PT2023 and Mac Studio disappointments

[Darryl Ramm]

Originally Posted by loud neighbor

900MB/s on some SanDisk 4TB Extreme PRO Portable SSD V2's (check the recall on the firmware if you have any of these)

I would not trust any SanDisk Extreme drive, or what Western Digital says about them. Western Digital has so badly screwed up the handling of this corruption problem. Let’s see what happens to them with the multiple lawsuits underway. Just great leadership at WD, and everything is going great: https://blocksandfiles.com/2023/10/0...t-disliked-ceo

One good way to get large storage on external Thunderbolt is multiple PCIe/NVMe M.2 SSD cards in a Thunderbolt 3 Chassis. Using dumb single slot M.2 adapter cards or switch based adapter cards if you need 2 or more M.2 cards per PCIe slot. Use Samsung 980 or 990 SSD. Thunderbolt 3/4 will be a bottleneck here. You can get faster aggregate performance hanging a single M.2 drive in Thunderbolt case/chassis off dedicated Thunderbolt ports but this does not scale well/becomes an unreliable cabling mess. All this has been discussed in many hundreds of posts on DUC.

—-

And folks please remember that what VI samples on a faster SSD will do for you depends on how the VI handles samples, if the VI caches samples it mostly may startup faster, if it streams samples it might run with less problems (or might not if the current drive is “fast enough”) if it has a hybrid setup you can see both behaviors. Look at sample config/tuning options for your VI. But especially compared to HDD you can see some significant changes, it is pretty wild how affordable fast NVMe/PCIe M.2 SSD are.

[massivekerry]

Originally Posted by audiolex1

While keeping everything as is? no change in plugins or freeze?

If it's a full-mixed track that I'm revising with a bunch of look-ahead plugins that cause latency (Ozone, Soothe, etc), it can be problematic, and I'll have to deactivate some things. But usually I'm tracking the live stuff pretty early on. Occasionally, I've forgotten to change the buffer from 256, and I honestly haven't noticed. The Carbon has been great with its hybrid DSP recording paths. You just need to make sure you look at the signal path of those tracks and make sure you've got a pretty clean (low-latency) route through auxes and the master, since you'll want to disable the DSP mode on them. Also, fwiw most of my orchestral libraries are run off a partition (sorry... "container") "disk" on my internal drive, and I run the sessions off it, so it's pretty smooth. Some low-bandwidth sample libraries I keep on NVMes (Samsung 970 Plus) in a Sonnet (Echo) TB 3/4 enclosure. When I got the Mac Studio, I decided to just get all new components and drives, because I've been burnt before, trying to incorporate "legacy" peripherals with new hardware.

[innerbooty]

Hi,
Referring back to your original post here. I had posted previously in this thread that it was making me nervous because I was about to do the same upgrade. Well, I did. I recently bought a Mac Studio M2 Ultra 64GB/1TB, and have spent the last few days starting to set it up - new install, Ventura, PT 2023.12.1 - and I am having the exact same experience you had. Performance seems barely faster on the M2 than my 5,1 Cheesegrater, and Overall System Usage actually much higher %-wise on the M2 than the Cheesegrater, running the exact same PT session on eaach. Did you ever get to the bottom of what was happening with your system?? I am thinking of returning my M2 because this is not really an upgrade at all at the moment, a VERY expensive non-upgrade. Thanks!

Originally Posted by juh

I just purchased a new Mac Studio M2 ultra, with 128 GB RAM, my main use is composing music for movies.
I preformed a clean install from scratch, with every lastest softwares / plugins versions. OS is Ventura. Previously I used to work with a Mac Pro 2013.
With my new Mac I thought everything would be so much faster, like launching sessions, loading samples... Actually not that much.
Moreover, I’m still encountering CPU issues (same as when I worked with my old Mac Pro 2013) even when my sessions only have 20 tracks (my main virtual instruments as Kontakt, Omnisphere…). Not systematically but it frequently occurs.
I don’t have any CPU nor RAM overload, but playback is interrupted with messages saying that the DAE buffer size has to be increased…

This is so disappointing and frustrating, it’s really not what I was expecting. So I’m wondering if those issues are due to my system configuration ?
Or maybe PT is not able to take advantage of the full power of latest M2 yet ?

Thanks

J

[WorldStudios]

I did the same transition as you did and I had the same initial reaction when loading up one of my heavier sessions that was straining the Mac Pro 2019, giving occasional overload messages. It was roughly the same thing on the new M2 Ultra and I was very disappointed. But then I saw that the playback engine was set to something very low and the old computer had it set to maximum (2048 for a 88.2 kHz mastering session). When I tried higher playback buffer settings, I could run it at 512 or 1024 with less measured total strain on the computer and no errors. I use a few plugins on the master bus that are total cpu hogs. Ozone 11, where some of the processes are using a lot of CPU is probably the most power hungry, but the Pulsar Modular ones, Gold Clip and FiDef are a few I can think of that makes the CPU meter really shaky. Still, it is running considerably better on the M2. Worth the investment.

[innerbooty]

Thanks for that. It's funny, in some other threads I'm seeing people say that on the silicon machines with Pro Tools, performance actually improves with *lower* buffer settings, counter-intuitively. I'm running pretty darn big sessions on my old 2009 Cheesegrater with buffers @ 256 and 512 (not with things like iZotope however). I would expect the M2 performance to be significantly better. When all is said and done this is going to be about a $6K upgrade, and right now I sure don't feel like I'm getting $6K worth of improvement

[RobertDorn]

Originally Posted by innerbooty

Hi,
Referring back to your original post here. I had posted previously in this thread that it was making me nervous because I was about to do the same upgrade. Well, I did. I recently bought a Mac Studio M2 Ultra 64GB/1TB, and have spent the last few days starting to set it up - new install, Ventura, PT 2023.12.1 - and I am having the exact same experience you had. Performance seems barely faster on the M2 than my 5,1 Cheesegrater, and Overall System Usage actually much higher %-wise on the M2 than the Cheesegrater, running the exact same PT session on eaach. Did you ever get to the bottom of what was happening with your system?? I am thinking of returning my M2 because this is not really an upgrade at all at the moment, a VERY expensive non-upgrade. Thanks!

Well this must be very specific to M2 Ultra Studio Macs then. I'm on a M2 max MacBook, which theoretically only has half the cpu power of the M2 ultra. I don't encounter a single glitch and then I'm talking about 100+ tracks mixes loaded and loaded with plugins and virtual instruments.