Skip to main content


Please be aware that much of the software linked to or mentioned on this forum is niche and therefore infrequently downloaded. Lots of anti-virus scanners and so-called malware detectors like to flag infrequently downloaded software as bad until it is either downloaded enough times, or its developer actually bothers with getting each individual release allow listed by every single AV vendor. You can do many people a great favor when encountering such a "problem" example by submitting them to your AV vendor for examination. For almost everything on this forum, it is a false positive.
Topic: Getting Real - AAC 320 vs. FLAC (Read 6840 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Getting Real - AAC 320 vs. FLAC

I have many FLAC recordings ripped from CD and also some higher bit stuff. And use Foobar 2000, updated.

Is there the slightest chance that unless I am sitting in a noiseless, soundproof room with a $5,000 headphone, a $4,000 player, and am doing nothing but A/B testings of each format with each song, that I will ever hear the slightest difference between AAC 320 converted using Foobar and FLAC?

Re: Getting Real - AAC 320 vs. FLAC

Reply #1

Re: Getting Real - AAC 320 vs. FLAC

Reply #2
I partly disagree - with the price of the audio equipment. I A/B tested some Ariana Grande tracks a few years ago with 256kbps AAC and FLAC with some $40~ headphones on my mid-range Android phone. When repeating the test multiple times (the same few seconds repeating over and over again) I was able to tell a very slight difference in the high frequencies, but it was extremely difficult and a rather pointless test. As much as I hate to admit it, unless you're very familiar with a particular piece of lossless audio (and a complicated or busy piece at that), I very much doubt you'd be able to tell the difference in less than perfect conditions.

With that being said, one time I was in the car and Tidal was playing a 48kbps track and I found it to be rather more pleasing than the FLAC equivalent copy I own - I'm not entirely sure why but I put it down to my car's audio system using some tricks  :-[

Re: Getting Real - AAC 320 vs. FLAC

Reply #3
When doing any comparison, especially codecs, you must be certain to get the levels matched - use software designed to do ABX testing and pay attention to preparation of the actual samples being tested to be sure there aren't any glitches or differences due solely to editing.

What that said, I have done (far too much) headphone-based ABX testing of some high quality music samples and once I get above 128k AAC it becomes very hard to prove I hear a difference.  I have never successfully ABXed a 256k track against the original on any music that I would normally listen to critically.

My hearing does tend to roll off above 14k, so perhaps there are some higher frequency issues that can be heard in the so-called killer samples.  The best answer is: do the tests yourself and know for sure.

Re: Getting Real - AAC 320 vs. FLAC

Reply #4
I remember participating in a 128 kbps mp3 listening test on here years ago and finding it surprisingly hard to tell a difference, outside of the low anchor.

And in theory, 320 kbps AAC should be at least as good as 128 kbps AAC, which should be at least as good as 128 kbps mp3, at least with a good encoder (read as one that's reasonably modern and isn't FAAC)

EDIT: And samples in a listening test tend to be ones that the encoder trips up on.

Re: Getting Real - AAC 320 vs. FLAC

Reply #5
Some tracks at 320 kbps can be ABX'd if you have great ears and training, but that's only because there's no ideal AAC encoder. Some encoders sacrifice higher frequencies too often, and some sacrifice side/stereo fidelity too often. (And you don't need to hear above 16KHz to hear a lack of stereo atmosphere.)

Re: Getting Real - AAC 320 vs. FLAC

Reply #6
To me it seems there is zero practical difference in the quality between say 192 and 320. it would be difficult for me to even determine which is better. To me, all these types of questions seem a little misguided, given that the difference between lossless and high bitrate lossy just seems more philosophical than anything. To me, a format is determined by the situation, compression was intended to make files smaller, so a high bitrate lossy files, become irrelevant. I just don't see the need for 320s at all. There's so little qualitative difference between even a 160 and a 320 that I don't see myself ever needing more than say 192; honestly, I don't even turn up my nose at 128s anymore. It seems that any modern lossy format is good enough for me. Despite all this, all my new music is FLAC and I never use compression any more. Is it dying?
Quis custodiet ipsos custodes?  ;~)

Re: Getting Real - AAC 320 vs. FLAC

Reply #7
I have noticed that some MP3 encoders (at least those available 5-10 years ago) sacrifice high end compared to the AAC encoders; when I re-ripped some CDs in AAC they definitely had more content above 5 KHz (visible in spectrograms too).  Not sure what they are compromising, because it wasn't audible even in headphones.  I didn't encode the MP3s, so I have no idea what encoder they were using.

Re: Getting Real - AAC 320 vs. FLAC

Reply #8
They're not usually sacrificing audible information. At least, not the latest encoders. Spectrograms aren't really a fair way to judge an audio codec, unless you can actually hear a difference and are looking for a visual representation of this issue.

Re: Getting Real - AAC 320 vs. FLAC

Reply #9
I have noticed that some MP3 encoders (at least those available 5-10 years ago) sacrifice high end compared to the AAC encoders; when I re-ripped some CDs in AAC they definitely had more content above 5 KHz (visible in spectrograms too).  Not sure what they are compromising, because it wasn't audible even in headphones.  I didn't encode the MP3s, so I have no idea what encoder they were using.

The OP isn't referring to spectrogram comparisons, but asks for possible audible differences. My first reaction still stands, with the assumption that a decent encoder is used of course.

Re: Getting Real - AAC 320 vs. FLAC

Reply #10
Is it not still the case that "all lossy codecs have some killer samples"? 

I didn't think AAC-LC had improved recently even if it's transparent (at that rate and below) with the cast majority of tracks.

If there are killer samples for AAC at 320kbps, and OP happens to select one, and be paying adequate attention at the right time, then sure there is "a chance" of hearing the difference.

Re: Getting Real - AAC 320 vs. FLAC

Reply #11
Quite a few years ago I did a test with a friend of lossless vs AAC vs MP3. What I found, for me, in that test was that for music other than classical it was effectively impossible to tell the difference between lossless and AAC at 256kbps. As a result I import everything other than classical at 320kbps (because it’s an insignificant difference in storage), and do classical lossless. I’ve imported around 1500 CD’s using this approach.

I find 160 sounds bad to me, but when I A/B my CD’s with the 320kbps files on my media server they both sound great to me.


Re: Getting Real - AAC 320 vs. FLAC

Reply #12
I have done ABX testing many times and cannot distinguish between high bitrate AAC and FLAC. Not at all. No matter how hard I concentrate.

I personally do not believe any "audiophiles" who say they can. I would bet good money they would fail in ABX texting.

Re: Getting Real - AAC 320 vs. FLAC

Reply #13
IMHO it makes sense to ask for the intent of the question.
If it's about your FLAC-archive - there's no reason to abandon this with degenerated audio, you will regret this once..

Re: Getting Real - AAC 320 vs. FLAC

Reply #15
@ OP... I would not obsess over it as 320kbps for any lossy audio is overkill and a waste of storage space. just pick a efficient bit rate and be done with it. basically one of the following for AAC(Apple)... 96kbps or 128kbps, maybe 160kbps or 192kbps TOPS for the more paranoid types.

also, here is something to think about... unless someone is really young etc, I am confident if you just start giving people random music files made @ q45 TVBR (96kbps) with Apple AAC and Opus @ 96kbps, very few would complain from a all around sound quality perspective (since you can't compare to the original lossless file and are just sitting back enjoying the music and there is no obvious hits to the overall sound quality at this bitrate on a typical pair of speakers/headphones etc) as the sound quality is easily high enough overall at that point. even for a little safety margin, 128kbps or so is plenty and still quite efficient. hell, I am fairly confident many would not mind Opus @ 64kbps since the since there is not any obvious hit to the overall sound quality. NOTE: although with MP3 I would avoid anything lower than v5 (130kbps) to be safe and with AAC (assuming we are talking standard AAC-LC) I would avoid anything lower than 96kbps as a general guideline (but with Opus, I feel 64kbps is in that conversation).

plus, once someone has to start really focusing to hear differences (especially if the core sound (like singers voice and basic sound of instruments etc) is not really effected) you know your sound quality is easily within the 'good enough' range since they are straining/nit-picking at that point. because I figure if one can't fairly quickly (say within 5-10 seconds) detect any obvious differences when sampling random songs, you know it's good all around sound quality and is not worth obsessing over the tiny details. in this regard with MP3 (which is the encoder that will fail the quickest at the lower bit rates), like where things become more obviously weak compared to the original lossless file, I feel that point is at a bitrate lower than v5 (130kbps), which makes v5 generally in the 'good/safe enough' range overall and even the general word around these forums considers MP3 @ v5 to be good, or good enough one could say.

so while I get the users around here are a bit more OCD with audio quality, I think what I said above is a more practical view of encoding bit rates for most people, if not the vast majority, especially those who are not really young, and are on average-to-above average speakers or roughly your average-ish set of headphones. this helps put things into perspective and gets one out of the OCD with absolute top quality sound when using lossy files (i.e. AAC/MP3/Opus etc).

p.s. I suggest you keep your FLAC files as you may need to re-rip to another format in the future. but if a person is sticking to only one lossy format, and not going to keep their FLAC files, then MP3 is the safest all around choice since it's the most widely supported. so if one is going to use MP3 as their only copy of music, on the side of caution from a sound quality perspective, I would just use v2 (190kbps) and be done with it as while it's not the most efficient you ain't got to worry from a sound quality perspective at that rate for pretty much everyone and there is still a little efficiency left at that bit rate. although even v5 (130kbps) is never going to be a bad choice especially for someone with a little age on them and we all are not super young for very long as our ability to hear the very fine details won't last all that long as I tend to see things more from a average or so level of hearing (and speakers/headphones) and in this regard MP3 @ v5 is never going to be a bad choice although when using bit rates around 128kbps or so and less I tend to prefer AAC or Opus if possible instead of MP3 since around that rate and lower those tend to pull ahead.

Are there more recent tests? Is it possible to have another test?

Someone correct me if I am wrong, but I think the reason no one has bothered is because things have not really changed in a long time (call it roughly 10+ years) on MP3/AAC (they are quite strong once you reach a sufficient bit rate) even though Opus has a bit in fairly recent memory. but even with Opus it seems to be mainly at lower bit rates that there have been some level of sound quality tweaks (like around 48kbps and less if I recall correctly). so anyone using sufficient bit rates for music (with Opus I feel anything 96kbps+ is quite strong and even 64kbps is usable for sure) is already high enough to where very few people would complain and has been this way for many years now.
For music I suggest (using Foobar2000)...
1)Opus @ 64kbps or 96kbps.
2)AAC (Apple or FhG(Winamp)) @ 96kbps.
3)MP3 (LAME) @ V5 (130kbps). NOTE: I am using MP3(V5) on AGPTEK-U3(8GB) as of Mar 18th 2021. I use 'fatsort' (on Linux) so MP3's are listed in proper order on AGPTEK-U3.

Re: Getting Real - AAC 320 vs. FLAC

Reply #16
I use fdkaac -v5 for the music on my phone and car.  Probably overkill, but storage is plentiful.  I was surprised how well 96kbps AAC-LC sounds.  However, when playing that low of a bitrate over Bluetooth, depending on your hardware, you can hear the quality degrade.

Re: Getting Real - AAC 320 vs. FLAC

Reply #17
Depends on how you want to use your music  For example, if you ever want to try your hand as a DJ and need to bend the pitch, it Ia a very bad idea to use compressed audio.

Unless you are bumping up against insurmountable storage constraints I really don't see why you would want to sacrifice flexibility.

SimplePortal 1.0.0 RC1 © 2008-2021