Use OGG instead of MP3

For those with CD or LP or cassette digital copies, try exporting to 128 constant MP3, and some other slightly higher bitrates.

Then try export to OGG at quality level 0, 1, 2, and 3.

Those bitrates are in averagr:





I’m guessing on what quality 3 is, but that’s what audacity forum is reporting.

OGG supports higher sampling rates past dvd level 48, smaller file sizes that sound significantly better, and is open-source

1 Like

If you are exporting for archival it’s better to use FLAC, which is also open source. FLAC is lossless which can be later converted into Vorbis or other smaller formats if you want.


I second this. For storage you want as close to master quality FLACs as you can then you can transcode what you stream to save on bandwidth.

Hdd space is much cheaper than bandwidth particularly when you get 25 dollars back on tmobile unlimited for staying under 5gb and considering unlimited data at home is still a paid option with most ISPs where a zfs array is relatively cheap for what you get

Putting my money where my mouth is. I have 5TB of master tier 24b96k FLAC audio for all the music and artists I like. I transcoded it with navidrome when I play it on tempo on my phone


Third for FLAC for archival purposes, but convert to whatever sounds good to you for bandwidth/mobility purposes.

I’ll be re-ripping my entire music collection in FLAC eventually (it’s a fucking lot; over 3,000 pieces between vinyl, cassette and CD), then making AAC 128kbps copies for my iPod Classic for on the go.


I haven’t heard of ogg being used since back in the early 2000’s when Unreal based games were trying to push it. lol. Sure it sounded a bit better, but the compatibility was terrible then so it never caught on

1 Like

That’s cool, but I would suggest at least 160 kbps–I know, AAC is vastly superior to mp3, but just seeing the number 128 bothers me. At least with 160, you’ll know for a fact that it will sound much better for not much more space used.

OGG on quality 2 (I haven’t tested 3 yet) is incredibly good, and isn’t even passing 100 kbps average! So I’m sure 128 must be right there at “good enough” but why not go a few steps past that just to have the additional quality?

Now I’m going to test aac against ogg if it’s in audacity.

1 Like

If you have any flacs, convert one to mp3 128 constant (average and variable 80-135 sound even worse) and then convert to OGG at quality 0 (64-74 kbps average).

It doesn’t just sound a bit better, it’s SIGNIFICANTLY better.

For more proof, run spotify with data saver enabled, which sets audio quality to low, about 64 kb average, but it’s VERY difficult to hear the difference between that and “high quality” level. I’n very surprised.

Now, after that, play the same artists on pandora with the free account which is 64 kbps mp3 files. They’re awful.

1 Like

Both OGG and MP3 are legacy formats. Today there are so much better alternatives - especially FLAC.

Sure, you can fit 5x-6x more space with OGG or MP3 compared to FLAC, but that is about the only drawback. If the average MP3 is 4MB, that means 1GB can host 250 songs. 1TB SSD drives cost $60 and up, that is 250 000 songs. With FLAC, that shrinks to “only” about 45 000 songs, still a ridiculous amount of songs. If each of those 45 000 songs are 3 minutes long on average, that is about six years of constant listening.

Drive space is cheap enough that non-FLAC storage no longer makes sense. Encoding still makes sense due to bandwidth reasons though.


Some additional explanation:

Ogg is a container format, it doesn’t affect audio quality.

Vorbis, mp3, speex, flac, opus, are codecs that do dictate the audio quality. They can all be stored within an ogg container.

Vorbis was a competitor to mp3 and was commonly found in an ogg container, it was notable for sounding as good or better (opinion) than mp3 while not being encumbered with patent or royalty concerns.

However vorbis and speex (a codec specifically designed for low bitrate voice) have been displaced by the better performing opus.

Opus would be my choice today for a lossy format, I note many (most?) of the YouTube videos I watch are using opus for the audio encoding.

Ogg, vorbis, flac, opus, and a few other codecs, are all under the umbrella of the foundation. It is one of unsung (or lessor sung?) heroes of the Internet.

Hydrogenaudio Listening Tests - Hydrogenaudio Knowledgebase was what I used to go-to for picking codecs, but I am getting rather old and I see it’s a bit dated. I also see apparently the mp3 patents expired years ago (Wikipedia), but trust me, they were a big deal back in the day!

Doing your own blind A/B tests is fun, as an aside, and a great way to make decisions on codecs.


Seconded, Opus is great and a worthy successor from Ogg Vorbis (I remember the days when to fit decent quality music on my Windows CE PDA was Vorbis at really low rates, or whatever Windows Media Audio’s (WMA) low-bitrate options were - something similar to how HE-AAC does thing).

1 Like

I did a lot of testing about 20 years ago with AAC vs MP3. AAC 128kbps was equivalent quality to MP3 360kbps at a significantly smaller file size. AAC 128kbps vs AAC at any higher bitrate didn’t sound any different to me, so that’s why I personally landed on 128.

1 Like

Audacity doesn’t seem to provide any quality level settinga for OGG opus. I am on audacoty 3.1.3, so maybe a newer version has updated this? Why not include quality options for both?

1 Like

At the risk of triggering everyone here, I use WMA lossless to rip CDs and used WMP to play them (through my - to me - extremely baller X-fi extrememusic sound card and 5.1 speakers). I looked into flac once, but it wasn’t compatible with anything I used (and the file sizes were bigger IIRC). Way back when I had those early MMC/SD card MP3 players I’d use max bitrate WMA for those, sounded much better than MP3. Once I got an ipod though, I just put the lossless straight on there. Plenty of space. Although it had to convert it to “apple lossless” because… apple things. That took a good couple days the first time converting all of the existing library.

These days since that sound card is no longer useable (PCI) and I have not found a replacement (everything seems to be about driving headphones these days), I have sadly not really used the PC as a kickass sound system like I used to. “Onboard sound is good these days” is effin LIES, tried that and it sounded like MUD.

1 Like

I remember using WMA lossless for similar reasons back in the day. Of course any lossless format can be recoded into any other without, well, loss…

Sound cards kind of lost their reason of existence due to any sound effects being done in software now. The best sound is gotten via USB or optical to a DAC. Or HDMI to a reciever/processor if you need surround. Just extract the digital signal from the PC and do the rest outside the case with less electrical noise going on.

1 Like

Not a very fair comparison there. Why not use a VBR preset for MP3 (lame) encoding as well? lame --preset standard will give pretty high quality. Even if you want a CBR, use a --preset to specify it, otherwise you need to specify other options like -q 2 as well to get decent quality.

If you want the highest-quality lossy audio, try Musepack at, e.g. the --standard preset.

1 Like

Things have changed. Soundcards are basically extinct. If you care about quality at all, you either use a digital audio output (SPDIF, TOSLink, or HDMI) from your PC to a receiver (or soundbar). Or a USB connection, which (like HDMI can) provides both digital audio AND power to your PC speakers/soundbar/headphones/etc. in one, and which completely bypasses any soundcard you may have in the PC.

1 Like

If quality is your only concern, use DSD (or some other PCM format)
If you care about quality but space is to remain reasonable, FLAC.

for “normal” people: MP3 320kbit CBR at 44.1kHz sample rate or OGG Vorbis with >128kbit/s at 44.1kHz sample rate.

Here, quick test to decide if you need to care at all:

1 Like

It was very validating to get a perfect score on that test just now :laughing:

In any case, I use AIFF because that’s what my iPhone and iTunes support! Linux of course supports any format. If you really must have a compressed format, I agree that AAC is going to sound noticeably better. But with mass storage as cheap as it is today, it’s more a matter of compatibility with your devices.

If you have a OSX machine and are looking for a good audio encoder that will transcode your metadata and album artwork, I always use X Lossless Decoder (XLD) for my files.

Although since Adobe Audition (which I use for digitizing vinyls), my phone, and my PC all support AIFF, I tend not to transcode at all and store my files as 24 bit 96 KHz AIFF.


The way I have to do it now is the fiber thing (I think that’s SPDIF, but I’ve also seen it called TOSLink so IDK) to the logitech speakers which I guess have a “receiver” of sorts, but it sounds like ASS for music, and does not actually result in surround for games. As near as I can tell it just clones stereo to the rears. To me it makes much more sense to have the computer that’s generating the sound decide what it should… sound like, and the speakers just play the signal they get.

1 Like

The problem is quite likely with your logitech speakers. Perhaps there’s some settings you can play with, or perhaps you just need something better.

You can still do processing on the sound before your computer outputs it. And whatever amplifier you fed it into, then or now, has a lot of control over how it will sound coming out of the speakers.

1 Like