CDs from the 1980s - high rate of "fake" detections

classic Classic list List threaded Threaded
5 messages Options
Reply | Threaded
Open this post in threaded view
|

CDs from the 1980s - high rate of "fake" detections

Umlaut
I'm a new fan of the app, and I certainly appreciate what it's trying to achieve. But I've consistently noticed that with CDs mastered in the mid to late 1980s, at least in the rock genre, it seems like a really high percentage of known genuine lossless audio files are labeled as fake.

I say known genuine because I bought the CDs in the US in record stores (not Russia or wherever) myself in the mid to late 1980s. Yes, I'm that old.

I notice it often when I'm using Exact Audio Copy or CueTools or even iTunes or Apple Music  to rip my personal CDs to FLAC or WAV audio. In the case of Exact Audio Copy or CueTools I'm told I got a 100% accurate rip. But it it's rock, and mastered in the early years of CDs, as many as 100% of tracks are tagged as fake.

Is this something others have noticed? Was CD mastering in that era for rock different in some way?


Thanks for looking
Reply | Threaded
Open this post in threaded view
|

Re: CDs from the 1980s - high rate of "fake" detections

Fake No Funk
Administrator
Yes, the mastering of records has changed a lot (see "loudness war" e.g. at wikipedia)...

Could you attach a frequency spectrum of such a file? Would be interesing :)
Reply | Threaded
Open this post in threaded view
|

Re: CDs from the 1980s - high rate of "fake" detections

Umlaut
Yes, I'm acutely aware of the loudness wars. The dynamic range of CD from 1983 I most recently ripped to FLAC averages around 14 db. I'm sure that if I were to purchase it at Apple or Amazon Music, I'd get a recent remaster, with a probable average dynamic range around 8-9 db. And if was something actually released as new music - current pop - more like 4 or 6 db.

This album happens to be Sweet Dreams (Are Made of This), Eurythmics, CD pressed and sold in 1983, RCA - US release but pressed in Japan.

I'm inserting the spectrum for the title track, ID'd as fake by Fakin' the Funk. Thanks for the prompt response. I see this 85-90% fake "effect" with virtually everything from the '80s. Decreases considerably in the '90s.

Reply | Threaded
Open this post in threaded view
|

Re: CDs from the 1980s - high rate of "fake" detections

Fake No Funk
Administrator
There IS a clear cutoff around 20.5 kHz, which indeed is an indicator that a kind of filter was applied within the mastering process (if you are sure that ripping it to flac REALLY was lossless), but the rest of the spectrum looks fine.

That's why there is an option "allow cutoffs above xxx Hz" -->

From my point of view, you can either "allow cutoffs above 20 kHz" or manually mark those files as ok.
Reply | Threaded
Open this post in threaded view
|

Re: CDs from the 1980s - high rate of "fake" detections

Umlaut
I just found another example in my CD collection. In 1985, Dire Straits released one of the earliest fully digital rock releases (DDD). Brothers in Arms, and I still have my copy, Phenomenal dynamic range compared  to almost anything in rock or pop now. - average for all tracks 16 dB. I just carefully ripped it to .wav using Exact Audio Copy - and I'm positive no other conversion was done. 100% accurate rip according to the Accurip database, likewise for the CueTools database.   7 of 9 tracks are flagged by FTF as 320 kbps "fake" CD audio.

This isn't meant to disparage the app. Not many of the users (probably) are working with pre-1990 releases. But it's definitely something to watch for on early releases.