Jump to content
IGNORED

Analysis of a particular spectrum


Recommended Posts

Hello, first please tell me if the question has been already asked ; I'll delete this thread.

 

I am wondering what really is quality of this track : http://hpics.li/b339bb9

It seems to be cutted at 15/16kHz like a "fake" 320kbp/s but the spectrum spreads to 20/21kHz like a true 320 kbp/s . Is it the same case for this one : http://grab.by/RX7E ? (except the fact it's lossless format (or not)) Does someone can explain me this please ?

 

PS: sorry for my approximative english, I have to take lessons

Link to comment
Share on other sites

  • 1 month later...

The "shelf" effect in these spectrograms are a sure sign that lossy compression has been applied to the files at some point in their history. The degree of information loss is due to the specific lossy encoder used and its settings.

 

If I understand you correctly, the second spectrogram is from a lossless format? If so it is not the original format, but has rather been transcoded from a lossy source - hence the "shelf".

Link to comment
Share on other sites

What you are seeing just below 16kHz is the effect of a brickwall lowpass at that frequency. Such filters are applied as a first step by MP3 encoders (lower cutoffs for the standard settings at low bit rates, higher cutoffs for high bit rates), but that is not the only use. It is fairly common for recording engineers to apply a 16kHz lowpass to tracks including vocals to reduce sibilance. I personally think it is a very bad practice to use a filter with such an abrupt cutoff for this purpose (a filter that rolls off over several kHz will have better time-domain properties), but that doesn't make its use any less common. There are basically two possibilities why you see the sharp cut-off there but also content above that frequency. Either the music was mixed from multiple tracks, one with vocals that has been brickwalled and another, perhaps a drum track, that has not, or the whole thing was brickwalled at 16 kHz and then put through a nonlinear filter, a peak limiter or clipping, that produced spurious content above 16 kHz. My guess is it's from mixing or a combination of the two effects, but I can't be sure without inspecting the wave-form and close-ups of the spectrogram. I do not see definitive signs of lossy compression in the second track, certainly not MP3. The content is too consistently present up to 22 kHz. It's almost certainly lossless as advertised. In the first file you see signs of MP3 or similar lossy compression at a high bitrate. Certain time/frequency windows have no content from 13 kHz on up, and others look blotchy. I could believe it's an original CD rip to MP3 320kbps, but MP3 160 transcoded to 320kbps wouldn't look all that different.

Link to comment
Share on other sites

  • 1 month later...

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
 Share



×
×
  • Create New...