Music Production - Poor performance of certain DAWs on M3Pro

[ProgRocker]

Yes, just logic. I don't use anything else. I've read Reaper may actually be better optimized for apple silicon, but without using it I can't comment.

It might be worth your time to try another DAW to see if that is a contributing factor. Yes, Reaper is supposedly better optimized for AS. You could also try Studio One as there is a free 'full' version available. Both are free to try. It may provide you with some addition information.

I'd have to re-read this thread but I don't recall you mentioning having a problem with any other apps other than Logic, that would then lend itself to point to Logic as a possibility.

 

[chrfr]

Is this just with Logic ? Have you tested other DAWs ?

For what it's worth, Avid has not certified Pro Tools to be compatible with macOS Sonoma, and the M3 Pro/Max laptops can only run Sonoma so there could be issues there that Avid has not yet resolved.

 

[3SQ Machine]

I'm probably going to throw in the towel on testing new rigs, outside of potentially an M1 Max. I've got to get something that works or just stick with the same gear for now. Talked to Apple Support yesterday and they don't have any answers yet (it's been a week waiting on engineering). Unfortunately, I'm about to blow past the return window.

This reminds me of the issue we had some years back during intel days when Logic wouldn't play nicely with certain screen resolutions and external monitors--and no one could figure out why. It was related to some coding between logic and the built-in intel GPU--even though the specs made it seem this shouldn't happen. I had to run an external GPU to avoid crackles at lower buffer sizes on some gear. I eventually ditched that and just went for a 27" iMac since everything was optimized and the problem disappeared. Maybe in this case there is something about the 14" resolution that taxes the GPU. I can only speculate at this point and may just have to spec up the gear to handle whatever inefficiencies are in the software/hardware.

 

[3SQ Machine]

Couldn't find a reasonably priced 14" Apple Refurb M1 Max within driving distance (2.5 hours 1-way!), so snagged an open box apple certified M1 Pro 16" refurb. Did the same test run of projects -- yup, buttery smooth on this 16" M1 Pro base model 16/512 with minimal GPU/CPU usage and no jitteriness while scrolling/zooming.

Assuming this new-to-me M1 Pro checks out, I'll probably keep it since the delta between it and trading in my old 16" M1 Pro to Apple is less than the price to replace the battery. Plus I really like the silver color (had space gray). I'll also get to re-up my warranty + add Apple Care+ again--and a fresh battery! This is also about 1/3 the price of keeping the 16" M3 Max holy moly.

Seems bizarre to go through all this just to come full circle back to a 16" M1 Pro for my larger rig. But it works.

Still need a 14" for certain road situations, so Best Buy also has a sale on the 14" M3 Max 30 core GPU right now + a special for 10% cashback off the purchase if I apply for a card. F-it, I've now got that rig to test + a new card. Not as great of a deal (about 17% off) but still better than nothing.

So if anyone is looking for an M1 Pro 14" 16/1tb apple certified refurb (should be $1200 or so), one should be popping up in Micro Center's inventory anytime now.

I'll let y'all know how the M3 Max goes--probably fine I'm sure.

 

[3SQ Machine]

EDIT: Unbelievable but I've now had to throw everything out from this work (and thread) because all signs point to the more recent Sonoma updates. Before I was able to migrate from my trusty old M1 Pro, I had to update its macOS to match a new machine. BEFORE the update, machine was smooth. AFTER, just as bad as the rest. Even the M3 Max -- which I had to update--is now choppy. Signs point to the February 8 update as a possible culprit.

Soooo.... what a rabbit hole. Was probably just chasing software updates this whole time and not catching it as a Sonoma problem. This is probably why it's sporadic and didn't make sense. If the M3 Max is having an issue--it's definitely software, not hardware.

I'll downgrade one of the new machines that shipped with ventura and run a test there and report back. Then, at least, we'll have a clue.

UPDATE: Just to close the loop on this thread--reinstalled ventura on the M1 Pro from a TM backup. ALL IS WELL.

 

[ProgRocker]

UPDATE: Just to close the loop on this thread--reinstalled ventura on the M1 Pro from a TM backup. ALL IS WELL.

Looks like you may finally have your answer.

 

[3SQ Machine]

Looks like you may finally have your answer.

Yes indeed -- for now. Hopefully macOS 15 squashes these bugs. I'll still be running the M3 Max on sonoma and updating (no choice). Crazy that logic is choppier on the M3 Max than on the M1 pro, but this is why some people are still rocking intel machines.

 

[MikB]

Then you must be totally unaware of those that successfully use DAWs that utilize efficiency cores with no problem. And believe somehow you are more knowledgeable than the software designers that make these products. The internet is full of 'experts' who think they know best for everyone.
You're welcome to your opinion, but it's just that... an opinion. Maybe you should get off the forums once in a while and make some music, as I do.

I have not questioned the validity of using e-cores. Fact remains people are also using DAWs that do not make much use of e-cores. Why you pretend no-one is using also those? Pretend all you want, but reality is much more interesting. And, yes, I include also those DAWs that make use of e-cores into reality, unlike you that exclude those that don't. Yes, fact is you accuse me for what you do.

Why are you making up straw dog arguments instead of focusing on the issues at hand? Why do you assume a bunch of other things including how often I make music? Not that this is any of your business, but that's every single day, thank you very much. You just look like you got nothing.

 

[ChrisA]

The reason none of this matters one bit is because the tests used extreme examples that are totally unlike the what most users do.

What if I were buying tires for my Toyota Prius, One brand of tire was found to explode when driven at 300 MPH and another performed well up to 450 MPH. Which tire is better?

Tests that use hundreds of tracks are just as silly as my tire example. The software can "freeze" tracks. This means it can pre-render them. Using this feature even the old and slow 2011 dual-core Intel Mac can handle hundreds of sampled MIDI tracks. But most users would never go over a dozen tracks, yes, I know some do and were able to do it even on low-end Macs 10 years ago.

 

[Chancha]

Well, track count tests is about the only usable reference to compare different DAWs on different machines, it is by no means directly simulating a real session / project, but what it does is to break the processing down into the smallest chunk so we can even begin to compare them. If you are a heavy plug-in user or work on symphonies then it can translate into like 30 tracks to 1 for your case.

And ideally you would want to spec your machine with LOTs of computing headroom to save yourself from potential jitter / sync issues. These tests are the best we can get, before you buy and actually use the machine with your workflow and projects.

 

[MikB]

The reason none of this matters one bit is because the tests used extreme examples that are totally unlike the what most users do.

This argument doesn't actually hold as it's trivial if you use demanding oversampled software synths on 32 sample buffers to bring most machines to their knees. It's only a matter of amounts. No-one can know from an extreme stress test much else than where the limits are, but this is still actionable information. Unless perhaps if one flies through the roof not refitting that data to one's actual needs.
Also, no actual producer cares about "most users". They care about their own actual needs, of which you know very little. Personally, I've adapted my requirements, but I want more power as well.