Note detection inconsistencies in CDLC

[albatross213]

Apologies if there's a similar topic elsewhere - I couldn't find anything easily with a search.

 

Basically, there are several CDLC where people have noticed some inconsistencies with how the note detection works. Here is an example of me playing through a song where I've noticed some inconsistencies in the verses. The first two repeats in RR are the second verse with no misses. The next two are the first verse, where I get ~6 misses each time despite the riff being mostly the same. It looks like in the first verse (later examples) the notes arrive somewhat before the song audio, particularly early on, but not so much that I think I'd be getting misses on notes.

 

So, does anyone have any ideas on what might cause this sort of inconsistency, and how to fix it? It would be great to know if there's anything you can do to fix these sorts of things!

 

Just to be clear, this example is from the recent update of Domination (and the issue was there before the update), which I'm assuming was done with more recent versions of EoF and the Toolkit. So you should be able to get at XMLs and such by unpacking that.

 

The main things I'd think of trying would be to use a new audio track or to re-import the tab and maybe re-sync the song. Anyone have any other ideas of things to do, or experience with these sorts of problems?

 

And I'll just throw out a couple of developer names in case they have expertise on the issue or know someone who might: @@raynebc, @@cozy1.

 

Thanks!

[PC Plum]

Could happen if your notes are slightly off grid, or some of them are off grid.  Open up the EoF project and resnap the entire arrangement (remember to resnap any tech notes too).  Then save and make a new build.

 

I'm not 100% this will solve the issue but its definitely worth a go.

 

 

Also, I get missed notes all the time on repeated open E notes on bass.  Think the game engine's most difficult job is recognising open E hits on bass.  You could try using a plectrum or striking the strings in a different position (I tend to hit the bass strings as close to the bridge as possible for max recognition).  Again, might not solve the issue completely but its worth trying and experimenting with.

 

 

 

Good luck :D

[albatross213]

 

Also, I get missed notes all the time on repeated open E notes on bass.  Think the game engine's most difficult job is recognising open E hits on bass.  You could try using a plectrum or striking the strings in a different position (I tend to hit the bass strings as close to the bridge as possible for max recognition).  Again, might not solve the issue completely but its worth trying and experimenting with.

I have a 6 string, so I could pretty easily just play that part on the low string and see if that makes a difference - it definitely makes a difference with how long RS recognizes the note sustains, so it might work there. But even if that works it would be a pretty unsatisfying solution, as it wouldn't give any indication of what is different between the two verses. Which is to say I'll play around with your other suggestions. And try to be a good scientist and do any changes one at a time.

[firekorn]

i would bet on off grid note myself but i'll admit the recognition system is far from perfect and seems a bit too easily off for not much so it might be a bit complicated to know what is really impacting him...

[Alex360]

It seems that RS depend on Rhythm, for detecting notes I think off beat does all mess up, so fixing this in eof could still have some issues, since it cuts precise scale from 6 or 18 points to 3 after exporting to xml, so It may cause some shit, even if all perfectly snapped in eof.

I think we should test this if Raynebc can change export behavior or any of us two to test we need guy with good gear(like that at vid) two devs one of EOF and one of RSTK, test files to play with.

I'm here for 2 days to solve things, cozy is up too, rayne is up at morning by my time)