Radio spectrum doesn't look right. Soundblaster did this?

[taygan]

While I'm analyzing files I came across this one too, doesn't lok lossy, but doesn't look great either.

Supposedly: FM -> Soundblaster live Line-In -> Wave -> Nero burning Rom -> CDR -> EAC -> Wave -> dbPowerAmp -> FLAC

imageshack.us still is down for me.

This doesn't have the nice even FM with the spikes, instead the spikes are cutoff (like an mp3) and there's some noise (analog?) up high. No carrier signal, but the top of the spectrum is scalloped a bit. Does this look acceptable? Could soundblaster do this?

8 sec sample:

http://s6.yousendit.com/d.aspx?id=0JONYJHKO7MC52XRQQYW63JQNY

[Five]

you're really posting the tough ones now

some stations broadcast in lossy, I can see some squarish blocks and of course the obvious cutoff at 16kHz yet the strange thing is there's no carrier visible. there's also noise in the highs which suggests an analog transfer. pretty much all the same things you said in your first post

http://img378.imageshack.us/img378/5792/louiseattaque200518so.th.gif (http://img378.imageshack.us/my.php?image=louiseattaque200518so.gif)

[taygan]

I think this has been MD'd. The little spikes seem consistent with the MD block parapets, and that would explain the lack of a carrier signal.

edit:
Oh, forget it. The upper noise rules that out. Digital broadcast? MD before FM?

[ssamadhi97]

Doesn't really look like MD to me (the spikes, they are different! :P)

Alas the yousendit link expired, so I can't get my hands on the sample right now. Mind reuploading it or putting it on rapidshare.de, taygan?

[taygan]

Alas the yousendit link expired, so I can't get my hands on the sample right now. Mind reuploading it or putting it on rapidshare.de, taygan?

Wow, I like rapidshare.de!

Here's a 5 sec flac sample:

http://rapidshare.de/files/12254234/lousseattaque2005-39-46sec.flac.html

Thanks.

[ssamadhi97]

All right, thanks for providing another sample.

The little "spikes" you see are indeed different from those on many MD sources - they occur every 20ms (as opposed to 12ms for MD), so 50 times per second. I'd suspect that this might've been caused by the power supply of an analog device in the lineage.

Be that as it may, the holes in the spectrum strongly hint at lossy compression somewhere down the line (as Five said).

My guess would be something like "recorded from the analog output of a digital radio receiver". The other alternative, a straight FM broadcast recording (so basically something where both lossy compression and decompression happened before the material went on air), would not explain the absence of a stereo pilot tone (like Five mentioned, too).