Transfering Vinyl to Cd

Posted: Mon Jan 28, 2008 12:15 pm

At present im in the process of transfering some of my vinyl collection onto my pc so that I can burn the tracks to disc for use with cd players.

So far i've just been using audacity as the recording device, having connected the output of my mixer into the mic input on my pc and saved each track on a seperate hard drive.

Most of the tracks i've done so far sound quite good as they are but some of them have a little crackling through them due to wear and tear.

Does anyone know if there is a feature that can be used to clean up the tracks in audacity? If not are there any alternative free programs out there that would do the trick?

Posted: Mon Jan 28, 2008 1:23 pm

Your first mistake is that your phonograph cartridge needs to be loaded into a proper 48,000 ohm input which a microphone input is not. BIG NOT. PLUS IT HAS DC on it which can actually destroy your phonograph cartridge. Microphone inputs are flat in their frequency response and phonograph records follow the RIAA preemphasis/deemphasis frequency response curve. So if you want your precious record collection to sound good? You need to use an actual phono-preamp which not only includes the proper 48k input impedance but also includes a very small value capacitor shunted across the cartridge/resistor combination to properly resonate the cartridge as its magnet and coil structure with resistor and capacitor creates a tuned circuit, which aids in a flat response after the deemphasis is applied to the signal. So just use your stereo system and take your tape outputs to your sound card line inputs for a proper record to computer system. The only thing that goes into microphone inputs on a computer is computer microphones. Not studio microphones. Not phonograph cartridges. Just $3.98 multimedia microphones. That's all that should be plugged into those anything else is WRONG WRONG WRONG.

In the middle of a 78/33 monster archive project and utilizing API & Hafler phono preamps.
Ms. Remy Ann David

Posted: Mon Jan 28, 2008 1:34 pm

Remy's right - unless your mixer happens to have a phono preamp input built in.

Even if it does though, it's probably under spec and likely to at least sound questionable and possibly damage the delicate cartridge.

That being said, I don't know of any free programs that are good at this, but Magix has Audio Cleaning Lab (or similarly titled) which is only $40-$50 and well worth it. It's designed for exactly what you're talking about and works quite well given its meager price.

How are you liking that Hafler phono preamp Remy? I'm using Creek myself and really like the sound. I just don't like that it doesn't have any other curves than RIAA. Oh well...

Posted: Mon Jan 28, 2008 1:45 pm

Hi, thanks for the reply. I guess that I need to try some other way then?

I'm pretty much limited to the mic input that is on the pc. Maybe if I give some details on the equipment i'm using someone could maybe suggest an option:

Technics 1210 M5 Turntable
Numark DXM 06 2 Channel 24 Bit Digital Mixer
Acer Aspire PC
Audacity
Numark Monitors

What i've been doing so far is connecting the turntable to one of the channels on the mixer via the phono input with the RCA connectors fitted to the turntable.

I've then connected the mixer from the master output to the mic input on the pc via a RCA/TRS Jack adaptor cable.

Then i've simply hit record on Audacity. My speakers are connected to the record output on my mixer.

If anyone can help further that would be great.

Posted: Mon Jan 28, 2008 2:36 pm

Is there a line input on the computer?

If not, I would seriously advise you to pick up even a cheap/modest external USB or Firewire device. It doesn't have to be fancy, just have a line input (stereo).

Posted: Mon Jan 28, 2008 2:39 pm

Additionally, you say you've recorded a few albums into the CPU already?
Were they mono recordings before?
I'd imagine they are now.

Posted: Mon Jan 28, 2008 3:00 pm

Upon further investigation I can borrow an external firewire sound card device from a friend. If I connect my set up through this would it be enough?

To answer bent's question. Yes they are mono recordings that I now have on the PC. Can I simply duplicate the track to make it stereo or is there something else i'd need to do?

Posted: Mon Jan 28, 2008 3:16 pm

Mono is mono, regardless how many speakers you play it back on, or number of tracks you copy the source to.
Stereo is a different animal altogether.

You'll want to re-record those albums with the FW Interface, either way - the quality will be a lot higher.

And, yes - the Numark is a DJ mixer so you will be just fine going this route.

Posted: Tue Jan 29, 2008 11:16 am

thanks for the help guys;)

Posted: Mon Feb 04, 2008 11:01 pm

The RIAA de-emphasis curve consists of two, 1st order (6 dB/octave) low pass filters. I forget the frequencies (look it up), but one could simply record flat into the PC and then put these two EQs onto each track and, voila, get the correct response (paying attention to Remy's note about cartridge loading - which, BTW, depends on the specific cartridge model. not all cartridges want 48K ohm loads.)

The downside of doing this is that the dynamic range of the recording will be somewhat compromised by the over-abundance of HF content being recorded. Easy fix, though - just give yourself LOTS (say, 20+ dB?) of headroom going into the PC and record using 24 bits.

Posted: Tue Feb 05, 2008 4:58 am

dpd wrote:

The RIAA de-emphasis curve consists of two, 1st order (6 dB/octave) low pass filters. I forget the frequencies (look it up), but one could simply record flat into the PC and then put these two EQs onto each track and, voila, get the correct response.

This would get you a really poor-sounding result. Equalisation curves such as those published by the RIAA for vinyl cutting/playback, are specified in terms of time constants. It's up to the equipment designer how (s)he implements these time constants in a circuit design, but they are typically done using a chain of first-order R-C networks in either feed-forward or feedback (or both) configurations around active amplifying stages. Specifying time constants not only specifies the frequency response, but also the phase response, and to get the correctly equalised sound, both are important.

EQ sections of mixers, whether implemented as hardware or software, are made up of elements of at least second-order, and cannot be set to implement first-order time constants in both frequency and phase. You may very well be able to approximate the frequency curve, but the phase would be wrong.

Posted: Tue Feb 05, 2008 8:20 am

Thanks for that info Boswell!

Damn, I learn something new here every friggin day!

Posted: Tue Feb 05, 2008 8:57 pm

- I tried converting LP's to CD a few years ago and got very good results. I started by putting in a new cartridge on the turntable and took the stereo out from the amplifier into a mixer and into the stereo line in on the sound card. I used Sound Forge to clean up some of the noise (it was a record from the 40's), and played with some EQ. It was tedious, but once you find the right formula it works, and you can re-use it.

I'd tried so-called automagic hiss and pop removal software that also claimed to auto save the recording into seperate tracks by judging where the song breaks were... neither feature worked very well. Especially if you have music with dramatic breaks in it.. it assumed these were seperate tracks and chopped them up! Hopefully someone has improved the software beyond this by now?

Otherwise the manual method works - but it was so tedious I dread the thought of attempting this again!

Posted: Tue Feb 05, 2008 9:51 pm

There's a german company called Terra Tec that makes a phono to usb interface that comes with noise reduction and burning software.

Its a little sluggish as an interface and the process is VERY time consuming but it works and creats reasonable cds. I don't use it for any high fidelity things but for most pop/rock things it works. You can get a cleaner cd but it does so by introducing phase issues through eq (surprise, surprise) Where's jp22 when you need him?

BTW-in the for what its worth department, my high school students don't know where to drop the arm on a vinyl recording- outside edge or near the label. I'm feelin' old - where's my whittlin' knife.

Phil

Posted: Tue Feb 05, 2008 10:23 pm

Actually, I remember building a passive RIAA de-emphasis network between two gain stages - and did the computer circuit simulations that matched the measured response. that's why I believe that this will work. Would I recommend it for serious vinyl transcription? not really

The time constants (t), by definition, set the -3 dB frequencies (f) in the Bode Plot of the equalization curve: f=1/(2*pi*t). t= R*C. These define (as you correctly stated) both the magnitude and phase (integrated group delay) responses. They are two, ganged low pass filter sections.

You can equalize in the feedback path or not. There are issues either way and a lot of it depends on the amplifier's open-loop bandwidth, phase margin, etc. Both can work and provide absolutely accurate RIAA inverse equalization.

I should just plug my TT directly into my ProTools rig, dial in the filters and see how it works - play some white noise into it and run it through a spectrum analysis plugin to confirm. But, it's easier to just grab the phono stage output of my SP-9. Very Happy

Boswell wrote:

dpd wrote:

The RIAA de-emphasis curve consists of two, 1st order (6 dB/octave) low pass filters. I forget the frequencies (look it up), but one could simply record flat into the PC and then put these two EQs onto each track and, voila, get the correct response.

This would get you a really poor-sounding result. Equalisation curves such as those published by the RIAA for vinyl cutting/playback, are specified in terms of time constants. It's up to the equipment designer how (s)he implements these time constants in a circuit design, but they are typically done using a chain of first-order R-C networks in either feed-forward or feedback (or both) configurations around active amplifying stages. Specifying time constants not only specifies the frequency response, but also the phase response, and to get the correctly equalised sound, both are important.

EQ sections of mixers, whether implemented as hardware or software, are made up of elements of at least second-order, and cannot be set to implement first-order time constants in both frequency and phase. You may very well be able to approximate the frequency curve, but the phase would be wrong.

This topic sponsored by: