Conversion Quality

I’m new to Audacity and the whole “A-to-D” process. I’m using an AR Turntable with a moving coil (MC) cartridge, which drove me to use my Carver C-1 preamp (MC section) with the RCA outputs hooked up to a Behringer UCA202 interface. I’m using the latest version of Windows 10.

I’ve followed the Audacity Sample workflow for LP digitization and have successfully generated .wav and .mp3 files of albums using recommended default quality settings. When I do an “A/B” comparison of the original phono-preamp sound and the 32-bit .wav file (played back through the same preamp and headphones), there is a notable loss of audio depth/fullness. Is this what I should expect from the prescribed A-D process? Any suggestions for improving the conversion quality?


Any suggestions for improving the conversion quality?

Buy the CD or MP3 (if it’s available). :stuck_out_tongue:

Is this what I should expect from the prescribed A-D process?

No. It should sound identical.* It’s possible to get noise into a USB-powered interface (through the USB power) but it’s rarely an issue with line-level inputs. Microphone & phono preamp circuits are more sensitive to noise. preamps actually generate some noise, but the worst noise is almost always from the record itself. And you didn’t mention noise.

Of course, use “CD quality” (16-bit, 44.1kHz) or better.** Audacity uses 32-bit floating-point “internally” for technical reasons but the Behinger is “only” 16-bits. I don’t know of any 32-bit analog-to-digital converters. There are some 32-bit digital-to-analog converters but even 24-bit converters aren’t accurate to 24-bits.

And make sure you’re not clipping (“trying” to go over 0dB and distorting). I usually simply run the Amplify effect after recording to check the peaks. Audacity has pre-scanned your file and Amplify will default to whatever gain needed for maximized/normalized 0dB peaks. i.e. If Amplify defaults to +3dB your current peaks are -3dB (you have 3dB of headroom before amplifying). If it defaults to 0dB (no change) you are probably already clipped and you may want to re-record.

…Low digital levels are NOT a problem. This isn’t analog tape where you needed a strong signal to overcome tape hiss.

Make sure your levels are matched before A/B listening. And since sighted listening tests can be misleading, [u]What is a blind ABX test?[/u] Of course you don’t have to do an ABX test.

there is a notable loss of audio depth/fullness.

I don’t know what those words mean. :wink: There is noise, distortion, and frequency response, plus with MP3 there can be compression artifacts which are harder to measure & define but can be one of those or a combination of those. You can also get dropouts/glitches while recording, but I doubt that’s happening with your “symptoms”. Of course when you record with a microphone there can be unwanted reverb or other acoustic effects. See [u]Audiophoolery[/u].


  • [u]Here[/u] is a little story about an informal blind listening test of direct vinyl vs analog-to-digital-to-analog. (It looks like he was digitizing at 24-bits/96kHz but 16/44.1 is generally better than human hearing.)

** MP3 is lossy but a good quality MP3 can often sound identical to the uncompressed original (in a blind ABX test). And, it’s technically better than vinyl or any consumer analog format. “Damage” does accumulate if you compress more than once so if you want MP3, compress to MP3 once as the last step. If you want go-back and make more changes, go back to the WAV.

Thanks for your quick and thorough reply. Your comment, “No. It should sound identical” is very helpful. It tells me I should be able to do better. As you noted, noise is not the issue. The digital copy I created is actually quite clean that way. I can hear a bit of record noise, but I would expect that and it seems pretty good otherwise.

I apologize for my poor description of the quality issue (“…notable loss of audio depth/fullness.”). I would liken the degradation to listening to a recording from a good source through a high-end stereo, then listening to it again through a low-end system. The “audiophile” system has a certain clarity and richness (sorry, more subjective terms) that the low-end system can’t replicate. The conversion I got is “good,” but doesn’t have the same quality as the signal coming out of the phone-preamp.

I used the 16-bit, 44.1kHz setting which, as you noted, is consistent with the Behringer’s capability. I was very careful not to get any clipping (I quickly learned how distorted clipped passages could be). I sampled the album in multiple locations and targeted the recommended -6dB. My final version had a single peak of about -4 dB, but never got above that. I did only minimal postprocessing of the file (e.g., high-pass filter at 20 Hz, Normalized, and silenced between tracks).

Interesting story about the blind A/B testing. I’m pretty sure I could tell the difference between the two, but maybe I should have someone else listen - or do a blind test on me! The digital file I’m comparing with the phono output is the 32-bit .wav file (no mp3 compression). I also created 16-bit and 24-bit versions and can’t tell the difference between the three digital files. I have listened to the digital conversion in three different ways. First, simply playing it on my computer and listening through the same set of headphones. Second, thinking my computer sound card might be the limiting factor, I put it on my cell phone and listened there (same phones). It’s essentially the same as when played on the computer. Third, I output the cell phone signal through a minijack-to-RCA cable, plugged that into my preamp AUX-IN, and played it through my preamp (same headphones). The digital version’s quality is essentially the same no matter how I play it and still doesn’t match that of the phono-preamp.

I’ll go back and review my workflow process and see if I somehow changed settings, but otherwise, I’m stumped.

Thanks again. Any other thoughts or suggestions would be welcome.


Have you tried listening to the “live” output from your album on the computer before it is recorded - basically just using the monitoring function of Audacity and/or the Line-In function of your audio card, and comparing that to whatever source you are thinking sounds better?

Obviously, if you are monitoring the sound output on the computer from Audacity BEFORE the recording, and then the playback of the recorded .wav (again from Audacity, and before exporting), and they sound different, then you can be comfortable that whatever difference you are hearing is being caused by the actual recording process in the program instead of something else along the way.

Good suggestion. I’ve monitored the Audacity feed via my computer, but wasn’t focused on the audio quality as much as setting the proper recording levels. I’ll give that a try. Thanks!