Cubase 13: random CPU spikes still here, is maddening

Cubase
145

I completely nuked the Stereo Out and both my Control Room outs (headphones & monitors) and rebuilt them, based on something I was seeing in another thread mentioning a bug that would somehow relegate them to ASIO Guard something or other. So far, no spiking when scrolling the MixConsole window horizontally. Progress!

I very much wish Steinberg would focus more on making sure old issues like anything that would necessitate “trashing the preferences” and any other such nonsense in 2024 would go away for good.

For instance, if I have a template built on, say, Cubase 10, it should work just fine in 13, because 13 should be smart enough to be able to automatically rebuild it if necessary, discarding any old formatting that may cause issues and/or recreating everything necessary in the new format.

These persistent bugs that are very, very hard for us to locate in some instances just kill the user experience.

146

hi, i have read the entire post today, because im aware of this spikes and i also tweaked the hell out of my machine to get things smoother without spikes, i have a question regarding your nvidia graphics card driver:

Have you installed the driver alone without the Nvidia Geforce Experience and also without the Nvidia Physix engine? just de driver itself without the extra components, also make sure to install the Studio Driver version and not the Game Ready Driver

This configuration helped my machine (an old i7 7700k with a nvidia 1050ti card)

Also have you check this program ? ( O&O ShutUp10++ - O&O Software GmbH) it really helps shutting down background tasks from win10/11 and makes the cpu ultra light in % of use, mine iddle will hit 0-1%, making it ready for cubase use.

Plus tip: Also check the power plan in Windows Control Panel, because the steinberg power plan sometimes is not the best, in my case it works worse than my own custom high power plan, thats why i created my own (this really helped to reduce spikes when i have 200 tracks during playback and some tracks are armed).

I hope this tweaks help you, as they really helped me in the past.

Have a great day!

147

Hey, thanks much! As I’ve said before, I really feel like these issues are the industry’s responsibility to solve and it should not fall on us, the users who are paying lots of money over many years to keep Steinberg and others afloat.

Nonetheless, I appreciate your willingness to help very much. That O&O program is definitely interesting. I’ve found so, so many deep dive tools over the years to the point I’m concerned about unintended consequences and I’ve shut down & disabled so many processes & services…

I agree about the Steinberg power plan: I have better luck with the Bitsum Highest Performance and Ultimate Performance plans, generally, for unknown reasons.

Yes, I’m doing NVCleanstall for my drivers (and selecting the Studio ones). I need the PhysX for occasional gaming but I install nothing else and disable all telemetry I’m able to. I’ve dug into ProcessExplorer and so, so many other tools I can’t even remember the names of.

But what it always seems to come down to is some obscure Cubase bug. In this case, I’ve been having an excellent session all day with no spiking ever since I deleted and recreated those inputs and outputs. Not a single errant spike!

THAT IS A SERIOUS BUG AND IT IS WHOLLY STEINBERG’S CONCERN.

You see what I’m saying here? We should not be the ones on the hook spending countless hours losing our minds and being our own sysadmins.

Steinberg needs to get their poopies together and test their software more rigorously under ALL possible scenarios, especially templates created in earlier versions, anything that normal musicians might actually do with a reasonable expectation of reliability.

I really hope that Yamaha starts to lay down the gauntlet and demand a higher quality bar, because this has just been too much. The last few years have been awful.

Either way, my sincerest thanks to all of you who have chimed in in this thread.

148

Yep, I’ve turned off Steinberg Power Plan and use W11 ‘Ultimate Power’.
Also, I just use the NVIDA Studio Driver only. No Nvidia Geforce Experience, or
Nvidia Physix engine, or indeed the Nvidia High-Definition Audio driver.
Just the basic custom install.

1
1893×457 70.7 KB

149

@deleted_by_self

after been on this forum and a cubase user since the 1980’s I can say that trashing prefs is one of the first things to do with a new Cubase version IF YOU HAVE ISSUES.

I’m really surprised you didn’t do this until now.

Yes, in an ideal world you shouldn’t have to but it’s been the case on windows/MacOS that ‘old’ prefs can and do cause issues on new systems/Versions.

I’m glad you’ve sorted things.

enjoy making music.

M

150

I did trash the prefs. Multiple times. Early on.

Totally removing the output channels and rebuilding them I did not do, though.

Either way, this kind of behavior falls under the bug category and should not be ours to try and pinpoint. It really amazes me what some of you old school users are willing to put up with (and I count myself as one).

We should be demanding a higher quality bar as the complexity of these DAWs has exponentially ballooned over the decades, what with all this new memory and hard drive space available, not to mention processors that should in theory be able to handle practically infinitely more than their forebears.

151

The solution for me has been to close the Chrome Browser in the background… seems it graps hold of some system capacity and caused the random CPU spikes… though sometimes they still occur randomly.

152

Thanks. Yes, I’ve tried this, too. I’ve tried virtually everything. I find it really upsetting that the designers of this software seemingly are not prioritizing working with all parties involved (Microsoft, Intel, NVIDIA, Asus, whoever else) to make sure they have a way to prioritize real time audio threads over anything that could break the flow state of a composer/producer by causing a dropout.

We are supposed to be allowed to be musicians and attain flow state to create, not act as our own system administrators for free. And even if it was our responsibility, we do not have any publicly available tools that would provide the requisite deep insight into what’s getting interrupted and where. LatencyMon ain’t it. Something else is going on that we cannot see.

I will continue to respond to these such threads until Steinberg starts to prioritize this in earnest. It isn’t fair to us as consumers to have to deal with this. I’m very happy for those who don’t experience these issues on their systems, but they are not us–and we have a real problem here.

153

Just for a test, I’ve been leaving Cubase up and idle with only a few VSTi’s turned on and only a couple VST effects on the project activated. I’ve tried this under various power plans, including Steinberg Audio Power, Ultimate Performance, and Bitsum Highest Performance.

The result is always the same: within about 2-5 minutes, the VST meter will have overloaded for absolutely no apparent reason. Reliably.

Again: why is any other process on my system allowed to interfere with whatever Cubase is trying to do in real time? That’s the crux of the issue here. Nothing else should have this priority, period. The fact that it demonstrably does demonstrates an engineering failure on the part of Steinberg, as outlined in my post history.

154

Sounds much like some sort of Windows/hardware/driver issue.

I read somewhere that the built in TPM stuff on some motherboards can mess with DAW performance. Particularly for Windows Pro users who use encrypted drives for any of the partitions, and I suppose it could also get involved for some of the numerous copy protection schemes out there common for DAWs and Plugins. Adding a new dedicated TPM module was said to clear things up.

Long shot, but if nothing else jumps out as the culprit it’s worth having a look?

155

Appreciate the suggestion, but again, I think my point stands: this should not be our problem as consumers to isolate and solve. We’re not the ones being paid to troubleshoot and test software.

If Steinberg were to roll out an ML-enhanced test suite that was capable of truly isolating these problems that we could install that would report its findings right back to the mothership so they could actually engineer around these issues like the engineers they supposedly are, that’d be one thing. But they’re not even doing that. They’re simply not making these performance problems so many of us are experiencing a real priority to solve vs. staying on their release schedule, much like many companies these days–and so, the general enshittification continues unabated.

(Also, I don’t have my drives encrypted, I am using PTT because I have a Z790 motherboard and i9-14900K, so I can’t imagine how this would be a problem. Again, these companies who are allegedly actually testing the software they write need to make diagnostic tools for said testing that actually work and can actually isolate problems that would affect the real time audio performance so they can work with their partners to address them properly.)

156

I’ve been testing the same basic setup on both Ableton 12.1 and Bitwig 5.3 all weekend. No bloody CPU spikes.

It’s Cubase.

It’s Steinberg.

They need to fix their audio engine on Windows. Period.

157

Why do do need to use cubase at all?
You have so many other working solutions just use them. Period.

158

@Hippo - nice gatekeeping.

I am trying to transition away from Cubase, but I’ve been using it for decades. It still has some extremely powerful features that aren’t available in other DAWs (and some very malformed, needlessly complicated ones, too).

Cubase has become top-heavy and Steinberg is ignoring basic functionality issues in favor of getting new features out the door at the cadence the executives want.

I would love to be able to use the features I paid for many times over by sinking untold money into this company over the decades.

THAT’S WHY.

159
  1. Are you still getting the same problems running Cubase on your M4 Max mac? I ask because I note you say Steinberg need to sort out their Windows offering.
  2. Did you take up the offer made in another thread of sharing a problematic project just to see what results other users had?

Steve.

160

  1. Not really: the Mac version seems much more likely to crash, though, to just go poof while I’m working. The CPU overload happens quicker than I would like at 24/96, which is how I routinely track on the Mac, though. It would be really, really nice if there were some at-a-glance way we could tell whether a given plugin was Apple Silicon ready so we could eliminate any less efficient plugins in favor of optimized ones. I don’t know if I have a mix of them going on much of the time.

  2. I tried to share a bunch of info with Steinberg like a year ago, but there were issues getting the files to them (which seems like a completely solvable problem if they’d just accept a link, but I remember it being frustratingly difficult). I don’t remember specific non-Steinberg folks asking for me to share a problem project, nor do I think that’d really work because one of the main issues I’m having relies on a specific external synth being controlled with a VST plugin–there shouldn’t be any significant load at all on the plugin, but yet it’s the one most likely to cause the spiking behavior currently.

Again, I really think Steinberg should focus on developing a suite of diagnostic tools to run inside Cubase to see what it’s actually doing and what other processes are allowed to get in the way of the audio stream on any number of bespoke PC builds. I think this is the only way to effectively solve this at this juncture and I can’t for the life of me understand why the whole industry has not moved in this direction. Deep insight into individual system behavior and automated telemetry back to the mothership (optional, of course) would be an absolute boon for quality.

161

one of the main issues I’m having relies on a specific external synth being controlled with a VST plugin–there shouldn’t be any significant load at all on the plugin, but yet it’s the one most likely to cause the spiking behavior currently.

haven’t you just given us the clue to what the actual problem is here?

If most of us are working fine and you say the above surely that would make yo draw a certain conclusion, no?

M

162

No. This is not the only one. I thought I made that clear already.

I’ve suggested a way Steinberg engineering can approach this effectively, but some of you don’t seem to want to hear it. You’d rather remind me that everything works just fine for you.

163

With the greatest respect (genuinely) I don’t think you’ve made that clear at all. What you’ve made clear is a general wish to make Steinberg, and Intel, and Apple, and Yamaha, and Microsoft work more closely to ensure music software is more stable and it is easier for users to identify their own problems if and when they occur. This is a Steinberg Cubase user forum so I’m not sure it’s an effective way to get your demands for industry-wide improvements to be heard or taken seriously. Users here can and will share their experience of problems and their attempts to address them in the hope it can help other users.
I’m not the internet police and you, like everybody else, can use this forum however you want (within the rules set by Steinberg). To get assistance in cracking certain problems, you’ll need to do the boring stuff and isolate each and every part of your system(s). To vent your frustrations with the world of software/hardware integration, then feel free to carry on.
Either way, good luck with achieving your aims.
Steve.

164

Thanks, Steve, but if you look back through my post history you’ll see that I’ve tried a great number of things. This is the point I’ve reached after years of travail: I’ve tried so, so many things. I’ll leave it there.