OK, I assume this was a delay in the mods system, transferring the mod to the server folder. I'll forget it for now but let us know if a delay happens again.
Thanks, I've emailed Victor about it.
EDIT: This should be fixed. Victor fixed it by editing my download link to https instead of http which seemed to cause the bug in Chrome.
-
(Scawen)
DELETED
by Scawen : edited previous post instead
Yes, hence it would need to be optional as a server setting.
It would only really be useful for organised events where you can enforce racers to have the *default* colour at the beginning of their name, which would mean the old colour behaviour (ahead/behind/lapped etc) would still apply to racers' arrows, but course cars with coloured names would show that colour instead.
Heres some videos about the weird clipping drifteris mentioned, for me it happens in 0.7D3, not 0.7D. I included both of the video clips and replays. The clipping exists even if I play the 0.7D replay on 0.7D3. The prevent clipping is set to no.
Edit: Like anything in the sound menu can make it worse or better, I can change the pulse tone to make it worse, but exhaust pipe tone and the MIXes also can. So its a mix of everything am guessing. Needs more investigation.
Edit 2: I can make default sounds also clip if I increase the header MIX
Edit 3: In 0.7D, the header mix can also make it clip, but the sound isnt as high pitch
-
(Viperakecske)
DELETED
by Scawen : off topic / spam
-
(superlame)
DELETED
by Scawen : spam
-
(Scawen)
DELETED
by Scawen : off topic / spam
-
(Viperakecske)
DELETED
by Scawen : spam
-
(Viperakecske)
DELETED
by Scawen : off topic / spam
Having looked in the code and got a better idea what the old LFS output distortion was really about, I think that it is important to keep it.
My current feeling is the old LFS sound should be reinstated. Maybe the experimental "Prevent clipping" option should still be there but now it would not be enabled by default.
To explain, for those who like formulas:
The old LFS output distortion uses a formula like this:
v = bound(-1, v, 1)
v = (3v + v^3) / 2
The thing about this is that the derivative of the final function is zero at -1 and 1 so it avoids the click associated with clipping.
The 'experimental prevent clipping' option uses a formula like this:
v = v / (|v| + 1)
This prevents clipping in a different way that is quieter than the traditional method.
The new D2/D3 default was simply this:
v = bound(-1, v, 1)
Which causes the clipping displayed in LV's D3 video because there is no rounding of the wave at the clip point.
Someone might have a better idea. But for now my thought is that the traditional LFS one was better than I thought so I shouldn't really have removed it. The new fully linear one produces a bad clipping effect.
Thank you for the feedback.
I've reinstated the traditional LFS distortion / clip softening function in Test Patch D4. https://www.lfs.net/forum/thread/102117