Announcement

Collapse
No announcement yet.

Flat Preamp, CNF and Impulse

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Flat Preamp, CNF and Impulse

    I transfer my records using the flat preamp, then I use the virtual phono preamp, de-click them and then I use the CNF as needed.

    Lately, I've experimented with using the cnf before the virtual phono preamp. It ends up with a nice clean sound, and the settings can be very gentle. The problem, though, is that it puts the impulse filters after the cnf has worked.

    Does anyone here use the cnf or ez-clean before the virtual phono preamp? It seems very easy to get distortion that way, but it seems like it should be do-able once you get used to it and develop some presets. You could have it in a multifilter with the impulse, then cnf, then virtual phono.

    Just wondering if anyone does it this way.

    Dan
    Last edited by Craig Maier; 11-16-2019, 04:29 PM.
    Dan McDonald

  • #2
    Just going on the appearance of impulse noise in the spectrogram - it usually covers the entire frequency spectrum. As such, I would be worried that impulse noise present in the audio during the CNF stage will allow through more unwanted wideband noise. The CNF will also apply it's attack-release parameters to the impulse noise which might make it harder for the INF/EZ-INF to detect the impulse later on?

    those are just my feelings so I may be wrong!
    At work I may look like I'm doing nothing, but at the cellular level I'm actually quite busy

    Comment


    • #3
      Hey Doug -

      yes, that's the problem with putting the cnf in before the impulse filter. I'm thinking more of the impulse cleaning of the signal, then the cnf, then the vpa. When I said EZ-clean below, I meant EZ-impulse.

      I don't see any reason this should be difficult to do if you can avoid having the ez impulse distort the signal. I may be wrong, but it seems like the impulse filters work mainly off the high end of the signal in detecting clicks, and the high end should be easy to detect before the vpa. I'm just worred about the distortion.

      Dan
      Dan McDonald

      Comment


      • #4
        Dan,

        Why does the order matter on the CNF ? Is it because the EQ action reduces high frequency sounds left from the CNF action ? I'm assuming that you are transfering LP records ?

        Regards,
        Marc

        Comment


        • #5
          Hi All,

          In general, the CNF has some effect on a signals time domain transient response, especially when high fft sizes are used and/or long attack and/or release times are applied. This could effect the discrmination capability of the various impulse filters, although I admit that I have never actually tried reversing the order between the various filters as recommended on this thread.

          I always use the following order after a flat transfer:

          1. VPA / VPP (Virtual Phono Preamp)

          2. Impulse Filters

          3. CNF

          4. Other processes / effects / EQ's / etc.

          So, I am not sure what happens with other filter sequences. I use the order mentioned here because that is basically how we developed the system (with a certain logic associated with that decision) and how we tested it, which does not say that other filter orders will not work.

          A single test is worth a thousand expert opinions, so give it a whirl.

          Craig
          Last edited by Craig Maier; 11-16-2019, 04:31 PM.
          "Who put orange juice in my orange juice?" - - - William Claude Dukenfield

          Comment


          • #6
            I've been testing and happy with putting the cnf in front of the vpa, but then the impulse filters don't seem to work as well or as easily because of the effect on the impulses. I was just thinking that maybe someone else had found a way around that problem. I would think the impulses could be just as effective in front of the vpa, but the settings have to be somewhat different, and I haven't been able to figure out what those settings should be.

            Dan
            Dan McDonald

            Comment

            Working...
            X