I've got two FLAC files of the same song.

One is 870 kbps, while the other is 864 kbps but ripped with EAC. I know that EAC does a good job at ripping CDs so I'm inclined to keep that copy. Yet, do the extra kbps matter more? In this case it's just 6 more kpbs so it wouldn't matter, but what if it was 100 kbps more, would that make a significant difference?

Is there any software that I can use to compare them?

 

you can select different compresion levels in flac. after uncompression, the audiostream is exactly the same, bit by bit.
so just think of it as compression-strength in zip or rar

 

Comparing audio sounds like you should first do a null test.
If you load both flac files into a DAW, perfectly align them, invert the polarity of one of the copies and play them back, whatever you hear is the difference between the files.

If you hear nothing and the meters don't show activity, the files are the same.
If you just barely hear something, you probably shouldn't worry about the quality of the 6kbps in/decrease.

 

6kbps could just be extra length at either end or both. the actual start/stop positioning. Agree with alesky about null test.

 

Yes just null test. Should be identical if 2 files are encode from the same source. (Think of you zip .wav file with 7-zip and winrar The result from 2 program maybe different but content inside after extract should be the same).

 

Flac is compressed lossless. It's like comparing a 32mp image with a 34mp image. What's the slight difference gonna do?
If you drag them into a DAW or wave editor and line them up and flip the polarity of one they will probably cancel out to silence (or very near silence).

 

Last edited: Dec 30, 2022

As other people have said: FLAC is lossless, so file size has no bearing on the quality. If you're FLAC compressing a single source, you can trade more CPU crunching time for a reduced file size, but the audio will be exactly the same as a quickly compressed FLAC with a higher file size.

That said: If these FLAC files are from two different sources (vinyl versus CD, etc.) then there will be a difference between the FLACs, just as there was a difference between the original source audio.

"Yet, do the extra kbps matter more? In this case it's just 6 more kpbs so it wouldn't matter, but what if it was 100 kbps more, would that make a significant difference?"

No difference at all. Go with the FLAC created from the source you trust the most.

 

Maybe Swap is talking about differences in mastering of the same track. For example if you take a pop CD single from 1989. Then compare to the greatest hits pressing from 2010 or something. Unlikely they would null.

About the flac encoder, note the maximum compression setting on earlier versions will create larger files then the newest builds. So the same source on different versions of the flac encoder will output different files.

 

If you can't ABx the files, it probably doesn't matter which you keep unless you're an archivist, and it's some extremely rare obscure media.

Extra kbps shouldn't really matter - FLACs can differ in compression level, and a lower compression ratio doesn't necessarily mean that version is higher quality - in fact, it might simply have a higher noise floor, and so compresses less.

I use Spek which shows the spectrum of a file. Though it takes a bit to familiarize yourself what to expect.

The null test people brought up, is also useful, it's a highly sensitive test (can easily show you the most subtle artifacts of a 320kbps MP3 compared to a FLAC). Though you may want to attempt to match amplitude and phase as close as possible first, since if you're off by even one sample, you just get a phasy version of the track.

 

If you have original wav or can rip to wav, just compare wav and flac file one, then wav and flac file two. Null test. Same bit depth, same sample rates. If you always get -inf dB, it means all three files are sonically same. If you have some difference, it means you do something wrong. Should not be difference there

 

When I couldn't hear an obvious difference or I had too many files to manually compare them all I've had good results with the Gar Software Similarity App giving me various acoustic fingerprint comparisons for quality, sameness, clipped samples, bitrates, true sample rate, duplicates etc.

 

Utilities to check whether a WAVE or FLAC file is truly lossless or not:
https://losslessaudiochecker.com/
https://fakinthefunk.net/en/

 

Most people can't tell the difference between a WAV file and 320 kbps mp3 file.

I wouldn't lose sleep over it. I used to get all my music in FLAC, and if possible WAV, but nowadays 320kbps mp3 is more than enough, and saves me tons of hard drive real estate. And I need it for my porn.

 

throw them both into a DAW, flip phase/polarity on one of them, and if they null perfectly, they're same, just slightly differently compressed

 

Does it really? Or is there some more underlying technical detail that allows to bypass the LP? i am quite sure I have worked with MP3s which definitely had 18k+ content.

 

The higher the bitrate the lesser LP filtering is usually applied. But it also depends on different encoder and their settings.

 

You can't bypass the lowpass filter after the fact, it's baked into the resulting mp3 file.

Here's how LAME (the only mp3 encoder that is still in widespread use these days) applies its LP depending on the bitrate target when encoding:

{BITRATE, LOWPASS_HZ},
{8, 2000},
{16, 3700},
{24, 3900},
{32, 5500},
{40, 7000},
{48, 7500},
{56, 10000},
{64, 11000},
{80, 13500},
{96, 15100},
{112, 15600},
{128, 17000},
{160, 17500},
{192, 18600},
{224, 19400},
{256, 19700},
{320, 20500}

so 20500hz is about as good as it gets.

 

Hold them both up to the light...the one that makes u sneeze is the best one.