More Ripping Vinyl Madness

I know this topic has been covered a number of times, and I've read quite a bit on the topic. But I'm not sure my methodology is optimal or if I'm just splitting hairs.

In the past I would rip rare albums that weren't available in any other format. I used my older Apple Power Book, Sansui SR-838, iMic and Audacity. Seemed to work easy enough and the filters would easily remove noise. The sound quality was good enough for mp3s.

The last couple days I've gotten serious about the whole setup/procedure. I'm now using a Apple MacBrook Pro, Pioneer PLC-590, Cary Audio tube pre-amp, Pro-Ject Phono Box II and the iMic. For software I started using Soundtrack Pro, which is an older version which doesn't have the click and pop filter so I bring the file into Adobe Soundbooth CS5 to remove the pops. But the noise reduction in Soundtrack Pro is fantastic. It lets you preview live, the music and noise separately while you make fine adjustments.

I experimented at capturing the music at various setting and what seems to be the standard. I tossed the files so these are approximations.
96/32 = I forget kbps, (AIFF-C, 200 MB)
96/24 = 6000 kbps, (AIFF-C,150 MB)
96/16 = 4000 kbps (AIFF-C,100 MB)
44/16 = 1536 kbps, CD Quality, (AIFF-C, 53 MB) Down sampled from 96/16.

Soundtrack Pro can capture at 196/32 which seems way overkill, so I didn't try it.

My ears can't tell the difference between these versions while using the Grado RS-2i headphones. These headphones are very revealing. I spent an excruciating time trying to reset-up the TT, because I noticed a little sibilance in the vocals. The song I used was Phoebe Snow's Poetry Man because I didn't have a digital copy and because the sound is clean and it covers a broad spectrum.

I think the 44/16 will work just fine for me, but am I missing anything? I'm asking because it takes me nearly 2 hours an album. Noise reduction is relatively quick, but taking out individual pops/clicks is time consuming.

Btw, the results are excellent. I'm going to have to A/B test ripped vinyl, 200 gm vinyl and CD.

Thanks!
Carl

you should get the software called vinylstudio and it will take a lot less time from you to record and edit. some people cant tell the difference between 16/44 24/96 etc. it may be hearing and it may just be what and how they listen to music. I myself can tell a big difference in 16/44 and 24/96 I had a hearing test last year and my results indicated I have exceptional hearing. not saying that has anything to do with it but it may. if you can't tell the difference and 16/44 works for you then you get to save a lot more space on your HD. haha

Elitemav,

I'll try the demo version of VinylStudio and see if it can streamline my procedure. My hearing probably isn't the greatest so 44/16 may be the best for me. I also don't want to have a bunch of redundant files at different sizes eating up HD space and I may want to fit them onto CDs.

Thanks!

IMO, ripping is a waste of time. It's probably wise to archive the music in case of loss or damage, but half the record playing experience is playing records. There are some recordings only available on vinyl which should be digitized, but life is too short to simple rip music that's available on CDs.

onepixel said:

I think the 44/16 will work just fine for me, but am I missing anything? I'm asking because it takes me nearly 2 hours an album. Noise reduction is relatively quick, but taking out individual pops/clicks is time consuming.

Click to expand...

I've transferred thousands of records to WAV, for paying customers and personal use. The end product for a customer is a CD-R, so out of necessity I'm creating 16/44.1 files. I also use 16/44.1 WAV files for the weekly radio shows I produce.

I have done transfers at higher sampling rates and bit depth. I've compared them to my standard 16/44.1 files. I cannot hear a difference. That's not to say someone else couldn't, just my experience.

IMO, you are not "missing anything."

OTOH, it is easy to hear the difference between a low resolution MP3 and CD quality WAV file.

illini said:

IMO, ripping is a waste of time. It's probably wise to archive the music in case of loss or damage, but half the record playing experience is playing records. There are some recordings only available on vinyl which should be digitized, but life is too short to simple rip music that's available on CDs.

Click to expand...

Not to argue with your personal opinion and preferred playback methods, but...

There's a lot of folks with R2R tapes, 8-Tracks, Cassettes, LPs, 45s, and 78s with no way to play and enjoy them. In many cases, the music was never reissued on CD. I've built a respectable home-based business doing audio transfers for folks like this.

Even if the music has been reissued on CD, there's no guarantee the CD will sound good. Master tapes can degrade over time and the remastering process for CD may have been sloppy and careless. I've had several customers compare the vinyl-to-CD-R I've made to the same release on commercial CD, reporting that my transfer was noticeably better.

In some cases, a commercially available reissue CD is dramatically different than the original LP release. I've seen/heard poor quality demo or live versions used, alternate takes, etc.... In other cases, you might only have a "greatest hits" CD compilation of a particular artist available.

There are big advantages in digitizing records. A 'mix' CD of favorite tunes can be very easily created. MP3 files can be loaded on an IPod for on-the-go listening. I produce three weekly radio shows that air every Friday, Saturday, and Sunday. I assemble the shows in advance in my home studio, recording the voice-overs and using WAV transfers of my music. There's no way I could do the same three shows live-in-studio every weekend, given my busy schedule and family life.

There's also audio restoration to consider. I can take a badly worn or damaged record, do a transfer to WAV and with careful use of pop/click removal, hi and low bandpass filters etc...., dramatically improve the sound.

Like yourself, I love records and the ritual of playback, reading cover notes & lyrics, etc.... But I've also embraced the many advantages of digital.

A couple of notes...

1) 24 bit gains you around 12db of additional headroom
This is really really valuable at the initial recording stage.
Once the recording has been cleaned up (clicks removed...) you can adjust the overall volume upwards to best use the available dynamic range - and then 16bit is plenty.
So I always record at the highest bitrate possible, even if my final result will be 16bit.

2) Process the files in 32 or 64bit - the mathematics involved in the processing have rounding errors, the higher the word size used, the lower the levels of the (cumulative) distortion created by the various processing stages.
At the final step dropping back to 24 or 16 bit truncates the errors leaving them behind.

3) No point recording at 32 bit or higher - none of the chipsets are actually capable of it - so although you will be processing at these bit levels, record at what the hardware is capable of

4) My personal measurements of a number of 192/24 capable ADC's showever that although they were capable of the higher bit rates, their optimal performance was at 96/24.... so no only am I wasting space recording at a higher rate than 24/96, but the result is actually a marginal step down in quality. (so I record at 24/96)

5) Brickwall filters kick in at the uppermost end of the real bandwidth of the recording rate (22kHz for 44, 24kHz for 48, 48kHz for 96 etc...) - along with these filters are a range of distortions related to the filters.
Setups running 88 or 96 move these distortions outside of the Audio range, and allow for the use of gentler filters with less distortion.
As part of the mastering process, you can then filter out the frequencies above 22kHz, eliminating some of the unwanted artifacts of conversion to 16/44.
I think part of the reason 24/96 sounds better than 16/44 is this filtering issue... the distortions from filtering are quite audible, raising the cut off frequency takes them out of the picture - and careful mastering keeps things clean when dropping down to 16/44 if desired.

6) If planning on dropping back to 44, the optimal recording frequency might be 88 rather than 96 - being a simple multiple, it simplifies the associated mathematics and reduces the amount of errors involved.
If the software being used is good, this makes no difference whatsoever (or very marginal), but there are plenty of software solutions that are less than perfect, where this does indeed make a difference.
The main "gotcha" with 88, is that many chipsets (hardware) are simply not well optimised for that rate... their sweetspot is 96 (as mentioned above).

Conclusions:
I record at 24/96 because that is the resolution at which the hardware I have gives the best results.
I process at 32/96 or 64/96 (depending on software) - to minimise errors.
I then either keep the resulting mastered file at 24/96, or drop it back to 16/44 if needed.

My setup is PC based so not much point talking about specifics in terms of software!

bye for now

David

cactuscowboy said:

.

Click to expand...

dlaloum said:

.

Click to expand...

Informative posts. Thanks. I need to rip some of my LPs, so instead of regarding digitizing as work I'll try to make it a hobby.

i've done a comparison recording. 24bit is considerably better than 16it, but 48khz vs 96khz i hear very little difference. slightly more treble info in 96k but too subtle to be significant. thus i record @ 24/48.

48k recording: http://www.mediafire.com/download.php?h9xusdp81dsb5uh
96k recording: http://www.mediafire.com/download.php?amqco7j7riabyy4

Thanks guys! I wasn't to far off.

I have Realtek High Definition Audio Player, common in PC's, and JRiver upsamples to 24/192 if you choose. My player does it as well, and before I knew this I thought I was hearing a big SQ difference but didn't know why! Once I figured out that this was going on, I made note of how to get back into the player to see if I was in fact upsampling, as on occasions it would inadvertedly switch back. I have the same setting on JRiver AND my player, redundant, yes, but it guarantees that if by some mistake my player reverts to 16/44.1 I'm still getting 24/192 (studio-quality sound). KILLER.

24 bit has no additional "headroom"! In any linear, PCM (pulse code modulation) scheme, 0dbfs (0 decibels, full scale) means all bits are exercised. Every one of the 16, or 24 bits is a 1 (as opposed to a 0).

With linear PCM, each bit represents six DB of dynamic range, but this additional dynamic range is at the BOTTOM of the volume scale, not the top. 16 bit digital audio has a dynamic range (range between the quietest sound that can be captured and the loudest) of 96db (16 x 6 = 96). 24 bit, on the other hand, has a theoretical dynamic range of (gulp) 144db! (24 x 6 =144). I say "theoretical" because there isn't a room quiet enough to take advantage of all that additional range at the bottom of the scale...not even an anechoic chamber. And recordings are made in rooms, with open microphones, each connected to a preamp with far less dynamic range than that. Real rooms, with real people (musicians) in them, are seldom quieter than -60 or -70db...and that would be an exceptionally quiet room.

The thing about the "enormous" differences people claim to hear between 16 bit 44.1khz and 24 bit 96khz recordings is that, under double-blind conditions, these differences disappear.

The typical audiophile "test" is to play two versions of a recording, announcing which is being played, then ask: "which sounds better"? What's wrong with that? A)-a person's biases are on full display...all things being equal, they'll always choose the one they BELIEVE is superior. And b)-it's asking the wrong question!

With double-blind testing, using what's called an a/b/x comparator, switch position A would be, say 16 bit 44.1khz, position B would be 24 bit 96khz, and position X would be, randomly, either A or B. NO ONE IS EVER ASKED "which sounds better?" Instead, they're asked, is "X" the same as "A" or "B". Because if the person being tested can't answer that question to a scientific certainty, it makes no difference which one they think is "better", they've proven they heard no real difference.

There's a reason 16 bit 44.1khz was chosen as a standard more than three decades ago. It's a very good match to the limitations of human hearing. If you can barely hear something at -96db, then a sound at 0dbfs would be so loud as to be beyond the threshold of physical pain! That was true then, and it's still true today. There's a very good argument that, for archival purposes, recordings should be preserved at the highest resolution possible. Since digital storage is cheap these days, there's certainly no harm in storing at a higher bit depth and sample rate. But for the end user, there's also no provable, scientific benefit to so called "high res" formats.

Now a dozen "golden ears" will argue that they can OF COURSE tell the difference. Ask them if they've ever participated in a double-blind test. If they say no, or can't supply the data from the test...and especially if they attempt to discredit scientifically quantifiable test results, shrug and walk away!

24bit has substantial additional headroom... - 12db worth of it!

It allows one to record 12db lower, allowing 12db for unexpected pops or clocks that can later be editted out...

Final mastering can then reduce the dynamic range to 16bit with no loss of resolution....

To really get the best out of vinyl and provide for the true peaks (a pop climbing up to +20db or more is not unusual, and generates quite a bit of distortion if it clips, which cannot then be edited out!).... 32bits would be really really handy (24db of additional headroom)

Unfortunately 32 bits is beyond the ability of current hardware, which in actual fact only achieves approximately 20bit resolution (yep we ain't achieving 24bit dynamic range yet!)

A well made record can easily have over 60db dynamic range - in addition to the ability of an analogue system to have audible detail extending below the noise floor (probably for at least a further 10db maybe more).
In addition to which the nominal 0db level is an RMS measure at 5cm/s, and peaks go well above that - in fact the bias test torture tracks are recorded at +18db - and produce a clean signal with only the best cartridges.

Recorded dynamic ranges (before taking into account pops/clicks) can therefore extend 70db below 0db, 18db above 0db (for a total of 88db) - and then you need margin for error - gain control, to allow for the unexpected flaws in the vinyl, and for the fact that you cannot perfectly gauge levels before making the recording. (for perfectly set levels record first to determine peak levels, then record again at the ideally set levels....)

So you provide 12db margin for recording overhead (fudge factor), and a further 6db over the max expected dynamic range - another 18db - add this to the existing 88db and we are at 106db - oh woops most current ADC's can only manage around 104db (for the better ones!)

Of course, I am a perfectionist... if less than perfection is the target, all this need not apply.... but then to some standards an edison cylinder is sufficient fidelity.

bye for now

David

"...allows you to record 12db lower", well duh! But why choose 12db? The additional dynamic range of 24 bit is FORTY-EIGHT db, not 12! (6db per bit...144 db for 24 bit minus 96db for 16 bit).

I would submit that if you choose to record 12db lower, YOU are allowing YOURSELF the "12db of additional headroom", rather than the format offering more. But there are distinct limits to this train of thought, my friend. While the medium you've chosen has additional dynamic range, microphones, preamplifiers, audio consoles, etc...all necessary for making recordings of live music, DO NOT! In lowering the level, you'll hit the noise floor of these analog stages LONG before you hit the noise floor of either 16 bit or 24 bit media. Increasing bit dept to 24 bits doesn't do a damn thing to make these other, necessary components in recording a single db quieter.

And when we consider recording from LP, which struggles to reach a dynamic range of 50-60db, the argument becomes almost laughable! After all, 8 bit recordings have a dynamic range of 48db! (And very coarse, low-fi sound...but I'm speaking specifically of dynamic range. From that standpoint, 8 bit is almost "good enough" for vinyl recording).

So...while lowering your record level by 12db will indeed allow you to GIVE YOURSELF "12db of extra headroom", the result will inevitably be that your recording is 12db more noisy...because increasing bit depth didn't do a thing about the BOTTOM OF THE SCALE, where the noise of your phono preamp certainly hasn't gone down by a single db!

Oh, and as for recording audio below the noise floor, digital can do that too. Dither, random noise which keeps the least significant bit from ever toggling off, allows audio to be recorded BELOW the 16th bit in a 16 bit system. This was demonstrated many years ago.

As for pops and ticks on records generating distortion that can't be removed if clipped, that's just plain silly. These single impulses have no harmonic content. "Clipping" simply squares the top of the waveform. One can zoom in, find that individual click or pop, and cut it, or as I usually do, simply mute that tiny space in time. While a very short duration click or pop can be easily heard, an equally short silence usually can't. If it can, I grab a few samples of audio from either slightly before, or slightly after the silence, and paste that into the silence. This completely covers the "hole", and if done properly, has no audible consequences.

I do quite a lot of this on my syndicated radio show "Saving the 70s", heard on 70 stations in 9 countries. I use vinyl quite often on my show. And for radio, recording at -12db ain't an option! If I did that, stations would laugh at me! My recordings must be competitively loud when compared to all other sources of audio! Sadly this "competitively loud" business can be taken to extremes. Today's recordings are TOO DAMN LOUD...they squeeze all the life out of the music. But that's a source for another rant!

You can hear some of my work here, but it's at a greatly reduced quality level from the actual files that stations carry.

http://www.savingthe70s.com

Here's a demonstration of how I just did what you claim can't be done: I removed two very loud pops, one of them all the way to 0dbfs, on the end of this song.

The file includes two fade-outs from "Red Red Wine" by UB40. The first is a straight transfer from my turntable. On the second, I have removed both of the loud pops...one just before the word "don't", and the second and LOUDEST one in the middle of the word "be" (don't let me be...)

Both loud pops are gone. And I'll bet you don't hear the very tiny silences where they used to be! I've been restoring vintage audio for a long, long time!

http://www.theproductionroom.net/realaudio/popremoval.mp3

Onepixel,

Just curious as to why use the iMic instead of the powerbook's sound card? I've found the input on my iMac's sound card to be very good for the little bit of vinyl ripping I've done. I haven't tried my daughter's Powerbook though. Does it come up short?

illini said:

IMO, ripping is a waste of time. It's probably wise to archive the music in case of loss or damage, but half the record playing experience is playing records. There are some recordings only available on vinyl which should be digitized, but life is too short to simple rip music that's available on CDs.

Click to expand...

Can't play LP's in the car (anyway I wouldn't. even if Elvis would) and I like some of my tunes to carry around while doing yard work, or out in the boat. With 2K plus of vinyl, I'm not about to buy single tracks or all those CDs. If you rip as you play, it ain't terrible as far as time

RadioGiant said:

I do quite a lot of this on my syndicated radio show "Saving the 70s", heard on 70 stations in 9 countries. I use vinyl quite often on my show. And for radio, recording at -12db ain't an option! If I did that, stations would laugh at me! My recordings must be competitively loud when compared to all other sources of audio! Sadly this "competitively loud" business can be taken to extremes. Today's recordings are TOO DAMN LOUD...they squeeze all the life out of the music. But that's a source for another rant!
http://www.savingthe70s.com

Click to expand...

Absolutely right, RadioGiant. FM stations today crank up the gain so d**n LOUD these days that all dynamic range is lost, making them unlistenable! Somehow they believe that if they sound "louder" than the competition, more people will listen to them. Never understood that philosophy, and maybe it worked back in the 70's. Satellite radio is only slightly better, but I don't know what sampling rate they use. It is touted as "near CD quality", but it sounds nowhere near 16/44. Must again be loss of dynamic range through high modulation levels. It has come down to the point that now I take my fantastic sounding little Cowan J3 portable DAP, and rip my LP's or download albums in FLAC or minimum 320kbps (I don't notice a big difference in MOST cases) and run it through my aux. input in my car. THAT is on-the-go music that is capable of putting a smile on my face!
I do the same thing at home, plugging the Cowan J3 into my aux inputs on my Little Dot MKIII, then output that into my Rotel gear. Glorious sound on- the- cheap, next best thing to my analog LP source!

It's loud so you can hear it while driving your 4x4 Chevy V8 truck with flow-masters and "mudders" (tires) down the highway with the windows open while your GF pops bubble gum and plays with her iPhone. Whadda ya think FM is for