Hey folks.
Artemio: I should have been more clear. The 3rd Fantom-X8 was one that had its motherboard replaced. The first 2 Fantom-X8s were original motherboards with different OSes. There's where the conclusion came from. The delay decrease seems to have been a result of a hardware change rather than an OS change since the difference between the first Fantom-X8 with OS 1.01 and the second with OS 1.03 had virtually no difference in delay. Again, the assumption is that all the Fantoms are representative of same model and assembly run (this may not be the case, but there's no way to tell without hundreds more samples to measure).
Watermelon: I don't know Roland's true reasons for recalling the original motherboards. Unless you have authoritative information to the contrary, I don't believe anyone here does either. The high level cracking noise seems to have been significantly reduced, at least on my Fantom-X8, with OS 1.01 and 1.03. I can't speak for other here.
"Scientifically, the only valid assumption you can make is the following: 'if some models show a bigger patch switching delay than others, then it means that something IS DIFFERENT'."
Correct. And by controlling 1) the procedure used to generate the output (described in detail on page 1 of this thread), 2) the precise playback and recording of the test sequence, we have eliminated a great number of confounding variables. However, this is not a scientific evaluation. Just because I attempted to implement some methods used in scientific research does not make this a scientific experiment. The goal, for me anyway, was simply to get an idea of how significant a difference there is between first run Fantom-Xes and the later Fantom-Xes with newer hardware. Our example WAV files are just demonstrations, not scientific research. It is impossible to do real scientific research in a completely uncontrolled environment. Without every one of the Fantoms in the same room with the same RAM, same expansion boards, same PC card expansion, and same person performing the same test, it really can't be considered scientific research, as I'm sure you realize.
"This difference might not be the bug at all (many things may get changed in the manufacturing process depending on the available parts on a given moment... manufacturing is kind of shopping
"
Again, I agree. And, as stated earlier, without hundreds more Fantoms to test in a controlled environment, exact differences can not be determined by simply measuring delays in switching Patches. However, such significant differences in Patch changing (67 percent increase in speed between old Fantom-X8 and new Fantom-X8 both with the same OS) demonstrates a fairly radical difference in hardware (since the software is the same). That's not to say that I know exactly what the differences are, but something other than software is different. The only thing left after eliminating software is hardware (yes, it's possible that installed RAM or expansion cards may have confounded the test as these were not controlled).
"...I just want to point out that nsheldon experiment, although interesting, is not sufficient to prove if the "fault when switching sound" (Service Note number: 101811) affects your device."
That depends on your interpretation of the specific fault identified by that service note. Do you have the exact text of that service note (I don't)? If the service note only identifies a loud popping noise occasionally when switching Patches, than, no, the previous tests would not necessarily indicate that your Fantom exhibits this fault. However, if the fault identifies "long" delays in sound output while changing Patches (that's too vague to be considered an operational definition, I know), or includes that as part of the fault, then the preceding test would be strong evidence that your Fantom-X exhibits the fault. If you have the exact text of the service note, please share.
Your reliance on anecdotal information (a Fantom-XR owner who had their main board replaced and no longer heard the loud noise), I think you'll agree, is not evidence of the service note content, nor does it demonstrate mutual exclusivity of the two problems (long delay and loud noise). It's possible that this FXR owner also had longer delays when switching patches which was also fixed by the main board replacement (though we'll never know). And, if I understand your point here, I agree that a loud noise during patch switching is a much more serious problem than simply a long delay in switching Patches. Without much more investigation, it is impossible to establish a correlation between the Patch delay and loud noise while switching Patches. That said, these tests that we've been informally performing, do indicate that the delay is improved upon with newer Fantom-X hardware, regardless of whether that hardware difference addresses the loud noise problem.
Regards,
Nathan Sheldon
http://www.nathansheldon.com/
The good news is that the newer Fantom-X line have greatly improved switching speeds. In fact, if the one Fantom-X with an updated board is representative of all newer Fantom-Xes, there is a 50 percent improvement over the delay in the Fantom-S, and a 67 percent improvement over the earlier Fantom-Xes without the motherboard replacement.
) I was fortunate getting someone that was definitely aware of this problem. He told me that the service note refers to loud cracking noise produced by the effect section when switching patches and told me that there was no mention about the delay problem. In other words, the reason of recalling early motherboards is linked to loud cracking noises when switching patches, not delays.
