There's a reason why I ordered a Famicom and a copy of King of Kings (one of the two N163 games that use all 8 channels). I tested it on a standard, unmodded Famicom with RF output, and the channel-switch noise is absolutely inaudible. Proof. In the video you can kind of hear a faint high-pitched hiss, but in real life it's really impossible to hear (the microphone picks it up better than a human's ears I guess). And even if you could hear it it would be indistinguishable from the CRT TV's own high-pitched hiss.
My guess is the RF box acts as a lowpass filter of sorts which suppresses the 10 kHz hiss the chip produces as it cycles through the channels.
That alone should be reason enough to make it possible to choose between the two emulations. Another very good reason is that none of the common NSF players actually implement this quirk, so it makes no sense having to endure the agony that is this chip's shortcoming, only to 1) distribute it in a format which itself doesn't, and 2) hypothetically play it on a Famicom which filters it out anyway.
Don't get me wrong, I think it's great that FamiTracker is the only NSF driver which is fully faithful to the original hardware, but I think for the reasons stated above it should be the user's choice whether the channel-switch noise is present during tracking stage or not. If I want to know how my tune will sound on a real Famicom (assuming it's AV-modded and doesn't filter the hiss out) then I'll turn it on - and I will, I'm curious about this stuff. But there should be a way to turn it off.
Oh and Pie, using beta 3 doesn't work because there's a flaw which causes frequencies to be calculated using the wrong clock (beta 4's tuning is much more precise).
I believe it's not really the RF signal itself, but the television's RF demodulator which by design requires a lowpass filter of some kind. The precise amount of lowpass on the audio depends on who built your TV, really.
Also, I find the switching frequency is pretty clearly audible in this clip. It's obviously reduced by the LPF, but it's still clearly there to my ear. Also there's a huuuge spike visible if you use a spectrum analyzer. (vis) This spike is just under 15kHz, which should be he NES CPU clock frequency / (15 clocks/channel * 8 channels), I believe.
Anyhow, this is why I think emulators are nicer than hardware. Options! Like when I listen to Mega Man II's ending I can turn off the phase reset so I don't have to listen to a bunch of nasty vibrato clicks.
I've done a few experiments with NEZPlug++'s N163 output and lowpass filters; they do cut down on the acerbity of the channel cycling sound, but you lose a ton of brightness in the process. For 2A03 stuff I would never listen via my TV's RF demodulator when A/V is an option. It's just an inferior sound.
I believe it's not really the RF signal itself, but the television's RF demodulator which by design requires a lowpass filter of some kind. The precise amount of lowpass on the audio depends on who built your TV, really.
That must be it then.
rainwarrior wrote:
Also, I find the switching frequency is pretty clearly audible in this clip.
I did say it was audible in the clip, but if you're listening to it in person it's definitely impossible to tell it apart from the CRT TV's own high-pitched his, and it's much, much less aggravating than it is in FamiTracker.
rainwarrior wrote:
Anyhow, this is why I think emulators are nicer than hardware. Options!
Couldn't agree more.
nicetas_c wrote:
i like how b4 actually cannot load certain files from previous versions
Yes that's definitely an issue. I found any FTM that's been hex-edited with values not supported by the program fails to open in beta 4. So forget about 600 bpm or 3600 Hz. Not sure if it's a bug or an intended feature though.
That .ftm still plays sound even with 7 channels enabled.
It doesn't play any sound when you're using 1 channel, or if the waveform is '0 0 0 0'.
I'm guessing this is a bug (the latter point surely is a bug - flat waveforms are supposed to be silent).
_______________________
iGotno_scope wrote:
im going to continue making this crazy stuff then after a while my style will be so sick that you will be like damn suuun that shit is so sick i dont even get it. i will be like bro its ok.. you dont have to.
The 8 channel whine is pretty high up in the range of human hearing, 7 channels it's in a range that a lot of people can't hear very well, 6 channels is at the extreme limit, 5 or less channels and nobody can hear it.
So... as you reduce the number of channels, the whine very quickly goes above audible frequencies. Also, the audio path on an NES itself does have a LPF of its own (much weaker than an RF demodulator) which cuts these higher frequencies down a bit. So, probably by 7 channels it's already not very noticable, esp. depending on how good your hearing is (17kHz is hard to hear for a lot of people).
No games actually use 7 channels, though, to my knowledge. Most use 4, some 8, if I recall.
And here I thought the frequency actually went down as the channel count was reduced... Thanks for clarifying.
And yeah, all N163 games used 4 channels, except for Erika to Satoru no yumebouken and King of Kings, both of which used 8. Erika and KoK's soundtracks were done by the same composer (Hirohiko Takayama), which explains the similarity in style, waveforms and channel count of these soundtracks.
I don't remember Rolling Thunder using all the channels either which was by Junko Ozawa. Though it's been a long time since I've played it, I can't remember.
