Lazy developers exposed by Apple M1 transition (lack of native updates)

[Ploki]

The last one that I specifically remember was the Tone2 dev saying there wasn't a huge difference vs rosetta (or whatever it's called not a Mac owner), I remember reading that multiple times because it was a slow moving thread. I saw some other devs saying similar things but I can't recall who or where because uhhh why would I? I don't care about Mac stuff.

If this is outside of the norm, and most plugins are seeing huge performance increases, then I stand corrected. But calling them lazy because they're not catering to your specific early adopter needs fast enough is braindead. It's also kinda sociopathic, the idea that because someone that works at a job isn't catering to your specific needs, they aren't getting work done.

Regardless of the specifics of the average cpu efficiency increase across the plugin industry, the thread title is still super gross and if someone is unsatisfied with the software available on a platform, they should probably not buy into that platform just because it's the new thing unless they want to experiment but have the expectation that their needs won't be met. Or in the case of M1 or any new platform, wait until a satisfactory software package becomes available if it's a big deal to you not to have all the software right away, as it obviously is to the OP.

I'd say what's actually lazy is buying a system without using your brain to understand the bare basics of how reality works, then projecting your own personal issues onto developers, like the OP did.

Fast enough?
Dev kits have been out since june 2020. it's 2022.
Yeah no, 2 years isn't fast enough.

Thread title is what it is

and as i said, i'm like 90% native, and i buy plugins from devs who made it work. there's enough options that i'm not missing out one bit

Diva: 19 instances rosetta vs 36 instances native.
that's 50% improvement, not 7%.

<nda> compressor: 100 instances rosetta, 170 instances native
that's 70% improvement, not 7%.

And Logic New Benchmark Test:
63 tracks Rosetta
110 tracks native

7% performance improvement get devs that didn't really do a good port.

[jens]

Dev kits have been out since june 2020. it's 2022.
Yeah no, 2 years isn't fast enough.

June 2020 to now is one year and seven months, which is not nearly two years - and if you don't believe me you may read up on how all this calender-stuff works - I admit it may seem a tad complicated to some at first, but once you get the hang of it a lot of stuff all of a sudden will start to fall into place
...

[jens]

7% performance improvement get devs that didn't really do a good port.

You may have overread the part of this thread where Aleksey of Voxengo (who's is known to really know his stuff and is certainly is at least one thousand times more competent than you in regards to anything computer related) mentioned that there is basically zero improvement unless you start using a calculation method that is prone to yielding calculation errors - but hey, perhaps sound-quality doesn't matter to you one bit - I won't hold that against you...

[wvshpr]

Rosetta 2 IS measurably slower. It‘s an emulation layer, it's bound to be.

People with Silicon machines do see a significant performance bump when using native software.

This discussion is moot. Apple is not going back to x86 for the foreseeable future. Software will eventually catch up.

[SoftSynthLover99]

The last one that I specifically remember was the Tone2 dev saying there wasn't a huge difference vs rosetta (or whatever it's called not a Mac owner), I remember reading that multiple times because it was a slow moving thread. I saw some other devs saying similar things but I can't recall who or where because uhhh why would I? I don't care about Mac stuff.

If this is outside of the norm, and most plugins are seeing huge performance increases, then I stand corrected. But calling them lazy because they're not catering to your specific early adopter needs fast enough is braindead. It's also kinda sociopathic, the idea that because someone that works at a job isn't catering to your specific needs, they aren't getting work done.

Regardless of the specifics of the average cpu efficiency increase across the plugin industry, the thread title is still super gross and if someone is unsatisfied with the software available on a platform, they should probably not buy into that platform just because it's the new thing unless they want to experiment but have the expectation that their needs won't be met. Or in the case of M1 or any new platform, wait until a satisfactory software package becomes available if it's a big deal to you not to have all the software right away, as it obviously is to the OP.

I'd say what's actually lazy is buying a system without using your brain to understand the bare basics of how reality works, then projecting your own personal issues onto developers, like the OP did.

Fast enough?
Dev kits have been out since june 2020. it's 2022.
Yeah no, 2 years isn't fast enough.

Thread title is what it is

and as i said, i'm like 90% native, and i buy plugins from devs who made it work. there's enough options that i'm not missing out one bit

Diva: 19 instances rosetta vs 36 instances native.
that's 50% improvement, not 7%.

<nda> compressor: 100 instances rosetta, 170 instances native
that's 70% improvement, not 7%.

And Logic New Benchmark Test:
63 tracks Rosetta
110 tracks native

7% performance improvement get devs that didn't really do a good port.

Thank you for posting that! Yeah the thread title is a but over the top I admit, but people just don’t get that M1 Native M1 plugins are a significant improvement over rosetta layer or intel versions. For developers who really did a good port the results are really clear that native M1 plugins outperform the intel or rosetta versions substantially.

I’ve noticed significant performance improvements in plugins like Klanghelm Mjuc as well. Although idk how to measure like you did the %.

[jens]

Rosetta 2 IS measurably slower. It‘s an emulation layer, it's bound to be.

Calling it an "emulation layer" is highly ambigious, to say the least. It translates the code before the program gets executed. It is not something that happens in realtime while the application is running. And hence your claim that it is bound to be "measurably slower" than native code may require some more in-depth explanation from you - especially given the fact that what you end up with when Rosetta 2's work is done is native code.

[briefcasemanx]

7% performance improvement get devs that didn't really do a good port.

Hey that's really cool info and all that I have no use for, I assume you're a Mac plugin dev and can state this with authority over the Tone2 dev or the Voxengo guy talking about their own plugins. I will not say in the future that the improvement is small, like I said in the post you quoted, I stand corrected.

The most relevant part is the OP's crappy attitude, which is the only reason I clicked on this thread, not some hyper specifics about cpu-utilization improvements in a specific plugin. I really, really, don't care. At all.

I'll say it again, I stand corrected.

(watch people still quote me with cpu usage stats though).

[SoftSynthLover99]

7% performance improvement get devs that didn't really do a good port.

Hey that's really cool info and all that I have no use for, I assume you're a Mac plugin dev and can state this with authority over the Tone2 dev or the Voxengo guy talking about their own plugins. I will not say in the future that the improvement is small, like I said in the post you quoted, I stand corrected.

The most relevant part is the OP's crappy attitude, which is the only reason I clicked on this thread, not some hyper specifics about cpu-utilization improvements in a specific plugin. I really, really, don't care. At all.

I'll say it again, I stand corrected.

(watch people still quote me with cpu usage stats though).

So you basically came into the thread to offer nothing about the actual topic..Lol and despite it’s challenges life is beautiful and I do my best to keep that in mind. However everyone needs a little rant every now and then

[Ploki]

June 2020 to now is one year and seven months, which is not nearly two years - and if you don't believe me you may read up on how all this calender-stuff works - I admit it may seem a tad complicated to some at first, but once you get the hang of it a lot of stuff all of a sudden will start to fall into place
...

I rounded years up because it's more than 1.5years

You may have overread the part of this thread where Aleksey of Voxengo (who's is known to really know his stuff and is certainly is at least one thousand times more competent than you in regards to anything computer related) mentioned that there is basically zero improvement unless you start using a calculation method that is prone to yielding calculation errors - but hey, perhaps sound-quality doesn't matter to you one bit - I won't hold that against you...

Calculation errors below -150dB FS, if i recall - i know because i had mail exchange with Alexey and tested TEOTE builds where he first implemented the methods he mentioned.

And no, i don't care about stuff below -150dB FS at all, and i trust Alexey that he picked the method that yields good results and doesn't compromise sound quality.

I'm sure some crazy-ass mofo somewhere runs TEOTE in Rosetta2 tho, so he can get better maths below -150dB FS.
And Voxengo improvement is more than 7% iirc anyway.

...

You can go over u-he forum and talk with urs as well to get another dev perspective

[Super Piano Hater 64]

The exact performance difference depends on so many details of the workload being benchmarked that it's no use trying to generalize. There are certainly programs that will run fast as native code (either x86 or ARM) but suffer in Rosetta, or even crash outright. There are also programs that will run much faster in Rosetta than as native code on a comparable x86 processor. You can argue forever and never work out a safe rule of thumb, because there will be more exceptions to any rule than programs that satisfy the rule.

The most useful things to benchmark are the host processes. Running these in Rosetta causes specific documented problems that in most cases only Apple can fix, and that's not exactly likely to happen. This is the single biggest reason to demand native ARM plugins... but it's not an excuse for SoftSynthLover99 to be rude.

[BONES]

This is the 1st generation M1 chip. I personally have a max spec M1 Max.

And, as I pointed out if you bothered to read it, Intel also have their 12th Gen chips available now, with the highest performance gains we've seen in many years. And even in those benchmarks, they were only testing laptop CPUs, Intel and AMD both have a range of far more powerful desktop processors so if you need more power, Apple is nowhere in the game.

But did you ready the part of the article that said this:
“ None of this bothers Apple’s M1 much though. Based on TSMC’s most advanced 5nm process, it’s a stone cold killer, with Macworld reporting no fan noise at all during the run. That can’t be said of the x86 laptops, which all vary from fairly quiet to a little rackety.”

Yes, and if the main reason I was buying a laptop was so I couldn't hear fan noise, I'd buy a liquid cooled jobbie like the Acer I had a few years ago. That's not why I bought that laptop, though, because I couldn't give a flying f**k about fan noise. Honestly, I can't think of anything less relevant to my buying decisions. I'd be more likely to be put off by the colour of the chassis.

Like I said you just have to try it to truly understand. Looking at benchmarks don’t do these machines justice.

No, I don't have to try it at all because I've been using way more powerful machine for years. I finished our 4th album on a 6th Gen Core i5 laptop and it was all perfectly fine. Audio isn't that hard if you knw what you're doing. OTOH, if you "learned" how to produce music via YouTube videos, you're likely completely clueless.

[SoftSynthLover99]

No, I don't have to try it at all because I've been using way more powerful machine for years. I finished our 4th album on a 6th Gen Core i5 laptop and it was all perfectly fine. Audio isn't that hard if you knw what you're doing. OTOH, if you "learned" how to produce music via YouTube videos, you're likely completely clueless.

How do you know you’re using “way more powerful machines” if you’ve never worked with the new M1 Max chips to compare?

Yes you can make an album on literally anything nowadays from an iPhone to an iPad with the power of current technology. We are truly in the golden age of music production equipment, as long as we aren’t held back by “lazy” developers lol

BTW I record a lot of vocals, acoustic instruments and guitars so fan noise is extremely important to my work flow. When the laptop is close by your mic and you are recording yourself or other vocalist this can be a problem if fans are going crazy. Low fan noise is essential for my kind of music production.

[AnX]

Dev kits have been out since june 2020. it's 2022.
Yeah no, 2 years isn't fast enough.

June 2020 to now is one year and seven months, which is not nearly two years

Actually, it is.... basically anything past zero to end point is 'nearly'.....but 19 out of 24 def counts as nearly....

[SoftSynthLover99]

Heres a few good articles from pro tools expert in the same vein as the topic for those interested: https://www.pro-tools-expert.com/produc ... ytime-soon

https://www.pro-tools-expert.com/produc ... ps-so-fast

Few highlights and quotes: “ On YouTube, I watched a Mac user who had bought an iMac last year. It was maxed out with 40 GB of RAM costing him about $4,000. He watched in disbelief how his hyper-expensive iMac was being demolished by his new M1 Mac Mini, which he had paid a measly $700 for.”

“ In real-world test after test, the M1 Macs are not merely inching past top-of-the-line Intel Macs, they are destroying them. In disbelief, people have started asking how on earth this is possible?”

“ The reality is that there are benefits that Intel and AMD will never be able to offer, even if they are dragged screaming and kicking into the SoC world, whereas Apple is able to offer the full deal because they control both the hardware and software.”

Interesting reads both articles. Seems iLok pace are indeed moving really slowly. And the M1 chips especially for music production are outperforming even the higher end chips from AMD and Intel. Not my words it;s all in the articles above

Edit: Interesting video as well showing the performance. 700+ tracks on a 32 buffer and a unsupported version of Pro Tools (M1 & Monterey) recording 16 additional tracks flawlessly. Insane performance and this is all on Rosetta not native

[BONES]

Pro Tools Expert is an Apple fanboi website, one of the absolute wankiest fanboi sites you will ever come across. Nothing he has to say or show is of any consequence whatsoever because his bias bleeds out of every line he writes or utters.

How do you know you’re using “way more powerful machines” if you’ve never worked with the new M1 Max chips to compare?

Because, as I said the other day, the workstation I use in the office is a Dual Xeon Gold machine with 16 cores and 32 threads, 96GB of RAM and an nVidia Quadro RTX5000. If M1 Max was better than that, wouldn't Apple be putting them into their MacPros? But Apple are still only putting Xeons into the MacPro because those processors are on a different level to consumer grade CPUs.

BTW I record a lot of vocals, acoustic instruments and guitars so fan noise is extremely important to my work flow.

Recording hardly taxes the CPU enough that the fans are going to kick in, does it? About the only times I notice my fans working is when I am rendering out mixdowns. They sure as hell don't when I am just playing back a piece. 10 years ago I used to leave my laptop outside the V/O booth where I record vocals but for the last two or three albums, I haven't bothered because it has not been an issue.

When the laptop is close by your mic and you are recording yourself or other vocalist this can be a problem if fans are going crazy.

The simple solution here is don't have your laptop close my your mic. It's not an issue for me because a) I never hear the fan when I'm recording these days, and b) it's always on the desk behind the mic and well out of it's cardoid sensitivity zone anyway. I do lots of things to minimise the amount of noise that gets captured when I record but fan noise hasn't been an issue for a long time. Maybe it's a Mac thing because it definitely seems to be mostly Mac people who are obsessed about it?