CD ripping: noise appearing in rip, (solved, pebkac)

Spanik · l33t Joined: 12 Dec 2003 Posts: 989 Location: Belgium

I'm ripping my CD collection with Audex (*). And I notice that sometimes when I listen to the rip I have some "crackling" sound in the ripped file. When I play the CD with a "real" CD player this noise isn't present. When I record the digital output of the cd player and play that from the pc, the noise isn't there either. So it looks as if the ripping isn't ok.

Audex uses CD Paranoia so it should rip "bit perfect". I have configured Audex to use "full paranoia mode" and "never skip on read error". I rip to .wav. Playback is with Jack, Aqualung, RME DIGI 96/52 and RME ADI-8 AE. Drive is an Asus one, can't remember the type exactly.

But what is going on? If I cannot trust CD Paranoia, then what is left?

(*) I use Audex because of getting into trouble with other rippers when filenames had accents and until now I haven't run into that issue with Audex.
_________________
Expert in non-working solutions

lars_the_bear · Guru Joined: 05 Jun 2024 Posts: 401

It's perhaps less easy to get a "bit perfect" CD rip than it ought to be. It's particularly a problem for CDs with an imperfect surface. The problem is that there's not much error-correction information in an audio CD track. Allowable read-error rates on an audio track are surprisingly high, compared to a data track, which has checksums and what-nots. In addition, the hardware will sometimes try to correct errors on its own, which will defeat what CD Paranoia is trying to do. CDP does things like reading the same data multiple times, and comparing the results; but if the in-built error correction tries to smooth over errors, CDP gets the same read every time, and assumes everything is OK.

Having said all that, for me CDP always gives better results than anything else with imperfect disks, even better than my very expensive stand-alone CD transport.

So I wonder: do you get the same problem if you rip an immaculate CD, with minimal load on the computer?

BR, Lars.

Spanik · l33t Joined: 12 Dec 2003 Posts: 989 Location: Belgium

Well most of the CD's I rip sound ok. But then I do not listen to each of them after ripping. And if it is a single track I can just record the digital output of the CD player.

My problem is one of trust. I thought that ripping them would enable me to have easier access so I would listen to more variation. But now I'll be wondering at each thing I hear if it is a ripping issue or something in the recording. (there are recordings where "outside" noise is present) I'll look for a way to display the spectrum in time so that I can see the bursts of high frequency noise better.
_________________
Expert in non-working solutions

lars_the_bear · Guru Joined: 05 Jun 2024 Posts: 401

FWIW I always rip CDs in a system that is otherwise under no load, and I do listen to the results if I have any reason to think that there could be a problem. If I look at the CD surface, and it's immaculate, I don't necessarily check every track -- I usually just listen to a few seconds of a few tracks.

I've ripped literally thousands of CDs with CDP, and I've only had one or two that came out badly. Weirdly, these aren't necessarily the most tatty CDs -- I've ripped CDs that look as if they've been used as frisbees, and the results have been fine. The ones that came out badly weren't really skanky, but they had a few superficial scratches when I looked closely.

If you're getting dropouts on rips of CDs that seem to be in good condition, that sounds like the result of an inadequate CD drive. I don't think you can rely on CDP to detect broken hardware, if the hardware is doing a crappy job of error correction, and saying everything's fine when it isn't. CDP is pretty thorough, but it can be defeated by bad hardware.

BR, Lars.

pjp · Administrator Joined: 16 Apr 2002 Posts: 20348

As lars_the_bear mentioned, I also experienced some issues when the system performing the rip was being used for other things. So I made a point to only rip CDs when I wasn't doing anything else with that system.

If you re-rip and don't have the problem, there's a good chance that was the issue. If it happens again in the same or close to the same spot, perhaps it is the CD itself.

I presumed it to be an artifact of a non-RTOS.
_________________
Quis separabit? Quo animo?

Ralphred · Guru Joined: 31 Dec 2013 Posts: 566

Spanik · l33t Joined: 12 Dec 2003 Posts: 989 Location: Belgium

Spanik · l33t Joined: 12 Dec 2003 Posts: 989 Location: Belgium

After listening to those 3 variations I think I found the issue. And it is not inside the pc... I loaded the 3 versions into Audacity so I could switch from one to another.

All versions have the crackling when played by the pc. Even the version that was recorded from the digital output of the audio cd player. So the issue wasn't the act of ripping because the audio cd player, when you listen to the analog output, doesn't crackle.

What was visible was that the crackles always coincided with the audio level hitting the top.

So I checked where there is another thing different between the analog audio and the pc replay. The only difference is that I use a different input of the amplifier. The analog output goes to the "cd" input, the DAC goes to the "cdr/tape" input. It was usual that the analog "cd" input had a sensitivity of 2Vfs and maybe the "cdr/tape" had less sensitivity, overloading on the peaks. But according to the manual all analog inputs have the same sensitivity.

This left the DA convertor. And yes, for some reason (undoubtly user error) this was set at +4dBu output... changing this back to -10dBV solved the mysterious crackling on high output levels.

Don't think I'll be listening to Moscheles "Grosse Sonata" any time soon.
_________________
Expert in non-working solutions