My collection of about 200 late-‘70s and early-‘80s punk, rock, and pop albums is near and dear to my heart, but I couldn’t justify their shelf space when I know they could all fit on my phone. The only thing holding me back from ripping my collection was the overwhelming nature of recording, labeling, and cleaning up well over 100 hours of audio. While I procrastinated, I looked at turntables, read articles, and talked to musician friends about digital recording. Many of them recommended Audacity.
Audacity is a free and open source software application for recording and editing sounds. It runs cross platform on Apple, Windows, and Linux. Audacity is well-documented at the Audacity wiki, where you can find many helpful tutorials, including tips on how to remove scratches and pops.
Setting Up
Go to the Audacity Main Page to find the Mac download link and installation instructions. I used the screen reader-accessible v2.1.1 DMG. Note that Audacity supports MP3 via a plug-in called LAME. A link to download LAME is available under the Optional Downloads section of the Mac download page.
Because I’m sentimental, I decided to use my old ‘80s-era Thorens turntable. Obviously this didn’t come with a USB connection, so I needed an additional component – a Scarlett 2i2 USB audio interface – to bring audio into my Mac. (I could’ve used a less expensive USB turntable, but went with the Scarlett because I had other audio projects in mind.) I simply set up my stereo system as I have for decades, including speakers for monitoring, but I cabled the amplifier’s tape monitor output to the Scarlett’s input, and connected the Scarlett’s USB output to my Mac’s USB port.
I went to System Preferences > Sound > Input tab to change the Mac’s input from the built-in mic to the Scarlett. What?! SIRI can’t hear me anymore! SIRI’s input defaults to the system input, which means she’s no longer listening to me, she’s listening to The Velvet Underground playing on my stereo through the Scarlett. Oops. To fix this, go into System Preferences > SIRI > Mic Input and select the built-in mic. Also note that, In spite of changing the Mac’s sound input, I still had to change Audacity’s input in Audacity. With Audacity open, go to the Audacity menu and select Preferences. Arrow up in the tree control to Devices, then find the Recording group and change the device to the Scarlett.
Once you’re familiar with Audacity, you might want to consider changing some of the default settings. Rather than detail those here, I’ll point you to the Audacity wiki tutorial on ripping vinyl.
Audacity Basics
Launch Audacity and arrow through the entries on the Transport menu. Note that you can use R to record, spacebar to play and stop, and P to pause and unpause.
Visually, Audacity displays the stereo waveforms corresponding to your recording, along with a vertical cursor indicating the current position of playback. To jump to the beginning or end of the recording, press Home and End. To move the cursor position incrementally, use left and right arrow. By default, left and right arrow move you a fraction of a second. You can increase or decrease the arrow increments with the zoom in (Command+1) and zoom out (Command+3). Command+2 returns you to the default view and arrow increment size. This info might seem superfluous, but you’ll need it later when you edit and label your recordings.
Audacity is a multi-track digital recorder. If you record, stop, and record again, the second recording goes on a new track. This is probably not what you want when ripping vinyl. There are a couple ways to work around this. Use Shift+R to append a recording, or use pause and unpause – both let you keep recording on the initial track. The solution that worked best for me was to never stop or pause, but to simply record the entire album in a single long recording session. I started recording with R, then played side A, turned it over and played side B, then finally hit spacebar to stop. This resulted in a single (stereo) track with the entire album plus some dead space that I trimmed later during the editing phase.
Using Audacity’s GUI
The Audacity GUI contains several nested groups and scroll areas. The main top-level elements are the Tool Dock and Selection scroll areas, and an area labeled “Track 1” containing the waveform.
Tool Dock Scroll Area
Within the Tool Dock, the Mixer scroll area allows you to set the recording level. Find the Recording Slider and hold down left or right arrow. This is especially useful if you’re using a USB turntable with no gain control. In my setup, I changed the gain with my Scarlett. You might want to mark this slider with a VoiceOver hotspot. (Use VO+Shift+1 to mark the control, and VO+1 to jump to it.)
Also within Tool Dock, the Device scroll area lets you quickly switch between the built-in mic and the USB audio input. You can do this in Preferences > Devices as I mentioned previously, but during initial setup, I found it handy to toggle this control frequently with a VoiceOver hotspot. (Use VO+Shift+2 to mark the control, and VO+2 to jump to it.)
Selection Scroll Area
Use the Selection scroll area Project Rate control to set the sample rate of your recording. The label is above the control, which confuses navigation somewhat. The default is 44100Hz, but a pop up control lets you select different sample rates. If you’re a hardcore audiophile, you might want a larger sample rate than the default, but keep in mind the Audacity project file is uncompressed audio, and increasing the sample rate will use more storage on your drive. At the default 44.1kHz, a standard 42-minute vinyl album occupies about half a gigabyte of storage.
The Selection scroll area also contains three time controls, labeled Selection Start, End/Length, and Audio Position. These controls come in handy for precise editing. On my v2.1.1, End and Length are radio buttons, allowing you to set the second time control to either the end of the selection, or its length. The End radio button is mislabeled as “selection start radio button”, so custom label that with VO+/.
I use the Selection Start and End/Length time controls for precision selecting, so I assigned VoiceOver hotspots to both of them so I can jump there quickly – VO+Shift+3 and VO+Shift+4.
I recommend changing the default display format to display (and enter) values in milliseconds. Go to any one of the controls, such as Selection Start. Open the context menu (VO+Shift+M – I can never remember that) and arrow through the options to “HH:MM:SS + milliseconds”. This changes the format for all three controls.
”Track 1”
If you made a test recording, you probably have an area labeled “Track 1”. Audacity assigns a default numeric name to each track. For vinyl records, you’ll usually have two tracks, one for your recording and another for labels (song names), called “Track 2”. You can rename these (for example, to “Music” and “Labels”). Bring focus to the desired track, open the context menu, and select Name.
Workflow Tips
After doing a couple test recordings and familiarizing yourself with the controls, you’re ready to try an album and learn about some of the steps in the recording workflow. So blow the dust off your copy of Metal Box by Public Image, and record it start to finish.
Trim Dead Air
After you’re done slam dancing around the room, let’s start by trimming the dead air at the start of your recording. The basic idea is to select the dead air and hit backspace to delete the selection. Go to the start of the recording, then right arrow to the beginning of the first track, which is also the end of the initial dead air. A trick for this is to right arrow for a second or two then hit spacebar to play. If you’re still in the dead air, repeat that, and if you’ve gone too far and are already in the music, then left arrow instead.
Once you find the start of the first track, you can select from the cursor position to the beginning of the recording with Shift+J. This should be all dead air, and you can verify that by hitting spacebar to play the selection. If you’re satisfied that your selection contains what you intend to delete, hit backspace.
Verify you’ve done this correctly by playing from the beginning – the first track should start immediately. As with any Apple program, you can use Command+Z to undo, so if you delete the wrong audio by accident, undo and try again.
Trim dead air at the end of a recording using the same basic technique, but you should use Shift+K to select from the cursor position to the end of the recording.
Selecting dead air in the middle of a recording is a bit different, as you can’t use Shift+J or Shift+K. Instead, move the cursor position to the start of a selection, then Shift+right arrow to select. Periodically hit play to test your selection. Note that Shift+right arrow moves the end of the selection later in the recording, and Shift+left arrow moves the beginning of the selection earlier in the recording. In other words, Shift+arrow key only grows the selection.
You can also use the time controls in the Selection scroll area, but these controls are very difficult to use. To change a digit, you must move a custom highlight to that digit, and because it’s custom, of course it’s not accessible. Here are some tips that might help.
If you set the format to “HH:MM:SS + milliseconds” as I described previously, then there are nine digits you can change, two for hours, two for minutes, two for seconds, and three to the right of the decimal place for fractional seconds. Press Home to move to the most significant digit, then type all nine digits. For example, to set a time control to 0 hours, 13 minutes, and 8.496 seconds, you would hit Home, then enter 0 0 1 3 0 8 4 9 6. VoiceOver is entirely silent while setting these values, so press VO+F3 to read the control to make sure it contains the desired value. Additionally, the End key moves focus to the least significant digit, left and right arrow keys move between digits, and up and down arrow keys increment or decrement a digit (with carryover if necessary).
Now that you’re an expert on Audacity’s time controls (ha ha), here’s how you might use them to select a segment of audio. In your audio track (Track 1), use the arrow keys to find the end of the region you want to select. Go to the Selection Start time control (I use VO hotspots to jump there quickly), use VO+Shift+C to copy the time that VoiceOver just read, and paste it into Notes. Next, change your cursor position to the beginning of the dead air, then go to the End/Length time control. Make sure the End radio button is selected, then enter the control and enter the time you noted earlier. Note there’s no way to paste. Sounds convoluted, but it works and should select the desired dead air. Hit play to verify then delete it with backspace.
Remove a Pop
When I first started this process, I agonized over removing every single audio defect my ears could discern. I regret this now. Some small pops here and there, along with the stylus up and down noise at the start and end of each album side, really add character to the recordings. You can save yourself a lot of work by selectively preserving some vinyl noise. If you’re going to aggressively attack pops, I recommend taking a look at the Audacity tutorial on Click Removal.
Audacity offers a couple ways to remove pops and scratches, and neither solution is really the paragon of pop removal, in my opinion. Effects > Click Removal is an automated solution. Audacity opens a dialog containing parameters to control the process. I’m no expert on these settings, and have never been able to remove all pops and scratches reliably with any values I tried. Let me know how it goes for you.
I used a more hands-on approach. After identifying the location of a pop in the waveform, I zoomed in about 10x until the individual samples were visible. Then I selected the samples corresponding to the pop, and selected Effects > Repair. This interpolated the surrounding waveform through the selected area. Besides being labor-intensive, I depended on my remaining crappy vision for this task, and never found a viable accessible approach. The selection area is very small, a fraction of a second, and needs to wrap the pop precisely. If you can do this totally blind, you have more patience than I do.
Labeling Songs
During editing, you’ll need to place a label at the beginning of each song. Audacity uses the labels to split the recording into MP3 files during export, and also fill in each MP3 song title. To label a song, position your cursor at the start of a song and hit Command+B. Type the name of the track and hit Enter. Repeat for each song on the album.
This actually adds a second Audacity track, which means your GUI now has a scroll area containing Track 1 (the waveform) and Track 2 (a list of labels). Review and edit your labels under Tracks > Edit Labels.
Metadata
Add album title and artist information with File > Edit Metadata. I filled in the album name, artist, and genre, and left the other fields blank.
Export to MP3
Export MP3 files with File > Export Multiple. This exports each labeled song as an individual MP3 file. The export dialog provides an option to precede each filename with a song number. I used this option so that alphabetized files would appear in correct song order, but it’s not necessary.
If Audacity can’t find the LAME plugin, you won’t be able to select MP3 as an export format. Check the LAME installation instructions and restart Audacity if necessary.
During export, Audacity presents you with the metadata for each song. This gives you a final opportunity to check for spelling errors. It’s also useful for albums containing various artists, in which you want a different artist for each song.
Importing Into iTunes
Add these MP3 files into iTunes as usual. In iTunes, go to File > Add To Library, and select the MP3 files exported by Audacity. Then go to the Recently Added playlist and verify that the album imported correctly.
The only caveat concerns albums with various artists. Audacity makes no provision for an album with artists per song and a separate “Various” album artist. If each MP3 file has a different artist, iTunes will import each song as a separate album. To fix this, open the iTunes Info dialog for each song and set the Album Artist field to “Various”. iTunes will combine those MP3 files into a single album.
Next, Rock Out!
Now that you’ve learned the basics, I recommend ripping a half dozen albums and listening to the results. You might find you need to tweak the gain or be a little more OCD about removing pops. It’s better to find out after a few albums than after 50 or 100.
If you try Audacity, please share your experiences in the comments below. I’m especially interested in hearing from people who can compare and contrast with other audio processing tools. But however you rip your albums, remember to enjoy the music and rock out!
Comments
Audacity, Amadeus Pro, and GoldWave
Glad to see this posted. I happened to just blow my weekend playing with Amadeus Pro and cruising through many of Pete Layne's really excellent BlindAbilities podcast tutorials about it. But, after realizing I would still have to use Audacity to automatically remove quiet passages (silences), and then further realizing that AP's speed/pitch effect delivered poor-quality audio results, and finally not really liking the fact that Amadeus Pro doesn't distinguish between cue-markers (labels) and temporary editing start/end marks, I started reading the Audacity manual. Audacity is what I'm planning to use on Mac for the present. As a lecturer, I teach some online classes and have to edit the heck out of about 45 hours of my own babbling each semester until it becomes coherent. I've been doing this in GoldWave for years (yes, that program is still alive and well!). As I shift more of my workload onto the Mac, and as Apple makes bootcamp virtually impossible on laptops, it was time to try for more of a switch. Audacity 2.2 is about to be released, and there's still no good news about accessibility. 2.1.1 was released July 2015, and it's the current end of the line for screen reader accessibility. Apparently, they put a lot of custom code into it for Jaws, NVDA, and Voiceover, and that special code for Mac broke when Apple changed some things. I wonder if this is a classic example of how hard accessibility retrofits are as compared to creating universal designs from the start. Perhaps Amadeus Pro lucked out and was simply programmed in a more Apple-native way than the cross-platform Audacity. If that's the case, I don't really hold out much hope for Audacity for the future. On the other hand, there's not a lot of new stuff you can dump into an audio editor. The needs and the formats haven't changed much in 20 years. So, the old version will likely do me fine until it breaks or I go back to Windblows.
FYI, here's a quick run-down of my podcast workflow, followed by a quick list of the Audacity keystrokes that should accomplish it. I'm really just starting to try it, so I may find things that don't work as I expect.
Workflow: * load an 80-minute wav file that I recorded in class using my BookPort Plus, which I like using better than the IPhone, both for reliability and microphone quality. * Remove all silent passages in one go, defined as -27 DB for more than a secord or two.
* I might normalize the file, either just the normal way that brings the total waveform up to -1DB (to avoid distortion) or else the more sophisticated RMS (root mean square) algorithm, which uses the average peak level for amplifying the volume. RMS normalization sometimes allows clipping, though. * Speed the whole thing up by about 8% while retaining pitch. Makes it more exciting. I can't really listen to slow talking podcasts. * Settle in to remove hems and haws, mistakes, move passages around, etc. * Run the Compressor effect on quiet passages, such as students asking questions or answering mine. Dynamic compression brings up their voices until they're the same loudness as mine without clipping. * apply pop removal on "cracks" where an edit has resulted in a popping sound. * Label each topic, because I like to break the long file into more digestible segments based on the idea I'm discussing. * Export multiple to highly-compressed mp3s, as described in the blog post above.
Here are some keyboard shortcuts. Note that key bindings can be changed by writing them out in a properly-formatted text file and importing them. Perhaps my favorite Audacity feature is the key to repeat the previous effect, since it applies a preset as well without ever bringing up a window. That's not available in GoldWave or Amadeus Pro. I jotted down this list of keys this morning. I haven't tried them all yet. This list highlights how similar the abilities of Audacity, Amadeus Pro, and GoldWave all are to one another, although each acts and feels different in significant ways.
playback
selecting:
other editing:
Re: Audacity, Amadeus Pro, and GoldWave
Excellent comment. Thanks, Voracious P. Brain. I wanted to keep the blog to a reasonable length, and was hoping the comments section would flesh out the details. Your post was exactly what I was hoping for. The shortcuts dealing with selection are new to me, and will be very helpful. Thankfully my vinyl ripping project is complete, but I still use Audacity for other audio projects. Maybe another future blog post.
Or, you know, maybe not...
So, like, virtually none of those keys are working now, man. I can't extend the selection with the shift arrow keys, previewing with C doesn't work, and I'm not sure about zoom. Did High Sierra break it? It's been awhile since I broke a brand new toy within 24 hours... Or is this user error? Try some of those keys, which come from the 2.1.1 manual, and let me know. Was wondering why Amadeus got so much love around here while Audacity has only been mentioned a few times with mixed reviews. Maybe this is the reason from Sierra onward. But maybe I just have to have the focus on something in particular or do something in the view settings.
Re: Or, you know, maybe not...
I've had a lot of trouble with the audio waveform losing focus and not having a good way to navigate to it.
Sometimes this works: Bring VO focus to "Track 1", then VO+Shift+down arrow to interact with the text. Then VO+Shidt+up arrow to stop interacting. I've found this bring focus back to the waveform, and the selection shortcuts you mentioned suddenly work again.
I think the Audacity project could benefit from a vision impaired developer contributing to it. I don't have the bandwidth for that, and have never liked posting to an open source project about bugs if I wasn't personally willing to fix them.