Sample workflow for LP digitization | Hacker News Reader

Sample workflow for LP digitization

manual.audacityteam.org

100 pointsdirwiz52 comments

As a club DJ who selects and plays in a digital format now, I like to leave clicks and pops in when digitizing my "white-labels" or vinyl-exclusive releases (some of the best or most coveted tunes are released on vinyl first in very small runs, months before they are available as downloads). I like to normalize then run the recording through Izotope's Ozone 7 on the "warm and transparent" preset to get it closer in loudness to digital releases I am mixing them with. There is a small community of us who share rips of this type of release from around the world, online, as the best place to pick a lot of this kind of record up is in smaller local genre-specific shops. Black Market soho in London was a Mecca of mine or Rooted Records in Bristol but I live in Cali now and have a hard time keeping up my share ratio but still get a rush when I get a great vinyl rip of a tune earlier than most. It is silly and matters very little to my livelihood but an old tradition in dj'ing that I am glad still survives (having tracks the next Dj doesn't, that is).

Why is this done? Tradition? Certainly it's an extra expense that releasing in pure digital format would not include -- and it's not like all the music is not mastered digitally now anyway -- nobody is mastering on tape and cutting and splicing....

White labels and dubplates in general hearken back to the sound system culture that originated in Jamaica [1]. Mostly a means of getting the newest songs fastest; record something in the morning, press your dubplate that afternoon, play it out that night. The unofficial nature of the recordings makes it easier to hop around copyright stuff and not clear samples, etc. The song Alicia by Mala jumps to mind [2], I don't think it's ever seen a proper release because of the heavy use of Alicia Keys samples.

1. https://en.wikipedia.org/wiki/Sound_system_%28Jamaican%29 2. https://www.youtube.com/watch?v=zpV7radKuwo

Again, I don't see how pressing a physical recording is making anything faster. It's a step that's entirely unnecessary if you distribute digitally. It must be more of a tradition, or a desire to mix on turntables, or possibly a way to do a limited/controlled release that is not easy to redistribute widely as a digital media file would be?

He did say this, but maybe it wasn't clear. It's faster because you can skip the meaningless bureaucracy associated with obtaining copyright permission for a tiny run - a lot of dance music samples other music. If you try that with a digital release, it can bite you if it blows up.

No, he means like if you write the flac to a flash drive. Just write it to a digital medium, rather than go to the trouble of pressing discs.

Combo of all of those, really. Particularly the exclusivity bit, songs that get pressed like that generally aren't things people are trying to share. For instance, dubplates use acetate and not vinyl, and are really only good for about 50 plays. Lots of exclusives becomes the reason to check out your party vs the one the people a few blocks down are throwing.

>Why is this done? Tradition? Certainly it's an extra expense that releasing in pure digital format would not include -- and it's not like all the music is not mastered digitally now anyway -- nobody is mastering on tape and cutting and splicing....

Actually lots of artists master on tape.

What most don't do is track and mix on tape -- but there are some that do that too.

(and mastering doesn't involve "cutting" and "splicing" much -- that would be mixing).

Dance music is typically released in a stepped fashion - a handful of DJs will receive exclusive dubs, followed by a small release on white label, followed by a general release.

This is a mutually beneficial arrangement for DJs and producers. DJs can increase their audience by playing music that can't be heard elsewhere; This exclusivity provides promotion for a track before general release.

Radio DJs have a habit of speaking over the middle of an exclusive track, to prevent it from being ripped and used by other DJs. Exclusive dubs often mention the DJ by name in the lyrics, a tradition originating in the Jamaican sound system culture.

Vinyl releases are also much more profitable, which can be very significant for niche music with a relatively small market. A typical track download nets the record label 50-70p, but a 12" white label can net £3 to £4. If you only expect to sell a few thousand copies in total, a limited vinyl run can make a big difference to the bottom line.

Thanks for the tips - I rip vinyl for the same purpose.

Your use case makes sense to me. Prior to reading your post, I was always of the opinion that ripping vinyl kills the warmth that vinyl offers in the first place, so why do it?

I didn't see it mentioned, but I wonder if Audacity inverts the RIAA equalization curve.

https://en.wikipedia.org/wiki/RIAA_equalization

You are right, there is no mention of the turntable/stylus/preamplifier configuration at all.

I would guess that the Audacity people assume that people would be using a turntable plugged into a preamplifier with RIAA equilisation built in.

http://midimagic.sgc-hosting.com/mixphono.htm

History before the 1950s is complex. Audacity has provision for defining a frequency response curve and applying that curve to a project.

Really surprised they left out those details. Needles have a huge impact on the frequency curve, a lot of standard club needles are bass heavy and/or sacrifice fidelity for tracking.

And on the digital side the ADC makes a noticeable difference. Plus is it isolated from the other circuitry, particularly power supplies?

Ripping vinyl takes enough work, it's worth knowing the few details that make the effort worthwhile.

As bondaburrah points out in the sibling comment, there are probably no consumer sound cards in existence that would be able to record a non-preamplified vinyl recording. So the RIAA equalisation issue is moot.

Yes, there are. Any sound card that supports XLR connector mics will probably support phantom power and have a built-in mic amp, a vinyl pre-amp is not much different. I use an E-MU 1212m with a microdock, making it equivalent to a 1616m, which offers a "Turntable Input (w/ground lug and hardware RIAA preamp)".

http://www.creative.com/emu/products/product.aspx?pid=19007

Interesting. I had no idea how cheap what I would consider to be pro-grade sound cards are. I'd have thought an XLR input soundcard with phantom power support was close to $1000.

Well, any card with mic pres can do it, including a lot of older SoundBlaster type cards. I wouldn't recommend mucking around with doing things that way, but it could be done.

Most people rip vinyl using a hardware phono stage (which is a preamp + RIAA eq), but there are some audiophiles who prefer to rip using a straight mic pre, which bumps the levels up without applying any eq, and then applying suitable eq in software.

In theory, this would let you play around with different variants of the eq curve, and it would mean not needing a phono stage but only a (cheaper) mic preamp. In practice, most people who rip vinyl are going to have a phono stage anyway, because they want to listen in addition to rip, and there's just not that much benefit of doing it in software vs. using a cheap-but-decent analog phono stage, so few people do.

The exception seems to be people who are into early 78s, which often have proprietary eq curves.

Back when I used audacity, I recall there being an RIAA preset in the EQ plugin designed for reversing the EQ on records. I never wound up using it though because LPs still require significant preamp before your average sound card can get a good recording, and any phono preamp I had of course had the RIAA circuitry in it already.

It should still be present in the EQ presets list, if the screenshot is reasonably up-to-date here: http://manual.audacityteam.org/o/man/equalization.html#manag...

One use the guide mentions is using it in inverted mode, if you want to undo the effects of incorrectly applied RIAA equalization (eg. when a 78 rpm record is being digitized with a different pre-emphasis method).

While I have no need of ripping vinyl, just wanted to say that Audacity is one of the best examples of indispensable FOSS (when it comes to working with sound) that I have ever come across in my life.

I remember I was using it well over 7 years ago, and was in awe of how much power it gave me for free, and how capable of a tool it was. Couldn't donate then, but definitely can now. While I know my time/PRs might be more valuable, regular money will have to do.

Has the conversion to 16bit been fixed? Dithering still wasn't working properly the last time I checked.

Word! I just like to add that it's especially indispensable for many community radios across the world. Maybe there are more powerful tools out there (FOSS or not), but I never saw an audio tool with such capabilities which people got used to so quickly like they do with Audacity. The absolutely perfect fit for those community radios where almost everything has to be self-taught.

Of course, the real way to do this is to optically scan the record and simulate a stylus tracking it. The Library of Congress does this.[1] They've even recovered records broken into pieces, by scanning the pieces and reassembling the 3D model.

[1] http://www.theatlantic.com/technology/archive/2014/06/how-a-...

>the real way to do this

What's "real" about it? The overwhelming majority of people doing it who are not Libraries do it with a needle.

There doesn't seem to be any free or commercial software that converts high res scans of vinyl records to high quality digital audio. I remember this post from 2002 http://www.cs.huji.ac.il/~springer/DigitalNeedle/index.html where a guy posts his results with this approach, but the quality is bad and he doesn't provide source code.

A less technical, step-by-step way (also using Audacity) http://transfermymusic.com/vinyl-to-cd-or-mp3/

When I rip vinyl I want as little processing as possible, so no messing around with high pass and pop removal. If it's new or decent condition vinyl the pops and hiss will be barely audible anyway.

Far more important to have a decent MC cartridge, and make sure it goes through a decent phono (RIAA) stage, which generally means an older hifi amp, or not-cheap hifi component.

Any phono stage of the last 20 years, or the "rip your old records" turntables generally use lowest quality crap going.

Eh, there are tons of EXCELLENT and not-very-expensive modern phono stages. Just avoid DJ stuff

Why avoid DJ stuff? Isn't a Technics deck going to be better than your run of the mill USB vinyl "recorder"?

I guess it depends where you place "not very expensive", £100-200 will get you a good hifi preamp phono stage with MC/MM. You can pay far more if you want silly esoterica. It's not very far removed from the quality that would typically come "free" with a decent mid-market older preamp.

Not just older amps can have good built-ins.

Perviously I was running a standalone phono stage (Can't remember the exact model offhand, but it was $250 or so - solid, well reviewed component.) I upgrade to a Rega Brio amp about 2 years ago which has a built in phono stage that was even better.

To those who would ask "why rip vinyl instead of getting the CD?", there are lots and lots of records that never made it to CD.

"lots of records that never made it to CD" is a great phrase! Of course we're still stuck using the lingo created at the start of the recording industry boom of the twentieth century, much as we're likely to continue saying "disk drive" although we've already moved away from spinning rust.

Many records are still never released on anything apart from vinyl. Many subcultures in music have returned to/continued releasing on that medium for aesthetic or commercial reasons: most often it's either to encourage physical DJing, the relationship with the music as an artefact, or because exclusivity makes business sense.

Great comment. You need only Google "vinyl only 2016" or any other year to see how relatively common this is.

And lately I see a lot of special editions and "bonus" tracks only on vinyl.

There are also other relatively compelling reasons around quality.

And I'm not referring to the idea that vinyl is an inherently better quality format than CD, but rather that many[0] actual pressings have been mastered in higher quality in practice.

[0] http://dr.loudness-war.info/

I'm firmly in the camp that mourns the victims of the loudness war, but it's important to keep in mind that "quality" is not something that can objectively measured, and every generation has its own perception of it. In MP3: The Meaning of a Format by Jonathan Sterne, he cites research done on "The MP3 generation" -- people who came of age when Napster was on the rise -- that found those listeners preferred the sound of 128kbps MP3 over any other format or bit rate. The technology of encoding created an aesthetic.

And then of course you have Brian Eno:

"Whatever you now find weird, ugly, uncomfortable and nasty about a new medium will surely become its signature. CD distortion, the jitteriness of digital video, the crap sound of 8-bit - all of these will be cherished and emulated as soon as they can be avoided. It’s the sound of failure: so much modern art is the sound of things going out of control, of a medium pushing to its limits and breaking apart. The distorted guitar sound is the sound of something too loud for the medium supposed to carry it. The blues singer with the cracked voice is the sound of an emotional cry too powerful for the throat that releases it. The excitement of grainy film, of bleached-out black and white, is the excitement of witnessing events too momentous for the medium assigned to record them.”

Eno was dead wrong about that. Outside of a marginal Kickstarter project - probably for a Japanese fanbase - I don't think anyone is going to pay money for a high-jitter 12-bit resolution retro CD player.

Comparing that to the stylised imperfections of a human performer is unconvincing.

I do think there's an interesting effect where as soon as a first-to-market technology becomes good enough to avoid being hopelessly terrible, its characteristic flaws set standard expectations for a medium. Hence retro anything - guitars, synthesizers, cars, recording equipment, tapes, vinyl, paperback books, computers...

How likely is it that the early designers were so awesome they hit a technological bullseye with these things - apparently almost every single time anyone developed a new technology?

I can't quite believe that's how it works.

Personally I have no love at all for the sound of vinyl, even on stratospherically expensive hardware.

I've paid money for software which emulates a 12 bit ADAC, with a knob to adjust the jitter etc. I have several such items in fact:

https://www.plogue.com/products/chipcrusher

https://tal-software.com/products/tal-sampler

That chipcrusher thing in particular does some crazy stuff.

Are they on the high or low end for pricing of plugins? I can imagine a full setup (excluding hardware) gets very, very expensive.

I think you're confusing hitting an objective "technological bullseye" with creating something that a specific niche audience will someday see aesthetic value in; the former of course doesn't happen with every new technology, but the latter most certainly does (albeit with different size niches for different tech: guitar distortion and 8-bit music are a lot more popular than 12-bit CD players)

Eno is spot on in this quote

But there are chip tune bands who emulate 8 bit. That people pay money to see. So I can well imagine a niche for emulating flawed CDs.

At the end of BT's Somnambulist there's cut up samples that sound exactly what you'd get with a skipping Discman. I'm not sure I'd go out of my way to listen to something like that but I'm sure there's someone experimenting with those sounds.

I love how people naively pass judgement on an aesthetic, as if everyone's tastes were like theirs.

Besides, it doesn't matter if it sounds human or not, that's not the point. Old hardware introduces a lot of weird harmonics that often affect future stages in the signal chain in a big way. Predictably enough that one can make art with it -- and it's hard to emulate that in software.

As a counter example, there are bands whose aesthetic is defined by what Brian Eno described - taking the sounds of old hardware & exploiting their points of failure. Boards Of Canada uses 80s hardware, preserving/amplifying the sounds of grain & hiss in VHS tape and other analog mediums, and uses it not just as an aesthetic but exploring the theme of society on the brink of collapse as well:

Reach For The Dead: https://www.youtube.com/watch?v=2jTg-q6Drt0 Album trailer: https://www.youtube.com/watch?v=e2lyYEUPat8

It's something else. Take the old synths for instance, their sound wasn't specially great by themselves, but guys like Kraftwerk and Telex, and Black devil disco club were excited with them and did commit a few good tracks. Then the unstable blur of these sounds became something. And more people try to get this sound back, with some success. Music id like the air, it will fill all the space available.

When I go to buy a title from bands that I like, but never got around to buying back when they first came out, I now shop at discogs.com instead of Amazon or Apple. That way I can specify what release it is (typically mid 90's or earlier) and not get a super-compressed recording.

Years ago my dad wanted me to listen to 'the classics'. So he handed me a stack of LPs and asked if I was able to convert them into CDs for him. Of course I had to listen to each one to do a good job.

I'm happy to see that the process has improved in Audacity over the years. When I used to do this there was not a good provision for adding labels to become tracks.

I'm sad that there's no scene standard for ripping vinyls (apart from a few mentions http://scenenotice.org/details.php?id=2006). I'd like to think they'd use this guide.

Are there any services which will do this for you, given crates?

one needs to sample the original at at least double the final finished frequency, or the finished product will have a metallic tinge to it.

eg if 44,100hz rate is what you want to end up with, the sample rate needs to be at least 88,200hz.

Reply on news.ycombinator.com