Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.

Poll

What lossy format(s) do you use on a *regular* basis?

AAC or HE-AAC v1/v2 (.m4a, .aac…)
LossyWAV + lossless (.lossy.flac, .lossy.wv, .lossy.tak…)
MP3 (.mp3)
Musepack (.mpc)
Ogg Vorbis (.ogg)
Opus (.opus)
WavPack Lossy/Hybrid (.wv)
xHE-AAC (USAC)
I don't really use any lossy codec on a regular basis

Voting closes: 2022-01-02 02:34:16

Topic: 2021 Format poll [Lossy Formats] (Read 7386 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Re: 2021 Format poll [Lossy Formats]

Reply #50
LAME MP3 @ V5 because it still works.

Can you switch my vote to Opus, I started using it @80kbps for stereo and 32kbps for mono. It sounds pretty good, really Impressed.

Re: 2021 Format poll [Lossy Formats]

Reply #51
@erick128
You should probably contact some of moderators for that.

Re: 2021 Format poll [Lossy Formats]

Reply #52
The forum software does not allow editing votes, it would require manually fiddling with the database.

Re: 2021 Format poll [Lossy Formats]

Reply #53
9439]
LAME MP3 @ V5 because it still works.

Can you switch my vote to Opus, I started using it @80kbps for stereo and 32kbps for mono. It sounds pretty good, really Impressed.
The forum software does not allow editing votes, it would require manually fiddling with the database.
As per Kode's reply, you can always vote next year.

As it's been agreed upon before by some of us, these yearly polls aren't meant to precisely defy this  community's opinion as a whole, to be even thought of as precise.

IIRC, even Igor C. (the OP) said so before.
Listen to the music, not the media it's on.
Wavpack -hb4.3

Re: 2021 Format poll [Lossy Formats]

Reply #54
I settled on Opus 256kbps for my mobile library. It automatically gets resampled to 48kHz at encode time which is perfect for Android, and the bug where any files using the Ogg container didn't have their year/date tags read correctly is seemingly fixed with Android 10, meaning I have no reason to use Apple AAC anymore. Besides even when it comes to high bitrate transparency, which afaik isn't what Opus was designed for at all, Opus still wins out over other lossy codecs according to IgorC's listening test.

Re: 2021 Format poll [Lossy Formats]

Reply #55
So far I generally stuck with MP3 320CBR and FLAC.
I played with 'killer samples' lately to find out if converting a few of the FLACs would be okay for saving space, and so far OPUS does not look good. Out of MP3/AAC/OPUS OPUS is the easiest to recognize for me so far, at high bitrates. MP3 is hugely better than I anticipated. I happen to enjoy music with a lot of 'killer samples' so my goal is to get those fully transparent. Gapless playback is also still an issue with OPUS that disappointed me (see the threads here on that). AAC's gapless is also a mess, and hardly works anywhere. FB2K can do it though. MP3's works very well.

At lower bitrates OPUS is great though, but I see no reason to go lower than 320 or maybe 256. 320 or V0 has been standard for quite a while now.

See this thread if you want to get anxious about lossy codecs: https://hydrogenaud.io/index.php?topic=120193.25

Re: 2021 Format poll [Lossy Formats]

Reply #56
"Out of MP3/AAC/OPUS OPUS is the easiest to recognize for me so far, at high bitrates."

What is the high bitrate, specifically. Numbers please. And give us some ABX results with the killer sample, please?

Otherwise your post violates https://wiki.hydrogenaud.io/index.php?title=TOS_8, i.e.

Many many multiformat listening tests on this forum showed for example that Lame MP3 is inferior to OPUS significantly. You cannot find one that contradicts that. So how can you pick out OPUS more frequently or more easily than MP3 at the same rate. If not the same rate you do not compare like with like. Unless you have a magic MP3 encoder we do not know about. Normally Lame is used here, but even Helix one is just narrow margin better than Lame, so we do not know a MP3 encode that is not significantly inferior to the stock OPUS encoder.

So all in all what you say sounds nonsense.

Re: 2021 Format poll [Lossy Formats]

Reply #57
No, it does not. There are hardly any listening tests at high bitrates, and it's well known that certain samples trip up encoders if you get out of your bubble.
Harpsichord and chiptunes for one opened my eyes. I can't un-hear the difference.
I can reliably ABX my test tracks at 320 OPUS VBR - the MP3 and AAC sound better. But other samples trip up AAC.

Even though you prefer to insult rather than to search I will provide one for free: https://hydrogenaud.io/index.php?topic=120193.25

After reading dozens of HA threads I found various technical explanations for that as well, and how it's impossible to be 100% perfect at all times. Depending on one's music choice that is a problem. There are modern chips that aren't shared as source for instance, and my format of interest, 320 OPUS VBR, does not like them at all.
Classical and baroque are nicer in AAC to me, but AAC trips up on other samples as demonstrated above.
Ergo: there is no perfectly transparent lossy format that will handle everything. And don't even think about experimental music. It's optimized for 'typical' music.

Homework: type 'chiptune' on HA search.

Re: 2021 Format poll [Lossy Formats]

Reply #58
As this forum does not allow editing posts after a while I have to post again. Here's a developer explaining one technical reason for this:

The killer tracks for Opus that I have found are harpsichord test tracks (probably something in the codec is just plain wrong for those, but I don't know what).
As far as I understand, It is not Opus implementation is wrong (in some ways) but it basically comes from various kind of technical limitations due to being a low delay codec. Especially, it's MDCT window size is limited. I'm surprised how well Opus is doing despite of those limitations.
Read chapter "Tonality Estimation" at  http://people.xiph.org/~xiphmont/demo/opus/demo3.shtml

Like I wrote in my initial post, my intent is to get everything I can throw at it perfectly transparent, hence my willingness to just use 320 kbit/s, something that is extremely rare to find in listening tests, as it's overkill for most music. It's not annoying for casual listening, but if I have to archive lossy I want all edge cases covered.

Also interesting to note: the random chiptune I ABX'd resulted in a bloated 431 kbit/s OPUS at 320 VBR, another indication how it does not like that track.

Re: 2021 Format poll [Lossy Formats]

Reply #59
You have not provided ABX where you prove that you can ABX 320kbps Opus. You are just peddling unsubstiatied claims. There is no room for that here Again, TOS8. I hope forum admin will shut you down.

Your ramblings on chips/AAC/Opus also does not improve your standing, either.  You do not come across someone who knows what he/she talks about.

Re: 2021 Format poll [Lossy Formats]

Reply #60
Like I wrote in my initial post, my intent is to get everything I can throw at it perfectly transparent, hence my willingness to just use 320 kbit/s, something that is extremely rare to find in listening tests, as it's overkill for most music. It's not annoying for casual listening, but if I have to archive lossy I want all edge cases covered.

Also interesting to note: the random chiptune I ABX'd resulted in a bloated 431 kbit/s OPUS at 320 VBR, another indication how it does not like that track.
Obviously with no idea of how long you've been perusing HA before creating your username quite recently, but as you see yourself in the position to not only make non-TOS-8-compliant claims plus recommendations of posts for others to read (who like most here, have quite possibly come across them already), most will agree that you should have known better by now that people will just skim/avoid your posts and dismiss such unfounded claims as a truckload of BS!

So, friendly advice: publish your ABX test results to back them up or else keep on behaving and being seen as the proverbial village madman.
Listen to the music, not the media it's on.
Wavpack -hb4.3

Re: 2021 Format poll [Lossy Formats]

Reply #61
Nero AAC SBR q0.18 (~41kbps). Yep, a prehistoric monster, down in the abyss. :D
Reasons: space constraints (1427 hrs in <32GB!), software support (bye bye USAC), disappointment with other codecs or encoders (SBR+PS, Opus, FhGaac, FDKaac, QTaac...)

Re: 2021 Format poll [Lossy Formats]

Reply #62
I use Spotify ("Very High" quality), so Ogg Vorbis 320 kbps.

I wish Spotify switched to Opus so I could change quality to "High" (160 kbps) and save extra space...
sox -e float -b 32 -V4 -D gain -3 rate -v 48000 norm -1
opusenc --bitrate 128

 

Re: 2021 Format poll [Lossy Formats]

Reply #63
From foobar:

Code: [Select]
Duration : 2wk 1d 7:49:30.373
Avg. bitrate : 412 kbps
Codec : AAC (37.2%); MP3 (29.9%); FLAC (24.1%); Vorbis (6.5%); Opus (2.2%); MP2 (0.0%)

Re: 2021 Format poll [Lossy Formats]

Reply #64
Apple TVBR 73 Highest Quality (slowest encode) AAC mp4. For convenience, codec support/compatibility and efficiency, I’ve been impressed by the compression vs quality that I’m reconverting my entire lossless album images.

Re: 2021 Format poll [Lossy Formats]

Reply #65
I'm using qaac (tvbr 91) for my car with applied replaygain. I'm also using qaac for audiobooks (mono and tvbr 36). I used to use mp3 V2 and V9 for that.
I still use musepack -5 for devices that support it, like my phone and rockboxed sansa zip clip. But i might replace it by qaac for convinience, even though i think musepack might still be slightly better.
Aac has completely replaced mp3 for me.

 
SimplePortal 1.0.0 RC1 © 2008-2021