FAQs (Vinyl): Difference between revisions

From Hydrogenaudio Knowledgebase
mNo edit summary
 
(3 intermediate revisions by 2 users not shown)
Line 2: Line 2:
* What sample rate to record with?
* What sample rate to record with?
** ([http://www.hydrogenaudio.org/forums/index.php?showtopic=43709&hl=vinyl thread] [http://www.hydrogenaudio.org/forums/index.php?showtopic=38016&hl= thread])
** ([http://www.hydrogenaudio.org/forums/index.php?showtopic=43709&hl=vinyl thread] [http://www.hydrogenaudio.org/forums/index.php?showtopic=38016&hl= thread])
** In general 44.1khz will work fine. There are not significant reasons to choose either 44.1khz or any higher sample rate. Some postprocessing tools for vinyl recordings only operate at 44.1khz.
** In general 44.1kHz will work fine. There are no significant reasons to choose either 44.1kHz or any higher sample rate. Some post-processing tools for vinyl recordings only operate at 44.1kHz.
** Additionally the dynamic range and distortion specifications for many sound cards are better at 44.1khz than at higher sample rates.
** Additionally, the dynamic range and distortion specifications for many sound cards are better at 44.1kHz than at higher sample rates.
** This does not preclude using higher sample rates if there is a clear objective reason to do so. Probably the most important reason to use a higher sample rate is if you are recording a CD4 album, which has signal all the way to 30khz.
** This does not preclude using higher sample rates if there is a clear objective reason to do so. Probably the most important reason to use a higher sample rate is if you are recording a CD4 album, which has signal all the way to 45kHz.
** Certain DSP operations may operate more effectively in the audible range if the sampling rate is increased, but this is not a general rule of thumb.
** Certain DSP operations (such as impulse noise reduction) may operate more effectively in the audible range if the sampling rate is increased, but this is not a general rule of thumb.
* What bit depth to record with?
* What bit depth to record with?
** ([http://www.hydrogenaudio.org/forums/index.php?showtopic=38016&hl= thread])
** ([http://www.hydrogenaudio.org/forums/index.php?showtopic=38016&hl= thread])
** Similar to the sample rate question above. In general 16 bits actually does work well.  
** Similar to the sample rate question above. In general, 16 bits actually does work well.  
** It's a common misconception that the quantization noise will significantly raise the noise floor of the recording; as long as the signal peak is -10dbFS or higher it is not an issue.
** It's a common misconception that the quantization noise will significantly raise the noise floor of the recording; as long as the signal peak is -10dbFS or higher it is not an issue.
** CDs have a lower noise floor than vinyl, period. Their noise performance exceeeds vinyl at all frequencies.
** CDs have a lower noise floor than vinyl, period. Their noise performance exceeds vinyl at all frequencies.
** If you plan to perform a very large amount of digital processing on the recording, the rounding errors could build up sufficiently that they impinge on the background noise level of the vinyl. In that case the recording should be made at (or converted to) 24 bit resolution prior to the processing stages to reduce the effect of rounding errors. That said, you would need to do a very great deal of processing before a 16 bit recording of vinyl became degraded in this way. A few passes of EQ, noise reduction, amplitude adjustment, etc. isn't going to bring the quantisation noise up above the vinyl noise floor.
** If you plan to perform a very large amount of digital processing on the recording, the rounding errors could build up sufficiently that they impinge on the background noise level of the vinyl. In that case the recording should be made at (or converted to) 24 bit resolution prior to the processing stages to reduce the effect of rounding errors. That said, you would need to do a very great deal of processing before a 16 bit recording of vinyl became degraded in this way. A few passes of EQ, noise reduction, amplitude adjustment, etc. isn't going to bring the quantisation noise up above the vinyl noise floor.
* How low of a frequency can LPs produce?  
* How low of a frequency can LPs produce?  
** ([http://www.hydrogenaudio.org/forums/index.php?showtopic=37509&hl= thread])
** ([http://www.hydrogenaudio.org/forums/index.php?showtopic=37509&hl= thread])
** The common wisdom is that they don't have much below 60hz, but there is evidence that the signal extends as low as 20hz and possibly lower on some LPs.
** The common wisdom is that they don't have much below 60hz, but there is evidence that the signal extends as low as 20hz and possibly lower on some LPs.
** Note that low bass on vinyl records will be mono, but this is not a problem because low bass is non-directional. (The reason for this is that low bass signals cause the largest groove excursions, and of course  the vertical (out of phase) component of the groove must never be so great that the groove rises above the surface of the record, so the out of phase component of the low bass is removed when mastering for vinyl).
** Note that low bass on vinyl records will be intentionally mastered in mono, but this is not a problem because low bass is non-directional. The reason for this is that low bass signals cause the largest groove excursions, and of course  the vertical (out of phase) component of the groove must never be so great that the groove depth becomes so shallow at a maximum peak of modulation that it will not adequately retain a stylus, or so deep that during mastering, the cutter stylus passes through the lacquer into the aluminum substrate. The out of phase component of the low bass is largely removed when mastering for vinyl.
* How much bandwidth is on an LP?  
* How much bandwidth is on an LP?  
** ([http://www.hydrogenaudio.org/forums/index.php?showtopic=35530&hl= thread])
** ([http://www.hydrogenaudio.org/forums/index.php?showtopic=35530 thread])
** ([http://www.hydrogenaudio.org/forums/index.php?showtopic=93998 another thread])
** This is a hard question, because the dynamic range has to be considered and it is frequency-dependent. It is almost certainly lower than CD, and certainly far lower if the theoretical bandwidth limits of the CD medium are considered, rather than the digital specifications.
** This is a hard question, because the dynamic range has to be considered and it is frequency-dependent. It is almost certainly lower than CD, and certainly far lower if the theoretical bandwidth limits of the CD medium are considered, rather than the digital specifications.
** Although some sort of signal above 20kHz can be recovered from an LP, it is highly unlikely to be related to the programme material; in other words it is noise. The majority of LPs probably contain nothing of significance above about 16kHz.
** Although some sort of signal above 20kHz can be recovered from an LP, it is highly unlikely to be related to the programme material; in other words it is noise. The majority of LPs probably contain nothing of significance above about 16kHz.
** CD4 (quadrophonic LPs) had signals cut to 30khz or above.
** CD4 (quadrophonic LPs) had a subcarrier at 30kHz, and signals cut up to 45kHz and above.
** CED disks had a usable signal extending to the Mhz range, but used a completely different playback technology that is completely incomparable to audio LPs.
** CED disks had a usable signal extending to the Mhz range, but used a completely different playback technology that is completely incomparable to audio LPs.
* What recording level should LPs be recorded at?  
* What recording level should LPs be recorded at?  
Line 41: Line 42:


[[Category:Guides]]
[[Category:Guides]]
[[Category:Vinyl Guide]]
[[Category:Vinyl]]

Latest revision as of 06:49, 2 October 2016

HydrogenAudio threads

  • What sample rate to record with?
    • (thread thread)
    • In general 44.1kHz will work fine. There are no significant reasons to choose either 44.1kHz or any higher sample rate. Some post-processing tools for vinyl recordings only operate at 44.1kHz.
    • Additionally, the dynamic range and distortion specifications for many sound cards are better at 44.1kHz than at higher sample rates.
    • This does not preclude using higher sample rates if there is a clear objective reason to do so. Probably the most important reason to use a higher sample rate is if you are recording a CD4 album, which has signal all the way to 45kHz.
    • Certain DSP operations (such as impulse noise reduction) may operate more effectively in the audible range if the sampling rate is increased, but this is not a general rule of thumb.
  • What bit depth to record with?
    • (thread)
    • Similar to the sample rate question above. In general, 16 bits actually does work well.
    • It's a common misconception that the quantization noise will significantly raise the noise floor of the recording; as long as the signal peak is -10dbFS or higher it is not an issue.
    • CDs have a lower noise floor than vinyl, period. Their noise performance exceeds vinyl at all frequencies.
    • If you plan to perform a very large amount of digital processing on the recording, the rounding errors could build up sufficiently that they impinge on the background noise level of the vinyl. In that case the recording should be made at (or converted to) 24 bit resolution prior to the processing stages to reduce the effect of rounding errors. That said, you would need to do a very great deal of processing before a 16 bit recording of vinyl became degraded in this way. A few passes of EQ, noise reduction, amplitude adjustment, etc. isn't going to bring the quantisation noise up above the vinyl noise floor.
  • How low of a frequency can LPs produce?
    • (thread)
    • The common wisdom is that they don't have much below 60hz, but there is evidence that the signal extends as low as 20hz and possibly lower on some LPs.
    • Note that low bass on vinyl records will be intentionally mastered in mono, but this is not a problem because low bass is non-directional. The reason for this is that low bass signals cause the largest groove excursions, and of course the vertical (out of phase) component of the groove must never be so great that the groove depth becomes so shallow at a maximum peak of modulation that it will not adequately retain a stylus, or so deep that during mastering, the cutter stylus passes through the lacquer into the aluminum substrate. The out of phase component of the low bass is largely removed when mastering for vinyl.
  • How much bandwidth is on an LP?
    • (thread)
    • (another thread)
    • This is a hard question, because the dynamic range has to be considered and it is frequency-dependent. It is almost certainly lower than CD, and certainly far lower if the theoretical bandwidth limits of the CD medium are considered, rather than the digital specifications.
    • Although some sort of signal above 20kHz can be recovered from an LP, it is highly unlikely to be related to the programme material; in other words it is noise. The majority of LPs probably contain nothing of significance above about 16kHz.
    • CD4 (quadrophonic LPs) had a subcarrier at 30kHz, and signals cut up to 45kHz and above.
    • CED disks had a usable signal extending to the Mhz range, but used a completely different playback technology that is completely incomparable to audio LPs.
  • What recording level should LPs be recorded at?
    • (thread)
    • As high as possible without clipping. -3db is usually safe and some sound cards may distort if driven past that. If you are recording on a sound card with a good noise floor (100db or greater) then you can go as low as -18db without ill effects.
    • Be sure to take into account pops, ticks, and records cut at higher volumes, such as singles.
  • How to remove sibilance?
  • How should mono LPs be recorded?
  • What are some examples of good and bad vinyl rips?
  • Can/should LPs be ripped faster than real time, ie at 45rpm or 78rpm?
    • (thread)
    • The short answer: No.
    • The long answer: you can use a reverse RIAA filter, speed up the signal and reencode in the equalizing filter of your choise (RIAA, AES, misc 78 equalizations). Or you can use a flat preamp. Both of these options are far more complicated and risky than just recording in real time.