[Alpina]

Hey guys, I got some music of flac lossless format and noticed strange thing. Why sometimes .flac takes very large amount of space and sometimes not, for example:
 
Stairway to heaven: 157Mb vs 34Mb, both lossless .flac format but very huge difference in size.
GnR - Estranged: 185Mb vs 58Mb, same as above.
 
Is there any difference in sound quality between those two? By the way my Cowon i9 can't even play that huge .flac for some reason, it just freezes, no idea why.

 

[davidsh]

Perhaps the bit depth (16 vs 24 bit) varies and the frequency bandwidth (44.1kHz vs 192 kHz)?

 

[Alpina]

You are right it's 24 bit 96Khz vs 16 bit 44.1Khz.
 
But does that have any impact on sound quality? I thought lossless format is lossless or?

 

[davidsh]

Doesn't impact the sound quality. 16/44.1 is just as good as 24/192.

 

[Alpina]

  Doesn't impact the sound quality. 16/44.1 is just as good as 24/192.

Cool. But then why it's different? 

 

[davidsh]

You don't want to know the explanation
 
The bit depth translate to the possible highest dynamic range (difference between lowest and highest possible sound). Actually it means that anything below a certain dB level will be noice. 16 bit is plenty sufficient. With 44.1 kHz you can produce any sound with lower max frequency than 22kHz perfectly.

 

[Alpina]

  You don't want to know the explanation
 
The bit depth translate to the possible highest dynamic range (difference between lowest and highest possible sound). Actually it means that anything below a certain dB level will be noice. 16 bit is plenty sufficient. With 44.1 kHz you can produce any sound with lower max frequency than 22kHz perfectly.

I see. ^^
 
Thank you very much davidsh for explaining this stuff

 

[TonyVier]

Yes, flac is lossless, just as there is the tif format for images, also lossless. However, also images can be lossless, but still have another resolution. Sort of the same with soundfiles. Tif vs jpg, flac vs mp3.

What is different between images and sound files are the instruments we use to perceive them. While images with a high resolution can be zoomed in quite a bit to reveal details, we cannot zoom in the same way with sound.

 

[hogger129]

Lossless just means all the original data from the source is there.  It just stores it more efficiently which is why it's able to pack into a smaller size.  When you play back a FLAC file, it should sound exactly the same as the original source.

 

[radiofrog]

Yes, flac is lossless, just as there is the tif format for images, also lossless. However, also images can be lossless, but still have another resolution. Sort of the same with soundfiles. Tif vs jpg, flac vs mp3.

What is different between images and sound files are the instruments we use to perceive them. While images with a high resolution can be zoomed in quite a bit to reveal details, we cannot zoom in the same way with sound.

 
Good analogy.  You can compare the 2 FLACs to 2 TIF files.  
 
Say you have a 300 dpi image (akin to original audio recording), then save it as a TIF (i.e. - a FLAC).  
You can also take that original 300 dpi image and increase its resolution to 600 dpi (you haven't actually added any information, just increased the resolution needlessly), then save that as a TIF, and the file will be much larger.  It won't be better, just more cumbersome.

 

[limpidglitch]

   
Good analogy.  You can compare the 2 FLACs to 2 TIF files.  
 
Say you have a 300 dpi image (akin to original audio recording), then save it as a TIF (i.e. - a FLAC).  
You can also take that original 300 dpi image and increase its resolution to 600 dpi (you haven't actually added any information, just increased the resolution needlessly), then save that as a TIF, and the file will be much larger.  It won't be better, just more cumbersome.

 
Just a bit of nit-picking. TIFF does not support compression, as FLAC does.

Equivalent pairs would be more like:
 
JPG ~ MP3
PNG ~ FLAC
TIFF ~ WAV

 

[TonyVier]

Nitpicking you back . Tiff supports compression.

 

[pookeyhead]

What's the point of ripping a CD to 24/96 though if the  original source is 16/44.  Surely this is utterly pointless?   Someone used a photography analogy further up the thread, and they are correct. You can increase the resolution of the digital image, but there will be no more detail in it.   However...  when you up-res the image, you can increase the anti-aiasing...  the transition between pixels can appear smoother.  Is this the same with audio?  IS there a difference?  I know you can't add what's not there...  but is there an audio equivalent of anti-aliasing?


Oh...  and not wanting to be pedantic..  TIFF files support both LZW lossless compression, ZIP, and JPEG lossy compression

 

[davidsh]

As I remember, yes there is something equivalent to anti-aliasing in audio, which is already done in the studio, often at high sample rates. I'm not sure 'bout this, though.
 
DACs can sound different with hi-res files due to implementation, but hi-res is not better.

 

[proton007]

  What's the point of ripping a CD to 24/96 though if the  original source is 16/44.  Surely this is utterly pointless?   Someone used a photography analogy further up the thread, and they are correct. You can increase the resolution of the digital image, but there will be no more detail in it.   However...  when you up-res the image, you can increase the anti-aiasing...  the transition between pixels can appear smoother.  Is this the same with audio?  IS there a difference?  I know you can't add what's not there...  but is there an audio equivalent of anti-aliasing?

 
In images, interpolation techniques do the work of a low pass filter. Audio equivalent of anti-aliasing....Hmmm...all DACs have a low pass filter at the output stage. So effectively they are doing anti-aliasing, all the time.
 
Sampling at higher frequency reduces the need of a perfect cut off filter and improves S/N ratio, so I'm guessing its more useful when recording.
 
However, you can find a lot of DACs which have the 'oversampling' option.