9 min read · June 20, 2023
In recent weeks, I've decided to follow in my Dad's footsteps by digitizing my record collection. I've been down a few rabbit holes and have had some amount of trial-and-error with different equipment and setups. But now, with the help of Plex, I can now enjoy my collection on the go, using Plexamp on my phone, complete with real pictures of album art.
I don't have what I'd say is a large collection; only somewhere between 120 to 150 albums, the majority of which were gifts I've received over the years. To keep track of my collection, and what I already have digitized, I have most of them checked into my Discogs collection, but I still have a stack sitting to the side that I haven't added yet.
I've also been keeping track of my wishlist in Discogs. Which has had the added benefit of giving me a way to create my own "record of the month" subscription, in a way. There's a "Random Item" link in your Wantlist that just directs you to a random album in it. A great feature for eliminating analysis paralysis of what to buy next, and also adds kind of a surprise element to what to buy next.
The process of actually recording and digitizing them has been slow but a lot of fun. At best, I'm only getting through one or two albums a day. And I've gone back and re-recorded some after learning some lessons and changing my setup a bit. But it's forced me to get out of my Spotify playlists and recommendations and revisit some things I haven't listened to in forever and, in a lot of cases, songs I never gave a fair shot. After all, you're recording an album in real time while it's playing and there's no skip track function on a turntable.
When it comes to the practical or objective reasons for digitizing my collection, I have to admit I don't think I really have any. None of the albums in my collection are so rare that they can't be found on streaming platforms like Spotify. And I have to admit that I don't honestly believe that vinyl inherently sounds "better" in any objective sense. But I think there's something to be said for the subjective "feels" of listening to something that you actually own (increasingly rare these days) or the additional "character" you get out of an analog format like vinyl, for better or worse.
I've been amazed by how good some of my older, used classic rock and blues records still sound after decades of use. But every once in a while you run into one that you think isn't salvageable. Enter record cleaning kits.
Even a cheap one works wonders. I started with this kit that I think Dad got me for Christmas one year. Turns out it's usually listed for less than $10 on Amazon and still works some crazy magic on even some of my roughest vinyls.
When the solution ran out, I "upgraded" to a kit with some extra tools. Specifically this one, which had a tiny brush for cleaning the needle.
There's not much you really need to get started. A turntable, some sort of sound device with a line in, and the right audio cables.
I record on my laptop, which doesn't have an audio input so I had to invest in an external soundcard. I managed to get an open box SoundBlaster GC7 for about half price. Originally I wanted something that could record stereo line-in at the highest possible quality. At 192kHz/24-bit, the GC7 will definitely get you there but as you'll see below, I eventually discovered that a sampling rate that high is overkill.
If your turntable is too far away from a machine you'd record on and aren't willing to rearrange an entire room, there are some things I tried that might be useful to someone else.
There are Bluetooth codecs that support higher quality, and in some cases lower latency, audio. LDAC or any of the aptX flavors will get you sampling and bit rates beyond the limits of human hearing, assuming both the transmitting and receiving hardware support it.
The first iteration of my setup used two of these 1Mii B03 Bluetooth Transmitter/Receiver units - one in transmitting mode plugged into my turntable via a 3.5mm audio cable and one in receiving mode plugged into my SoundBlaster GC7 via an optical cable.
This mostly worked really well. And using aptX HD or aptX Low Latency (LL), I was getting sampling rates well above what our ears can detect. But I ran into some snags with interference from other devices that would occasionally cause signal blips and cause me to have to re-record an entire side of an album.
Also it eventually just got annoying walking back and forth across the room to start recording on my laptop and then hit start on the turntable. So I rearranged the room a bit so I could just directly wire my turntable directly into the GC7.
The 1Mii units weren't a total waste, though. The aptX LL codec has such crazy low signal latency that I was able to repurpose them to give myself a pair of rear speakers wirelessly without any real noticeable audio delay. So now I just have one in transmit mode plugged into the GC7 via a 3.5mm splitter that also goes to my "front" speakers. The other is in receiving mode with a separate pair of bookshelf speakers and a subwoofer plugged in behind me. Gives a pretty nice, faux surround sound sensation.
There's a lot of pretty impressive apps out there, for various operating systems. Some are super expensive. Some are free. I saw a lot of people just use Audacity to record and split their tracks, but that just seemed like it could get tedious for a variety of reasons.
I opted for VinylStudio. It works on Windows and macOS. The interface is admittedly pretty 2000s-y, but it has some great features that make it really easy to split the album into sides and tracks. It also has some correction features that can smooth out some stubborn pops and scratches that even have some guardrails to not eliminate percussion noise. My favorite feature is that you can just import artist, track, and genre metadata from third party sites like Discogs and it will write them into the files.
The few main TL;DR points here are:
1. You do not lose quality inherently by the act of converting analog audio to digital. See stairstep myth below.
2. A sample rate of 44kHz/16-bit is pretty standard for music production and beyond the limits of human hearing. Anything higher is probably just wasting disk space and/or bandwidth.
3. Where quality does become an issue in this process is if you choose to record to a lossy format with high compression like MP3. If you choose to use MP3, I would say just make sure you're using a bitrate of at least 320kbps. Else, use a lossless format like FLAC.
I think the YouTube below probably explains this in more depth - and with an actual demonstration - than I possibly could. A lot of it was over my head. But my simple understanding is that this myth comes from a mischaracterization of digital audio signals by incorrectly graphing them. There is no inerent audio quality lost by going from an analog audio signal to a digital one.
When I did some light research to see if this was actually correct, I found cases of other audio engineers that have been shouting about this for at least a decade. And yet the myth persists; I definitely believed in it before I came across this video.
There's not much more to say other than what I outlined in the TL;DR above. I mentioned that originally I was recording at 192kHz/24-bit. But after some reading, it turns out this is just overkill and, as the stairstep video above briefly mentions, turns out it could actually have adverse effects.
44kHz is beyond what humans can hear. There's a lot more you can read about this magic number on Wikipedia, but using anything higher than that for this purpose is probably just creating massive files for no reason.
While legitimate debates exist about using 24-bit in certain music production scenarios and lowering noise floors, for the average listener 16-bit is typically going to be enough.
Also consider that, especially for more modern music, it was probably recorded and pressed at 44kHz/16-bit. You can't really create a "better" quality than what you're recording from.
In my opinion, lossy formats like MP3 are totally fine as long as you're using a bitrate of at least 320kbps. There are a couple blind tests out there where you can listen to clips of the same audio in different formats and bitrates to see if you can hear a difference.
I'll be honest, I failed the lossless test when compared to 320kbps lossy audio. I can't hear a difference. But I was able to detect a very slight difference between 256kbps and 320kbps. You have to listen really closely, though. Lower than that I would say most people would be able to hear a difference with two audio samples side-by-side.
The main alternate format option to a lossy format like MP3, then, is FLAC. I personally use FLAC because it makes me feel better. I know full well that I could probably be saving even more disk space by just using 320kbps MP3 and not be able to tell the difference. But sometimes we do idiotic, irrational things just because they make us feel better. Y'know, just in case.