Well, I guess I'll trot it out again: No one could accuse me of having a digital bias, since my vinyl collection positively dwarfs my CD collection. But the fact is, of the material I own in both formats, some recordings sound better on LP, some on CD. This includes vintage as well as new vinyl pressings, although I own relatively few new pressings by comparison. And in certain areas of sonic performance, such as transient definition or background quietness (otherwise known as signal-to-noise ratio), CD always has an advantage.
I'll be the first to admit that neither my analog nor my digital rigs are state-of-the-art, but IMO it's a lot like real estate, where they say the three most important things are location, location, and location. With recording formats, the three most important things are mastering, mastering, and mastering (plus pressing, especially for the vinyl record). If your vinyl record is better mastered than your CD of the same material, then it will sound better on a decent rig, and the same thing is true in reverse if your CD contains the better mastering job. If both formats are very well mastered, then in my experience it doesn't make a whole damn lot of difference which format you listen to as long the playback gear is of roughly comparable quality.
With that in mind, CD's are more convenient for their potential of extended, uninterrupted playing times, programmable/remote track access, and less needy care requirements, so for new issues, if the price is equivalent I generally buy the CD where both formats are offered, and always if the CD contains more bonus tracks (or the LP if it contains more bonus tracks). Of course, there are still many CD's out there where the mastering is of questionable quality, but the same is true for LP's (pressing quality too), so oftentimes you really have to buy and listen to know which format is best for any individual recording, unless someone can tell you in advance.
Anyone who's interested can click on my 'Threads' and go to the one entitled "Upsampling Put to the Test", to read about a test I did where the signal from my analog rig was digitized into Redbook format, then decoded by my DAC and instantaneously compared in a level-matched A-B test to the original unprocessed analog feed. The upshot was that there was very little difference to be heard between the straight analog signal and that same signal as encoded/decoded by my digital rig, and part of what small differences there were in favor of the pure analog feed must be chalked up to the extra cabling involved in transmitting the digitally processed signal. These test results surprised even me, as I had always thought my digital front end did have some sort of slight but identifiably 'digital' character to it that I didn't hear from my analog front end, but the test revealed that the losses were both smaller and of a different sort than I had always assumed, despite the test construction unavoidably favoring the analog a slight bit by default. It forced me to conclude that the vast majority of what I had thought were 'digital'-sounding flaws present in regular CD playback were actually the fault of the software (the mastering and/or the manufacturing) or the disk-reading process, but not the Redbook format itself or my decoder's digital-to-analog conversion process.
Bottom line: If your gear's not handicapping the race too much one way or the other, then the quality of the software (meaning how well it was mastered and manufactured) - rather than its type, LP or CD - will be the most important determinant of the resulting sound quality, IMO.
