From: geoff@nospamgeoffwood.org   
      
   On 16/09/2025 7:25 am, Liz Tuddenham wrote:   
   > Scott Dorsey  wrote:   
   >   
   >> Liz Tuddenham  wrote:   
   >>> Tobiah  wrote:   
   >>>   
   >>>>> I could start a thread on an analogue declicking computer for 78s,   
   >>>>> which I am currently designing, but the number of participants would   
   >>>>> probably be very limited.   
   >>>>   
   >>>> You might be surprised.  Is it working?  How does   
   >>>> its performance compare to digital solutions.   
   >>>   
   >>> I'm pleased to report that the MkIII version which I have been working   
   >>> on recently has now worked for the first time.  The results are very   
   >>> promising and it is extremely easy to operate.   
   >>>   
   >>> Unfortunately, in trying to track down a common-earth hum problem, I   
   >>> accidentally blew up one of the power supply regulators and I didn't   
   >>> have a spare.  A new one is on order, so I hope to resume testing in a   
   >>> few days.   
   >>>   
   >>> The basic principle is to split the sound spectrum into ten bands, each   
   >>> about an octave wide.  In each band a click or crackle is identified by   
   >>> comparing the vertical and horizontal vectors of the stylus movement -   
   >>> if there is more than a certain proportion of vertical movement, a   
   >>> switch opens and momentarily disconnects the signal.  The band is then   
   >>> filtered again, so that the harmonics due to switch clicks are removed   
   >>> and damped resonance in the filter covers any short gaps in the sound.   
   >>> All ten bands are then recombined to restore the original spectrum.   
   >>   
   >> That's not a declicker!  That's a broadband noise reduction system!   
   >   
   > It was primarily intended to de-click and de-crackle 78s, but it can   
   > also de-thump cracked ones, so I suppose it is more broadband than I   
   > originally suggested.   
   >   
   >   
   >> Compare with the Dolby Cat 43 system.   
   >>   
   >> The declicker has a delay and looks at the first derivative of the input   
   >> signal, and when it is too high it replaces the signal with the delayed   
   >> signal or it holds the existing level, or it fades from one to the next,   
   >> in order to blend around a single impulse.  Compare with the SAE 5000.   
   >>   
   >> I think that the Audio Cyclopedia has schematics of both designs, both   
   >> pretty primitive but showing the techniques.   
   >   
   > Finding the clicks and crackle is realtively easy on a mono recording:   
   > they show up in the vertical component (which obviously wasn't   
   > recorded).  The big problem is what to do about them.  The Marantz and   
   > other noise-reduction systems simply faded out a click - which   
   > substituted a 'bloop' instead; cross-fading between before and after   
   > could be better if it is done well.   
   >   
   > The only system that seemed to do it really effectively and in real time   
   > was Cedar: it claimed to synthesise a 'gap-filler' from the Fourier   
   > transform of the sound before and after the click.  Even that wasn't   
   > really intended to handle the continuous crackle of some of the worst   
   > gritty solid stock material (particularly inter-wars U.K. HMVs).   
   >   
   > I wondered if some sort of resonator could 'ring' and fill the gap, but   
   > it would need quite a lot of them to cover the whole audio band.   
   > Octaves seemed the logical way to go but that involved either a large   
   > number of ferrite pot cores or a lot of op-amps in state-variable   
   > filters.  I tried the pot-core approach many years ago and gave up   
   > because it was becoming too unweildy.   
   >   
   > This time I opted for state-variable filters and my design finished up   
   > with 170 op-amps and several other I.C.s.  It was a bit of a gamble   
   > whether it would be worth building it, as the principle was untested, as   
   > far as I knew.  It has paid off, the results are stunning; as the   
   > sensitivity is increased, the crackle just fades into the background and   
   > the music comes through absolutely unaffected.   
   >   
   > There are still a few things that need attention, as I can hear residual   
   > artefacts at a very low level - but I exhibited it to a group of   
   > gramophone enthusiasts at the weekend and they were completely bowled   
   > over by the sound quality.   
      
   Isn't this all now trivial to do, better, in software ? Or is the   
   project more of an intellectual exercise ?   
      
   --   
   geoff   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)