Hi Jeff, if you’re on windows and potentially moving GP files between machines I’d highly recommend making use of Rig Manager - and setting up your rackspaces to use the midi input from the Rig alias, rather than a physical device (or worse, Omni). That way when you open the GP file on a different PC, it’s super easy to use Rig Manager to update GP with the devices windows says is now connected, rather than having to update all your rackspaces individually.
This is particularly the case if you end up accidently plugging a usb device into a different port than last time, as windows will give it a different name which then breaks your rackspaces - Rig manager avoids this.
Thanks, Matt. Yep, I have come to use the Rig Manager pretty regularly because I use a different set of (duplicate) keyboards for gigs/rehearsals.
Originally I had an issue where my two keyboards were switched and I learned I could quickly correct that using the Rig Manager,
I have also had issues that widgets were not recognized even though the proper keyboard was.
So, now at the beginning of each gig I open the Rig Manager and check that the proper keyboard is responding and that my 3 buttons and 9 faders are properly responding (if not I reset them).
Hello everybody, I am new to GigPerformer and obviously late into this topic.
I am trying to switch from a full hardware setup to a fully software setup around GigPerformer and a Mac Mini M1. To get tolerant to potential computer issues, I was planning to add a second Mac Mini and the conversation above has been very valuable.
The question I have is before the two computers how do you recommend to manage and switch the (master) keyboards themselves. I have two of them, currently connected through a USB cable to a USB hub connected to my Mac Mini. If I run two Mac Minis, one option I was looking at is to use the MIDI DIN outputs instead, and add MIDI Thru boxes to duplicate them for each keyboard and then MIDI Merge to USB boxes to connect to each Mac Minis. Looks like a lot of plumbing, but it is the only solution that came to mind.
Do you have better recommendation to solve that part of the game?
What I understand is that you want a redundant setup in a hot-standby configuration, for the computers and you look for a way to connect the keyboard simultaneously to both. Well, technically, splitting the MIDI signal and send it to both computers could work. However when there is flowing data towards the keyboards/controllers things start to get messy: which of the computers is allowed to communicate back?
Main thing is that you might be trying to solve a problem that has little chance of occurring. And from my point of view, when you doubt the computers, you should also doubt the USB hub and the keyboards itself (I recently repaired a keyboard of which the A4 had broken out of its mounting). And let’s not forget using a UPS for the power (according to other people that wouldn’t be far fetched btw).
Of course, it also depends on your usage: if your living depends on it, then additional costs and configuration to get this done can be justified.
I don’t have personal experience with these kinds of double computer setups but I’ve seen that Radials KL8 offers redundant USB input and you can switch between computers by pressing button on the rack unit or with footswitch.
Yes full redundancy can be expensive and should probably be justified by the chance of error and how fast you want failover to occur. Cheap solutions involve manually plugging and unplugging things but the failover time is slower. I did a bit of work with redundant switching over a network but have not really thought that much about redundant keyboard switching. I’ve had no experience with these USB switch devices but I consider that they need to be able to fake the device on the in-active system while working on the active system otherwise the USB drivers on the host systems may cause havoc.
I can envision 2 BomeBoxes each as a USB host with separate USB keyboards attached… The BomeBoxes would be networked to 2 Mac’s over Bome Network. But again, that solution might be a bit costly given the cost of a BomeBox. Although for professional musicians that are successful another $500 USD for a couple of BomeBoxes might be a nit considering the risk of killing a performance at a large venue. Bome has some musicians running 4 BomeBoxes or more BomeBoxes in various stage configurations.
Also, you might consider we would now move the single point of failure to the network (unless they are on separate networks).
Thanks to all for your inputs - a lot of food for thought regarding investment vs. actual risk of failure. And a couple of interesting hardware to look at. Much appreciated. Pierre
Are there any new insights in this area? This discussion has just flared up again for me:
I’m currently programming my setup on GP. Until it’s finished and ready for live performance, I’m still performing with Logic, and have been for 20 years. I currently have 35 songs programmed in GP. The loading time is already 1.5 minutes (I prefer to have predictive loading disabled). During my last gig (with Logic), the computer crashed completely for the first time. Normally, Logic crashes, but not the entire computer. The computer restarted due to a problem. The process until I was ready to perform again took 5 minutes. For live performances, this is an eternity. If I perform with GP and I’ve programmed 150 songs, it can take up to 10 minutes to be ready to play again. Therefore, I’m also looking for a solution where two computers run in parallel and share the audio interface. I don’t need a solution that automatically switches systems within milliseconds so the audience doesn’t notice. That would be nice, though—we’re a live band. But I need to be able to switch to the second system within a few seconds so we can keep going.
As a layman: Both computers are running. My iPad switches the songs between both computers simultaneously. Only one computer is connected to the audio interface (I have the MOTU 828 from 2024). If one computer fails, I just plug the USB cable into the running computer and we’re good to go. Sounds simple, but is it technically feasible?
I also know that you should ideally have everything redundant, but in my opinion, the computer is the weakest link in the chain.
As for crashing, you need to see what happened. Gig Performer cannot crash your computer, it is an application. If your computer crashed then you either have a faulty device/kernel driver (perhaps your audio interface driver) or your computer has a hardware problem.
I’ve been considering these things to be prepared for the future. Perhaps there’s an idea I can implement. My goal is simply to be more relaxed on stage, without worrying about when the machine or software might crash next.
I’ll look into predictive loading in a live setting. I’m currently in the programming phase of GP. However, it only makes a small difference to the loading times. And as long as I haven’t found the error (possibly a loop), it diminishes my confidence.
I need to continue monitoring the computer crash issue. Regardless of whether Logic, the audio driver, or the computer hardware is the cause, a second system would simply be much safer.
Okay. That’s too small for me. I just checked with MOTU. The new 828 (but also the 848 and pre-2019) also have two USB inputs. I guess I bought my 828 (2024) too early…
Really? I just looked at MOTU’s site? I don’t see two physical USB inputs — I see a single USB input that can handle USB2 or USB3 but not both at the same time
My current approach to Gig Performer loads quickly, regardless of the number of Songs. The bottom line is that I put almost all of my plugins in the Global Rackspace. I control and enable/disable them in the Local Rackspaces with Variations.
The signal flow is as follows:
From the audio interface to the Global Rackspace for any pre-processing. I put my instrument plugins here.
To the Local Rackspace for any one-off instruments or effects. I keep this to a minimum. Mostly, it’s just straight wires back to the Global Rackspace.
In the final stage of the Global Rackspace, I have a looper, mixer, and limiter, then back to the audio interface.
My MIDI routing goes from the hardware to the Rig Manager. The Local Rackspaces process and route the MIDI to Virtual MIDI ports/channels. I can do splits and connect pedals as needed here. The Virtual MIDI feeds the instruments in the Global Rackspace.
The upshot is that I load my plugins just once, rather than over and over for different songs. This makes load times very quick. It scales really well. If I find that a big song uses too much CPU, I’ll add a second instance of Gig Performer with just a Global Rackspace and some instruments.
I’m finding that this approach makes it easy to create new songs. If I need a new set of instruments, I just create a new Variation and set the switches to enable them.
One thing I do is heavy filtering and constraining of MIDI messages to ensure no unwanted messages or feedback loops. MIDI feedback loops and stray Note On messages are the most likely reasons I’ve had the odd problem in the past. My current approach loads super quickly and is rock solid.
I bought my new setup about a year ago: a Motu 828 (2024) and an iConnectivity Mio XL.
The backup computer feature isn’t urgent for me. It would simply give me more peace of mind on stage. Once I’ve finished programming Gigperformer (next year), I’ll buy a backup computer and try out my idea (simply swapping USB cables). I’ll report back here.
Perhaps the two USB ports aren’t compatible with simultaneous operation. I have no idea. But it’s irrelevant anyway, since I already have my equipment.
Wow. That’s rocket science. It’s way beyond me. I’d need to watch videos to understand it. I guess I’ve taken the classic approach.
In my Global folder, I only keep things that are always the same.
In Rackspace, I have a separate Rackspace for each song. I only use the Variants folders if I’ve made minor changes to the same song. That was the easiest way for me. However, I also want to avoid always having the same instruments. The more songs I program, the more problems arise that I’m now trying to solve.
I really like the principle of GP. When everything is running smoothly and I haven’t enabled predictive loading, I can switch between songs super fast.
