3 new? JJOS2 -v1.11 bugs need verifications

hi, program edit sample level not scaled properly:

i have a beat running, 2 kiks, in program edit, when one kik is turned down to zero [SAMPLE-F1] [LEVL=0].

but i can still hear it at a low level, its not quite at zero.

in fact any sample in the program editor isnt fully off at zero level...

as there are two areas in program edit where level is concerned , can someone explain why, cos one works ok for setting zero level (at the "out/fx" section but the other doesn't (at the "sample" section)

any ideas?

also, tempo Qlink control: range not as specified...

Range: Low:50% High:200%, doesn't factor by %, but SETS TEMPO TO 50bpm and 200bpm respectively

another : slider reset is not valid after 'slider' page (mode+pad1) is left.

when the qlinks are reset (mode+pad1, F6). even though the values are reset in the software, IF the slider is actually not at center (reset position), then the slider value is still dominant once the user has left "slider" page, ( say when going back to the main screen).

ugh, re:WHOLE......id like to be able to select the range of "whole" by hitting the pads to get the lower and upper limits

this is why it's good to have fresh blood in beta testing
i'm so used to what i use most often that i miss this kind of stuff.
let me look at it and see, will do it on my lunch break...

as there are two areas in program edit where level is concerned , can someone explain why, cos one works ok for setting zero level (at the "out/fx" section but the other doesn't (at the "sample" section)

one relates to the mixer, and the other to the sample.
this is good and important, because on drum programs i do most of my level setting at the sample level since you can set the level per layer.

the level setting is ALL layers, simultaneous.

if i want to turn down just 1 layer, i go to sample-level.
leveling whole pad (all 3 layers) = mixer level.

though i agree at sample level playback should be silent when = 0, the logical question would be why the need? when do you have a sample loaded playing back at dead silence, unless you were actually trying to max polyphony without any volume?

also, tempo Qlink control: range not as specified...

Range: Low:50% High:200%, doesn't factor by %, but SETS TEMPO TO 50bpm and 200bpm respectively

was your original tempo 100 bpm?

Nym wrote:

also, tempo Qlink control: range not as specified...

Range: Low:50% High:200%, doesn't factor by %, but SETS TEMPO TO 50bpm and 200bpm respectively

was your original tempo 100 bpm?

aahahahah what a fool, yup was 101 actually hehhe

didnt clock that

though i agree at sample level playback should be silent when = 0, the logical question would be why the need? when do you have a sample loaded playing back at dead silence, unless you were actually trying to max polyphony without any volume?[/quote]


nah, i dont want samples at zero for the sake of it, unless i'm doing one of those artsy "silent trax" that are very trendy at the mo'

it was purely to hear just one of two sounds in a pattern, a kind of solo really.....

i was setting filter settings, and wanted to focus on one sound at a time, but i could never get the other sound fully out of the mix, but did it in the other field in the end, but had to query that setup it just incase!

yo, again, thank you for doing your part in this.
you've inspired me to finally pin down/create a demonstration of a filter bug that's in os2

here's the short of it:

http://nymantics.com/comparison/

os2 filter sweeps are "chirrupy"

Nym wrote:yo, again, thank you for doing your part in this.
you've inspired me to finally pin down/create a demonstration of a filter bug that's in os2

here's the short of it:

http://nymantics.com/comparison/

os2 filter sweeps are "chirrupy"

amen to getting this one fixed - well spotted ... hope there's a fix soon.

Nym wrote:yo, again, thank you for doing your part in this.
you've inspired me to finally pin down/create a demonstration of a filter bug that's in os2

here's the short of it:

http://nymantics.com/comparison/

os2 filter sweeps are "chirrupy"

i would say that there seems to be a slight "envelope" on the filter response that is only more apparent when small slices are used and filter swept......i tried to recreate the same with a break over here,

if u flatten all and any adsr's filt and amp to be sure that the envs have NO part in the equation (ie: no A, no D, 100 S, no R, then i switched the LPF for a BPF set using the qlink (realtime cutoff - for quick manual focusing at various freqs - move it around, i found best audible results at lower values ), then if u turn up (slowly up to say a max of 90-95, dont want to be deafened btw!) the resonance value, u will definitely hear the exaggerated effect that the "chirrup" behavior has........in my experience this treatment would create a steady resonant "whistle" not a pulse-like resonance.

so it looks like theres some effect of the filter corresponding to the phrase of audio data, or at least some "peak - shaped" response at the start of any sample/filter event......


funny 'cos in the 90's kids would pay for that effect , giving that filter/gate effect that pads n voices went well through! .....but yeayh it shouldnt be hapnin.


nym , i thought it could be the short slices to blame in your example, but it does seem that the OS's flt is responding with a shape even when all adsr shapes are set to flat.


hmmm

there's definitely a shape on the filter, even with "no" ADSR shaping setup in GPE......

and only really apparent on the attack portion of the samples, long one's still do it the same, but they're just less obvious than when in sliced up bit repeatedly - like in yer demo nym.

ive even been compelled to do my own comparison file, but just in OS2 - not bothering with OS1 as i knew it would be fine, since i've not encountered it in OS1 before.

http://www.rob-b.f2s.com/arc/AmenFilterTest.zip

when loaded, use track mute's "solo" F6 to flip from track one to track two to get the difference of the loop as a whole whole sample thru a sweep filter, compared to the sliced up loop thru the same filter treatment, but you'll notice the filter has a noticeable 'shape' on the attacks of each slice instead of sounding as smooth as the whole loop sample.

thanks for the corroboration
i submitted this last night. it'll be nice to have a smooth filtersweep back.

jj wrote:The bug of 1 and a sound level corrected.
0-100 ==> convert to Log curve 0-32767.
The bug was shown in the table of Log curve.

thanks, arebee.

jj also mentioned that he's looking at the filter bug. however, i was unclear whether he was immediately able to fix it, as he said he needed an oscilloscope but his was broken. just in case, i sent him a link to a couple freeware oscilloscopes.

^ this all goes to show A, the speed of correction, and B, the power of an uploaded demonstration.

these are the bugs i'm finding in os2, all the while making some beats. multitasking.
http://nymantics.com/nym-loosecannon.wav
very early cut of some boom bap i'm doing, mind you it is 24 mb