Skip to main content

Just purchased an ERA 100 and Roam.  As I already have a gen 1 setup with multiple Sonos devices, I created a 2nd Sonos account/system for the ERA 100 and Roam.  I can successfully stream Pandora, SiriusXM and Sonos Radio to both devices with no problems.  However, when I try to stream my Apple Music Library to either device, most songs are highly distorted - sound is muted, vocals are very faint.

I don’t actually own an Apple Music account - I simply use it to locally store my music which consistes of ripped vinyl and CDs ( > 800 songs ).  I’ve noticed that songs in the library encoded as MP3s stream OK, but those encoded as m4a - AAC codec, 44,100 sampling rate, 32 bit sample depth, Apple MPEG-4 Audio (m4a) container - are distorted.  Sonos says this format is supported, but my experience says otherwise.   I have the latest IOS versions on an iPhone 13 and the latest Ventura on an M1 MacBook Pro.  Further, my entire music library streams flawlessly from either IOS or Ventura to my Bose noise cancelling headphones and to my Marantz NR1200 receiver over AirPlay 2, which then puts out sound to a Sonos gen 1 ceiling speaker.

The web appears to be full of unhelpful information.  I’m hoping someone here can help

Check out the supported audio formats in this article and make sure your music files are within the maximum supported bitrate and bit depth by file type:

https://support.sonos.com/en-us/article/supported-audio-formats-for-sonos-music-library


Bitrate is 260kbps, sample depth as noted in my original post is 32 bits


Convert the files to 24- or 16-bits.


Any suggestions on how to that, other than re-ripping everything?


Download Audacity for free. Open the music file in Audacity and save it at 24- or 16-bit. Test one music file first to see if it works before doing all of your music files.


The confusion mounts.  I’ve just learned that the AAC codec, as well as other lossy codecs like MP3, don’t have a “sample depth” or “bit depth”.  Instead, for lossy codecs what’s important is the maximal bitrate (Sonos says 320K).  My audio files are all < 320K bitrate, AAC encoded and stored in an m4a container.  So what’s wrong?


The audio files that play distorted on Sonos ERA 100 and Roam speakers used an AAC codec, wrapped in an m4a container.  Since I no longer have the original transcription from vinyl (created at 32 bit depth, sampled at 44,100 Hz), I tried modifying the m4a files to other formats, including MP3 and AAC/m4a, both at a bitrate of 320K or less.  Nothing worked.  I have concluded that to play on my Sonos speakers, I must re-rip and re-process all my vinyl-originated music.  What makes this frustrating is that these files play perfectly (no distortion) when streamed to bluetooth headphones (Bose noise-cancelling) or a receiver equipped with AirPlay/Bluetooth (Marantz NR1200) with wired output to a ceiling speaker.  Sonos is clearly inferior to the competition in supporting AirPlay/Bluetooth technologies; they need to up their game.  BTW, all my audio processing was done on a Mac using Amadeus Pro.


Hi @LifeinSlowLane 

Welcome to the Sonos Community!

Although lossy codecs have no bit depth internally, in order for them to be played they must be decompressed back into PCM (in memory only, not on disk), and the codec will generally (probably always, but I don’t know for sure) decompress to the same bit depth that was present in the source PCM audio. Therefore, I recommend you follow @GuitarSuperstar’s advice - the software mentioned will decompress the files into PCM, resample that 32 bit PCM to 24 or 16 bit PCM, then re-encode the result into a lossy compressed format again - one that can be played successfully on Sonos - and save it to your music folder. Assuming you use the correct application settings, that is.

I hope this helps.


Hi @LifeinSlowLane 

We posted at the same time!

32 bit audio is basically a waste of storage space and computation. The difference between 16 bit and 24 bit is barely perceptible, and the difference between 24 bit and 32 bit will be even less so - arguably completely imperceptible.


There are technical reasons why 32 bit recordings are convenient in the studio, but no one has been able to prove in a peer reviewed study that 32 or 24bit uncompressed distributions to the end user sound different from 16bit. Many listeners are sensitive to the differences between methods and levels of compression and uncompressed files.


Thanks @buzz - I should have mentioned that there was no point in using 32bit at the end-user stage.