There is a very short but nonetheless noticable release envelope on all FDS sound when played in most nsf players (nsf live, foo_gep, nezplug/play, nsfplug/play, virtuansf and g-nsf all have it, didn't check any of the winamp plugins but i assume they do too.) however, famitracker doesn't have it. it's particularly noticable if you are trying to write ultra staccato notes (1 frame long) on the disk system channel. in export they'll always be more legato.
example attached, comparing S01 on FDS and 2a03 chips. they're the same length in famitracker, but not in the other nsf players. unfortunately i'm assuming the players have the correct behaviour and famitracker needs to be brought in line, unless it's just an nsf export error, which i'm doubting. another thing i'm not delighted about, as it's seeming more and more like using the FDS channel as a 3rd 2a03 channel is not as practical as i hoped. no funky basslines either :B
oof, another thing i forgot to mention. FDS pitch-up wraps on export but maxes out at A-5 in the tracker. this happens with portamento, arpeggios, pitch envelopes and blank notes as well. the tracker behaviour is fine, it's alike how 2a03 pitch-down stops at A-0 as opposed to cutting out entirely like 2a03 pitch-up/FDS pitch-down do (obviously because of their respective ranges, being extremely high/low.)
doubt an example is necessary but i made one anyway.
i found or a problem of compatibility with the command 7xx.
If you try 7B7 in the noise channel and open it into the 0.3.0 version it sound different when you open the same file into the 0.3.6 version. Dont know what version is wrong if the old one or the new one. Let me know if this is a bug or is a solved problem. So i can change or modify my new songs to the new sound/command version.
---- Another bug ----
When you use the use the command Dxx in whatever place and you want to delete the command, sometimes the command Dxx modify the Song Settings cuttin the ROWS to the number you had the command. It happend to me sometimes, but not all the time.
2a03 tremelo is definitely better now because it's the same between tracker and export. however tremelo on the FDS channel reacts differently to the same values. you need to use a higher depth value to get the perceived depth equivalent to 2a03. i don't know why that is. it means you have to spend a bit of time experimenting, instead of a simple copy+paste job from 2a03 channels.
still, at least there's no more export inconsistencies (knock on wood!)
1) Hxy/Ixy seems to be very different between 2A03 SQ1 nd SQ2, as in FamiTracker. For example:
C-3 00 - H73
If I put it in both 2A03 SQ1 and SQ2, there should be something similar to chorusing towards the beginning.
2) Hxy/Ixy gets faster as notes are triggered in the other channel. Is it also intended?
The FDS period table works inversely (a higher register gives a higher pitch).
Thus I am sure that jsr makes the Pxx work properly in VRC7 channels, but forgot them in the FDS channel.
myself, i think Pxx could be fixed; it'd be a minor inconvenience for people to update their song that use it right now (i doubt many have used it anyway) and it might cause confusion to other people down the road as well as it is, especially if it is undocumented.
The FDS period table works inversely (a higher register gives a higher pitch).
This could just be because of how the FDS works. I'm not sure about it really, but it's entirely possible that it behaves like the N106, which has inverse pitch behavior (low frequency registers correspond to low notes, and high ones correspond to high notes). This makes things like vibrato sound stronger on low notes than high ones, which is very strange to manage.
...I'm not sure about it really, but it's entirely possible that it behaves like the N106, which has inverse pitch behavior (low frequency registers correspond to low notes, and high ones correspond to high notes)...
It is true.
A440's period is around 3C6B in N106 while that of a C-6 (C-5 in FamiTracker) is 8F33...
Looks like I have an unfortunate time fine pitching if there is once I do an N106 cover...
In 0.3.6, if you try to copy pattern information (notes, effects, etc..) from one instance of Famitracker to another, it doesn't work. Also, after that, it won't let you copy anything saying that it "cannot open clipboard." You can resolve this by closing both Famitrackers and re-opening them.
Has the issue of not opening 0.30 and earlier FTMs been recognised yet on here? I try to open something like TILT's 1aaa, and it says that it needs to have been made with the newest version of famitracker.
