Dropouts after a certain amount of time and high CPU usage with “just” piano & pad — corrupted gig file?

Hello everyone,

I hope you’re all doing well and having a musical Sunday! I’m currently facing an issue and wanted to ask if anyone here might be able to help.

I play in a cover band using Gig Performer (around 30–40 gigs per year) and I’m still amazed by all the possibilities it offers. Unfortunately, for the past 3–4 weeks I’ve been having an issue with my existing gig files: whenever I use a rackspace with piano (Pianoteq 8) and a pad (a combination of Arturia Matrix-12 V and Session Strings 2), I start getting audible dropouts after about 20–25 minutes.

In Gig Performer, I can see the CPU usage suddenly jump from around 20% up to 35%, which already seems fairly high for that setup. In the Global Rackspace I’m also running an EQ from FabFilter and a reverb from Arturia (Intensiv).

Initially, I was running fairly demanding settings on my MOTU UltraLite-mk5 (sample rate 44.1 kHz at 256 samples with a reported latency of 5.8 ms). Even after increasing the buffer size to 320 samples during yesterday’s gig, the dropouts were still audible. The MacBook itself wasn’t particularly hot or warmer than usual.

In Activity Monitor, CPU usage during that set was around +/- 60%. Luckily, we happened to play a lot of piano-heavy songs that night, so I opened a completely new file from scratch, loaded only the piano, and everything worked perfectly. That makes me think the issue must be related to my gig file itself.

The frustrating part is that I play in several bands, and because some songs overlap, many of the files are built on top of each other. So I’m afraid the problem may not affect just one file, but several different ones.

Here’s my setup:

  • MacBook Pro M1 Max

  • 64 GB RAM

  • MOTU UltraLite-mk5

  • Pianoteq 8

I’m also running an amazing script that pianopaul and ztones created for me in this forum. It’s still absolutely incredible — thanks again for that! It’s one of the reasons why I use Gig Performer in the first place. :slightly_smiling_face:

I’ve attached a file for reference — maybe someone has an idea for troubleshooting?

Due to the size, here is the file:

So GP has worked fine for years but has only started failing 3-4 weeks ago? Did you have that Pianoteq 8 and pads there before 3-4 weeks ago and were they working or did the failure start after you added that stuff? In other words, what has changed in your system at the point where you started seeing an issue?

Also, have you checked whether you have enough RAM on your machine so that it’s not swapping?

Thanks for your help and feedback! Regarding the plugins, I actually hadn’t changed anything in Pianoteq, Arturia, or the Native Instruments plugins (although it’s possible that I installed an update sometime last summer). In the meantime, I had also installed a few additional plugins such as Omnisphere or Keyscape, but apart from that I haven’t really made any major changes. I did have my mobile interface with me for a gig abroad (a somewhat cheaper MOTU), but I’m now running the gig file again with the original interface. Could that somehow be the cause?

The challenge is that the dropouts only happen after some time. So during a live gig, where a song might only last 4–5 minutes before I switch to the next rackspace, it might not even become noticeable. That’s why I can’t say for certain whether the issue has actually been there for a longer time or not.

So in theory, it shouldn’t really be a RAM issue either if I’ve never had any problems with it before, right?

I don’t know - if you are using more plugins (e.g. Omnisphere and Keyscape) which also load more samples then you’ll use more RAM. Might also depend on what else you have running.

With ram (unless you are using Predictive Loading) it all loads when the Gig File is loaded. So, if you have added other rackspaces with other plugins (Omisphere can be very ram heavy, from what I understand) it will use ram in the Gig File, regardless of whether you are using those plugins in a particular rackspace.

But, this is a Mac, so I won’t even fake knowing anything (I’ll occasionally fake knowing something regarding Windows machines. Hah!).

My old MacBook had 32GB of RAM. With over 150 songs, that wasn’t enough. My new computer has 127GB – it handles it easily. However, I also think that the RAM fills up during a performance (I just posted about this here) and thus replenishes my purged samples. I’m currently still in the testing phase and haven’t performed with GP yet. But I’m considering restarting GP during breaks. That would eliminate the need for refilling during the gig.

In your case, though, I don’t think it’s a RAM problem. Pianoteq isn’t sample-based. Perhaps the M1 is underpowered. My old M3 also had occasional dropouts, especially at the beginning of newly played sounds.

Thanks everyone for all the feedback and suggestions — really appreciated!

Also, apologies for the late reply. We had almost 10 live shows already in May, which, besides work, definitely consumed quite a lot of time.

In the meantime, I re-enabled Predictive Loading again, although interestingly I had been running without it for almost two years with essentially the same amount of rackspaces and never had issues before. I also updated a few plugins and moved to a newer macOS version.

So far, throughout May, everything has been running smoothly again and Pianoteq hasn’t caused any further problems. I’ll definitely keep you updated if anything changes.

Best regards,
Michael

Uhm, if everything used to work and now it doesn’t, then something changed on your computer - if you added more plugins, did you possible get closer to running out of RAM? That would certainly explain why predictive loading fixed the problem