[ultramage]

Hi. I've been tinkering a bit with cd image contents, and noticed that whenever I make an image of an image mounted onto a daemon tools virtual drive, only the initial sector of a track contains '1's in the P subchannel. Everything preceding it is zeroes.

On real cds, every track is preceded by a 2+ second (149+ sectors) interval where P = 1, ending with the first sector of the track, so 150 sectors in total. This also applies to the final track, followed by the lead-out track (which is not accessible).

According to the spec,
"The minimum length of a Flag (i.e. a continuous sequence of ONEs in the p-channel) shall be 2 s (i.e. 150
Sections). The p-channel of the last Information Track in the User Data area shall end with a Flag of 2 s to 3 s (i.e. 150 to
225 Sections). Its end shall indicate the beginning of the Lead-out Track."

Current DT behavior can be exploited to identify images that were made from a daemon tools virtual drive, and to identify that a drive is a daemon tools virtual drive (assuming that a real drive would have to conform to the standard to get a license). I can provide a proof-of-concept DT virtual drive/image detection tool if you're interested.

Also, burning of such cd images in DAO96 mode might cause playback issues in drives that might, for some reason, rely on the P subchannel. Not sure if drives like that still exist, but if hardware still conforms to the standard, then why not make DT do it too?