I’m developing a workflow for digitizing my vinyl collection, and I’m wondering about how to export a WAV file after I’ve captured the signal. Here’s a rough sequence of steps:
Start a recording session in Ardour, with two channels, left and right to capture left and right channels from phonograph.
In the mix interface, left channel panned 100% left, right channel panned 100% right
My next step is to export a WAV file for further signal processing using Brian Davies’ software. When I export, should I be selecting the “Master”, or should I deselect that and select the individual channels? I’ve tried it both ways, and can’t hear much difference. My ultimate goal is to save the processed WAV files as master files to archive, and create good quality MP3s for everyday listening.
I made a start on this on my (then) many-thousands of vinyl albums. Two observations from that:
command line tools made much more sense to me (at least then). I ended up using jack_capture
In the end it wasn’t worth it except for a handful of material that I could not find in digital form. I stopped ripping after 10’s of albums: the work of dividing the recording into tracks, adding meta data, cover scans etc. just did not add up to me.
I have a similar experience as @paul, and only digitized a few select albums, but I did use Ardour for the process.
This allowed me to add a noise-gate, and some slight compression to clean up the recording. After defining ranges for each song, Ardour can export them all in one go, directly applying Loudness Normalization to each range, and produce both flac and mp3 in the same step.
I cannot find a currently active link with the Davies software, do you know if it accepts floating point files as input, or just integer PCM wav files? If the software only works with integer files it may be easier to use software which records directly to integer files rather than recording into Ardour (which saves as floating point files), then having to export for processing.
The software accepts 16, 24 or 32 bit integer, or 32 bit float WAV files, or AIFF or AIFC files. I’ve decided to work with 32 bit float WAV files, sampled at 96K, which is admittedly overkill, but I have the storage space and processor speed to do this.
After some years I found that the quality of sound will mostly depend on the stylus, turntable (mainly motor noise and tone arm resonance) and on the cleanliness of your stylus and your records. Digitizing and exporting details will make no big difference.
Even if you record and re-record a record immediately 3 times all the three will sound somewhat different compared to each other (in headphones). The low-mid-high balance and the stereo details will differ somewhat. And not necessarily the first run will be the best, often the third (for me, to my subjective ears).
I start with a microfiber cloth + ipa for cleaning rough dust, and this is followed by a Degritter record cleaner (available for me in a local record shop), with a medium run. Actually this makes the biggest difference.
After this, cleaning the stylus after every play.
A month ago I had to replace my stylus because it become worn out, and the new one sounds different than the old when that was new 3 years ago (Ortofon Concorde MkII Club with VTF ~3.75g). Noticeably different. For me it is not the question of worse or better anymore it is just good in a different way.
The best “de-clicker” is said to be an ultrasonic cleaner. Better to avoid the noise at the beginning than to use software to repair clicks later.
I have also seen a recommendation if using de-click software to record “raw” with no RIAA equalization if possible. That makes the clicks more obvious so the software can detect the click area more easily, and also helps to hide any artifacts of the modification by heavily filtering after the modification.
That may not be possible easily, depending on the equipment you have available, so just suggesting for consideration.
I have a handful that have been waiting until I have a little more time and space available for setting up new hobby equipment, but generally fall into a couple of categories:
1a: completely out of print, never made available in a digital format
1b: completely out of print, rare enough that used CD copies are in the $200 range but LP is much cheaper
2: Available as CD or digital file, but only in a modified (“remastered”) version with trendy heavy multi-band compression which changes the sound considerably from the original vinyl release
An ultrasonic cleaner? Like the ones you can put your glasses in? I have one of these but never came up with the idea of cleaning my records in there. Anyone has any ecperience with that (time, temperature)?
Well, same technology, but devices made to fit your glasses are probably not likely to fit an LP.
Just search for “ultrasonic record cleaner” on Google or whatever search engine you prefer and you should find enough info to keep you reading and/or shopping for the afternoon.
With 96k and also with even higher samplerates check with a spectrum analyzer what you are really recording above 20kHz-22kHz. In my case I found that I’m just recording ultrasonic noises and distortion, and with higher levels than I normally have in the 20kHz-22kHz range.
As far as I know only Voxengo’s Span spectrum analyzer can be set to show stuff above 20kHz up to the actual nyquist frequency. Also this is the only one that can be set to show stuff below 20Hz to 0Hz, so you can check only with this analyzer if you have some bad low frequency tonearm resonance that should be high-pass filtered (somewhat).
Yes!
In many cases using higher sample rate for analog signals results in worse quality, since most transducers do not support it and ultrasonics may fold back into the audible range.
–
Instead of using an analyzer plugin, you can use Ardour’s built-in one: select the region and then Menu > Region > Spectral Analysis (or right-click on the region, region-name > Spectral Analysis…)
it is just good in a different way.