Audacity was recommended to me by a friend when I told him I was looking for a program to use to convert a large WAV library to FLAC and MP3
I should mention many of the CDs were imported at 16bit 48khz instead of 44khz due to my ignorance at the time.
I have left all Audacity project settings at default except for turning off dither since the import files and export files will both be 16bit (32bit Float, 44100, Real-time Conversion Best Quality, Dither None, High-quality Conversion Best Quality, Dither None). I am looking to get the best possible audio quality out of each export. Here is my planned workflow,
1st Question - Is Audacity actually changing the bit depth from 16 to 32bit (32bit float) then back down to 16bit for my FLAC exports? If YES should I turn Dither back on to Shaped or leave Off for best quality?
2nd Question - Even though the source WAV files are 48khz since cd files are 44khz am I okay to lower to exports to 44khz (for compatibility) or should I leave files at 48khz to preserve quality?
I just wanted to get some clarity from some users before I start converting these files, I have about 1900 albums to convert and I don’t want to start and realize I was doing something wrong during the process and have to start all over.
If you’re on Windows I’d recommend Kabuu Audio Converter or TAudioConverter. If you are just converting without editing, these are a lot easier to use. You just set-up your output folder and output-format and you can drag-in a folder full of files and just “click” to batch convert.
Audacity doesn’t support embedded artwork so it won’t copy it over to the new files.
Any other metadata (“tags”) may not be copied over “perfectly” for a couple of other reasons - Metadata for WAV is not well standardized, and every format has different metadata standards/formats. If you want to add or edit metadata to the new files, I use Mp3Tag (which also works with FLAC and all other standard audio formats) or there are other similar programs.
Set dither to none and will convert to 32-bit float and back losslessly. (If you decide to use Audacity.)
I’d leave them at 48kHz. You shouldn’t hear any difference either way, but it’s best to avoid unnecessary conversions and converting-back doesn’t restore the exact same data.
The 44.1kHz FLAC files will be proportionally smaller, if that’s important to you.
It won’t make a difference for MP3 file size which is determined by the bitrate, not the sample rate (320kbits per second is 40kbytes per second). Of course, any embedded artwork will make the file larger than that.
You MIGHT want to re-rip the CDs, leaving the original 44.1kHz sample rate and if you use a ripping application that supports AccurateRip, the FLAC files can be bit-perfect (as long as AccurateRip shows no errors).
Thank you for your detailed answers. I will check out the other programs later today .
I have considered re-ripping the cds though many are from the early 2000s and were kept in a large CD case in my car and over time have been scratched, in some cases badly. I will grab some of the worse off discs and try the program you recommended and see how it turns out. If it throws a bunch of errors I will just stick with the current library I have and not worry about it.
Another suggesttion for batch audio file format conversion is dbPowerAmp Music converter, which I use to convert my completed audiobook files from WAV to MP3; though it can do much more than that.
It’s very reasonably priced, also includes their phenomenal CD ripper* if you want to go that route, and has a licensed-from-Fraunhoffer MP3 converter.
(* it has arguably the best error handling of anything on the market and will often rip even badly-damaged CDs successfully)
I have no connection with them other than as a very satisfied customer, BTW.
No need to do that in recent versions of Audacity as it can now detect that case, and does not apply dither when it’s not needed.
Yes Audacity does convert to 32-bit float, but the conversion is lossless because all 16-bit values can be represented exactly in 32-bit float. Also (recent versions of Audacity) as long as you have not processed the audio (the sample values are unchanged), exporting back to a lossless 16-bit format will also be completely lossless (dither automatically disabled).
If you change the sample rate (resample) , then the new samples will no longer be exact 16-bit values - they will be more accurate than can be achieved with 16-bit. Of course this extra accuracy will be lost when you export back to 16-bit, and at this point dither should be applied so as to retain the best possible quality at the new sample rate.
In theory, not changing the sample rate (keeping it at 48kHz) is the best option for sound quality as the sample values in the exported file are exactly the same as the imported file. In practice, the difference is likely to be undetectable (still much better than any MP3).
Using a dedicated batch conversion program is likely to be quicker than using Audacity as there are less intermediate steps involved. If you are on Windows, Foobar2000 is excellent at batch converting audio formats (and it’s free, and it’s a very good audio player).
Thanks! After sleeping on it I have decided to Re-RIP my CDs with the proper format (16bit 44100Hz) with either EAC or dBpoweramp (since they have AccurateRip). I have multiple machines so hopefully it won’t take too long.
I will keep my old WAV library as a backup and a short term solution if I need to throw some music on my phone or tablet.