Your Vinyl Transfer Workflow (sharing best needledrop practices)*

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

  1. On a stereo file, that's perfectly normal. The left channel is not a mirror of the right or vice versa.
     
    c-eling and miguelfcp like this.
  2. BrilliantBob

    BrilliantBob Select, process, CTRL+c, CTRL+z, ALT+v

    Location:
    Romania
    Besides clipping, distortion matters too. First, is important to plan the recording destination. Personal use, trading and so on. I read somewhere the mastering engineers said any audio digital file with peaks above -6dB will generate distortion in whatever cheap audio gear (amps, DACs, speakers,...). For quality recordings they recommend more room (-6dB to -3dB max peak). Any audio gear has the volume control knob to play as louder you want.
     
  3. miguelfcp

    miguelfcp Well-Known Member

    Location:
    Portugal
    Now I'm confused, should I normalize to -0.3dB or -0.6dB?
     
  4. BrilliantBob

    BrilliantBob Select, process, CTRL+c, CTRL+z, ALT+v

    Location:
    Romania
    Additional headroom reduces artifacts used by all data compression formats which use the masking effect. The lower the data rate of the target format, the more headroom is necessary. For highly compressed music, 0.3 dB headroom is not sufficient in order to avoid distortion. In this case, a headroom of 5 dB would be theoretically necessary. Recordings which have the DR logo are fortunately not so strongly compressed; a headroom of 0.3 dB prevents unwanted artifacts in most cases!

    That means if you compress the audio file for more loudness and lose some dynamic range and additionally you want later to resample to mp3 or wav 44/16 you need more headroom. The peak value to -5dB!
     
    Last edited: Jan 15, 2019
  5. He's not talking about normalizing to -0.3 or -0.6 dB but -3 or -6 dB, which is a much lower peak level. That's another approach suggested by some experts to completely eliminate any possibility of distortion, depending on your playback device. It's "the safest bet", but -0.3 dB has always worked fine for me. All the needledrops in my YouTube channel (link in my signature) are peak normalized to -0.3 dB (I know my recordings aren't a "reference point for audiophiles" or something like that, but they definitely aren't bad either).
     
    miguelfcp and BrilliantBob like this.
  6. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
    Sorry I meant-0.3
     
  7. BrilliantBob

    BrilliantBob Select, process, CTRL+c, CTRL+z, ALT+v

    Location:
    Romania
    Subscribed on your YT channel. Nice stuff.
     
  8. I'm glad you like it. Thank you! :)

    P.S.: Right back at you!
     
    BrilliantBob likes this.
  9. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
    I confess over the years I have heard that running using DC offset correction software is good but I have never used it. Sound Forge has a process for but RX does not or it's buried somewhere inside and I never looked for it. That and phasing issues I have no idea how to tell. As far as I'm concerned my system is fine and all my drops sound as good as the records without the pops.

    So here is a sample of Steely Dan Pretzel Logic it's 2 tracks around 60 seconds each can someone test it to see if I'm having unknown issues.
    Steely Dan Pretzel Logic.flac (37.62MB) - SendSpace.com
     
    BrilliantBob likes this.
  10. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
    I'm not sure what and or why it happened for you but I will occasionally renormalize the lower channel to get it in the ballpark. Sometime the record level is a tick off. It's always a judgement call may be how the engineer wanted it to be mastered.

    I found this in RX and used it on the sample.
    [​IMG]
     
    Last edited: Jan 15, 2019
  11. Machiventa

    Machiventa Forum Resident

    Location:
    Salida, Colorado
    Can someone please give a few examples of peoples setup for needledropping? I read through many pages and have yet to see any.
     
  12. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
    I think like you most or all my equipment is listed in my profile page. It's pretty basic for me TT preamp soundcard software. I record and do most everything in RX6 advanced.
     
    c-eling, BrilliantBob and Machiventa like this.
  13. harby

    harby Forum Resident

    Location:
    Portland, OR, USA
    [​IMG]

    The Denon receiver is a bit unique in that you can set the tape out/rec out to an input different than the one you are listening to, so the receiver will always output phono while you choose what to monitor. Excellent phono pre for what others charge for just a phono stage.
     
  14. miguelfcp

    miguelfcp Well-Known Member

    Location:
    Portugal
    Then I will keep -0.3dB. I will use the "rips" with my portable equipment (1More Triple Driver and DragonFly black) so it's ok, it won't clip or distort I think.
     
  15. Grant

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

    I do my phase DC offset correction in Audition. You can check for DC in RX simply by running the statistics.
     
  16. Grant

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

    You're in good shape! In phase, and your DC is at 0.
     
    ghost rider and BrilliantBob like this.
  17. FrankieP

    FrankieP Forum Resident

    Here’s My setup for needledropping. I’m using a Music Hall PA2.2 phono pre which outputs both analog and digital streams at the same time. I record the analog signal into my Tascam DR-100MK3 and the digital signal gets fed via USB into my MacBook Pro running ClickRepairRT for real-time click and pop removal (a poor mans version of SweetVinyl’s Sugarcube SC-1 or 2). The 24bit 96k audio is fed from my MOTU Microbook IIc via coaxial SPDIF and recorded into a MOTU 828MK3 Hybrid.

    Here’s a diagram of all the components that I’m using…

    [​IMG]

    [​IMG]

    And here’s a sample of a needledrop…

    Abandon Ship (Instrumental) - April Showers
    Raw unprocessed audio (recorded with a Tascam DR-100MK3
    Dropbox - Abandon Ship (Instrumental)_Raw Unprocessed.flac

    ClickRepairRT processed audio (recorded with a MOTU 828MK3 Hybrid)
    Dropbox - Abandon Ship (Instrumental)_ClickRepairRT Processed.flac
     
  18. BrilliantBob

    BrilliantBob Select, process, CTRL+c, CTRL+z, ALT+v

    Location:
    Romania
    The recording is well dressed with mid-low frequencies and sounds balanced.

    I needledroped and processed with the Mid/Sides technique the "Rush - 2112" new LP. Soon I will put this on YT. I used only the TT, the desktop PC, the speakers and my ears. I am interested in your opinion about the quality of the recording.
     
    Last edited: Jan 16, 2019
    Grant and ghost rider like this.
  19. c-eling

    c-eling They're made of light,We never would have guessed

    Quite true Hail, my input's have two separate left/right adjustable in's so I just use a mono record to set channel balance.
     
  20. c-eling

    c-eling They're made of light,We never would have guessed

    Just got done doing this one.
    From a 86 US Dennis King Cut
    French Canadian synthpop band Trans-X
    Technics 1900-->Lounge MK III-->Focusrite 2i2 Gen2-->PC
    This album has some sibilance issues, I found the AT7V did the best job on it
    (compared to an Orto Blue, Nag 110 and a AT 150MLX)
     
    Last edited: Jan 16, 2019
    arisinwind likes this.
  21. Chris Schoen

    Chris Schoen Rock 'n Roll !!!

    Location:
    Maryland, U.S.A.
    Simple: Turntable to phono pre-amp to CD Recorder "in" (CD Recorder "out" to Receiver or Amplifier "in", so you can hear what you are recording).
    (Option: add a Schiit Loki equalizer between pre-amp and CD Recorder for custom tone control). :righton:
     
  22. harby

    harby Forum Resident

    Location:
    Portland, OR, USA
    That should be fine if you are listening to WAV or FLAC. However, if you are encoding to a lossy perceptual codec such as MP3 or AAC, you should allow further headroom, such as -3dB. The filtering of stuff you can't hear by the codec can actually increase peak levels and cause clipping.

    For example, below,
    • "Lossless" is the source material from a CD - it is brickwalled, but not technically clipping.
    • "Peaks truncated" is when the mp3 is decoded at the same volume as the source, and is clipped.
    • "Peaks preserved is when the mp3 is encoded at a -5dB lower level, preventing clipping.

    [​IMG]

    Source: Why does my mp3 Clip?

    The LAME encoder has a command-line option --clipdetect. It will show you how much lower you have to set the --scale setting to prevent clipping. You can fix the source WAV instead, by not normalizing to the maximum.

    --
    Back to another topic - one thing that does survive lossy encoding is the subsonic rumble. Filtering it out yourself will leave more bits for accurate encoding of actual music. Here's a vinyl rip that's been mp3 encoded without high-pass (subsonic-cut) filter, from my archives:

    [​IMG]

    Halfway through the clip, I removed everything above 30Hz with a Chebychev 12th order filter. We can see in the second half, there is nothing musical about the sub-bass that is remaining - it is a repeating pattern of rumble and tonearm resonance that corresponds to the disc rotation.

    How loud is the subsonic noise? Let's look:

    [​IMG]

    To demonstrate "musical" bass vs turntable noise, I applied first a 60Hz low-pass filter, leaving bass drums and bass notes. You can see the musical rhythm between 15-30 seconds. After that, 30Hz lowpass leaves only noise, and again we see a repeating pattern from the disc rotation.

    RMS analysis reveals subsonic noise (not audible, but you can feel the subwoofer cone move) is only 20dB below the RMS mean of music. It is significantly louder than the turntable's rumble spec, both because that specification is A-weighted (corresponding to our perception of noise), and tonearm resonance is excited by the music on non-blank vinyl.

    (also note the low peak levels at the start - set by the mp3 clipdetect algorithm)

    PS Here's the audio depicted in the first spectral view; enjoy 30 seconds of quality needledrop (after two mp3 encodings) and 30 seconds of rumble. abdul-subbass.mp3 (1.5MB MP3)
     
    Last edited: Jan 16, 2019
  23. atoxique

    atoxique Forum Resident

    Pioneer PL-200 turntable
    Audio Technica AT-95E cartridge
    Yamaha AV-M99 amplifier

    This is a simple effective workflow: all I do is connect my amplifier's record out to my soundcard. Then I use Audacity to record the entirety of each LP side at 48kHz (I don't bother with 96kHz) then use Audacity's de-click function whenever it's needed. Next, I use "Effect\Amplify..." on the entire recording, using the default option which is how loud the audio can go without clipping so that the loudest part is as loud as possible without distortion and all the quiet parts stay quiet. After that, I use "Effect\High pass filter..." to get rid of all frequencies below 25Hz with the roll-off set to 24dB per octave. I also use "Effect\Low pass filter..." to get rid of all frequencies above 2okHz, again with the roll-off set to 24dB per octave. Audacity has a neat label function, so after trimming the silence before and after each LP side, I go to "Tracks\Add new...\Label track" and then simply click and type the song's name and track number at the point where each song begins (like this: A1 Song Name, A2 Song Name, A3 Song Name, B1 Song Name, so on so forth). Lastly, I use "File\Export\Export multiple..." to export each song in AIFF at 24-bit, 48kHz. I do this by doing this in the menu for the export: I select the folder I want my files to go into and then choose Format\Other uncompressed files with the options being "Header: AIFF (Apple/SGI)" and "Encoding: Signed 24-bit PCM". I use AIFF because it is uncompressed and supports metadata unlike WAV. In the "Split files based on:" section, I choose the "Labels" option and in the "Name files:" section, I leave it as "Using Label/Track name". Finally I click Export and add in metadata tags. Now I have an entire LP ripped to my computer, split up and tagged ready for importing into iTunes or any other media player.
     
    bever70 likes this.
  24. ghost rider

    ghost rider Forum Resident

    Location:
    Bentonville AR
  25. harby

    harby Forum Resident

    Location:
    Portland, OR, USA
    I was hoping you would discover the link I provided, with instructions detailing how to enable the waveform view in RX for yourself.

    The spectral editor looks like it does a good job, and it should have 0 phase shift if implemented correctly. The magnitude of the 25Hz and 30Hz tones at the end is preserved while the subsonic tone is eliminated.

    [​IMG]

    I think Izotope RX, both for their spectral viewing and visual spectral editor, must use a sub-sampling scheme, where the audio is broken into parts, and the low frequencies are down-sampled to allow more precise FFT.
     

Share This Page

molar-endocrine