It seems to be the rate either averaged up to that point, or perhaps averaged over the last few minutes. At least, this is what I surmise based on how the per cent completed changes compared with how the reported write rate changes. Especially notable is that at the layer change, nothing is being written for several seconds, but there’s very little, if any, change in the write rate.
Much the same applies to the read rate, and for reading it’s very noticeable when the change in per cent completed suddenly slows to a crawl due to reading problems (such as a dirty/scratched/defective disc), but the reported read rate drops off slowly.
Okay, I admit that I also watch the “bytes written per second” to my hard disk when reading the DVD, and the “bytes read per second” from my hard disk when writing the DVD, which isn’t an exact method, but with almost no other applications running, ought to be reasonably close, and indeed seems to agree with what I see in DVDFab’s per cent completed.
Perhaps Fengtao will give us more information regarding the reported rates?
The rate averaged over a fairly long time may be better for estimating the remaining time required to complete a task, but it seems to me that for user information, the “current” (well, averaged over a few seconds at most) rate would be more useful. That said, the most important thing is the total time to complete a task, so I wouldn’t lose any sleep over temporary drops in the read/write rates.
Of course, as often noted in this forum, for a given angular velocity, the read/write rate is lowest nearest the center of the disc and highest nearest the outer edge. Assuming a constant angular velocity read/write strategy, for a single-layer disc, this means that the rate starts low and increases, and for an “opposite track path” double/dual-layer disc, starts low, increases until the layer change, and then decreases. Some pressed dual-layer originals have a “parallel track path”, and for these the rate starts low and increases, then suddenly drops and increases again. I suspect that when the drive is writing at less than its maximum rate, it changes its angular velocity (in steps) to keep the write rate somewhere near the specified rate.
Of course, the rate may also be limited by other “bottlenecks” in the system, such as the maximum rates for the bus, hard disk drive, processor, or memory, other tasks running at the same time, perhaps any copy protection method, the DVD drive itself re-trying a read at a lower speed after a read error, or adjusting its laser power or focus while writing.