|
Post by david2 on Jun 17, 2018 3:42:39 GMT
To help you understand the test track better and how stuff should look after being filtered into separate tracks for different notes/pads, I’m including an annoyed version of the pic of the original recording: Again, I hit each part of each pad in turn about 4 times, sometimes doing it again (like when I got hung up on the high hat for a while). Most pads have two sections, the main part and a secondary "rim" or "edge" part, others such as the ride and especially high hat have more parts. The kick only has the one section though, so I only hit it for 4 hits at the beginning of this track, and then again a few times as part of a short combined-pads beat at the end
|
|
|
Post by david2 on Jun 17, 2018 3:44:13 GMT
‘Annoyed", lol. You probably are annoyed by now, i’m sure! I meant "annotated" or "labeled"
|
|
nic
Soapbox Supremo
Troublemaker
Press any key to continue
Posts: 2,011
|
Post by nic on Jun 18, 2018 9:03:00 GMT
Hi @david2,
Annoyed? No, not at all! That's all really helpful. First a couple of points:
1. I would disconnect the line between event monitor out and midifire in. That will result in the kick being doubled (assuming it was working, of course)
2. The event monitor (the part I can see) shows that only the kick events are coming from the Stream Byter, so we do know the filtering is working. Your rules look fine too. However, it would be worth checking that all pages of the events show only note C1 on and off events and nothing else. I'm pretty sure this will be the case but we should check.
Your shot of the setup for the armed track looks spot on; in theory, Auria should only be recording the events coming from the input you have selected, but I wonder if it actually is.
I think we should simplify things here and concentrate only on the kick to debug further:
- disconnect all lines in MidiFire coming from the MidiFire source to all the Stream Byters *except* the kick one - record a source track with just a couple of kicks and say the snare (something other than the kick) - send that source track to midifire and then create a new armed track with midifire as source (kick) - record play and then examine the newly recorded track in detail.
Assuming both the kick and snare events are in it, are any of these events doubled up?
Now, in the kick Stream Byter rules, comment out the NX 24 = XX +C line, leaving just the XX +B line. This will block all events.
Do the record test again. There should be *no* events in the resultant track.
There must be something basic we are missing. The above two tests might assist in pinpointing where we need to look.
Regards, Nic.
|
|
|
Post by david2 on Jun 25, 2018 5:50:37 GMT
Hi, Nic. I’m back after a bit of a break.
Concerning your above points 1 & 2:
1. I disconnected...Oops. Typing it out just now, I see I didn’t actually do exactly what you specified. Instead of disconnecting the line between the Event Monitor and MidiFire In, I disconnected the original line from the Stream Byter labeled ‘KICK’ to ‘MidiFire’ In so that I was only left with the line from the KICK Stream Byter Out to the Event Monitor In and then from the Event Monitor Out to MidiFire In. So, all events ran through the Event Monitor on their way from the Stream Byter to the MidiFire virtual port.
2. There were only C1 on and off events recorded in the Event Monitor’s log
Then, following your above advice in relation to debugging further, I - disconnected all lines in MidiFire coming from the MidiFire source to all the Stream Byters *except* the kick one - recorded a source track with just a couple of kicks and the snare - sent that source track to midifire and then created a new armed track with midifire as source (kick) - recorded, played and then examined the newly recorded track in detail.
Okay, so the result was that just the Kick events filtered through to the newly created Kick track in Auria. So far, success! But...
Okay, reviewing your above debugging instructions, I see I never got to the last step of “comment out the NX 24 = XX +C line”. I’ll do that later (I’m at work now), but for now I want to tell you the rest of my findings in a separate post.
Oh yeah, I saw in your App Store update that you raised the number of MidiFire virtual ports to 16 — you’re definitely on the job!
|
|
|
Post by david2 on Jun 25, 2018 8:25:05 GMT
Okay, so then I additionally connected the SNARE StreamByter to MidiFire1 (with an Event Monitor in between) and, after deleting the above-mentioned filtered Kick track that I’d previously recorded in Auria, recorded both the KICK and SNARE in respective tracks in Auria.
However, once again, both tracks, both the Kick track and the Snare track, recorded all events — no filtering occurred. The StreamByter for KICK showed all events for C only, while the StreamByter for Snare showed all events for D only.
I later recorded a new track (drummed again and recorded my playing in Auria) in which I hit all pads (all notes available on my drum set) and then I tried running that track through MidiFire and recording just Crash2 and the Ride simultaneously. No Event Monitors were used in MidiFire this time. In this case, filtering did occurr as only Crash2 and Ride were recorded (no events from other pads got recorded), but both tracks recorded all Crash2 and Ride events. The level meters moved identically in the two tracks and the resulting two recorded tracks were identical.
Going back, I disconnected the RIDE StreamByter and the associated MidiFire7 output port and was able to record just the Crash2 events in the Crash2 track, and then I went back once more, disconnected Crash2 in MidiFire, connected the Ride back up, went back to Auria and set that up to now record the Ride and not Crash2 and was able to successfully record the Ride events in a Ride track. In this manner, I went back and forth between the apps and recorded separate tracks for all 8 of my groups of pads (I’d really like 22 or so to get each available trigger a separate track).
So, I was able to dissolve the original MIDI track to separate tracks for each note (well, small groups of notes such as having both snare and snare rim in one group), but I had to do it manually. If I had a project of numerous 3 to 5 minute long songs with events spread through 20 notes (pad triggers) each, such method would take an inordinately long time
|
|
|
Post by david2 on Jun 25, 2018 8:51:31 GMT
Quoting myself from the previous post:
“I later recorded a new track (drummed again and recorded my playing in Auria) in which I hit all pads (all notes available on my drum set) and then I tried running that track through MidiFire and recording just Crash2 and the Ride simultaneously. No Event Monitors were used in MidiFire this time. In this case, filtering did occurr as only Crash2 and Ride were recorded (no events from other pads got recorded), but both tracks recorded all Crash2 and Ride events”
Regarding the above, I left out that *before* I got to trying to record Crash 2 and Ride simultaneously, I had first gone through each of my other groups (Kick, Snare, Hi hats, Toms, Crash1 and Splash) and individually filtered for each of them in turn through the MidiFire app and recorded individual tracks for them in Auria, one at a tie. *Then* I tried to simultaneously record separate tracks for the last two (Crash2 and Ride), but both tracks combined the Crash2 and Ride events (while filtering out the previously recorded events)
|
|
|
Post by david2 on Jun 25, 2018 8:53:14 GMT
“One at a tie” = One at a TIME
|
|
|
Post by david2 on Jun 25, 2018 13:43:28 GMT
Regarding your troubleshooting instructions, “Now, in the kick Stream Byter rules, comment out the NX 24 = XX +C line, leaving just the XX +B line. This will block all events.
Do the record test again. There should be *no* events in the resultant track”, okay, I’ve now done that.
As expected, when the NX 24 = XX +C line was removed from the StreamByter, no events were recorded in the corresponding track in Auria. Also, there were no comments in the Event Monitor (which I have now hooked up like you originally instructed, as a terminating separate branch off the StreamByter that is not inline with the path from the StreamByter to the MidiFire virtual port that sends to Auria.
Next, reinserting the NX 24 = XX +C line in the Kick StreamByter lead to a successfully recorded track of only kick events in Auria, with the snare hits filtered out. Only kick events (one ON message and one OFF message per kick hit in the short test track) were logged in the Event Monitor
|
|
|
Post by david2 on Jun 25, 2018 14:06:13 GMT
So, then I re-did the test of also hooking up the SNARE StreamByter (plus an Event Monitor) and then trying to simultaneously record filtered Kick and Snare tracks in Auria.
Unfortunately, both tracks show all combined events (no filtering occurred in either of the resulting recorded tracks).
Going back to the MidiFire app and looking at the Event Monitor logs, I can see that the KICK StreamByter recorded only the appropriate ON and OFF C1 events (there were 4 kick hits in the test track, so 8 messages total), while the SNARE StreamByter recorded only the appropriate ON and OFF D1 events (there were 6 snare hits in the test track, so 12 messages total).
It seems MidiFire is filtering, but somewhere after the StreamByters the events are somehow getting recombined
|
|
|
Post by david2 on Jun 25, 2018 14:16:30 GMT
And, just reconfirm, I unconnected the KICK StreamByter and it’s ‘MidiFire’ port in the MidiFire app, leaving only the SNARE StreamByter connected (to MidiFire 1) and re-recorded in Auria (with only the Snare track armed) and got the desired filtered track with only snare events
|
|
|
Post by david2 on Jun 25, 2018 14:31:13 GMT
"just reconfirm" = just TO reconfirm
|
|
nic
Soapbox Supremo
Troublemaker
Press any key to continue
Posts: 2,011
|
Post by nic on Jun 25, 2018 17:15:47 GMT
I set up a little test where I hooked up a Stream Byter generating 4 note events into 4 filtering versions (same as yours) and then on to 4 separate outputs, MidiFire to MidiFire 3.
I then used a another monitoring app to observe the events coming from the 4 virtual ports. This app also registers which port each event comes from.
Each port correctly received just the one filtered event in the other app on each port.
I can see no other place that these could be combined except in Auria. I am loathe to push the blame onto others, but I'm pretty sure we are going to need help from the Auria side of things.
I suggest you drop Rim an email and maybe with a pointer to this thread. If he needs any input from me I would be happy to assist.
Regards, Nic.
|
|
|
Post by david2 on Jun 26, 2018 1:23:59 GMT
Okay, I’ll look into how to contact Rim
|
|
nic
Soapbox Supremo
Troublemaker
Press any key to continue
Posts: 2,011
|
Post by nic on Jun 26, 2018 8:01:58 GMT
If you would like me to get in touch with him, let me know.
|
|
|
Post by david2 on Jun 27, 2018 10:56:23 GMT
Okay, thanks for the offer.
As I just sent him an email earlier today through the Auria Discussion Board contact tool, perhaps it would be best to wait a few days and see if he responds?
If, however, you think it would be best for you to just go ahead and contact him directly and talk developer to developer, I’d be completely comfortable with that. Whatever you think would be most effective
|
|