I thought I'd pass this on.
a unknown cd wouldn't mount and apparently had no filesystem. I let ddrescue loose and it ran for 3 hours before I stopped it (Ctrl_C). Then even Thunar could pick it up as an audio cd. I transferred the single 413MB .wav file without error

So the dvd drive being warm improved the sensitivity. I hope this helps hardware folks.

BTW, what options do guys use on ddrescue?

Could be just a coincidence?

In this particular case an audio CD is not mountable and therefore ddrescue can't read it. Did you run ddrescue on purpose just to warm it up? It would depend on what was "wrong" and either warming affected the underlying disc structure or some marginal component in the drive hardware. I assume it was the disc.

ddrescue has a bunch of options so it just depends on the situation. Luckily, the only time I needed/wanted to use it the drive had already died...

makes sense.
thanks for the tip.
but i will sure to use it only on CDs that i do not intend to keep afterwards

Or DVD drives....

I don't think it's coincidence. My son had failed with this, I had failed, The disk looked in poor shape and came with some of vintage 2005/2006.

Interesting about ddrescue needing a mount. That cuts out half the dodgy disks, doesn't it?
I would expect CD reflectivity to be fairly constant.
The laser emitter is digital, but the forward voltage drops as it warms, slightly increasing current.
The received light in the opto receiver and comparator are the only truly analogue components in an otherwise digital drive. From there, it goes to something that makes a 1 or 0 out of it. High and low are rarely marginal, but light reception can be very temperature dependent, as can light emission to a lesser extent. Take it from someone who had an optoelectronics project in 2013/2014. As the drive warms, you would increase laser emission slightly and receiver sensitivity markedly.

I post to repent in dust and ashes. K3B picks it up on a cold drive. Mind you, there's the best of detection stuff in k3b, but it could rip it. It does a good job ripping, but assigns a subdir with some very funny permissions. Mind you, I stand over what I said about the Electronic stuff, but that can be compensated for to a great extent (e.g by ensuring a constant current through the laser, taking only A.C. from the light receiver - all good design anyhow).

My advice for dud CDs is to hang them up on trees in your garden in order to scare birds away from eating the berries.

i'm with business_kid.
every now and then there's this one CD that you really want to save.
you probably already know it's scratched; a normal cd player will skip over it somehow, but the linux ripper is being prissy about it.

and that's what i meant with my last post: if just this once i can get it to play, only to rip it and then throw it away.
and if you don't have dozens of cds like that, the player will survive the 3h of heat.

but - i had better experience with abcde. it needs to be set up properly through a well-commented config file, where i told it to be as paranoid as possible and (almost) never give up re-reading a sector etc.

I wanted to test what could be done with the tools at my disposal.
Thunar read the audio cd while warm; K3b didn't need to be warm. Nothing else read it. Funny, but the other 2 discs in the same handful were an update for WinXP to MCE(?). Disk 1 was read fairly well, but disk 2 mounted, but after that nothing. README.txt (37k) showed nothing. Autorun.inf (23 bytes) came through. Data3.cab (511M) gave me 5.6M and ddrescue could add nothing.

I ended up underwhelmed by ddrescue, impressed by k3b, and with a temporary allergy to disk recovery of any sort.