Youāre right in the sense that the output cannot be monitored by the GP MIDI monitor. But in the current version, GP opens all available MIDI inputs, including loopBe1 of course. So you will not see the outgoing messages in that instance but you do see the exact same messages as incoming ones, as in the screenshot below. So there is no need for a secondary instance in that case.
Regarding the redirection of MIDI messages:
In this case, not having an extra instance is very likely to be problematic any PC messages assigned to rackspaces in that instance can not be redirected.
You are right I focused on the MIDI monitor plugin in the connection view, but the one used in the screenshot is the general MIDI monitor which should also display the Loopbe1 in. You are right.
For testing purpose, I just created two rackspaces in two different GP instances and I perfectly synchronized the two instances. When I change the rackspace in the first intance, the corresponding PC is send to the second instance and the rackspace changes accordingly. No problem.
Perhaps I misunderstood what @FabryMusicMan exactly wants to doā¦ thatās why I sometimes like to summarize.
But you are generating the PC messages for the second rackspace in the first one, right?
As I understood, @FabryMusicMan wanted to change rackspaces via PC messages from an external device in 3 different GP instances. I figure that in this case, the straight way is to have those 3 instances react to the PC messages from the foot controller (and on Windows use loopBe1 as a proxy because of poor driver support). Having a single of these instances being different in a sense that in each rackspace, it must contain a MIDI out block which on activation sends PC messages to the other one (presumably the same by which the rackspace in this one was activated) introduces another layer of configuration and a potential source of error.
Sometime in the future, proxyMIDI will become available and let you work around those multi-client MIDI woes on Windows more easily
If one would want to do more sofisticated multi-instance switching (e.g. trigger arbitrary combinations of rackspaces in the 3 instances with a single PC command without the need of all 3 having the same PC number), that can be achieved easily by extending the master/first rackspace.
Thatās exactly what I did. No extra instance, but the first one is a bit different because it receives the MIDI messages directly from the MIDI hardware while other instances use Loopbe1 MIDI input.
And only one Loopbe1 MIDI out block with the Program Change option checked,the right PC number AND the āDonāt block Program Change Messagesā checked has to be added to synchonize a rackspace in another GP intance. And of course, if you add more Loopbe1 MIDI out block with different settings you can send different PC to different instances (the good idea is then probably to use a different MIDI channel for the Program Change Messages, one per GP instance). But if I remember well the initial idea was to split a single instance rackspace, into rackspaces in different GP instances, so I supposed @FabryMusicMan wants all rackspaces to be synchonized according to the same PC.
The thing that I like about having an extra instance just for the MIDI control stuff is that you donāt need an extra block in each rackspace that you have to remember to modify if you change the program change number that activates that rackspace. Iām fairly sure, that with a decent number of rackspaces, I would overlook or forget something and introduce an error
But maybe this just comes down to taste (Got to admit, I like the symmetry of having the 3 instances respond to MIDI the exact same way)
ok, now that thatās been resolved, I have a question related to cpu usage, as the thread title mentions. If someone wants to move it to its own thread, fine by me.
Last night I was working with IK B-3X and having some cpu issues. When the UI was open and I played at same time I was getting spikes up to 100% (when UI was closed it was down in 30s and playable). Anyway I happened to see in task manager that my cpu usage was around 7% while GP was maxing out. Now I understand the above statement, I thinkā¦but this seems out of whack. Am I not getting as much power out of my laptop as I shouldā¦ or am I misunderstanding something?
Open or closed plugin should never really impact its performance and the audio CPU usage.
Basically - it would be ok for you to keep seeing the 30% of audio CPU usage but your task manager showing the slightly higher usage when UI is open.
Also - there seems to be some issue with some plugins where they somehow impact the audio cpu usage calculations and the number jumps to solid 100% when it obviously isnāt. Weāll investigate.
ok, well this morning I returned to the same exact session left open from last time, so absolutely nothing changed. While I am still getting similar results they are not nearly so extreme as before.
Playing B3X UI not openā¦ GP 30%, task manager 7-8%
Playing B3X UI openā¦ GP 50-70% (crackling) task manger 20%
Playing Repro5 (8 voices) same with UI open or closedā¦GP 48-50% task manager 20%
Iāve sent a message to IK, but also just wondering if that is normal for task manger and GP to be showing such different %
Yes it is. The task manager shows your CPU utilization as a compound percentage of all physical and virtual cores you have. This has been discussed in detail already.
This is obviously an issue with the plugin. A plugin display should never impact itās audio performance not have anything to do with how well the plugin performs audio-wise.
Okay, I apologize if my eyes glazed over a bit reading through this threadā¦ and Iām still not sure I really understand whatās being said. So if I have a laptop dedicated to just running GP, would I theoretically be better just having one single core? So GP is then able to take advantage of all the power the laptop has?
Various plugins do and can take advantage of multiple cores.
For example Kontakt has a setting that allows you to run it over specified number of cores (take a look at the Kontakt settings) so no - you would not be better off with just one core.
Additionally - many other things are happening like graphics processing etc which is most likely done on a different core
yeah stupid question, I guess. I just did a little quick research to better understand cpu, cores, threads, In my naivete, I was seeing it as all this unused potential processing power in my laptop, which was upsetting to meā¦ and was seeing it like having dual core was splitting the processing power. I think I am understanding it is just about speed and obviously having just one core doesnāt mean I would get twice the speed.
Hi! Can you tell us the model of your cpu? A dual core seems to be an old processor. Talking about laptop CPUs, everybody who works with live music should avoid a āUā processor and take a āHā or āHKā processor (like i7 8750 / 9750h or i9 9980HK, talking about the best).
Thatās because the āUā processor are designed to work in tablet and ultrabooks: they have a TDP of 15W instead of 45W thatās the standard for performance/gaming/workstation laptops.
Thermal throttling is unavoidable on a not powerful laptop after hours of intense usage, even for us that donāt use GPUs a lot.
Talking about your specific situation, I think that dual core is not enough anymore. You should have at least a quad core i5, better if 8th gen or further.
Modern CPUs have hyperthreading, more cache and they donāt run too much slower when in multicore.
@FabryMusicMan Oh believe me, that much I understand!! I think I bought my current gig laptop back in 2011, maybe earlierā¦Intel Core i5-2520M dual-core processor (2.5GHz, 3MB cache, 3.2GHz Turbo frequency, 8gb ram.
My needs are pretty simple however, I am not generally running a whole bunch of things at the same time. In the last year I added Amplitube 4 and Repro, both of which kicked up cpu usage significantly, but Iām running them without problems. So for me its a matter of if it aint brokeā¦ and money always an issue.
But yeah, having problems with B-3Xā¦ but not sure what thatās about at the moment, not seeing anyone else complaining of UI affecting performance. And it doesnāt appear that B-3X is a must buy for me anywayā¦
IK is quite renowned for their first generation plugins to beā¦ unstableā¦ and cpu intensive. In my experiences with IK products, they are especially unstable on PC. I used SampleTank3 on windows 7 with no problems. Migrated to a windows10 laptop and it was crash city; and over at KVR many people have said the same.
I do miss my Swiss Army knife samplerā¦
IK is also, unfortunately, rather unresponsive to fix issues on older products (like ST3) ā this is totally just my experience though and ymmv. For all I know, itās just a quirk of my system!
(Got to admit, I like the symmetry of having the 3 instances respond to MIDI the exact same way)
