In the past I’ve chosen I’ve often kept AC3 audio tracks thinking that their substantially higher bitrates made them better than the AAC tracks I compared them to. As I’ve since learned that AAC can be comparable to AC3 at a substantially lower bitrate, to have a means of comparing the two codecs, what would the AAC-equivalent bitrates be for 224kbps and 640kbps AC3?
You must log in or register to comment.
Definitely needs some more testing as those numbers in the table look low to me. Archiving vs listening is a key point and storage is relatively cheap.
Listening on cheap Bluetooth headphones then you can get away with a low bitrate. If you are using better gear and/or music that has a lot of dynamic range and comes from strings (piano/classical/etc) then I would recommend 320 but your mileage may vary.
Listen, compare and decide. I have a library of about 110k tracks (musicbrainz-Picard is amazing) and I wish I had:
- Focussed more on encoding albums rather than just songs and
- Had 320 or flac
I’d honestly suggest using your ears and deciding if you can tell the difference. My music library is under 20GB so I do not fuss too much, but I am not truly sure I can tell the difference.
Hi Zedstrain, I’m guessing we’re talking about movies’ audio tracks. Passthrough is always best obviously, but if compressing, why not opus? Should be at least slighly better than AAC, it’s constantly improving and thanks to its licensing the encoders you’ll find included in any up-to-date software like Handbrake should work really well.
With AAC (which does sound great and is very close to opus) you’re gonna see bigger differences depending on the encoder. I only used it for music and podcasts (when I got into modding classic iPods 😂), I used to use fre:ac because it included good encoders like FDK.
Anyway normally 224kbps is plenty with AAC, for stereo. Around 200 with opus or AAC I don’t hear the difference with flac (music).
For movies you have to consider the number of audio channels, if you have a surround sound setup. Migh be safer to go a bit higher for 7.1 or more channels, but I have no direct experience.
As I don’t want to reduce the quality of an already lossy codec, I’m instead comparing identical audio tracks of the same release that differ only in their codecs and bitrates. For instance, would a stereo 224kbps AC3 audio track be equivalent to a 128kbps AAC audio track, or is one of the two better than the other?
Sorry I edited my other reply heavily because I noticed later that you were interested in some exact bitrate numbers… I don’t know enough about AC3 to know an equivalent number, all I can say is those numbers I’ve written for opus and AAC are in my experience enough to enjoy any movie.
From a quick look on wikipedia, looks like AC3 does not support VBR. That is enough to make AAC twice as good at least, especially since movies have a lot of silence in them, so your ratio of 1:2 equivalence seems right to me
As others mentioned having a good encoder is an issue for AAC. And some skills in using it, tuning, etc.
Nearly all quality releasers now use AC3/EAC3 or FLAC. Tigole is the last one who uses AAC to my knowledge and the rest of the QXR group rolls their eyes at it.
You’re not going to get a meaningful reduction in bitrate and file size with AAC over EAC3/AC3 without loss of quality. We’re talking maybe you can shave 2-300kbps off an AAC version versus an AC3 5.1 track. And it’s tricky. So much so no one other than that one person I mentioned bothers. At least no one accepted in the higher echelons as competent in creating acceptably transparent encodes.
If a source has EAC3 (itself capable of up to halving the bitrate required vs AC3) or AC3 I’d recommend keeping it as they tend to already be efficient. They’re also universally compatible as codecs. Re-encode those big 1500kbps DTS tracks and those even bigger monster lossless Dolby and DTS tracks but I’d leave efficient codecs like AC3 alone.
That said it’s up to you what sounds good. If you’re using lower end stuff and can’t tell the difference after trying a few different test videos with different types of sounds then go for it.
