Pro Tools 2022.4 on Mac Studio M1 Ultra

[Pedro Garcia BR]

I'm waiting (and paying) forever for native m1 support.

[Selle Sellink]

Originally Posted by K Roche

I would also consider ("all the plugins I use") one or more,, 3rd party especially,,,, could be slowing things down ??

As I understand it. While Pro Tools runs via rosetta 2, there are no individual plugins running extra compute power via rosetta 2.

That's going to be an issue once Pro Tools runs natively on Apple Silicon and you make use of non silicon native plugins. Since it has to run Rosetta for every individual plugin. But we'll see how that turns out once Avid decides to support Apple Silicon natively.

[Darryl Ramm]

Originally Posted by Selle Sellink

As I understand it. While Pro Tools runs via rosetta 2, there are no individual plugins running extra compute power via rosetta 2.

That's going to be an issue once Pro Tools runs natively on Apple Silicon and you make use of non silicon native plugins. Since it has to run Rosetta for every individual plugin. But we'll see how that turns out once Avid decides to support Apple Silicon natively.

You have this backwards.

All plugins in Pro Tools today on Apple Silicon run under Rosetta 2, along with the actual Pro Tools code, they have to since it's all Intel code running in the same address space and like most DAWs the plugin code and Pro Tools code is all an intertwined mess.

Unless Avid pulls a big surprise out of their back pocket(*), an Apple native version of Pro Tools will only run Apple silicon native ports of plugins. Some folks are going to have to pay to update plugins etc. to get those native port versions.

And this is likely to mean an increase in instability for a while as new native ports/updates of plugins get to run in a new native port/update of Pro Tools.

Native plugins running in a native port of Pro Tools should run faster and consumes less resources became Rosetta 2 is not involved at all. However because of lots of changes and the complexity of a multi-vendor ecosystem there will be a long-tail here with vendors needing to work on fixing new bugs, and improving performance of the native ports, and there will be some AAX plugins that are just never ported over to apple silicon, and just won't run on an Apple silicon version of Pro Tools. So for a while here you'll have to continue to be paranoid about stuff, test things with your own workflows/workloads etc. But if you have legacy Intel only plugins you will be able to launch the Intel code version of Pro Tools embedded in the app and have that run with Intel plugins (and ideally all the plugins will include both Apple silicon and Intel executable code which will help make the installations and transition easier here.

Separate from whatever is native or not or running under Rosetta or not. There are lots of things changing/in flux now. As with any new version of Pro Tools or any new operating system, if you are having problems you need to troubleshoot what is happening, and you always suspect (and try to quickly exclude/include) plugins. And nobody should be leaping in and buying new Macs right now and just assuming stuff will work for them, be paranoid, test, and be prepared to return purchased is needed.

(*) That would require Avid make significant architectural changes to Pro Tools to run intel plugins alongside native code. A complex effort. Apple has done this with Logic Pro. I suspect Avid will not, and it's potentially yet more complexity/destabilization and would slow down delivering a native port even more. Personally not something I think is even a good idea for Avid to try.

[StudioTalk]

I’ve been doing extensive testing since I got the Mac Studio Ultra. It should be 5x faster than my M1Pro MacBook Pro.

If the max is double the Pro and the Ultra is twice the Max, then shouldn’t the Ultra be 5x faster than the M1Pro.

Well because single core is dominant, in my testing it is only 3 times faster. And barely that.

So considering that, I can’t possibly justify the Ultra, at least at this point. I’m also testing in Silicon Native apps with both Logic and Studio one where you can get so much more from M1 Processors.

I don’t think Avid has any intention of making Silicon Native a priority. If anything it goes against their new pricing structure. Think about it, if they get the performance smoking like other DAWs at this point, wouldn’t that make users seriously abandon Ultimate or whatever nonsense they call it today.

Based on my experiences with the M1 Pro, and now the Ultra, there is no way I’d stay Ultimate as I never use the DSP on my Carbon, I simply have never needed it yet.

So to wrap my perspective and that’s all it is, I would not waste my money on a MacStudio Ultra today, not the right time. The M1 Max would be a much better choice.

[Chevron]

Originally Posted by StudioTalk

I’ve been doing extensive testing since I got the Mac Studio Ultra. It should be 5x faster than my M1Pro MacBook Pro.

If the max is double the Pro and the Ultra is twice the Max, then shouldn’t the Ultra be 5x faster than the M1Pro.

Well because single core is dominant, in my testing it is only 3 times faster. And barely that.

So considering that, I can’t possibly justify the Ultra, at least at this point. I’m also testing in Silicon Native apps with both Logic and Studio one where you can get so much more from M1 Processors.

I don’t think Avid has any intention of making Silicon Native a priority. If anything it goes against their new pricing structure. Think about it, if they get the performance smoking like other DAWs at this point, wouldn’t that make users seriously abandon Ultimate or whatever nonsense they call it today.

Based on my experiences with the M1 Pro, and now the Ultra, there is no way I’d stay Ultimate as I never use the DSP on my Carbon, I simply have never needed it yet.

So to wrap my perspective and that’s all it is, I would not waste my money on a MacStudio Ultra today, not the right time. The M1 Max would be a much better choice.

The M1 Max is not double the Pro, I think they are almost identical in CPU terms apart from the Max being able to host more RAM and be configured with more GPU cores.

[Darryl Ramm]

And the increased memory capacity of the Ultra may be the critical item for some users, including folks who just want to have a system more likely to have a long useful life, with the forever increasing memory needs of new software.

[PhilSchroeder]

I'm running a new Mac Studio with the M1 Max chip, PT 2022.4 on Monterey and I gotta say, there's absolutely NO significant speed or performance increase over my 2013 trashcan. And a LOT of the same problems I'd hoped would be improved by a bump up on processing power. Very disappointed. Waiting for M1 qualified, or, as usual, looking at alternative DAWs.

[StudioTalk]

Originally Posted by Chevron

The M1 Max is not double the Pro, I think they are almost identical in CPU terms apart from the Max being able to host more RAM and be configured with more GPU cores.

That's not how they advertise it. You seem to only be looking at Single Core benchmark scores, which I can confirm do not mean as much as everyone says.

The Single Core scores on the M1 Pro and the M1 Ultra are almost neck-in-neck with each other, the Multi Core is much different. If single core were the dominating measurement, then I should have not seen any improvement over the M1 Pro as they are literally almost identical in Single Core Scores, so there is more to it. In some of my testing I have used Audio only tracks, where memory is a non issue, hardly used at all.

Now as to the multiplier, sorry, but Apple absolutely advertise the M1 Max as twice that of the Pro as all they did was double the processor. The Ultra is advertised as twice the Max. You were seemingly taking a single thing to compare to and that is not what I was talking about.

[Eric Lambert]

This isn't adding up. How do we account for those who are experiencing great gains in ProTools with Mac Studios? Some are reporting a slew of CPU horsepower with very high track counts. Others seem stuck behind a wall of some sort.

So, what *really* constitutes the wall, and how is it not an issue for many?

I'd love to see an Activity Monitor report for each of these cases to compare all that's happening under the hood, along with a list of plugins being used. There are so many factors, any one of which could be the sole culprit for unexpectedly bad performance.

[Bob Olhsson]

My understanding is that multi-core performance doesn't buy you much other than bouncing speed and video performance.