Your Vinyl Transfer Workflow (sharing best needledrop practices)*

Discussion in 'Audio Hardware' started by Vocalpoint, May 11, 2011.

  1. Tim Lookingbill

    Tim Lookingbill Alfalfa Male

    Location:
    New Braunfels, TX
    No, I just named it that to distinguish between the edited version. What I did to the original which starts out at -30dbRMS is first applied Amplify which gave it about 4db then I applied a 3db limiter increase to bring it up to the same RMS level as the the edited version which varies between -16db to-13dbRMS according to Audacity Wave Stats.

    If I didn't do this I'ld have to use the MacMini volume slider to make it louder which changes the sound and doesn't allow much headroom between low and high frequencies on top of becoming too fatiguing being that loud. I keep my volume slider in the middle for a normal listening level across all music I listen to which most fit.

    IOW I applied the posted EQ to the 4db amplified version.
     
    Last edited: May 11, 2021
    arisinwind likes this.
  2. mantis4tons

    mantis4tons Forum Resident

    Location:
    Denver, CO, USA
    I get this sensation all the time when I'm manually de-clicking vinyl rips - the sudden starts and stops of the audio create that weird pressurized/depressurized feeling. It's part of the reason why I find it important to take breaks every 15 minutes or so during a post-processing session.
     
  3. john morris

    john morris Everybody's Favorite Quadron

    Location:
    Toronto, Ontario
    Whatever it takes to make your needle drops 120 % Grant. :edthumbs:
     
  4. john morris

    john morris Everybody's Favorite Quadron

    Location:
    Toronto, Ontario
    Very smart.
     
  5. Grant

    Grant Life is a rock, but the radio rolled me!

    After manually declicking, I start to hear clicks where they don't exist. When that happens, I take a break.
     
  6. john morris

    john morris Everybody's Favorite Quadron

    Location:
    Toronto, Ontario
    You should get someone to have a listen at that point. A sort of unbiased ear.
     
  7. john morris

    john morris Everybody's Favorite Quadron

    Location:
    Toronto, Ontario

    Have you noticed that the majority of this guy's posts seem to be lecturing you? Who does he think he is?

    Oh sheeet.....That's me....Point still stands though.
     
    elvisizer likes this.
  8. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
    Some time moving my head causes the headphones to shift and that can sound like a click. The big thing that happens to me from declicking to long, my tinnitus starts blaring.
     
  9. john morris

    john morris Everybody's Favorite Quadron

    Location:
    Toronto, Ontario
    My Mother has that. In Canada they say there is no treatment for it. Although it has been successfully treated in the States for the last 30 years. We are 25 years behind the times in medical care here in the Great White North. R & D is different story. Good luck.....
     
  10. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
    There is no treatment here either. Protection and acceptance are your best options. Actually many people have it at some level. It can sound like crickets or a swishing sound when you're in silence. Can be an old age thing or life long exposure to noise. Or even a single very loud exposure.
     
  11. john morris

    john morris Everybody's Favorite Quadron

    Location:
    Toronto, Ontario
    There have been known treatments for the condition for decades. I guess it depends on your doctor. Many people in the States have been successfully treated for the condition. Unfortunately the medical community calls the treatment "experimental." Which is bizzare since it had been used for 30 years. May I suggest you do a little research.

    Pete Townsend was successfully treated for it (1988 - 1992). It was a big issue on the Tommy 25th Anniversary Tour but not after that.

    Don't give up. If your specialist says there is no treatment then get a second opinion. Don't give up dude!
     
    stetsonic likes this.
  12. harby

    harby Forum Resident

    Location:
    Portland, OR, USA
    Here I show how I transfer slow - and also demonstrate scripting the SoX audio processor.

    You know what should make an easier-tracking LP record? Reduced velocity and acceleration... so why not slow down the transfer, in the spirit of half-speed masters?

    So I modded my turntable. Its speed is not quartz-locked, but by a frequency generator run by a voltage reference, which is set by resistors and internal trim pots (the same way a Technics pitch slider works). 50k trim pots were too fussy to adjust in small amounts, so I measured their value when set, and replaced them with 5k and a new resistor.

    Then I added a toggle switch on the back (replacing the "syncro remote" port), which adds more resistance still, for a third, even slower, speed setting. Now my turntable both runs at 100.0098% speed -- or runs at 80.1227% speed at the flick of a switch.

    But then, I'd have to use a non-RIAA preamp for the slower speed, wouldn't I? We'll let software take care of that...


    Thought I'd automate some of the required processing. SoX batch file command to "fix" my reduced-speed transfer:

    Code:
    sox --multi-threaded %1.wav -V3 --comment "" %1sped.wav ^
    rate 192k ^
    biquad 1.0000000000 -1.9312626096 0.9313724325 1.0000000000 -0.8796912298 -0.1023703458 ^
    speed 1.248086 ^
    rate 192k ^
    biquad 1.0000000000 -0.8796912298 -0.1023703458 1.0000000000 -1.9312626096 0.9313724325 ^
    highpass 20 ^
    reverse ^
    highpass 20 ^
    reverse
    exit /b 0
    
    (a caret ^ character continues the command on the next line)

    Can you figure out what it's doing?

    sox INFO sox: effects chain: input 96000Hz 2 channels
    sox INFO sox: effects chain: rate 192000Hz 2 channels
    sox INFO sox: effects chain: biquad 192000Hz 2 channels
    sox INFO sox: effects chain: speed 239633Hz 2 channels
    sox INFO sox: effects chain: rate 192000Hz 2 channels
    sox INFO sox: effects chain: biquad 192000Hz 2 channels
    sox INFO sox: effects chain: highpass 192000Hz 2 channels
    sox INFO sox: effects chain: reverse 192000Hz 2 channels
    sox INFO sox: effects chain: highpass 192000Hz 2 channels
    sox INFO sox: effects chain: reverse 192000Hz 2 channels
    sox INFO sox: effects chain: output 192000Hz 2 channels

    In detail:
    • open file
    • upsample
    • apply inverse RIAA filter
    • speed up the audio (increase sample rate)
    • resample back
    • apply RIAA filter
    • subsonic filter (2nd order 20Hz)
    • reverse entire file
    • subsonic filter (2nd order 20Hz - cancel phase delay of first)
    • reverse it back


    A batch processor could do more (SoX has a silence-signature noise reduction) but so far I haven't gotten SoX processing for mid-side to work right. The below should convert to mid-side per docs (and then we can do effects on the individual channels after splitting to files) but levels and mix are wrong...

    Code:
    sox sped.wav -V3 --comment "" midside.wav ^
    remix -m 1v0.5,2v0.5 1v0.5,2i0.5
    
    sox midside.wav -V3 --comment "" midtemp.wav ^
    remix 1 ^
    
    sox midside.wav -V3 --comment "" sidetemp.wav ^
    remix 2 ^
    highpass 60 ^
    reverse ^
    highpass 60 ^
    reverse
    
    sox --guard --combine merge midtemp.wav sidetemp.wav test3out.wav remix -m 1,2 1,2i
    
     
    stetsonic likes this.
  13. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
    This is a better explanation to my current work flow.
    Record 32BF /96 in RX8, manual declick and archive 32/96wav file for future use should I change my denoise flow.

    I'm usually mixing in -85db of noise but I don't really know if it's even audible one way or another. Using one of @Stefan denoise profiles I do a 1st pass. I spend some time setting up the final profile then I scan several places as show in image one. Outputting to noise only and listening to the noise for bass that might have been removed. Once I am happy I denoise the whole file.
    [​IMG]
    I do basically the same thing in a 2nd pass with different sliders in the setup.
    [​IMG]
    Then a 3rd pass just on the gaps. I'm more than confident this is conservative enough to not harm the low frequencies.

    The 3rd image shows how I can use Gain, Waveform stats and simple math to hit -18 RMS
    [​IMG]

    Last is a image of both samples.
    Dropbox - Steve Miller Fly like an eagle.flac - Simplify your life
    [​IMG]
     
    arisinwind likes this.
  14. What is the sample? The finished work? I hear lots of static at the beginning along with inter-track noise.
     
  15. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
    Track 1 has not been denoise.
     
    arisinwind likes this.

  16. I get static at the beginning of both samples. Turn the volume up a little and it is very predominate. I downloaded the sample twice with the same result.
     
  17. harby

    harby Forum Resident

    Location:
    Portland, OR, USA
    Well, got a chance to try out low-speed transfer on something that I thought might show improvement. Result - sounds the same.

    Here we have the vocal intro to "What A Feeling" by Irene Cara, off the same album, 1983 reissue #1 (which didn't reach CD until 1997). It has quite sibilant vocals - apparently also quite baked-in. Have a listen to some needle-drop with this technique:

    feeling-80.flac (2.9MB)

    Here's what you'll be hearing - ghostly traces of airy vocals. However, the broad vertical near-solid lines are SSSibilance, broadband. This is within the first 20 seconds of the LP side.

    [​IMG]

    60kHz bandwidth is apparent, but mostly I see my electronic noise background signature has been shifted higher in frequency by 20% with the later speed-up.

    The reward of all this digitization and processing is sound quality one would never achieve by just taking the 38-year-old record out of the jacket and playing it.
     
    Last edited: May 16, 2021
    stetsonic likes this.
  18. anorak2

    anorak2 Forum Resident

    Location:
    Berlin, Germany
    Does "low speed transfer" mean you play the record at a lower speed such as 16RPM, digitise, and then speed up digitally? Be careful with that. Assuming you have a preamp in the chain, its RIAA curve will be applied in the wrong place of the spectrum thus implicitly und undesirably EQing the outcome.
     
  19. harby

    harby Forum Resident

    Location:
    Portland, OR, USA
    Five posts back is the processing. We are applying a 25% speedup to a 26.7 RPM turntable mod. Much more than that, and the 12Hz tonearm resonance becomes closer to audio.

    First I reverse the calculated frequency response of my cart load into capacitance and spice simulation of the actual phono preamp.

    [​IMG]

    Then inverse RIAA with 192kHz's biquad poles and zeroes applied. That gets us to equivalent of a "no-eq" flat phono preamp.


    After processing, we compare a low speed transfer to the same audio passage at normal speed:
    [​IMG]

    So yes, I was careful with that.

    Besides potentially better tracking, one plausible benefit is that the cartridge recovers faster from dust pops after a speed-up, and those transients and cartridge resonance are higher in frequency from the free increase in sample rate.
     
    Last edited: May 16, 2021
    stetsonic and anorak2 like this.
  20. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
    I guess you are referring to a few light clicks. I'm listening to the needle drop looking for artifacts. To me it sounds good. I don't listen loud enough to hear the clicks. Someone with better hearing might not be happy.
     
    arisinwind likes this.
  21. I don't know what was going on but I downloaded it a third time and it sounds fine. In fact it sounds real good.
     
    ghost rider likes this.
  22. Stan94

    Stan94 Forum Resident

    Location:
    Paris, France
    Any idea about the best way to remove a scratching sound in RX8? Or a noise that sounds like rice krispies, if that makes any sense to you.
     
  23. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
    I have had good luck highlighting the section use M/S and hit it with a strong declick convert back
     
  24. De-crackle might work
     
    Grant likes this.

Share This Page

molar-endocrine