Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › "Digital Vinyl"?
New Posts  All Forums:Forum Nav:

"Digital Vinyl"? - Page 7

post #91 of 152
Quote:
Originally Posted by Tarkovsky View Post
The issue is that you will easily and unknowingly clip the original recording in that the re-recording because RMS meters don't show transients. That's why RMS levels should be set to under a third of the bandwidth on your meters at 24 bit.
PPM meters?
post #92 of 152
Quote:
Originally Posted by GuyMe View Post
You say "most", is there any advantage to transferring vinyl above redbook standards?
I did some tests capturing direct to disks to a 24 bit ProTools workstation. By the time I bumped it down to redbook and burned it onto a CD, it sounded exactly the same as the capture through my normal setup. The only difference was filesize and the slowdown in handling in Peak.

LPs don't contain a wide enough dynamic range to benefit from high bitrates. Most LPs fall under 45 or 50 dB and the ones that do get up that high start losing detail to the noise floor long before the low level resolution advantage of 24 bit cuts in.

See ya
Steve
post #93 of 152
Quote:
Originally Posted by Tarkovsky View Post
Tarkovsky - Be careful of how you record because digital clipping is absolute and easy to do as RMS meters don't show transients which can easily be lost in the process
I skipped that one because I don't use analogue needle type meters. I use digital meters with clip monitors on them. I can adjust the level exactly. If it peaks into the red even for a tiny fraction of a second, it shows red on the clip meter. I just back it off a bit and start the side over again. If I'm not sure, I can go into the file and blow up the wave form and make sure it isn't flattening out. The problem you mention there just isn't an issue with a computer.

Quote:
Originally Posted by Tarkovsky View Post
Now if you take a wave, make it a vinyl, make it PCM, then make it an mp3 or the like you get all these SQ problems. And it'll sound like cack.
Assuming that all the steps you mention are done properly, the only step that will have any significant audible affect on the music is the MP3 encoding (and that's only if you do it at too low a rate). Inaudible alterations to the sound aren't particularly cumulative.

When I first got my CD burner many moons ago, I did a test and ripped and burned a CD to 20 generations. The 20th generation CD was identical to the first generation CD. I would imagine that capturing adds a bit of distortion to the signal. I haven't tried going through digitizing and playing back and digitizing again. (Mostly because it would be difficult to separate the conversion from analogue to digital from the conversion from digital to analogue.) But for one generation, the difference is completely theoretical. I've done an A/B and there is no audible difference.

See ya
Steve
post #94 of 152
Quote:
Originally Posted by bigshot View Post
I did some tests capturing direct to disks to a 24 bit ProTools workstation. By the time I bumped it down to redbook and burned it onto a CD, it sounded exactly the same as the capture through my normal setup. The only difference was filesize and the slowdown in handling in Peak.
LPs don't contain a wide enough dynamic range to benefit from high bitrates. Most LPs fall under 45 or 50 dB and the ones that do get up that high start losing detail to the noise floor long before the low level resolution advantage of 24 bit cuts in.
I know we've been over this before ad nauseam and disagree about some aspects of it but I think it's worth emphasisng in the light of the OP that you will get different recordings of the same lp from different turntable set-ups and that the better your turntable the better your recordings.

Really obvious things like the quality of your cartridge and how well isolated your deck is from acoustic feedback will have a massive influence. But how well engineered your motorboard is and the quality of your tonearm will also determine how well your stylus will track and how much detail it will pick out and how much background noise you will hear.

That is not to say you can't get a perfectly good recording from any decent quality deck which is Hi-Fi enough for most everyday uses, but the better your deck gets the better your recordings will be just in 16 bit in my experience.

As for capturing at higher bitrates or via DSD for instance, there is more information on the audio band outside of what CD can resolve and yes a spectograph of a record will generally reveal this but whether it's worth the extra resources of capturing and manipulating files this size for the arguably slight audio benefits gained is moot.

I personally would prefer to listen to the original records and concentrate on improving the playback of these first and formost.
post #95 of 152
Bigshot-
If I'm not sure, I can go into the file and blow up the wave form and make sure it isn't flattening out. The problem you mention there just isn't an issue with a computer.

Yeah that's a perfectly acceptable method. You could probably just back off your levels during the tracking as you won't loose much resolution and up the volume before mixdown, but each to their own! The issue is that a lot of old timers are used to hitting analogue gear with hot levels whilst digital requires a bit of a different approach.

When I first got my CD burner many moons ago, I did a test and ripped and burned a CD to 20 generations. The 20th generation CD was identical to the first generation CD. I would imagine that capturing adds a bit of distortion to the signal. I haven't tried going through digitizing and playing back and digitizing again. (Mostly because it would be difficult to separate the conversion from analogue to digital from the conversion from digital to analogue.) But for one generation, the difference is completely theoretical. I've done an A/B and there is no audible difference.

Well that's suprising because the CDs will have very different levels of jitter. Of course you may have a nice dac that handles jitter well...

But jitter isn't my qualm. You have outlined the audible limitations of vinyl as a medium and I have done my best to explain the audible limitations of red book. Running between these two we end up with the worst of both worlds.
post #96 of 152
Quote:
Originally Posted by memepool View Post
Really obvious things like the quality of your cartridge and how well isolated your deck is from acoustic feedback will have a massive influence.
When I make a vinyl transcription to my recorder at 24-96, I just monitor via the headphone out of the recorder and don't use speakers at all. The ambient noise levels in the room when I record are thus extremely low which I feel maximises the potential quality of the transcription. The other advantage with the headphones is that if there is some sort of tiny glitch that might go un-noticed when using speakers, it won't go un-noticed using the headphones.

Personally, I don't see any compelling reason for someone to use speakers when actually making the transcription of the vinyl itself to digital, since at this stage one should be doing nothing more than a straight capture. If they want to play around with the file after that, that is their choice and they can monitor it to their hearts content, ideally through a number of different sets of speakers.

If one has already got the equipment choice, setup, technique and procedure down pat, then I would personally suggest having all sources of noise in the room eradicated as much as possible during the actual transcription process. It is of course entirely different when one is at the experimental stage and trying to get the setup that achieves the best transcription in the first place - but once that is done, in my experience it does not need to be fiddled with after that.
post #97 of 152
Quote:
Originally Posted by memepool View Post
I think it's worth emphasisng in the light of the OP that you will get different recordings of the same lp from different turntable set-ups and that the better your turntable the better your recordings.
Absolutely. If you're going to go to the trouble of digitizing an LP, at least use a clean copy of the record and a decent turntable. Digital is a lot better at reproducing clean sound than it is at fixing messed up sound.

See ya
Steve
post #98 of 152
Quote:
Originally Posted by Tarkovsky View Post
Well that's suprising because the CDs will have very different levels of jitter. Of course you may have a nice dac that handles jitter well...
At the level it occurs in even the most humble CD player, jitter is 100 times below the level of audibility. The error rate of the capture card is MUCH more likely to introduce distortion by the accumulation of generations than jitter ever would be. Even so, a single generation of capture would be well within the range where no one would be able to do an A/B comparison and be able to tell the difference.

Quote:
Originally Posted by Tarkovsky View Post
You have outlined the audible limitations of vinyl as a medium and I have done my best to explain the audible limitations of red book. Running between these two we end up with the worst of both worlds.
Actually a digitized LP is *better* than a physical LP. It takes up less space. It's easy to duplicate and back up. You can play it without risk of damaging it through mishandling. It's possible to improve the sound through judicious noise reduction. It's editable and appendable. The only advantage an LP has over a CD is the size of the cover, but you can always digitize that too.

I'm a big fan of LPs. I have tens of thousands of records in my collection. I think records can sound terrific- with the potential to sound just as good as most CDs. I just don't think they are in any way superior to digital as a format... with one exception. I have a record of a band playing Auld Lang Syne that was recorded a century ago by Columbia. I can still play it and it sounds as good today as the day it was made. I'm not at all sure that there will be machines to play audio CDs in 100 years. I sure hope so though.

See ya
Steve
post #99 of 152
Quote:
Originally Posted by ADD View Post
When I make a vinyl transcription to my recorder at 24-96, I just monitor via the headphone out of the recorder and don't use speakers at all.
You're right. Transcription is an entirely different process than playing records for pleasure. It requires different techniques and different equipment configurations. It's a skill all its own. That's why engineers like Mark Obert-Thorne get credits on the CD jackets.

See ya
Steve
post #100 of 152
Less than 100th the level of audible jitter? Really? Not to accuse you of lieing but that's totally out of whack with everything I've read, that is of course unless I've miss read my sources. Where did you read/hear/extrapolate this?
post #101 of 152
Quote:
Originally Posted by Tarkovsky View Post
Less than 100th the level of audible jitter? Really? Not to accuse you of lieing but that's totally out of whack with everything I've read, that is of course unless I've miss read my sources. Where did you read/hear/extrapolate this?
Hunt down these papers. The results depend on whether you are talking about random jitter (much like noise) which has the effect of degrading resolution or signal-correlated jitter which appears as specific distortion products. The overall result is that for a pure tone a signal-correlated jitter signal of 10ns is just audible, when music signals are used the threshold becomes about 20ns. For random jitter the detection thresholds are no less than 250ns. Put this in context my old (1998) DAC has jitter of 157ps, an Oppo 970 has jitter of 3ns and a typical mid-range Marantz CD player runs to about 680ps.

Simulation of sound degradation due to time jitter on digital audio",
Ashihara et al., J. of the Acoustical Society of Japan, 58, 232-238 (2002))

Acoustical Science and Technology
Vol. 26 (2005) , No. 1 pp.50-54
Detection threshold for distortions due to jitter on digital audio
Kaoru Ashihara, Shogo Kiryu1, Nobuo Koizumi, Akira Nishimura, Juro Ohga, Masaki Sawaguchi and Shokichiro Yoshikawa5

Acoustical Science and Technology
Vol 24 (2004) , no 4
The maximum permissible size and detection threshold of time jitter on digital audio
Ashihara Kaoru Kiryu Shogo Sato Sojun

Benjamin and Gannon (1998), AES preprint 4826

Nishimura et al. made attempts to measure actual jitter of various DA systems during reproduction of music signals. They could not detect any jitter larger than 3 ns in their measurements (“Measurement of sampling jitter using a musical signal”, Nishimura et al., AES paper 5797)
post #102 of 152
Quote:
Originally Posted by bigshot View Post
Actually a digitized LP is *better* than a physical LP.
Something we agree on! Though I should add I only feel this is mainly the case on sonic grounds. Even when I listen to the digital transcriptions, I still have the record cover in front of me and enjoy holding it and reading it.

I know that comment prima facie goes against my love of vinyl, but quite honestly when it comes to listen via headphones, it's a whole new ballpark out there. Equalisation for staters becomes a big issue not only on account of the forensic capabilities of even a $300 pair of headphones, but also because everyone's ear canal (and pinia for full sized phones) makes a difference to the way we perceive the sound balance, as the sound is amplified and bounced around between headphone diaphragm and ear drum.

If I had the luxury of having a great amp, speakers, listening room and double-brick house on a big block of land, I would probably just plop the vinyl on the platter and just do it that way. Even so, I see so many advantages of a top quality transcription. Not only the wear / damage factor that has been mentioned but also the ability as you say to digitally remove significant flaws (which are even adundant in audiophile quality new vinyl) with pretty much no audible impact except for a positive one (I prefer to do this manually, since usually it's not surface noise that is the issue - it's just the ubiquitous surface flaws).

Then there is the huge variability of the master tapes, which gets back to my EQing. The balance and soundstage of a 1957 European Decca is nothing like a DG from the same era, a Mercury, RCA or anything else. Or even a 1961 Decca. They are all so different. I am transcribing a 1957 Decca today (Curzon Emperor Concerto) and quite honestly it needs some judicious manipulation to get the sound right - those old tapes were sometimes a bit brittle and tinny sounding. But a Kenneth Wilkinson effort from the same era sounds much better balanced and eerily modern. Some recordings are starkly bright, some are relatively dark, some sound "hard" because of an early roll-off in the high frequencies and subsequent lack of "air". Some have a soundstage that seems like you are the conductor, some sound like you are halfway down the hall. These are all things that can be subtly manipulated in the digital domain so as to obtain a bit more sonic consistency from one's collection, but there is nothing you can do in the analogue domain. But if I put something on like Billy the Kid on Mercury from 4 years later, I can leave absolutely everything completely alone - it sounds right with virtually no interference on my part.

But I buy the vinyl because I still honestly feel these new audiophile LPs are done a lot better than the CDs which were done in high volume by the big record companies in the 90s. I now have duplicates of many CDs and LPs and in all cases I much prefer the LPs, even before making any adjustments in the digital domain for usage on headphones.

I have far more to choose from on vinyl too. The big companies were haphazard at best - they'd start a series of releases and stop a tenth the way in. The original Mercuries on CD from the 90s were a concerted production efort, but quite honestly soundwise they are miles and miles behind the Speaker's corner LP reissues. The Mercury SACD series didn't get past about 20 titles. Decca never got enthusiastic about their archives either. Philips did some great CD reissues but their classical catalogue wasn't really that huge back in those days. And the Living Stereo CDs from the early 90s were better done to my ears than the later SACD remasterings which seem to have lost the soul of these great recordings. But now Bernie Grundman has done the job on vinyl that these Living Stereos deserved right from the beginning. And just in time, since I wonder how much longer these 50 plus year old tapes that have been played over and over again will remain useful.

Anyway, I could go out and buy over 100 brand new LPs tomorrow of exactly what I want if I had the money. And that would just catch me up to the latest releases - not anything released from today onwards.

I just feel lucky that I now have a great turntable / cartridge / phono amp combination, as in my bitter experience cheaper turntables do more to put people off vinyl than they do to encourage it.
post #103 of 152
To OP:

Check also that the system you're going to get has a subsonic/rumble filter on-board (or get it separate). A steep (18-36dB/octave) subsonic filter is needed because of bad frequencies around 0.55Hz (33.1/3 rpm).

Quote:
Frequencies below 20Hz are usually not able to be reproduced, and with the exception of synthesisers and pipe organs, are not a wanted part of the audio spectrum. This is especially troublesome with phono systems, since many of the vinyl discs you treasure (or wish to transcribe to CD) will be warped to some degree. Any warp in a vinyl disc will cause large outputs in the subsonic region, typically well below 20Hz.

For example, a 33 1/3 RPM album with a single warped section will create a signal in the pickup at 0.55 Hz (33.3 RPM / 60 = 0.555 Hz). This is a signal that will cause significant cone movement, but is undesirable in the extreme. Not only will vented subs be completely unable to handle such a signal linearly, but sealed subs will also be stressed. Large amounts of available power will be wasted trying to reproduce a signal that was never intended to be there in the first place.
Source: Subsonic / Rumble Filter for Phono preamps and Sub-Woofers

If your target resolution is CD Audio then just use 16-bit/44.1kHz to avoid additional SRC/Dither needed to get to CD audio format.




Quote:
Originally Posted by IPodPJ View Post
Yep. If I'm not mistaken, a sound wave at 20Hz is about 120 feet (as opposed to a 10kHz sound wave which is about 2 or 3 feet). And to hear a non-reflected 20Hz sound wave you need a listening room that has 120 feet between the drivers and your ears.
So, you need to have a huge head to hear this low frequencies w/ headphones (IIRC, most headphones are rated to start around 5-10Hz...).


Quote:
Originally Posted by nick_charles View Post
I still have about 100 LPs, which I never play, but if I were to transcribe them I would not bother going above 16/44.1.

I am skeptical about the value of doing high res transfers on a medium that is so technically compromised (no offense to LP, but it is what it is and no 5000 quid TT will change that)

First 16/44.1 already has superior noise, dynamic range and distortion levels than LP so it isn't a limiting factor in terms of a transfer.

Second , so far as I know , nobody in anything even remotely resembling a carefully proctored test has been able to tell the difference between high res recordings and downsampled high res recordings. When you consider that so many people cannot tell Mp3 from wav it is hardly surprising.

Third 16/44.1 A/D/A chains have been shown to be undetectable in playback as far back as 1988 when the Boston Audio Society showed Ivor (I will never make a CD player) Tiefenbrun that he could not detect the presence of a Sony PCM-F1 (nominally 16 bit) inserted into the chain.

Fourth CD-R recordings of LP vs LP tests have been done - no difference found.

Fifth the sheer space requirements of uncompressed 24/96 or 24/192 for dubious payoff. 3.26x the space for 24/96 or 6.53x for 24/192

It just doesnt seem worth the effort...
There are noticeable differences between 16-bit/24-bit/32-bit and 44.1, 48, 88.2 and 96kHz 'resolutions'. Last nite I recorded using only 24-bit resolution @ 44.1, 48, 88.2 and 96kHz and the clearly best quality came from 96kHz even I encoded the data to WMA 24-bit/<samplerate>/440kbps ... file sizes are equal (~18MB each) so, the data "loss", because of compression, is over twice for 96kHz data compared to 44.1kHz ... and still the 24/96 sounds clearer/more detailed than the others.

In this case, there aren't much data above the 22.05 kHz.

http://jiiteepee.fortunecity.com/tes...-riaa-emu.html


My testing equipment was:
Technics SLQ2 (DD)
Technics 205CMK3 cardridge+Tonar stylus
E-MU 0404 USB (Hi-Z inputs)
self programmed software RIAA stage (3rd-5th order filters ( < .000005dB error in reproduction curve)
No subsonic filter active
Hermann Seib's VSTHost software (I/O, recording)
Voxengo SPAN (levels)

Turntable -> E-MU -> PC -> VSTHost -> RIAA-> SPAN -> disk.


jiitee
post #104 of 152
Great post, ADD.

See ya
Steve
post #105 of 152
Quote:
Originally Posted by jiiteepee View Post
There are noticeable differences between 16-bit/24-bit/32-bit and 44.1, 48, 88.2 and 96kHz 'resolutions'. Last nite I recorded using only 24-bit resolution @ 44.1, 48, 88.2 and 96kHz and the clearly best quality came from 96kHz even I encoded the data to WMA 24-bit/<samplerate>/440kbps ... file sizes are equal (~18MB each) so, the data "loss", because of compression, is over twice for 96kHz data compared to 44.1kHz ... and still the 24/96 sounds clearer/more detailed than the others.
What your test also shows is that 96khz has over twice as much perceptual redundancy as 44.1. That is because perceptual encoding works by removing data that is perceptually invisible, it doesnt care what the bit-rate or bit-depth are. Your test shows that 96khz has twice as much perceptually unnecesary data.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Dedicated Source Components
Head-Fi.org › Forums › Equipment Forums › Dedicated Source Components › "Digital Vinyl"?