Normal for a FM broadcast?

[Tate]

Since there seems to be a lot of experts here, thought I'd inquire here......I got this boot wich looks MP3-sourced to me. It cuts off at 16K with no activity above it. When looking at the spectral view, you can see that "haircut" line that's suppose to be bad.

But the lineage says 'stereo sound from an FM broadcast'. I'm not sure if this is typical of FM broadcasts. I've attached the frequency analysis & spectral view. Any help on this would be appreciated! Thanks!

[Five]

no, that's not normal for an FM... normally the top wouldn't be pure black. Strangely, it doesn't look like the usual mp3. Whatever this is, it doesn't look good. Please post another screenshot of the spectral view zoomed in on 1.5-2 seconds so I can tell if it's lossy or not.

[Tate]

Here's a close-up of the spectral view. I've noticed that a lot of times, when inquiring about a lossy source, people are asked to zoom in ~2 seconds, why is that? What do you look for? I really wanna learn as much as I can about spotting lossy sources.

My initial thought was that this is a bad source, cause whenever you got no sign of life above 16K (and it's not a mono or MD recording for instance), that's never a good thing. Thanks for your help!

[oldbrokentapes]

When's the broadcast from? A lot of people are still labelling digital radio broadcasts (DAB/DVB) as FM, even though they have absolutely nothing to do with frequency modulation. A 128kbps DAB/DVB stream would look pretty much like this. Digital radio is obviously still fairly new technology, so the older the broadcast is, the more likely it is to have been lossy-compressed post-broadcast.

[Five]

Originally Posted by Tate

Here's a close-up of the spectral view. I've noticed that a lot of times, when inquiring about a lossy source, people are asked to zoom in ~2 seconds, why is that? What do you look for? I really wanna learn as much as I can about spotting lossy sources.

My initial thought was that this is a bad source, cause whenever you got no sign of life above 16K (and it's not a mono or MD recording for instance), that's never a good thing. Thanks for your help!

You're welcome... what I'm looking for is the telltale "blocks", when you zoom in and these are visible then it's positive that it's a lossy source. eg:
http://www.thetradersden.org/forums/...achmentid=1521

Yours is very strange, it seems to cut straight across twice (??!!), once at about 15.2kHz, then again at 16kHz (the blue part). I can't see the blocks here, so I can't say for sure that it's lossy. Perhaps this was over-nr'ed? I'm stumped. It looks a little like FM, but normally there'd be noise at the top of the spectrum and the reddish parts don't cut off so abruptly. Anybody got some 2nd opinions on this one?

If you want, you could also attach a 4-5 second FLAC in a .zip here, too, I might be able to tell more if I can hear just a little of it.

[Tate]

I uploaded an audio clip (in WAV file) through Yousendit.com:
http://s20.yousendit.com/d.aspx?id=3...S10R4AS4RNK5I5

It's an Aerosmith show from 1994 (so it's not digital radio ) When listening to it, sounds legit.

[Five]

Listening to it, I can't hear mp3 artifacts, so this is probably a good FM source with some heavy noise reduction used on the higher frequencies. That's my best guess.

The train that kept a rollin' all night long of rock-n-roll
That you can not kill
It will live forever... is here again

I love the 'smith

[Tate]

Thanks for your help Five. Guess I'll keep this boot (but I'll always hold a little tiny doubt about it )

Originally Posted by Five

no, that's not normal for an FM... normally the top wouldn't be pure black. Strangely, it doesn't look like the usual mp3. Whatever this is, it doesn't look good. Please post another screenshot of the spectral view zoomed in on 1.5-2 seconds so I can tell if it's lossy or not.

On the contrary, that's pretty much what FM broadcasts look like. It wouldn't take NR to do this.

The frequency cutoff may vary a little, though 15khz is most typical.

[oldbrokentapes]

Originally Posted by buzzy

On the contrary, that's pretty much what FM broadcasts look like. It wouldn't take NR to do this.
The frequency cutoff may vary a little, though 15khz is most typical.

I have to disagree, and I think I'm with the majority. An FM signal should always show something across the entire range because of the noise/hiss inherant in the medium. I don't know how the sample this thread is about got so completely killed, but there's no way that's a standard FM capture.

The attached images are SBD > FM > WAV (and then > mp3@160kbps) for reference. There is a dip around 16kHz but nothing close to what is being discussed here.

edit - ok, how do I make my image attachments show automatically rather than needing clicking?

edit again, follow up thought - are there radio stations that apply a high frequency band filter pre-broadcast? I can't figure out how that featureless line stays so high around -96dB

[oldbrokentapes]

The same FM>WAV file with a harsh 16kHz band filter applied in Goldwave. Looks exactly the same as the mp3 and still nothing like the troublesome original.

ah I get it, the image shows if you only post one!