I finally registered, I couldn't resist this thread !! I'm a long time reader of this site, first time poster !!
Anyway, to the issue:
This topic is something that's caused great debate over the past few years. As usual, most of this debate is rooted in what people "think" they hear, or what makes sense to them, instead of actual testing and facts. It usually boils down to those that have "The Golden ears" ears and those that don't. (No offence to anybody here that may possess such a set of ears)
Here's a fact - The only real thing will cause deviations in the sound quality of cd's are the number of errors on the cd. Either the player can read a valid 1 or 0, or it can't. All cd's contain errors, even pressed ones. CD players attempt to 'fill in the blank" when it encounters an error. The more blanks it needs to fill in, the higher the potential for it to effect audio quality.
Talk about pit/land leading/trailing edges not defined well enough in burned cd's are pointless, it's still about error's. If the pit/land isn't well defined, but the player can still properly determine it's a 1 or 0, then it doesn't matter since the data is valid and correct. If a pit/land isn't defined properly, and can't be determined, either because of burning, media choice or software used, it still becomes an error. So, let's focus on that.
So, the there are two factors at work here, the number of errors on a cd, and the cd players ability to overcome these errors. The former we will get to in a second, the latter is wholly dependant on your hardware/equipment.
So, we need to see some actual test measurements on errors generated by burning at different speeds. It just so happens that people have actually done these tests to find out the truth. Go have a look here:
http://www.emedialive.com/EM2000/starrett5.html
I think that pretty much speaks for itself, so I won't summarize much other to say that it's clear that using quality hardware and media, the number of errors at higher burning speeds aren't any more than at lower speeds, and in some cases higher speeds generated fewer errors since the burner was designed to operate at higher speeds.
I'd also like to raise a few more points -
First, What kind of music are you burning exactly ? If you are making cd's from MP3 files then it's really a moot point since the audio has already taken a huge fidelity hit in the MP3 conversion process.
Second, What are you playing these back on ? If it's your PC or average/generic home stereo, you probably aren't going to tell the difference either since 'average joe' home stereo's and PC speakers have anything but a flat frequency response and color the sound anyway (sometimes in a way we personally like, but not always).
Lastly, and most important, does it sound ok to you ?? After all, you are the one listening to it. Do you like what you hear ? You are what really counts in all this. If you enjoy what you are hearing, and can't "hear" the difference between different burn speeds then you might as well burn at the highest your hardware supports.
Me, I burn at 12x (fastest for my burner)
Regards,
Dukey