Posts by piotrmaj

    I'm not really good at debugging via forum, but at first glance you're using INPUT_PULLUP, which I think makes your button normally HIGH and LOW when pressed. Your code appears to assume your button is normally LOW. Check your output using serial monitor or oscilloscope to make sure that correct values are sent.

    I was out of town for a few days - checked today and this sequence of CC works for me (I'm sending them from Logic to KPA). You might have a bug somewhere else in your code.

    If you replace line 13 with something like:


    MIDI.sendControlChange(99, X, 1); // this will set MBS of parameter (on Kemper it is address page)

    MIDI.sendControlChange(98, Y, 1); // this will set LBS of parameter (on Kemper it is address number)

    MIDI.sendControlChange(6, V1, 1); // this will set MBS of parameter value (first 7 bits of 14-bit value)

    MIDI.sendControlChange(32, V2, 1); // this will set LBS of parameter value (last 7 bits of 14-bit value)


    This will effectively be an NRPN message.


    You have to look up X and Z in Kemper's MIDI manual (look at Downloads section) and also read in manual how V1 and V2 values are interpreted by KPA.

    Instead of sending one MIDI.sendControlChange you'll have to send 4 CC messages, one after another. Each of those 4 messages will contain a quarter of information needed to form NRPN message. Paste some code - will be easier to help.

    NRPN is nothing else but 4 MIDI CC messages in a row. Two first CC choose parameter, two latter value. Two messages are used to jump from 7-bit values to 14-bit values. In both cases first message sends most significant bits, second message sends least significant bits. When you already built your own MIDI controller which can send MIDI CC all you need to do now is to send 4 MIDI CC messages on a single key-press with correct values. Values for CC and parameters are described in document: "PROFILER MIDI Parameter Documentation", which can be fetched from Downloads page, User Manual section. Good luck!

    Nice, congrats! Reminds me of me buying KPA 5 years ago. :) Also couldn't stop playing for hours! The good thing is that even after 5 years I still enjoy it like on the first day. Gear for life! Have fun, WillemNL!

    Tutaj to juz raczej niczyja opinia Ci nie pomoze. Musisz sprawdzic samemu - mi nie podpasował - chyba za dużo lat gralem na minitorach studyjnych i Kabinet brzmi mi obco - fajnie na czystych, ale na przesterowanych brzmieniach do absolutnej d.... Uzywam go teraz jako wzmak do klawiszy i jestem zadowolony.

    Stage boots 3x faster than Toaster / Rack. Would be great to have instantaneous boot, but KPA is pretty stable - it just doesn't reboot itself. The only situation when that happened to me is when too much MIDI information is sent to it (for example: when you scrub a project in DAW and have MIDI Chase enabled). Otherwise solid as a rock - you have much bigger chances to experience full stage power loss than KPA going bad in its own. Power glitch which affects just one devices on stage is very unlikely (typically on stage there is a metric ton of gear connected to the same circuit), unless you kick the plug or KPA unit itself - don't do that :)

    Latest RM (I'm on beta channel, don't know if it is released as stable already) solved issue with performance management being unrelible - it now works as advertised. I'm always switching to perf mode when managing performances because I typically want to try them out immediately, switch between slots to level volume, etc. Don't know of it is necessary to be in Perf mode to avoid issues but sure it is more practical.

    Technically speaking in full text search indices you are typically searching for base forms of words. Input string is split by whitespaces, each word it brought to its base form and the search is performed for each token separately. Last stage combines partial results and returns an intersection of them. What quotes do is they change this default behavior and input is not split into individual words but you perform "search by phrase". This means the search engine will try to find documents where there is word "Be" followed by "sure" followed by "to"... This is different matching algorithm and it this case it worked better.


    In your case there probably is a token in your query which returned zero results (can happen if some works are blacklisted or not indexed for some reason), so intersection was also empty.


    I know it is technical and you should not be bothered with it as a user. But "quotes" trick apply to pretty much all search engines out there.

    why is not a great idea to suggest someone to use social media?anyway is up to you guys if use or not social media ,and it was just a suggestion not to rely on social media.

    It is because social media, as they operate today, are platforms created almost exclusively to get information about you and sell this information to advertisers (among other bad things they contribute to, and which are way outside of scope of this forum etc). They are useful, can't deny it, but the price you and societies are paying for using them is higher than you think. Vendors are free to choose their communication channels but relying solely on social media for communication with customers is just not wise (you can find a lot of examples where companies were screwed by social media lock-outs) and excludes people who choose not to use them for this or other reason.

    interesting. care to share details?

    Sure. Open Terminal and type: $ leaks Rig


    About 2% of memory is leaked. This might not necessarily be cause of RM instabilities unless some of leaked object hold on to important resources, like file descriptors. Among leaked objects there is:


    1 (240 bytes) ROOT LEAK: <KPAComm::GetGlobals 0x6000002b23a0> [240]


    which is the most suspicious object - could this leak be responsible for RM not being able to reconnect to KPA after KPA was powered down and turned on on the next day (while RM is kept open all the time?).


    2 (224 bytes) ROOT LEAK: <Access::Dialogs::StandardAlert 0x600003cfda20> [32]

    1 (32 bytes) ROOT LEAK: <Access::FloatingWindow 0x600003ce8120> [32]


    tons of: 1 (128 bytes) ROOT LEAK: <CGPath 0x60000105c880> [128]

    tons of: 2 (144 bytes) ROOT LEAK: <dx::exception 0x600001b103c0> [80]

    Pretty much all setlist performances: 39 (68.5K) ROOT LEAK: <Access::SetlistPerformance 0x7f976730b690> [320]

    and rigs: 8 (7.92K) ROOT LEAK: <Access::ProfilerRig 0x60000c9e8160> [176]


    (Leakage of rigs might be a fallout of leaked performances).


    The way I use RM is I turn it on use it until it works. There are typically two outcomes:


    1) RM is hanging after some time with spiny spiny thingy - in which case I can't use it anymore and need to force quit.

    2) RM is no longer able to connect to KPA (I do power off KPA for night, RM is always open, computer itself might go to sleep).


    2) has two variants:

    a) RM just doesn't see KPA and that's it

    b) RM sees KPA but doesn't interact with it.


    Some other issues:


    Sometimes RM opens, sees my KPA, called Toaster, but number of rigs next to it is... zero. In this case I need to restart RM to recover.

    Recently, with latest update, for the first time I have to unplug USB cable to be able to reestablish connection between RM and KPA, so far I could always recover just by restarting RM.

    Recently, with latest update, for the first time I experienced RM hanging right after it started (Spiny, spiny thing). This never happened before.

    All this is happening on macOS: 11.3.1


    I suggest trying my workflow - don't ever close RM manually. Just keep it open and see what milage you get. My milage is pretty low - need to manually restart RM every single day - sometimes forcefully. But at least performance management works now without issues - took years to fix it but I'm really thankful and happy.


    Bonus bug:

    Open RM, slide right pane all the way to the right, making it skinny. Load different Rig. Make right pane wider (slide left). You'll get what is in the picture with no ability to resize columns (one has to resize the whole window to recover).