Topic: FLAC 1.2.1 compression ratio?

[terrapinj]

just FLAC'd two different shows using Frontend 1.2.1 level 8

all settings were identical

show A was just over 2 hours
show B was 1hr 45min

show A folder is 323MB - shows compression ratio around 4:1
Show B folder is 623MB - shows compression ratio around 1.8:1

any reason why the compression levels would be so different?

[flipp]

I've found a singer/songwriter show (just amplified acoustic guitar and vocals) compresses much more than a rock band. Type of music you are compressing will have a big difference in the resulting flac. If you were compressing two similar types, I have no clue why the flacs would have such a different ratio.

[terrapinj]

well music is similar enough, both loud electric full bands

A - Radiohead
B - WSP

[dallman]

They may be similar, but chances are their sonic signals are different. If the wave looks very fat, the file is usually bigger. If there are lots of peaks, but the middle of the wave is skinny or smaller, it will be a smaller file. If you have quieter music, but very high levels, you will have a larger file, so band loudness may not be much of an indicator.

This is not scientific, just based on my own observations after FLACing a lot of files...

[terrapinj]

waveforms look similar - both had volume adjustments so they often peak at or around 0db

perhaps it could account for some difference, but it just seems over twice as small is a huge variation

thanks for the comments so far

[JasonSobel]

if you still have the whole file for each set, one thing to do would be to check the average RMS level of each set.  in general, the lower the levels, the higher the compression.

[terrapinj]

if you still have the whole file for each set, one thing to do would be to check the average RMS level of each set.  in general, the lower the levels, the higher the compression.

thanks Jason - i'll try this as well

[rustoleum]

Is one mono?

[terrapinj]

Is one mono?

i thought about that but both were def recorded in stereo and i don't know how it would have been converted to mono

[mmedley.]

if you still have the whole file for each set, one thing to do would be to check the average RMS level of each set.  in general, the lower the levels, the higher the compression.

This is correct. As the dynamiic range is reduced (increasing average RMS) via compression or whatever method there is more data for FLAC to compress and not as much room for saving space to obtain a greater file compression ratio. Not saying you personally are using compression, normalization, etc. but if the PA is outputting a compressed mix you will notice this phenomenon. On the other hand, an increased dynamic range will usually yield a much better compression ratio.

PS...Rabold uses pretty heavy compression in his mix, apparently much more than Radiohead.

[terrapinj]

found the problem!!!

somehow in samplitude a box got checked to "force mono processing" so even though the waveforms looked unique in each channel and the file reported stereo it was definitely mono - as soon as I unchecked the box you could def hear the difference

thanks for all the suggestions! glad i checked the files again before torrenting