Marketing considerations aside, there are two problems which are likely
to coalbox the whole thing as a practical proposition. One is the simple
fact that material worth reissue was generally worth listening to when
it was first released - so people played the copies they bought, with
the pickups of the time - and the surviving copies will nearly all be
worn to some degree. The second follows from the first - the presence of
noise renders servo correction of alignment difficult, but declicking
and other processes may induce errors which, whilst audibly
insignificant, will also cause problems. The material for the Hicks
experiments was transferred as well as I knew how; meticulously centred,
with speed carefully monitored, from copies rather better than the
average, but even then it was a Herculean task to get less than a minute
of audio synchronised well enough to prove the point. It's hard enough
to find one decent copy of a lot of significant material, never mind two...
On 22/12/2020 05:43, Tim Gillett wrote:
> I feel the one turntable approach could be better. Then the hardware
> and adjustments would be the same down to the stylus and setup. If
> changes were needed for say a slightly different stylus, that could be
> I see it as really an information game, regarding errors. Errors are
> OK so long as that error information is captured and can be used to
> correct the errors, within the system limits. There are potentially
> multiple errors so each type of error would need to be captured and
> fed into an appropriate feedback loop.
> 1. A turntable with the necessary short and long term speed stability
> may not be practical. Capturing the turntable's *speed error
> information* synced to each audio capture seems a far better way to
> go, if not the only practical way. Software would apply the time base
> correction to the high sample rate audio file of each disc.
> 2. Off centre disc error could be captured, perhaps via an optical
> 3. Even the timing error caused by a slightly buckled disc could be
> captured and translated into speed and pitch error, via capturing the
> vertical displacements of the tone arm. Obviously a very buckled disc
> would be ruled out as a candidate.
> 4. Then when each of those error types had been corrected as well as
> possible we still may have timing errors between each disc side which
> might best be corrected (as well as is possible) with software
> comparing the two disc audio files and measuring the remaining timing
> difference error and attempt to correct it. Software could be based on
> a standard tape azimuth corrector, as per CEDAR, RX etc.
> As with any audio restoration software, the order of use can be
> critical. So we dont HPF before we declick etc. Maybe dont even
> declick until all the timing corrections have been made.
> Ted Kendall says that some years ago Christopher Hicks managed to
> successfully sync up 5 discs. So it's possible to do it, on more
> than just two discs and that was just one project. Compare the
> long timeframe and setbacks along the way it took to develop decent
> quality commercial videotape recording. As here, successful videotape
> recording and playback very much revolved around accurate timebase, on
> multiple fronts. Back in 1956, they finally achieved close to the
> then broadcast quality video record and playback, but without the huge
> benefits we now have of today's technology. It was even totally
> valve technology.
> Maybe sound quality aside, most consumers arent much interested in
> purchasing for example the recorded performances of the late 1920's
> until around 1950. Maybe the reason this sort of thing hasnt been
> developed to a high level is because there's mostly not enough income
> to be generated from the improved audio.
> ----- Original Message -----
> From: "Association for Recorded Sound Discussion List"
> <[log in to unmask]>
> To:<[log in to unmask]>
> Sent:Mon, 21 Dec 2020 18:35:59 -0500
> Subject:Re: [ARSCLIST] Noise reduction on mono records using two
> separate coherent sources
> So my humble idea FWIW would be to have 2 identical turntables, same
> tonearm, cart and stylus, TF, etc. both driven by a common source;
> some sort of servo controller that would output identical power,
> timing, locking, etc. to both tables. You’ve now theoretically got
> rid of most of those variations and you’re left with just the disc
> anomalies. They don’t necessarily need to be run exactly
> simultaneously, but the time base would be common and sync could be
> corrected (?). If I have the idea correct, this is different from the
> cake platter concept.
> Others seem to be suggesting a sort of M/S processing which, even if
> successful on one disc, I would never even bother attempting with 2
> Email sent using Optus Webmail