1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

FLAC vs. 320 Mp3

Discussion in 'Sound Science' started by icedup, Sep 7, 2011.
17 18 19 20 21 22 23 24 25 26
28 29 30 31 32 33 34 35 36 37
  1. Achmedisdead
    If you have space concerns, you should really consider using LAME v0 (highest grade VBR) over 320kbps. You'll never hear a difference, and you'll save space. CBR is so 20th century! 
    When I am doing my lossy conversions of the CD's I own (and other FLAC files I've come across in the wild) for my portable devices, I use LAME v2 for most, and v0 for the absolute favorites (and that's more for peace of mind than anything else). That's how I have 8,648 tracks on my 80GB iPod with a few GB of space to spare!
  2. chewy4
    Just keep in mind if you're doing testing with 128kbps - volume matching using replaygain is necessary. The gain on a 128kbps file is brought down at least a little over a half decibel usually.
    In my experience some files it can be surprisingly difficult to tell the difference, others there are some things that give it away fairly easily once you find out where to look. Overall sound is still surprisingly good either way.
  3. bigshot
    It's really good to read up about how codecs work, read accounts of controlled testing and try to determine if you can hear a difference yourself. If you do that, there won't be any doubt. You'll know the answer.

    I learned the most about codecs when I took the same piece of music and encoded it at a wide range of bitrates. The way a 96 file is compressed is quite different from the way a 192 is compressed. Below 192, high frequencies are cut to reduce file size. At 192 the frequency response goes almost up to the limits of hearing. By 256, it selectively allows frequencies that go all the way to the limit. Above 192, there is no overall filtering going on. It's entirely dependent on momentary artifacting. If you get past the point where the sound goes splat, you're home free.
  4. waynes world
    Really good info - thanks! I have seen reference to LAME v0, but hadn't yet gotten around to figuring out what it was.
    The sadly ironic thing is that when I last ripped my cd's a few years ago using my (now gone) linux box and grip/lame, they were done using VBR (and it might have been LAME v0 for all I know). But I recently got the grand notion that my mp3's weren't at the highest quality, so I very recently used foobar2000/LAME3.96r. I couldn't seen an option for VBR, so I used the "highest quality" setting which was 320kbps. I wasn't thilled at how much more space they were taking up, but I figured that's the price one pays.
    I just now went back in and figured out how to set quality down a notch and that magically comes up with - you guessed it - LAME v0 (approx 245kbps). Sigh! I wouldn't care except that I have a stinking small 32GB microSD card in my clip zip, and I can't fit everything I want onto it. I figured I'd have to get a 64GB card, but that wouldn't do it either. But possibly ripping at LAME v0 would get me quite a bit closer???
    I just re-ripped Jeff Beck's Who Else album at v0. The 320kbps version uses up 123 MB. The v0 version uses up 90MB, which is 73% the size. That's significant.
    As long as I can be sure that the LAME v0 VBR versions sound the same as the 320 CBR versions, then I might see some re-ripping in my future. And then I will have to be asking myself... should I also rip them in flac at the same time for archiving/listening on my PC purposes? No no no - I've already gone through that discussion with myself enough times lol!
    Anyway, thanks again.
  5. Dillan

    I would highly recommend keeping a FLAC copy of everything. If for no other reason than for pure archival. It is a way to keep the entire album in tact, without cutting sound waves or tampering with the original recording. It is also a way of being future proof, in case some new technology comes out later on. This is me giving advice without having any influence in opinion whatsoever.
  6. waynes world
    Well, I found my older VBR versions on a backup, so I won't have to re-rip if I want to use them. Cool.
    The older VBR versions are MP3 VBR V0 (LAME3.97) (done via linux and grip). The ones that I just did are MP3 VBR V0 (LAME3.964) (done via win7/foobar2000). What I don't quite understand is why the older ones have slightly higher kbps than the ones that I just ripped. This shows both versions for one of the songs:
    And I also see a pesky "joint stereo" turned on - time to learn more about that as I see it has been discussed in this thread.
    In foobar2000, the next setting down from highest 320 is "~245kbps" as per this image:
    Where is the "~256kbps" option! Possibly the difference doesn't matter. Probably though these are questions for another forum lol!
  7. bigshot
    There really isn't any reason to encode CBR below 320. VBR organizes the file better to dynamically assign extra bitrate where it's needed and cheat it from places where it's not (like silent spaces). When you encode 320, VBR will save a little bit of filesize, but it won't help the sound quality, because the bitrate can't go above 320. However, if you encode 256 VBR and the sound could use a little more oomph to encode properly, VBR will boost it up to 320 as long as it needs it.
    In my experience, LAME 256 VBR v0 is totally transparent. If that bitrate gains you a little extra space for music on your DAP, you should use it.
    P.S. To answer your question... It's reading approximately 245 because of the VBR. It's calculating a little bit of bitrate savings you gain over 256 from having VBR dynamically assign bitrate as it's needed. That's approximate. If you look at each file individually, you'll see the exact average bitrate.
  8. Achmedisdead
    The difference doesn't matter. The bitrate will go up or down as needed by the music. 
  9. julian67

    You issued a condescending command to me personally, also an unfounded assertion that one of the differences I described was due to not choosing correct encoding settings. Instead of addressing the points made, you instead suggested I was acting in ignorance. That is ad-hominem. It was also a strawman and an evasion, with other logical fallacies implicit. I pointed this out and now you're using it as the basis for yet another evasion (wrapped up in a threat to decline to engage) while asserting that I'm making a personal attack. That's pretty rich. If you don't want to address the points that's up to you.

    Not true. Discs are easily damaged by repeated handling. They all cost money, some being quite expensive new. I also have CDs which are no longer available to purchase new and which fetch high prices used. There is a huge practical advantage in having a lossless rip on hard disk. It serves as back up. If I want to use a particular lossy format for any purpose or device I don't need to touch the CD, I can conveniently and quickly make a transcode from lossless source. It doesn't matter what new formats or protocols might appear in the future I always have the lossless files to hand. If next week some new lossless algorithm is invented that can save me 100GB of disk space I'm ready. No Ripping Required! And best of all I actually listen to the music knowing that what I'm hearing is guaranteed the same experience as if I was playing the CD. That isn't a claim, an assertion, a wish, supposition or belief. It is rock solid, copper bottomed, actual fact and is irrefutable. No prizes for spotting the difference.

    Not true, only a claim repeated. There are samples which present problems for all lossy encoders. Nothing is guaranteed with lossy encoding. You can't claim universal tranparency, you can only claim that you personally don't hear a difference on your equipment. Your tests can demonstrate nothing more. That is their absolute limit. Mere assertion carries no weight, and repitition does not alter that.

    Next is The Big One! Telling people how great your ears and sound system is.....Oh dear...an assertion of the legendary condition of Golden Ears! Golden Ears are not only useful for those people who can hear which direction electrons move down their $1000 cables, they can also be used to make claims about lossy audio. Who knew? You have my grateful thanks.

    Now you're simply talking again about convenience. The supposed "fact" that flac playback is somehow inconvenient on a Mac doesn't speak to sound quality at all, it only says something (perhaps) about the Mac OS. This is simply about your personal situation and says nothing that can be extrapolated to the general. All my personal players support flac and play it using less battery power than any other format. So my particular situation is not your particular situation and each person sees a different range of positive and negative qualities when it comes to convenience and hardware support.

    If versatility and convenience are king then that's fine, it makes sense to do what you do. But your idea of convenience is not necessarily the same as the next person's. Nothing supports your claim that "there is no loss of sound quality". The furthest you can go with your claim is that it sounds golden to you, with your "good ears and a kick ass sound system".
  10. bigshot
    i'm sorry. I'm not paying attention to your comments any more. I won't be trolled. You can cheerfully ignore me.
  11. julian67

    I didn't notice this until just now.


    Wow! Again!

    That is so wrong it's actually incredible. Laugh or cry?

    If it was true that "no audio is lost" then the reconstructed waveform on decoding would be identical to the original waveform present before encoding. Anyone can check this for themselves in the space of a minute or two.

    It is audio data and only audio data that is discarded. It's not dark matter or mystery padding cunningly sneaked onto the discs at the CD pressing plant. This isn't complicated and it isn't opinion, it's simple, plain, VERIFIABLE, fact.

    The whole basis of a psychoacoustic model to make a lossy encode is that you do discard audio (lots of it!) but in such a way that (you hope) the difference between the original and the compressed result is not apparent when listening. Lossy is a perfectly good description: it isn't (or shouldn't be) a pejorative term, it is descriptive of the process and the result

    Anyone can verify how lossy compression works. There is a very concise description at Hydrogen Audio Wiki - Lossy or for longer description see https://en.wikipedia.org/wiki/Lossy

    I notice that since I started writing this post I have been issued a memorandum by bigshot (great name btw, really fits). For objecting to being addressed as a subordinate and also for that most heinous of debating crimes, pointing out deficiencies in the other party's arguments, I am to be punished by his refusing to engage with me. I think that's OK because a debate should at minimum be grounded on plain, verifiable fact, and a useful exchange requires good faith. If these are absent then there isn't any possibility of a conversation, only some hot air.
  12. Dillan
    As rough and passionate as Julian67's posts are, I have to say that I agree with everything that he says. I love discussing things like this, I just wish people wouldn't act like their opinion is fact. I love hearing opinions and learning new things, but at the end of the day I am a very down to earth, matter of fact kind of guy.
  13. streetdragon
    Audio data is indeed lost in lossy. Just normally an unnoticable amount that we don't hear any loss.
    Isnt that the defintion of lossy and how it is different than lossless?
    Currawong likes this.
  14. Achmedisdead
  15. Dillan

    Yes, completely agree. Data is indeed lost. Normally you cant, well I cant, notice a difference. I can agree to that.
17 18 19 20 21 22 23 24 25 26
28 29 30 31 32 33 34 35 36 37

Share This Page