Dynamic Difficulty Creator

[Alex360]

Can you point me to ODLC with chords that have ignore status?

ignores system built upon SOAD song for bass so it may have uncertain logic for chords ignores :\
even tho it's fine with NDD version of the song

[Chlipouni]

@@Alex360,

 

The link is in the post from Billkwando ...

[Alex360]

@@Chlipouni not sure if Cary on my wayward son has ignores, i need official example tho...

you think it's because of those mutes have douplicate ignore statuses?

it may be the mastery too I think

[Chlipouni]

@@Alex360,

 

By testing the song, I saw that a few sections never leveled down when I let the song running without playing a note. It maybe the case when the only notes in difference between two levels have all the ignore status.

Another useful information is that when you use DD, the percentage of played notes depends on the playable notes inside the current level ...

[Billkwando]

@@Chlipouni not sure if Cary on my wayward son has ignores, i need official example tho.

You weren't confusing my signature with the link were you?

[Chlipouni]

Update v3.1 :

   - The order of XML attributes has changed for note, chordNote and handShape tags

   - The feature that automatically adjusts the sustain of notes has been removed

   - Fixed bug : DDC may crash during the check of overlapping sustains

[Billkwando]

I think maybe Remastered has broken DD. Or, something.

 

I turned the difficulty override on, played my CDLC Kijun, by KREVA, and saw myself miss at least 4 or 5 notes (not that I should've but....it always fails me a couple times) and my 98% score jumped up to 100%.

708

[Chlipouni]

 

I think maybe Remastered has broken DD. Or, something.

 

I turned the difficulty override on, played my CDLC Kijun, by KREVA, and saw myself miss at least 4 or 5 notes (not that I should've but....it always fails me a couple times) and my 98% score jumped up to 100%.

708

 

 

I have just tried a few songs :

- Official DLC : works normally

- Official CDLC : works normally

- Custom CDLC without DD : works normally

- Custom CDLC with DD : 100% bug

 

For your information, XML arrangement files don't contain these kind of informations, so DDC doesn't manage them.

May be a solution in the other XML or JSON files managed by the RSToolkit ?

[Alex360]

Possibly overmastery bug with mastery values, now RS take those values in account i guess.

It usually off like this, Officials mastery about 0.6
customs like 1.3-1.6 values... 100% granted if this is multiplier for scoring mastery per note I guess?

[firekorn]

Note : IT IS NOT LINK TO DDC, my custom made with DD inside EOF have the exact same issue and some custom with DDC don't have the issue at the same degree (not go to 100% directly but still mastery superior to accuracy even at 100% difficulty)

[Berneer]

Bonjour @@Chlipouni,

 

I am trying to weed out my final imperfection in final testing of my latest custom, Blaze Bayley - Stare at the Sun. It's not a big deal, but since I've spent so much time on it  I'd like to try to remove this blemish.

 

See this passage where an 1/8 of a measure with no notes shows a floor-FHP of 5. I pause it in the animated gif.

 

http://i920.photobucket.com/albums/ad41/Berneer/CustomsForgeStuff/Questions%20for%20CF/Temporary%20half%20beat%20%20floor%20FHP5%20when%20should%20be%204/Temporary%20half%20beat%20wrong%20floor%20FHP5_zpsfceodniu.gif

 

In EoF, it is authored like this. Sorry Customsforge is bugging and won't allow me to post an image other than via a URL.

 

 

You could see I force the FHP to 4 a quarter beat before the b5 note, and FYI I place a new section and phrase at the beginning of measure 247, which contains no note until a quarter beat later.  I am suspecting DDC doesn't is not comfortable with this but I don't see any indication in the DDC log. All I could find in line 10960 and 10961 in one of the difficulties, a peculiar absence of anchors between the B5 note and the previous anchor being a huge 15 seconds earlier with an FHP of 5....

 

Here are the lines of code in question:

        <anchor time="398.692" fret="5" width="4" />
        <anchor time="413.475" fret="4" width="4" />

But as you could see when I pack the CDLC without DDC, it works fine as shown below:

http://i920.photobucket.com/albums/ad41/Berneer/CustomsForgeStuff/Questions%20for%20CF/Temporary%20half%20beat%20%20floor%20FHP5%20when%20should%20be%204/PreDDC%20works%20as%20desired_zpsuars7sft.gif

 

I am using DDC 3.1 from RSToolkit 2933f3b1 (dated 24Oct2016).

 

Would you know how I can avoid this issue?

 

I have made a Google Drive folder with all pertinent files, in case you needed to inspect more closely.

 

Thanks.

[Rockfirstlast]

Berneer:

 

I didn't look at your files (sorry if I'm wrong) but I bet there is a start of a section/phrase where that little glitch is.

 

Try placing a fret hand position of 4 at the start of the section/phrase, even if there isn't a note there.

[Berneer]

Thanks rockfirstlast.

 

Just tried it. It doesn't fix the issue unfortunately. It looks like this instead. Early in the song, in the first of 3 iterations of this same phrase, it looks slightly different but still with the issue. Again slightly different but still with issue in 2nd iteration of this phrase. Interesting the first two iterations highlight the floor as if FHP was 2. THe 2nd iteration shows a floor FHP of 2 but with width 5? The inconsistent look suggests it is a game glitch?

 

Aside: Is anyone else having the same issue of not being able to post images on Customsforge for at least a week now?

[iminashi]

I was just about to make a post about this myself.  :)

 

Here's a screenshot from my custom. A section starts where the FHP 6 is defined.

 

http://i.imgur.com/uGbjH9B.png

 

Importing the XML made by DDC back into EOF shows that DDC moves the FHP definition to the first note in the section, removing it from where the section begins. In this case it makes the anchor zone disappear in game. 

[Chlipouni]

@@Berneer,

 

  I am not sure to understand your exact problem.

 

  First two things, it is necessary to define :

   - The finger positions displayed on the strings are managed by "handShape" tags in the XML file.

   - The frets range displayed with yellow numbers under the fretboard are managed by "anchor" tags in the XML file.

 

  My understanding, about your issue :

  - the displayed finger position "2" is normally associated with the previous A5 chord, and its durability is defined by its "handShape" ("endTime" - "startTime"). So we still see the "2" finger, but it is not related to the next single notes, but simply to the previous played chord.

  - the next sequence of single notes starts with a new section and a new phrase, but without any "handShape" (no chord or no FHP defined for this sequence of single notes). So this situation may explain why the previous FHP is remanent (there is no new FHP to be displayed instead of it).

 

  So, to resolve this you should either :

  - Reduce the durability of the previous FHP (a shorter "endTime" for the associated handShape)

  - Define a new FHP for the next sequence of single notes

[firekorn]

@@Chlipouni The problem is that the FHP (or anchor) move in a way that isn't actually define anywhere in EOF by the user. The anchor move to the 5th position (with ddc) while it should move to the 4th (as define in EOF).

 

DDC doesn't care about FHP that are define inside EOF if there's no note on that position which can lead to difference in anchor placement after DDC.

[Berneer]

Yes, exactly as Firekorn wrote.

 

I did set FHP of 4, for the single notes. I tried shifting it a bit to the left and to the right but the same problem occurred with DDC and did not occur without DDC.

 

Thanks a lot for your attention. I really appreciate it!

[Chlipouni]

@@firekorn, thanks for your explanations.

 

DDC doesn't care about FHP that are define inside EOF if there's no note on that position which can lead to difference in anchor placement after DDC.

 

When DDC loads each note and each chord, it has to synchronize it with the current "handShape" and the current "anchor".

As the anchors don't have an end time position, DDC uses the start time of the following one to define its range time.

 

So, DDC is able to manage anchors that have a different time position than the associated notes or chords.

With this logic, only one anchor can be associated to each note or chord.

 

@@Berneer, the animated gifs were inverted in your post, weren't they ?

 

In your specific case, DDC adds a new anchor for the first single note in the new section (with a correct fret position of 4), but between the start of the section and the first note, there is no anchor and RS2014 uses the previous one of the previous section.

I will try to force the time position of this new anchor to the start time position of the phrase if the note is the first one in the current phrase.

 

Edit : New psarc file with DD generated by DDC v3.2

[Berneer]

Thanks for looking into this @@Chlipouni!!

 

I will try to check out your new psarc tonight but it's tight for me this weekend and I might only be able to check Sunday evening.

 

I did not invert the animated gifs in my earlier post.

 

Top animated gif --> After DDC

Bottom animated gif --> Before DDC

 

When I say "floor FHP" that was in contrast to yellow highlighted fret numbers below the strings which are goverrned by the anchor XML code. Well both should be governed by the anchor XML code. The 'floor' refers to the actual note highway that changes from black to dark blue in the fret lanes with upcoming notes. It is the "floor FHP" and not the yellow numbers that are momentarily wrong in the space with no notes between section start and first single note in the section (B5). 

 

When you click the link to the 3 screenshots from post 338 where I did as per Rockfirstlast's suggestions of moving the FHP=4 to the beginning of the section instead of being at the first single note, B5,  after section start, each screenshot behaves a bit differently. The 1st link (3rd occurence in the song) shows a momentary floor FHP=5, width 4, the animated gifs and EoF screenshot  are based on this occurence. The 2nd link shows the 1st occurence of this same phenomena in the song, in this case with a "floor FHP = 3,width 4". The 3rd link shows the 2nd occurence in the song and it also has a "floor FHP = 3, width 4. Sorry if I wasn't clear on that earlier. In each occurence, the preceding event is the same power chord with same FHP. But really we only look at the 3rd occurence screenshot for which I provided animated gifs and EoF screenshot and solving that will solve the issue. Perhaps you've already solved it. :)

 

Regarding: " but between the start of the section and the first note, there is no anchor and RS2014 uses the previous one of the previous section." 

 

There is an FHP definition, and by extension an XML anchor initially at the B5 note,but now at section start since Rockfirstlast's suggestion and shown with in-game footage  in post 338.

 

Regarding: "I will try to force the time position of this new anchor to the start time position of the phrase if the note is the first one in the current phrase."

 

That is what I did in EoF upon Rockfirstlast's suggestion and the problem persists. So I'm unsure this will solve the issue. I will try my best to confirm late tonight. Soccer party till 3am & waking up at 5am to drive out of town 3 hours away..... but I am burning to solve this so sleep be damned. :) 

 

Bon weekend!!

[Chlipouni]

I was just about to make a post about this myself.  :)

 

Here's a screenshot from my custom. A section starts where the FHP 6 is defined.

 

http://i.imgur.com/uGbjH9B.png

 

Importing the XML made by DDC back into EOF shows that DDC moves the FHP definition to the first note in the section, removing it from where the section begins. In this case it makes the anchor zone disappear in game. 

 

Can you share the psarc file without DD inside ?

I would like to check if the new release of DDC solves your issue ...

 

Thanks

[Chlipouni]

The 'floor' refers to the actual note highway that changes from black to dark blue in the fret lanes with upcoming notes. It is the "floor FHP" and not the yellow numbers that are momentarily wrong in the space with no notes between section start and first single note in the section (B5). 

 

Ok, I never thought to look at these level of details  :D

So, now, if we consider the changes of patterns of the carpet, I can affirm that DDC v3.2 fixes the issue.

It fixes too the yellow number issue (in your second animated gif, you can see that the minimum fret is 4 but changes to 5 in the middle of the single notes sequence).

 

Finally, we can add an other rule for RS2014 :

- the first anchor of a non-empty phrase should always start at the time position of the phrase.

[iminashi]

 

Can you share the psarc file without DD inside ?

I would like to check if the new release of DDC solves your issue ...

 

Thanks

 

 

I would, but I already went ahead and made the DD manually.  :)

 

It is the same issue that Berneer is describing, so it should be solved.

[Berneer]

Great stuff and fast work @@Chlipouni. Your fix resolved the issue completely and I am thrilled!

 

Here is the 3rd occurence of my original issue:

New animated gif

http://i920.photobucket.com/albums/ad41/Berneer/CustomsForgeStuff/Questions%20for%20CF/Temporary%20half%20beat%20%20floor%20FHP5%20when%20should%20be%204/DDC%203.2%20fixes%20my%20issue%20-%20thankyou_zpscy342cwt.gif

(trying posting image from Google drive, doesn't work, any special procedure to post an image from Google Drive? )

 

I also prefer how the yellow number FHP below the strings behave. So it seems you have somehow improved that too.

 

When I come back in town I'll have to ask you some questions about handshapes working differently from top level to lower level, such as starting later a lower level than a higher level or even splitting up and making a handshape hole in the middle of an intended handshape. I was just going to accept it and post my custom, but I thought maybe this was an issue you'd like to know about and why not make it perfect? :)  I'll post more details later, but they are the same details as I had noticed while testing chordify in my latest long chordify post I made a month ago. In fact if you check my custom I posted a few posts above in that Google Drive it's already present in that custom in the intro when you level down 1 or 2 levels. But I'll show specifics ASAP.

 

Hmm, 4:13am. Supposed to be waking up in 47 minutes.... screw it, staying awake and will watch newest MacGyver episode! At 41 years old, it should take me 12 days to recover from this.... :) It's ok, we are the soccer champions and it feels great!

[Berneer]

So using your DDC 3.2 psarc you linked above for my Blaze custom, I wanted to show you the following.

 

1. I did notice DDC 3.2 seems to fix the "Handshape starting later than in a higher difficulty" problem I reported in the chordify testing thread and which I thought I noticed in my DDC 3.1 packaging of the custom linked in Post 340 but can't seem to find anymore. I guess I'll confirm it when I resume chordify testing.

 

2. In the intro of my custom I noticed when leveling down 1 step the disappearing of every 1st of 2 concatenated handshapes (by concatenated I mean separated by 2ms, to give semblance of a partial handshape spanning the multiple concatenated handshapes). See below:

 

Here it is in EoF:

 

Can you please refresh my memory by confirming the following:

DDC intends to keep handshape marking (lane highlights and all fingerprints, even those of notes not appearing in lower levels) all the way down to the bottom level? If yes are there circumstances when it is expected that DDC will not do this. I have found some instances in my custom at lower levels where some or all handshape markings are dropped.  Just curious the criteria to know if the behaviour I see in-game is expected.

 

Thanks!

[Chlipouni]

DDC v3.2 Beta is available in the first post for testing purposes :

   - New process to force the density status of chords depending on their position in a sequence

   - New parameter in the configuration file to adjust or not the density status of chords

   - New message in the log if the time position of an anchor is adjusted 

   - Fixed bug : Wrong time position used when generating a new anchor at the beginning of a phrase

   - Fixed bug : HandShapes and anchors may be splitted at the beginning of a new phrase

 

@@Berneer, can you test again with the new beta release of DDC v3.2 ?