(Hopefully not making a fool of myself – no warranty for correctness)
The difference between 71min and 74min seems to be ATIP wavelength, forcing the CD writer on a higher speed when locking on the 22.05kHz frequency. Data Position Measurement (DPM) – unfortunately to my knowledge still only possible with closed source software – reveals the size of the angle a single sector needs on a specific position of an optical disc.
Graphics created by DPM software aren't fully self-explanatory (at least not for me). Took me a while to understand. X-axis is simply sector number. Y-axis shows degree of arc (for a single sector). This means: Higher values correspond with lower data density. Of course the degree value goes gradually down as the read/write pickup moves further to the outside of the disc in all cases (bigger circle, more space, less degrees per sector).
Enough talk! Picture:
View attachment 397055
The first picture shows a comparison between the legit Spyro 3 CD and a copy on 74min CD-R. At the beginning of the disc a sector on the original takes ≈38° for the original and ≈36.5° for the copy. Difference in sector size – supposedly caused by slightly higher rotational speed during writing – has almost the same value as 74/71. An astonishing precision for just measuring read timings!
Following the curves to the right we notice that the distance between the two shrinks. At first this left me confused. Thinking about it for a while the reason for that became clear: The original CD, almost 71min, reaches the end at the diameter 11.8cm while the CD-R has still some minutes of capacity free; we are filling the spiral track slower due to "smaller writing". Hence the our optical pickup is not at the outer boundary when finishing writing the disc. Smaller diameter → smaller circumference → less space per rotation. This effect partially compensates the difference caused by the two different "sizes of writing"
Next picture shows it even more clearly:
View attachment 397056
On common 80min CD-R with tighter spiral (smaller track pitch) we are still quite far from the outer boundary when reaching minute 71. At the beginning difference in arc degree between 80min and 71min is almost the same as 74min vs 71min. But at some point staying further inside thanks to narrow track fully compensates for the differences in pit size.
Using oversized, non-standard CDs would probably make the green line cross the red one (I will have to order more 90min and 100min blanks)
Note that the two pictures do not (directly) show absolute data position. Although the two lines seem to be more similar for the 80min CD-R, the absolute difference for
"At what diameter can a random sector be found?" is in fact bigger when using 80min media.
To clarify: I do not have any PlayStation Master Disc CD-R and just assumed it would behave just like the pressed disc. If only they weren't that ridiculously expensive…