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


Discussion in 'Sound Science' started by theawesomesauceshow, Aug 16, 2011.
1 2 3 4 5 6
  1. ev13wt

    People still do that? Didn't know.
  2. bigshot
    I use AAC 256 VBR in both my home system and on the road. Works fine. MP3 LAME 320 would probably work the same.
    NorCal likes this.
  3. NorCal
    Sounds reasonable to me.
    Ripped again to double check ... same results, using flac-1.3.2-git-20171020-x64\flac.exe
    Parameters: -s --ignore-chunk-sizes -0 - -o %d

    Used "Highest BPS mode supported" 16 when using "Output bit depth:" 16-bit,
    and "Highest BPS mode supported" 24 when using "Output bit depth:" 24-bit.
  4. Roseval
    Thanks for doing so.
    But this is not what I had in mind.
    If you take the original 24 bit recording. Is there any information below -96 dBfs?
    If there isn't , the result wil be the same as if it was a 16 bit recording.
    You can check this using a spectrum analyzer on the original file like https://www.xivero.com/musicscope/

    The reality is that I want to use the same file format in my vehicle and in my home system

    I have a different approach.
    I prefer lossless where possible.
    I use MP3 on my phone
    I don't want to maintain 2 libraries.
    I solved this by using transcoding
    My media player syncs a play list with my phone, converting on the fly
    NorCal likes this.
  5. NorCal
    All rips are from CD, so a 16 bit original recording.
    It's interesting that the size of the ripped 16 and 24 bit flac files stayed the same, (which seems like the correct way to go), while the 24 and 32 .wav files grew, even with the 16-bit recording.
    Thanks for the link. That spectrum analyzer does look nice.
    The way you're using transcoding sounds really cool, (maybe someday I'll get a smart phone). But since I'm currently using a 128G USB stick, I just copy the files from my computer's hard drive, and update, for variety, ever once in awhile.
  6. castleofargh Contributor
    your numbers make a lot more sense now. when you convert a 16bit file into 24bit, all you do is add a bunch a zeros at the end of every sample(no math required). FLAC being a compression boy, will look at the data and write the equivalent of "all the stuff at the end is zero" in the file. which doesn't take much space to write down.
    while WAV that doesn't compress anything will dutifully keep writing down each extra zero for all the extra 8 bit multiplied by all samples. in short your WAV file doesn't know that you've been recording only 16bit of signal, and the file is effectively as big as a 24bit one. it would be the same if you recorded silence.

    but in any case, there is little meaning in converting a 16bit file into a 24bit one in the first place ^_^. so that was not a realistic example. your 16bit results are more relevant to show the typical compression of FLAC.
    NorCal likes this.
  7. 71 dB
    Yes, do that. Rip .wav-files at 16/44.1 and use those files "everywhere."

    People like FLACs because they take less disk space. That's all.

    These numbers are exactly as expected. The sound quality of ALL these files is identical, so that makes 24 and 32 bit WAVs rather stupid waste of disk space with added zeros (silence).
    NorCal likes this.
  8. Roseval
    There is another advantage.
    Tagging support in WAV is haphazard.
    The WAV standard lacks a couple of tags like album and cover art.
    The alternative, writing ID3 tags in a info chunk is not a formal standard.
    A bit more detail can be found here: http://www.thewelltemperedcomputer.com/KB/WAV_KB.htm

    In case of FLAC you have both lossless and excellent tagging support including custom tags.
    NorCal and colonelkernel8 like this.
1 2 3 4 5 6

Share This Page