that's unavoidable, at least with any digital audio workstation that I know. IF in your software the audio really still has 16-bit after every step of sound-processing, it does not imply that no truncation noise or dither noise has been introduced; instead, it means that such noise is added with every step of processing. In our case, that would not be hurtful, since our process has only one step anyway: resampling from 46370 Hz to 44100 Hz. In other cases, where serious mastering or remastering is involved, we usually have many steps of processing, so we would end up with lots of unnecessary noise. That's of course undesirable, which is why iZotope and most other qualified software increase bit-depth to 32-bit float once at the beginning and allow us to do the 32-bit float > 16-bit step when WE want and apply the dither settings WE want - which is usually at the very end of the process. Put shortly and crudely, the advantages of processing in 32-bit float more than outweigh the apparent drawback of added noise when going back from 32-bit float to 16-bit at the end of the process.) 
- Exported to WAV file (no further processing done)

Nessun commento:

Posta un commento