ProffieOS v6.x ALPHA testing

NoSloppy · October 12, 2021, 5:48am

Yes indeed.
OK Done for tonight, but last 2 things to report before I go.
At the moment pairing is not working using this in the config.ini and having same # of files for preon-out and pstoff-in

ProffieOS.SFX.preon.paired=1
ProffieOS.SFX.pstoff.paired=1

2nd, These both compile fine, but using WavLen causes nothing to show on the blade. This works fine:

TransitionEffectL<TrConcat<TrInstant,Green,TrDelayX<Int<700>>>,EFFECT_POSTOFF>,

This does nothing:

TransitionEffectL<TrConcat<TrInstant,Green,TrDelayX<WavLen<EFFECT_POSTOFF>>>,EFFECT_POSTOFF>,

I tried with WavLen<> no specified EFFECT same result.

Good progress night!

profezzorn · October 12, 2021, 5:56am

This won’t work, because the pairing needs to be done on the sound that is first. So to link OUT to PSTOFF, you would want ProffieOS.SFX.out.paired=1.

The first one ought to work though. I will check that out.

I’m guessing that the WavLen issue is because for some reason the postoff wavlen is returning zero. Will check that too.

profezzorn · October 12, 2021, 6:47am

WavLen/WavNum should now work for EFFECT_PSTOFF I think.

profezzorn · October 12, 2021, 7:38am

Found & fixed a bug in the variable parsing.
Haven’t actually tested that paired/volume it works properly yet though.

NoSloppy · October 12, 2021, 1:13pm

So if this is system wide, should pairing OUT to HUM work?
And therefore even BGNLOCK to LOCK , LOCK to ENDLOCK etc…?

profezzorn · October 12, 2021, 4:21pm

Currently you can only pair things that play “gapless”.
In monophonic mode, “poweron” has a gapless transition to the hum, as do most effects in fact. However, in polyphonic mode, the two effects overlap, so pairing them is not currently possible.

I suppose we could potentially change the way this is configured to something like: ProffieOS.SFX.hum.paired=out; That would also have the advantage that it would be possible to do things like pairing in to out, or pstoff to preon.

NoSloppy · October 13, 2021, 4:03am

Tested EFFECT_POSTOFF - now working.
WavLen<> within the TransitionEffectL without EFFECT specified also works fine.

Pairing sounds - test conditions:
Same #of preon and out files.
Same # of in and pstoff files.
ProffieOS.SFX.preon.paired=1
ProffieOS.SFX.in.paired=1

result no pairing.

Changing the default here to true, it does work, so it’s close to working.

github.com

profezzorn/ProffieOS/blob/a023672e32840e1d7d8b513d4c040ec153bd61f0/sound/hybrid_font.h#L41

    
      
                strcpy(name, "ProffieOS.SFX.");
                strcat(name, e->GetName());
                strcat(name, ".");
                char* x = name + strlen(name);
          
          
      struct PairedVariable : public VariableBase {
          	Effect* e_;
          	PairedVariable(Effect* e) : e_(e) {}
          	void set(float value) override { e_->SetPaired(value > 0.5); }
          	float get() override { return e_->GetPaired(); }
          	void setDefault() override { e_->SetPaired(false);  }
                };
                
                strcpy(x, "paired");
                PairedVariable var1(e);
                op->run(name, &var1);
          
          
      struct VolumeVariable : public VariableBase {
          	Effect* e_;
          	VolumeVariable(Effect* e) : e_(e) {}
          	void set(float value) override { e_->SetVolume(value); }

Pairing EFFECT_PREON blade effect to selected preon.wav file - test:
( the above temporary setting default to true still set for testing)

ColorSelect<SELECTION,COLOR1,COLOR2....> where SELECTION is WavNum<> and no EFFECT is specified, random COLOR is selected along with random sound.

When giving it the EFFECT as in ColorSelect<WavNum<EFFECT_PREON>.... , the COLORs are chosen correctly and match 1:1

This is likely expected, as ColorSelect doesn’t have an EFFECT already, it’s just that it compiles, but no warning that it’s not going to work.
I can’t think of any kind of check to prevent this, it just needs to be done right.

profezzorn · October 13, 2021, 4:41am

Updated top post.

Will take a look at this:

profezzorn · October 13, 2021, 4:58am

Discovered that readVariable doesn’t accept dots as variable names.
That should be fixed, so maybe it works now?

NoSloppy · October 13, 2021, 5:34am

Working! 123

NoSloppy · October 13, 2021, 5:39am

Testing Gapless preon-> out again.
Slight pop at beginning of preon, at the seam, and at the end of out.wav.
Maybe that’s the 500Hz sine test tone I used? It does have zero crossings and loops seamlessly unto itself. Typical saber sounds and even testfont spoken files sound fine.
Or did microfades on sounds’ beginnings and ends need to change at all to make gapless work?

profezzorn · October 13, 2021, 10:46am

Gapless just means that the samples are played as if it had been one file. It doesn’t blend, fade or modify the samples in any way.

KlubiGan · October 15, 2021, 6:30pm

Regarding Edit Mode using FETT263 prop file…
It seems the maximum setting for clash threshold is stuck at 4.
I want to increase it to 6 to allow for a larger range of clash threshold settings available in the edit menu, but I haven’t had any luck. I have tried deleting global.ini/global.temp from my SD card and reuploading my config with CLASH_THRESHOLD_G 6.0

Here are the #defines I am currently using in my config file in case I have a conflict going on that I am not aware of:

#ifdef CONFIG_TOP
#include "proffieboard_v2_config.h"
#define NUM_BLADES 1
#define NUM_BUTTONS 2
#define VOLUME 3000
const unsigned int maxLedsPerStrip = 135;
#define CLASH_THRESHOLD_G 6.0
#define ENABLE_AUDIO
#define ENABLE_MOTION
#define ENABLE_WS2811
#define ENABLE_SD
#define IDLE_OFF_TIME 60 * 5 * 1000
//OS6 defines
#define FILTER_CUTOFF_FREQUENCY 250
#define FILTER_ORDER 12 // Higher filter orders also uses more cpu has to be an even number between 2 to 16 and defaults to 8
//#define DYNAMIC_BLADE_LENGTH
#define DYNAMIC_CLASH_THRESHOLD
#define NO_REPEAT_RANDOM
//FETT263 prop defines
#define FETT263_CLASH_STRENGTH_SOUND // enables Clash Strength selection for sounds
#define FETT263_MAX_CLASH 16 // optional define to set the hardest clash on saber range 8 ~ 16, defaults to 16 if not defined
#define ENABLE_ALL_EDIT_OPTIONS
#define FETT263_EDIT_MODE_MENU
#define FETT263_SWING_ON
#define FETT263_SWING_ON_NO_BM
#define FETT263_SWING_ON_SPEED 250 // 250 ~ 500 recommended
#define FETT263_TWIST_ON // automatically enters Battle Mode
#define FETT263_TWIST_OFF
#define DISABLE_BASIC_PARSER_STYLES 
#define DISABLE_DIAGNOSTIC_COMMANDS
#define KEEP_SAVEFILES_WHEN_PROGRAMMING 
//#define FETT263_SAY_COLOR_LIST // requires .wav files
//#define FETT263_SAY_COLOR_LIST_CC // requires .wav files
//#define FETT263_SAY_BATTERY // requires .wav files
//FETT263 prop battle mode defines
#define FETT263_LOCKUP_DELAY 200 // delay in millis to determine Clash vs Lockup 200 default
#define FETT263_BM_CLASH_DETECT 6 // max value to use clashes in Battle Mode Range 0 ~ 6
#endif

Fett263 · October 15, 2021, 6:57pm

If I recall 4.0 was the max value possible, is this right @profezzorn? All the max values are built into the menu system, you can’t override them. If it can be set higher then I’ll update the menu accordingly.

profezzorn · October 15, 2021, 7:00pm

Technically it depends on what the hardware is.
However, for Proffieboards, the maximum was essentially 4 up until proffieOS 5.x
In 6.x it is 16, assuming I wrote the code correctly.

Fett263 · October 15, 2021, 7:02pm

OK, so change max to 16.0?

profezzorn · October 15, 2021, 7:03pm

I should also point out that clash thresholds might need to be tweaked a little between 5.x and 6.x.
The acceleromer is getting read more often, which will change the sensitivity a little.

(And yes, I think you should probably change the max to 16, but I’m not sure if it’s useful to be able to make it that high.)

Fett263 · October 15, 2021, 7:07pm

OK, how about 8.0 (or something else)? Seeing as CLASH_THRESHOLD_G becomes the minimum value when calculating for the Clash Strength Sounds, if you set it too high there’s no range to calculate seeing as 16.0 is the maximum level detected.

profezzorn · October 15, 2021, 7:44pm

If you’re going to make it lower than 16, then you should also test it to make sure that whatever maximum you choose is reasonable.

Fett263 · October 15, 2021, 7:51pm

Fair enough, I’ll do some testing and update.