Your Vinyl Transfer Workflow (sharing best needledrop practices)*

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

  1. Stefan

    Stefan Senior Member

    Location:
    Montreal, Canada
    Sorry but this makes no sense. Any noise that's below 20kHz will remain below 20kHz. Recording to 5.6 MHz DSD then converting the file to 384/24-bit PCM will not move noise from one frequency to another (if it did then the music would move too!). The noise above 50kHz you're removing with your next step is shaped noise added by the DSD format. Any background noise from the original vinyl will remain at it's original frequency and level (otherwise you wouldn't have to sum the bass below 30Hz to remove rumble!).
     
    stetsonic and Grant like this.
  2. BrilliantBob

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

    Location:
    Romania
    DSD = less noise in BASS & MID area, more noise in UPPER TREBLE area (>15000 HZ). Rumble is considered part of the recorded music signal. Focus on the quantization noise.

    If I record in DSD 5.6 -> convert to PCM 384/24 -> cut above 50 KHz -> resample to PCM 192/24, the resampled wav 192/24 sounds way better than the recorded wav 192/24. Try and compare.

    [​IMG]
     
  3. Stefan

    Stefan Senior Member

    Location:
    Montreal, Canada
    this may well be true, but the focus of this thread is vinyl digitization. The quantization noise you're taking about in a 24-bit file is by definition -144dB, at least 24dB lower than the lowest level ever measured by science as perceptible by humans. It's mushc lower than the real world noise added by the physical components (thermal noise from chips, etc.). Even if we could hear that noise in a vinyl rip, it would be masked (and naturally dithered) by vinyl background noise, ambient noise in the recording, etc. These are all well above the lower limit of human hearing.

    If you feel your method sounds better to you, great. I would say however that if you notice a significant difference between your DSD recording converted to PCM 192/24 and a recording done directly to PCM 192/24 then there's either something seriously wrong with your PCM ADC! (or else you're a victim of expectation bias)

    Although I didn't always agree with mastering engineer Barry Diament when he was on here, I did respect his opinions on quite a few points and he was of the strong belief that the finest digital recording format he ever heard (including DSD), was PCM 192/24. If I recall correctly, he stated that it was the first time what he heard coming back from the recording was indistinguishable from the the microphone feed. Mind you, he was using an $8000 Metric Halo system that most of us with mortgages, etc., to deal with won't be using. However, if the difference really was as clearcut as you claim, why wouldn't Barry and all the other recording and mastering professionals out there be massively switching to DSD. There are some who like DSD granted, but if it really sounded "way better" don't you think a lot of professionals in the industry would be switching?

    Besides, think about the logic for a moment. If PCM adds quantization noise that reduces the fidelity, it stands to reason that once you convert your DSD file to PCM, the quantization noise will be added by the conversion process so in total you'll have quantization noise from the PCM plus all that noise from the DSD (which will be greater above 15kHz as you stated). Perhaps all that noise in your converted file is affecting your listening experience euphonically!

    Anyway, let's not turn this useful thread into a debate on the merits of DSD versus PCM. Most of the folks participating in this thread are not using DSD recorders, which are only now reaching somewhat affordable levels for home recordists. We're more likely to be using PCM-based ADCs.
     
    ascot, stetsonic and ghost rider like this.
  4. BrilliantBob

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

    Location:
    Romania
    The built-in ADC my TT is a Burr-Brown PCM4202. I can record in both formats, DSD 5.6 MHz or WAV 192/24. I was surprised too when I heard that DSD5.6->PCM384/24->PCM192/24 sounds better than the simple PCM192/24 recording.

    DSD Output 5.6448MHz
    Total harmonic distortion + noise: −105 dB
    Dynamic range: A-Weighted 118 dB

    WAV 192kHz 24-bit
    Total harmonic distortion + noise: −103 dB
    Dynamic range: A-Weighted 117 dB, no weighting 108 dB

    https://www.ti.com/lit/ds/symlink/pcm4202.pdf
     
  5. Stefan

    Stefan Senior Member

    Location:
    Montreal, Canada
    Yes but your turntable's signal-to-noise ratio is only -50dB so its noise floor is a whopping 55dB louder than the DSD distortion+ noise or 53dB louder than PCM. So I can't see how a 2dB difference that's more than 50dB below your turntable's noise floor is going to make one format "way better" sounding than the other.

    Anyway, if you're happy with it, more power to you!
     
    Robert C and ivor like this.
  6. BrilliantBob

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

    Location:
    Romania
    My TT noise floor have total RMS level -77.83 dB with some rare and random peaks up to -64 dB, in the inaudible range, but I remove it anyway because you know, any "quiet" garbage give mud to the audible range by signal combination.

    I remove three noise layers to make the song as transparent as possible; the TT noise floor first (when the tonearm is in its rest but TT is running), then the stylus-on-groove noise removal applied twice, until the whole TT and record noise floor is dumped below -120 dB. The noise profile samples I put in the spectral de-noise don't include the quantization noise existing inside the musical signal.

    I don't know how these DSD->PCM conversions work, but the "DSD5.6->PCM384/24(brickwalled at 50KHz)->PCM192/24" method give better quality sound.
     
    Lucca90 likes this.
  7. marcob1963

    marcob1963 Forum Resident

    Have to agree with Mr. Diament, 24/192 is King. At least for a great facsimile of analog.

    Pure DSD may compare, but for the purpose of needle drops, I don't think you can edit in DSD.
     
  8. Old Zorki II

    Old Zorki II Storm Watcher

    Location:
    near Tampa, FL
    Gentlemen, I have a question. Still thinking about what equipment I need to purchase (without computer connected to my system).
    A lot of people here emphasized importance of proper recording level. Yet some phono pre and processors (Sugar Cube, PS Audio, Puffin, sound processors such as miniDSP, etc) routinely convert to digital and back, and most of reviewers claim that sound is very transparent.
    If they can move to digital without significant distortions - what stops vinyl digitizing equipment from doing the same?
    Again, not a rhetorical question, I simply do not know the answer.
     
  9. Pythonman

    Pythonman Forum Resident

    Location:
    Florida
    My own practice of recording “needle drops” is to record as faithfully as possible straight to CD with no further processing. My only concession is I will bring up the levels to just a bit under the redline. If I pick up a tick or a bit of the sounds of the stylus going through the grooves then it’s fine as long as it’s not worse than what I hear just playing the record.

    Afterwards I’ll make a note on each CD the turntable, cartridge and phono preamp used to make that recording. My thinking is if at some much later date I’m unable to spin the records I’ll at least have an archive of them and an account of the different carts, etc. I once enjoyed.
     
  10. stetsonic

    stetsonic Forum Resident

    Location:
    Finland
    I got Standard and it's more than enough for me. The extra features of Advanced are mainly dialogue related, save for "Deconstruct" which seems like a great way to REALLY mess up a needledrop. :D
     
    Old Zorki II likes this.
  11. stetsonic

    stetsonic Forum Resident

    Location:
    Finland
    That strategy has one possible caveat - if you store your recordings as plain audio CDR's, they are more prone to go bad over time. The reason being that an audio CD has way less error correction compared to a data CD or DVD. Writable CDR is a volatile media, the ink inevitably fades over time, making it harder for the laser to read accurately.

    Years later, they might still play flawlessly or near flawlessly on a CD player but when you try to rip them to back your computer you might find the process is riddled with small read errors. Especially if the CD/DVD drive is not the exact same model the disc was burned on. Audio CD suffers more when "bit rot" occurs; sturdier error correction of a data disc format makes it easier for the drive to deduce the missing/unreadable information.

    What I'm saying is, if you've invested a lot of time making those discs and/or the contents are valuable to you, consider making image file backups of those CDR's now, and store them on an external hard drive or a cloud service or something. I don't trust an audio CDR as a backup medium, even if it was burned onto a super azo Verbatim.
     
    Grant and Old Zorki II like this.
  12. miguelfcp

    miguelfcp Well-Known Member

    Location:
    Portugal
  13. JohnO

    JohnO Senior Member

    Location:
    Washington, DC
    It is probably ok, but I recommend this as better, from Audio-Technica
    https://www.amazon.es/Audio-Technica-AT6012-Record-Care/dp/B0169N7KOK/

    After that original fluid is gone, the 8 oz Vinyl Styl "Record Cleaning Fluid" is an almost identical replacement fluid at a better price for 8 oz. They also make a "Record Wash Fluid", but that is not the right thing, that one would be ok for a Spin Clean type thing.
    https://www.amazon.es/Vinilo-Styl-Record-líquido-limpieza/dp/B01ARJMW3G/
     
    miguelfcp likes this.
  14. Lucca90

    Lucca90 Forum Resident

    Location:
    SouthAmerica
    Hey Bob, how did you determine that?
     
  15. Old Zorki II

    Old Zorki II Storm Watcher

    Location:
    near Tampa, FL
    Totally agree. As price of storage drops a NAS with RAID will serve as perfect storage for rips, as well as media library ready for streaming, for very little money (in comparison with what we spent on equipment and records) .
     
    Grant, stetsonic and Stefan like this.
  16. Stefan

    Stefan Senior Member

    Location:
    Montreal, Canada
    Good question. Sony rates the PS-HX500 as having "more than 50dB signal-to-noise ratio" but it's probably somewhere in the low 50s or else they'd state more than 55 or 60db, which would be better from a marketing point of view. Mind you both Project turntables I owned were rated at 70dB yet they were nowhere near that quiet!
     
    Grant likes this.
  17. Lucca90

    Lucca90 Forum Resident

    Location:
    SouthAmerica
    I believe Sony only gives reference points, not to be taken literally, just minimums.
     
  18. Stefan

    Stefan Senior Member

    Location:
    Montreal, Canada
    Strange marketing strategy. I would tend to stay away from anything with less than 50dB and I know a lot of folks around these parts who'd want at least 70 or more.
     
    Grant and Robert C like this.
  19. Pythonman

    Pythonman Forum Resident

    Location:
    Florida
    Good advice. All my recordings, most anyhow are uploaded as AIFF files to a “ cloud”.
     
  20. Vocalpoint

    Vocalpoint Forum Resident Thread Starter

    +1

    While I have burned the occasional rip to CD - the actual final "masters" of every needle drop I do are saved as FLAC files at the end of my transcription process. All files are stored in three places along with the standard info and documentation should I need to go back to them. For me - a CD burn is non-trusted disposable item that should never be considered the master.

    File based storage is where it's at - and those should exist in three unique places to be considered truly back up.

    Cheers,

    VP
     
  21. BrilliantBob

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

    Location:
    Romania
    1. TT noise incl. hum = TT is running but tonearm stays in its rest location
    total TT noise RMS = -77 dB with peaks up to -64 dB (I damped the tonearm and the platter)
    - I detached the tonearm and I isolated it below with 4 rigid anti vibration gaskets;
    - I wrapped the tonearm with teflon tape insulation for water pipes;
    - I put on the platter a cork+rubber suplimentar mat and I glued the outer area inside the platter with Titebond II Premium wood glue).
    2. Record noise = Stylus is dropped on the outermost empty groove of the record
    3. The music area = the stylus start recording the music

    [​IMG]

    Now the needledrops sound like this (VG+ 1981 record):

     
    Last edited: Aug 15, 2019
    Lucca90 likes this.
  22. BrilliantBob

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

    Location:
    Romania
    How it looks the cleaned needledrop:

    [​IMG]
     
    superstar19 and Lucca90 like this.
  23. Lucca90

    Lucca90 Forum Resident

    Location:
    SouthAmerica
    Marketing strategy? I don't think so.
    It seems to me that they (Sony I mean) are just covering themselves, just in case. For example, W&F spec of 0.25, which I think is pretty high, in real life is more like 0.12 or 0.13. so, in this way, no one should be able to claim that the turntable they receive is defective 'cause everything is within published specs.
     
    arisinwind and BrilliantBob like this.
  24. marcob1963

    marcob1963 Forum Resident

    Why the Low Pass at 20,000? I used to do that and it sounded cleaner, however it had also lost ambience and snap. Wouldn't recommend it.
     
  25. BrilliantBob

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

    Location:
    Romania
    Worn record with audible distortion in 20KHz-24KHz range. When the needledrop is recorded from a new vinyl with no sibilance and distortion I only put a brickwall at 32.5 KHz about to remove aliasing and other unwanted HF noise.
     

Share This Page

molar-endocrine