Deprecated: mysql_connect(): The mysql extension is deprecated and will be removed in the future: use mysqli or PDO instead in /storage/content/49/145849/famitracker.com/public_html/forum/classes/dbHandler.php on line 29
FamiTracker
I still see a problem involving VRC7 & FamiTracker 0.3.5 latest stable.
Whenever I createa VRC7 track where the VRC7 channels have different volume settings, the volume sounds fine in the FTM module, but are totally ignored when compiled into an NSF file.
So the VRC7 channels' volumes aren't exported into the NSF file, making all VRC7 channels at their maximum volume.
See for yourselves.
Will this problem persist until the next major update to the tracker?
_______________________
Technology: the one thing that's hated & cursed at by all engineers, technologists, scientists & technicians!
I've tried it in your home-made NSF player, Nestopia, VirtuaNSF, AudioOverload & NotSo Fasto for WinAmp. In all cases, using the default volume settings for all sound channels & expansion chips will yield the same results.
In the FTM, the volume settings play correctly, but it breaks on NSF export. Even if I've set a volume for one of the VRC7 channels to a value of 8 or 9, the exported NSF will still play the channel as though its volume were still at maximum.
And thanks for looking at this.
_______________________
Technology: the one thing that's hated & cursed at by all engineers, technologists, scientists & technicians!
That's funny. It does work properly when the volume is decreased & exported to NSF.
That said, the third VRC7 channel (the one at maximum volume) still sounds overpowering when played from the NSF, even though it doesn't sound as loud in the tracker. So I guess my only work around would be to normalize the volume of all VRC7 channels to C or A.
Now, is this something normal (I mean the fact that the audio's maximum volume plays softer in the tracker than it does in the exported NSF)? Or is it just me?
_______________________
Technology: the one thing that's hated & cursed at by all engineers, technologists, scientists & technicians!
Nope that's probably normal, I imagine different players have different mixing volumes for the expansion chips (I haven't seen any levels documented anywhere). But in this case it could also be the tracker that's wrong because I haven't measured the VRC7 levels yet. All other chips are adjusted according the real hardware, but there might still be differences in different players.
Well, I have a feeling it's the tracker (possibly a simple mistake in a calculation).
My understanding is that the maximum peak-to-peak amplitude of any given sound channel (without the use of special effect commands or the addition/subtraction effects of other channels) is typically 5 V.
So I decided to stick my computer's headphone jack to an oscilloscope while playing a true VRC7 rip & a FamiTracker-made NSF in all of my players.
Here's what I found; the default maximum amplitude for an individual VRC7 sound channel is indeed 5 V peak-to-peak. This occurs on a true VRC7 rip.
However, the FamiTracker-made VRC7 NSF outputs a maximum amplitude of 7 V peak-to-peak (a difference of 2 V).
As I calculated, the volume steps up in increments of 1/3 V per hexadecimal value increase (1=1/3 V, 2=2/3 V, 3=1 V, etc.). In a FamiTracker-made NSF, the true maximum volume is actually 5V+2V= 7 V, or F+6=15.
This is why I have a feeling that it's a simple miscalculation; perhaps you set the correct volume output for the tracker's regular playback function, but not for the NSF export function.
As for the differences in each player, I set all their volumes to play at their default values (either all at 100% or all at 50%). The changes were quite negligible with respect to the readings themselves (the variations were only cause by the fact that I was measuring from my computer's headphone jack at 50% volume).
_______________________
Technology: the one thing that's hated & cursed at by all engineers, technologists, scientists & technicians!
First, you only have 16 volume levels from the software's point of view and those are accurately preserved when exporting. So there are no differences like that between ripped NSFs and exported ones, it's only the mixing levels coded into emulators.
Second, the only thing that would make sense to measure is the real hardware (the Lagrange point game cart), as emulators are not guarranteed to be correct. This is what I've done for the other chips but haven't yet with VRC7. (And the voltage levels are usually around 1-2V.) And, you cannot measure the absolute levels, it has to be compared to the internal channels (because only the relative mixing level are of interest).
Well, I guess it's back to the drawing board for me. But at least I know now that the volume as played through the tracker itself would be equivalent to 6 hexadecimal values less of that in a compiled NSF.
Thanks for your consideration, JSR.
_______________________
Technology: the one thing that's hated & cursed at by all engineers, technologists, scientists & technicians!
) 
