My M4 Pro MacBook Pro Testing video is here! Watch it ➡th-cam.com/video/hccy19Hm6M8/w-d-xo.html ▶️My Analysis & Breakdown of the New M4 iMac, Mac Mini, and MacBook Pro for Music Production th-cam.com/video/POKZlRo-Lgo/w-d-xo.html ▶ My $12,000 Video/Music Production Desk Setup, Powered by an M1 Max MacBook Pro: th-cam.com/video/cA3ms_2Yn-A/w-d-xo.html
Thank you for watching! 🙏 1⃣A popular comment is that e-cores are too weak to make a difference. They actually are not! In my M3 Pro MacBook Pro video where I tested 7 different DAWs, test results showed significant better performance when a DAW can fully use the e-cores. (Watch that video here ▶th-cam.com/video/FSqX4bt9to4/w-d-xo.html) 2⃣Another recurring comment is that e-core should be reserved for system tasks/light tasks. OK, the problem is that when you have a chip where almost half of the cores are e-cores (e.g. 11-core M3 Pro or the 8-core M2), it makes a HUGE difference to be able to make your computer to pull out all the stops (aka use all the cores fully) to handle CPU intensive tasks as needed. 3⃣Lastly, the point of this video (and my other DAW testing videos) is to help fellow musicians, audio engineers and music producers make more informed decisions when determining which Mac to buy, because it’s not as simple as “buy the latest because it’s better.” What you use as your primary DAW is a critical factor here due to different levels of core utilization. I appreciate your support and stay creative!
on Macs, performance based DAWS like Bitwig Studio and Ableton forces all the realtime audio processing to the performance cores, as the efficiency cores are not suitable for it (they get aggressively forced to sleep etc.). Some linear arrangement hosts make use of the efficiency cores to process other things ahead, which are not realtime dependent, but with Bitwig Studio and Ableton Live being a non-linear DAW that has to be able to start any clip at any given time with sample accuracy, there's currently not much it can offload to those cores. Possible that logic is the same. Have you reached out to these DAWS to ask them? This is what they will tell you.
What WAS the case with multi-core Intel processors (and may still be with Apple silicon), is that where music production is concerned, high single-core clock-speed performance was more important than total core count of the processor. This was true because while a high core count MAY be able to allow your Mac running any DAW to handle a 'larger', more demanding project file, the workload of a single plug-in instantiation cannot be distributed across multiple cores. So as many users have experienced, a very cpu-intensive plug-in could halt your song-file with the dreaded processor overload dialog when the majority of cpu cores were just chillin'… because the demand of just one plug-in instantiation was more than the assigned core could handle. Since through successive iterations of Intel processors we saw that models with fewer cores generally had higher single core clock speeds, while the models with the most processor cores were clocked at lesser (slower) slower speeds. For audio/music applications, if you want to run one or several virtual synth plugins with high polyphony (for example), then you'd likely do better with a processor having fewer cores, but able to run at a higher clock speed. I've said all this because I don't have any knowledge that this radically different ARM-based SoC architecture has changed the importance of looking for high single-core clock speeds. This is important work you've done in discovering the need (for logic users at least) to focus the 'performance core' count of the processor when shopping for a Mac, but AFAIK, high single-core performance is still more important than total 'performance core' count for this kind of work.
Clock speed is irrelevant, you need to look at the performance of the core. With intel and amd, there are examples of cpus with higher clock speed and fewer cores compared to a model of the series above, and yet in single core tests, the model with more cores and lower clock speed in the higher series still has higher single core performance. Also when you are running more tracks than the core count, you really can't know that the single core performance will be the thing that overloads the cpu, as each core is running several tracks and their plugins. I have not seen a demonstration of any recent cpu, where it was obvious that higher single core performance trumps multi core performance. I have seen a lot of people claiming that. But I am very sceptical to that still beeing true, it might be something that people still belive because early multi core chips did have single core performance issues. And it might be the case for some tasks that can't utilize multi core systems, which was the case even long after the single core performance issues were dealt with.
@@5urg3x but that has nothing to do with my comment. Apple's CPUs has great single core performance and good multi core, especially considering the wattage. But the tests are always or usually based on the performance off all cores, and with their mix of p-cores and e-cores, it is difficult to know what kind of performance one will get, without someone actually doing both kinds of tests.
Developers have had over a decade to learn how to optimize performance over multiple cores. If Steinberg Cubase can do this, there's no excuse for developers who cannot.
@@GG-zv9ku that’s only true to a point.. Bitwig is better optimised, has a faster workflow. It is what Ableton was back in the day! It’s the new comer with all the improvements that the older ones fails to improve on. I have used Logic since the 90’s & I won’t be changing any time soon. Bitwig is the superior DAW.. & besides your eyes won’t bleed looking at it. Ableton is programmer art.. 0 definition it’s all homogeneous & thus strains the eyes more
Bro you're a Legend! Keep on grinding! As I am using an M1 Macbook Pro and regularly hit a limit. I'll probably switch to reaper as the daw uses all the cores...
There might be a good reason why the efficiency cores aren't utilized...you want them running background tasks, and do what they're good at. Logic needs the heavy lifting of the performance cores. That being said, a good compromise might be a toggle in the preferences to enable the e core utilization (hey and why not add how many are available).
Pro Tools supposedly added support for efficiency cores - I don’t suppose you’d be able to test the difference between the previous versions and the latest?
Your videos are highly appreciated. This gives me a good idea of which hardware configuration to go for. This whole situation with the CPU cores is a massive let down when you consider that Logic is run, maintained and developed by Apple... or is it? It would be nice to get an official statement from Apple, I don't feel like they'd be inclined to comment. At this stage in the game, the discussion should be about the utilisation of GPU cores. It definitely raises eye brows.
I can't believe they never fixed this for LP11. Interesting that on the CPU meter within logic, it is showing 12 processing threads, but on activity monitor it is obvious the performance cores are the ones doing the heavy lifting while the e-cores are not doing much to assist at all.
another great videos of yours. actually Logic 11 runs smooth on my Macbook Pro M3. I've never seen a system overload pop up, as it did all the time with my Macbook Air M1
I'm glad you said this. I'm thinking of diving in to Mackbook M3 Pro. What CPU and Ram do you have. How many tracks to you typically record and play on?? Thanking you in advance.
@@Darrin.Crawford I had a Macbook Air M1 from 2021 and switched it for the 14 inch Macbook Pro M3 with 18 gbRam. Try to get it with a student discount (which I did, I just found someone who was a student and they gave me 200 us$ off at the apple store) I only do instrumentals with 1 vocal track, however, I noticed that whenever I had to switch on low latency on the Macbook Air, now I can run tons of plugins almost withotu any latency due to the smaller buffer size which can handle more plugins at once. Usually I have tons of saturation, compression, delay, reverb plugins of all kind, all running through busses with sends and returns, with parallel side compression, parallel comp bus, etc... tons of routing and still the Macbook Pro M3 can handle everything. Never heard the Fan even! never! it has. Even though I did lots of research because the M3 models have less perfomance cores than the maxed out M1 or M2 models, there is a video on YT showing the differences in regards of perfomance cores, that's all you need to care about. You could also get 8 cores of which 4 are perfomance cores but 32 Gig Ram however I once read that heavy plugin processing and live recording benefits more from more perfomance cores rather than more Gig. So the 18 gig ram are completely fine. I hope I didn't confuse you too much, it's easy to get lost, just get the macbook Pro M3 model with 18 gb ram and you're good for the next 5 years at least. It's an investment for sure!
This is by design for consistency. The E-cores are significantly weaker than the P-cores and are reserved for non-audio application/OS processing. Every time playback is started in Logic it redistributes the tracks across the cores to balance out processing, not the individual plug-ins on them. Since all P-cores are equally powerful it can freely do so without having to worry about moving a very plugin heavy track to a weaker E-core and causing an overload.
it's true, but the efficiency cores are actually quite powerful. And as meislit9217 points out, other DAWs can and do use all cores. Perhaps it would require too much of a Logic rewrite, and the performance as it is, is enough. Perhaps Apple knows something we don't, and if ALL cores are maxed out, the extra heat makes the clock speed drop a llittle, and the benefits are negliable. No idea. But it's interesting for sure.
@@EricGoetzMusic Why exactly would it be a bad idea? If almost half of my computer's chip is e-cores (e.g. 11-core M3 Pro has 5 e-cores and 6 p-cores), I would absolutely want to have the option for them to be all maxed out when I'm doing intensive tasks. My tests showed that you can get significantly more performance from a chip when a DAW can fully use the e-cores.
Thank you for taking the time to conduct these tests and for sharing these results. What core settings do you recommend for an M1 Mac Studio Ultra and M1 Mac Pro?
Key factors to note here - these are laptops. The efficiency cores are there to optimize for laptops. It is rumored that going forward with M4, the desktop versions like the Mac Studio are going to have fewer efficiency cores, and more performance and GPU cores. So that will change these equations significantly. Should they adjust Logic to use efficiency cores more, even if that’s not the what they’re intended for? I’m not sure. For the DAWs that use the efficiency cores more, does that impact battery life and/or thermals?
Thank you for these videos - very clearly explained and deeply tested. I'm about to upgrade from intel to silicon (Logic and Ableton user) and very interesting that M1 Max with 64gb ram might be the best option and considerably cheaper than an M3.
Hi james. Do you think new CPUs (Qualcom Snapdragon X Elite and Intel Lunar Lake) that eliminates hyperthreads, taking as a priority single cores instead of Multi-cores, would present better results for DAWs performance in terms of more channels with more plugins?
I just opened a "real world" session randomly and looked at the CPU usage and it basically used a lot more of the efficiency cores than your example, so I'm not sure your selection of plugin doesn't matter.
My guess about chromaglow is that it is heavily using the neural engine, which is improved/bigger on the M2 chip, and that’s why you can get more tracks than on M1.
Thank you for the tests, and they are valid -- but wondering if you've tried to record enable an audio track with a few plugins, and few plugins on the master buss at the same time. There is some strangeness in how Logic has always handled this - it's very easy to get an overload because Logic forces those processes onto one core only. It's weird and not sure why this is the case, and I doubt that has changed in LP11 - but it's a very interesting and important thing to consider when considering which Mac to buy or when to upgrade. This is why I'm waiting to upgrade my M1 -- because single core performance could make the biggest difference when it comes to recording tracks while playing back a lot of existing tracks (and while keeping the buffer low to decrease the latency of live input tracks). thanks again for your videos...
Oh here we go! I can’t wait for the test on the hopefully coming new Mac Studio with an M3 or M4 chip. I’m running a mid 2014 MacBook Pro and been waiting for a year for this new Studio.
Mac Studio M1 base model is cheaper and likely better given the performance core count is higher. Not every daw and plug-in are utilising the e cores. Like others have said, when you upgrade the m1 you can keep it and run audiogridder.
When they 1st released Logic Pro 11, there was talk that certain things wouldn't work on an Intel chip because of it's lack of a Neural Processor. I wonder if Chroma Glow is utilizing that?
Many thanks to you James for reviewing these detailes, it's a shame Apple doesn't care about such optimisation, when Reaper (much smaller software producer) did.
This is absolutely ridiculous! (Not your test or results, but Apple's decision to have its own software fail to take advantage of its own computer's efficiency cores!) If Apple itself will not generate software to run on efficiency cores, why would any other developer do so? And why sell us computers that are "more powerful" only to find out they actually are not unless the software is developed in a very specific way?
Well, the thing is many other developers really *do* make their software utilise the efficiency cores. It's miserable that Apple doesn't, but they're not necessarily the best programmers for their own hardware.
You can take a look at my M3 Pro chip testing video! There's no point testing every single chip (nor do I have the financial ability to do so, haha), because the core utilization will be the same.
@@JamesZhan so I know that the M3 Pro has less performance cores and more efficiency cores but that doesn’t seem to affect Reaper as much (the DAW I use) as it does other DAWS.
@@nerdpunk8922 reaper can use all cores. And all M pro chips still has the same amount of p-core or more compared to the M3 in air. The m3 non pro/non max model in the entry level mac pro is the same as in the air, it comes with a card reader and a hdmi port, but only support 1 external monitor, while the air supports 2 if the internal screen is off and you have one display adapter per usb-port. But any M3 pro/max, or m2 pro/max will outperform m3 air with reaper, but at a higher cost. This test is not intended as a real world test, so you can expect the track count to be much lower in real usage, so it doesn't tell you what you can expect, so it isn't a real life performance test, thus there it would not tell you exactly what the real life performance difference would be, so it can't really guide you in terms of determining if the price difference is too high.
Logic appears not to even distribute among the performance cores well once you make track stacks and start to sidechain. It seems to overload a single core in that case. Make two relatively cpu heavy track stacks and try side-chain one over the other (very real world) and Logic will error with a single core spiked.
Really appreciate this analysis. "Efficiency cores" seem kind of useless, if they're only being used to offload low-power tasks then having a whole core sitting mostly idle basically all of the time seems like a waste of resources.
Although it's kind of disappointing that Logic Pro cannot benefit from using all cores I still don't see this as a bigger problem, I've worked entire production and mixing sessions on a simple M1 chip and haven't had any problems yet, Logic is just so well optimized that it can handle (almost) everything that you throw at it.
I was hoping for this test, just to be sure they had not addressed that issue. I saw no information suggesting that they had. But I still wonder if avid really has. They have had some buggy releases, so it would be interesting to see a comparison, to see if they are showing the same to be expected perprmance boost, in your test scenario. I have seen requests for ableton 12 and latest bitwig, but if they have not announced support for e-cores, I dont think they have implemented it. Apple could have, with logic, without making an announcement, but that is apple, you mever know what they announce and what is just implemented without a mention.
thanks for the video, James. disappointing, yes..Apple’s decision(s). But not very surprising..from them. i’m sure they’ve got thermal and & sustainable load performance reasons..for choosing this approach…but i am also thinking they will be glad one will need to buy the more expensive machine, in order to get the performance Logic ‘deserves’ .. happy holidays ;) thanks again
Do you know how the plugins on one channel are distributed to the cores? I sometimes get an overload signal with what does not seem like many tracks and I wonder why. Is it possible that one channel with many plugins is causing it all by itself? Also - do you (or anyone here) know how UADx plugins compare to the average plugin use of processing power? They seem to be quite heavy on the CPU.
Could anybody explain or tell the reason, why the Logic Pro goes fluently on MacBook 2019 I9 on Catalina, and does not go without a lag on BigSur and more recent macOS... many thanks in advance 🙏
Nice vid, James. One thing I noticed about the Chromaglow screen capture: there is significantly more economy core usage on the M1 than in the other M1 clips, or vs the M2 chromaglow clip. Could that indicate an update or something was happening in the background on the M1 during the chromaglow trials, giving the unexpected result?
Pointing out that a benchmark is not realistic is silly, that's not it's purpose and it's near impossible to make a realistic 'project test'. keep it up James, you got the best channel on youtube on this topic!
Hi James, I’m in a little bit of a pickle and cant decide between a refurbished M1 Max MacBook pro with 64G Ram and 4TB SSD and a new M3 Max with 36 GB Ram and 1TB SSD. Both are in the same price range. I mainly want to use it for film scoring and I work on Logic. What would you recommend?
For film scoring, I'm assuming you will need to use a lot of sample libraries in your projects? If that's the case, I would get M1 Max with 64GB of RAM because that's what you will need if you use tons of sample libraries.
@@JamesZhan Hi James thank you so much for replying. I will be using logic or Ableton until I find my preferred daw. And I know my course will involve Pro tools. I will be making a range of music and curating soundscapes for short films. Would you recommend the same 2021 M1 MacBook Pro? I am a little bit worried about buying the older Mac for it to be unusable in a year. But I don't want to pay more money for a worse performing Chip.
Thank you so much for your reply James! Basically Ram is is the key here, rather than a newer chip and more cores. I also like that the older mac has a lot of SSD. Mac charges such ridiculous prices for upgrading your storage
Have you tried leaving the processing threads at Automatic? Read somewhere that this made Logic use the efficiency cores but I don’t have a silicon Mac to test yet
Once you throw video editing into the equation, it's a lot simpler: just get the latest chip. The M2 and M3 chips all have more powerful media engines that would be useful for video.
HI HOW FAR CAN I GET ON AN M1 8GB WITH REAPER USING SOFT SYNTHS LIKE ARTURIA WITH ABOUT 5 PLUGINS ON EACH TRACK PRO Q3 , UAD 1176 ECT - I NEED LOW LATENCY FOR OVERDUBBING SOFT SYNTHS AND KEMPER
So what does it mean when my 2020 13" MBP with quad core i7 can do the same stress test you did and running up to 146 tracks using Logic 11? Used a mono gtr track with Saturn 2 set to high quality. I even added the Townhouse Buss Compressor and Ozone 10 EQ on the master buss to boot. Are we saying I have the same performance as a Apple Silicon chip? I did let it run for about 10 mins and the fans were definitely spinning hard but it kept going. Obviously mileage will vary since I usually have a mix of plugins and some that might use more CPU.
I'd imagine that the efficiency cores aren't really meant to handle heavy workloads like this and so the incremental amount of performance gain they'd add might not be worthwhile for the system scheduler to make use of them. It does look like a bunch of system stuff is getting offloaded onto them, at least, otherwise the efficiency cores would be reading 0%. Also I'd expect that most plugins (especially third-party ones) would be single-threaded, which would explain the ChromaGlow discrepancy. I think ChromaGlow would also be using the neural cores on Silicon, which is also pretty uncommon for non-Apple things to make use of.
In my other Mac testing videos featuring DAWs that can fully make use of the e-cores, the performance gain you get from fully using e-cores is NOT "incremental." Testing data showed that it's actually quite significant-enough for users of those DAWs to be able to get a "lesser" chip while getting MORE performance out of their money compared to a more expensive chip.
As a Logic user with an M1 Max chip, I tried to reproduce your test. I loaded a normal pop production into Logic. Nothing particularly many plugins. The efficiency cores kick in first. Almost exclusively. So most of the work runs through these cores. Then I reproduced your test. I got 370 mono tracks with Saturn and Chroma Verb. Then the performance cores are at 100% but the efficiency cores are maybe at 30%. You would expect it to be the other way around. I don't know why that is. But the tests with maximum load do not reflect reality. I'm a bit confused.
I think it's good that you're doing the tests. Thanks. I'm interested to see how the M4 Max will perform. I wish you continued success with your videos.
what about single core performance? pretty sure m3 pro should have much shorter bounce times, opening heavy plugins, etc. It's weird that you only have one metric to measure 'performance' as if amount of tracks running at once is the only metric for measuring performance / user experience
The biggest bottleneck for Apple M chips is ram and not the CPU. This is especially true for virtual orchestra templates for film scoring where its typical to employ hundreds of sample tracks which are memory and SSD intensive. CPU is more important for things like video editing where speed of rendering and transcoding video uses the CPU to its max.
*RAM size, not RAM Speed. It is an issue on 512gb models because that ram is slow as hell, but other RAMS are fast. Still if the ram size is low, it will use caching, but that happens on every machine. The issue is that Reverb and VST Synth weight a lot. Ableton tried to fix that with max for live. But still...
Obviously they’re doing this on purpose, it’s their own software. Are they doing this to keep these cores available for background tasks? Or to extend battery life? They should allow the user to enable use of all cores.
we just opened 2 different sessions on M1 Max and M2 Max Mac Studios with Logic 11.0.1. and in both cases all CPU Cores are used ... so is the problem fixed? why are there different results on that? thats strange
Hmm interesting. "Used" as in fully maxed out? Because I'm currently doing more Mac testing in preparation for the M4 Pro review video and the latest Logic 11 is still unable to maximize the efficiency cores.
Hey! Very nice video. One question... my iMac 27" (2019) is maxing out with bigger projects (Ableton). It's an Intel i5 (6x 3.0 GHz) with 64GB RAM. The RAM is not the problem here. I think 32GB would actually be more than enough. I think it's more about the CPU itself... If I want to upgrade (and if I want to get some portability) which is better? 1. MacBook Air M2 16/512 2. MacBook Pro M2 Pro 16/512 I'm using very CPU intense plugins from iZotope (Ozone 11 etc...), Kontakt, Fabfilter etc... is the M2 enough or should I go with a M2 pro? Thanks in advance...
My M4 Pro MacBook Pro Testing video is here! Watch it ➡th-cam.com/video/hccy19Hm6M8/w-d-xo.html
▶️My Analysis & Breakdown of the New M4 iMac, Mac Mini, and MacBook Pro for Music Production th-cam.com/video/POKZlRo-Lgo/w-d-xo.html
▶ My $12,000 Video/Music Production Desk Setup, Powered by an M1 Max MacBook Pro: th-cam.com/video/cA3ms_2Yn-A/w-d-xo.html
You win the Internet every time you post a video. You have saved me personally over €5000 over the last 2 years. Thank you for your deep analysis.
Yes please try on live 12 😊
Thank you for watching! 🙏
1⃣A popular comment is that e-cores are too weak to make a difference. They actually are not! In my M3 Pro MacBook Pro video where I tested 7 different DAWs, test results showed significant better performance when a DAW can fully use the e-cores. (Watch that video here ▶th-cam.com/video/FSqX4bt9to4/w-d-xo.html)
2⃣Another recurring comment is that e-core should be reserved for system tasks/light tasks. OK, the problem is that when you have a chip where almost half of the cores are e-cores (e.g. 11-core M3 Pro or the 8-core M2), it makes a HUGE difference to be able to make your computer to pull out all the stops (aka use all the cores fully) to handle CPU intensive tasks as needed.
3⃣Lastly, the point of this video (and my other DAW testing videos) is to help fellow musicians, audio engineers and music producers make more informed decisions when determining which Mac to buy, because it’s not as simple as “buy the latest because it’s better.” What you use as your primary DAW is a critical factor here due to different levels of core utilization.
I appreciate your support and stay creative!
on Macs, performance based DAWS like Bitwig Studio and Ableton forces all the realtime audio processing to the performance cores, as the efficiency cores are not suitable for it (they get aggressively forced to sleep etc.). Some linear arrangement hosts make use of the efficiency cores to process other things ahead, which are not realtime dependent, but with Bitwig Studio and Ableton Live being a non-linear DAW that has to be able to start any clip at any given time with sample accuracy, there's currently not much it can offload to those cores.
Possible that logic is the same. Have you reached out to these DAWS to ask them? This is what they will tell you.
What WAS the case with multi-core Intel processors (and may still be with Apple silicon), is that where music production is concerned, high single-core clock-speed performance was more important than total core count of the processor.
This was true because while a high core count MAY be able to allow your Mac running any DAW to handle a 'larger', more demanding project file, the workload of a single plug-in instantiation cannot be distributed across multiple cores. So as many users have experienced, a very cpu-intensive plug-in could halt your song-file with the dreaded processor overload dialog when the majority of cpu cores were just chillin'… because the demand of just one plug-in instantiation was more than the assigned core could handle.
Since through successive iterations of Intel processors we saw that models with fewer cores generally had higher single core clock speeds, while the models with the most processor cores were clocked at lesser (slower) slower speeds. For audio/music applications, if you want to run one or several virtual synth plugins with high polyphony (for example), then you'd likely do better with a processor having fewer cores, but able to run at a higher clock speed.
I've said all this because I don't have any knowledge that this radically different ARM-based SoC architecture has changed the importance of looking for high single-core clock speeds.
This is important work you've done in discovering the need (for logic users at least) to focus the 'performance core' count of the processor when shopping for a Mac, but AFAIK, high single-core performance is still more important than total 'performance core' count for this kind of work.
Yes, the M2 and M3 are faster. Maybe the core of Logic isn’t that optimized yet, but ChromaGlow demonstrated that.
Clock speed is irrelevant, you need to look at the performance of the core. With intel and amd, there are examples of cpus with higher clock speed and fewer cores compared to a model of the series above, and yet in single core tests, the model with more cores and lower clock speed in the higher series still has higher single core performance. Also when you are running more tracks than the core count, you really can't know that the single core performance will be the thing that overloads the cpu, as each core is running several tracks and their plugins. I have not seen a demonstration of any recent cpu, where it was obvious that higher single core performance trumps multi core performance. I have seen a lot of people claiming that. But I am very sceptical to that still beeing true, it might be something that people still belive because early multi core chips did have single core performance issues. And it might be the case for some tasks that can't utilize multi core systems, which was the case even long after the single core performance issues were dealt with.
@@TheJonHolsteinApple has some of the best single core benchmark scores in the industry
@@5urg3x but that has nothing to do with my comment. Apple's CPUs has great single core performance and good multi core, especially considering the wattage. But the tests are always or usually based on the performance off all cores, and with their mix of p-cores and e-cores, it is difficult to know what kind of performance one will get, without someone actually doing both kinds of tests.
Developers have had over a decade to learn how to optimize performance over multiple cores. If Steinberg Cubase can do this, there's no excuse for developers who cannot.
chromaglow uses the Neural engine. new version of Neural engine is better would be my thought
Can you please do the same test with Ableton Live 12?
you'd be dissapointed.
Better off with bitwig.
It’s a better DAW but that aside it’s way better optimised.
Bigwig is not a better daw. The daw you enjoy is a better daw than every other daw on the market.
@@GG-zv9ku that’s only true to a point..
Bitwig is better optimised, has a faster workflow.
It is what Ableton was back in the day!
It’s the new comer with all the improvements that the older ones fails to improve on.
I have used Logic since the 90’s & I won’t be changing any time soon.
Bitwig is the superior DAW..
& besides your eyes won’t bleed looking at it.
Ableton is programmer art.. 0 definition it’s all homogeneous & thus strains the eyes more
@@fallenleaf24 can you already work with audio tracks without warping in bitwig?....tested it a year ago and it was not possible...crazy
Love these videos! Please throw Ableton Live 12 into the hat as well (if possible) 🙏🏾
On the ChromaGlow issue: I think it might run at least partially on the Neural Engine, which is much faster on the M2 generation than on the M1 chips.
😂🤣
Bro you're a Legend! Keep on grinding!
As I am using an M1 Macbook Pro and regularly hit a limit. I'll probably switch to reaper as the daw uses all the cores...
m1 pro with the standard m1 chip or the m1 pro chip?
Upgrade to m1 ultra 😳
Your work is important! Thank you. Waiting for the M5 Pro chips ;)
i'm still running my studio based on Imac Core i9 10cores. Waiting for the M4!
There might be a good reason why the efficiency cores aren't utilized...you want them running background tasks, and do what they're good at. Logic needs the heavy lifting of the performance cores. That being said, a good compromise might be a toggle in the preferences to enable the e core utilization (hey and why not add how many are available).
What happens if you run the tests while setting the computer to battery efficiency mode?
Pro Tools supposedly added support for efficiency cores - I don’t suppose you’d be able to test the difference between the previous versions and the latest?
Second this- if AVID can do it the really Apple should be able to as it means that logic and ablation will become the slowest of of the biggest daws
Your videos are highly appreciated. This gives me a good idea of which hardware configuration to go for. This whole situation with the CPU cores is a massive let down when you consider that Logic is run, maintained and developed by Apple... or is it? It would be nice to get an official statement from Apple, I don't feel like they'd be inclined to comment. At this stage in the game, the discussion should be about the utilisation of GPU cores. It definitely raises eye brows.
great video! I really appreciate you doing this .... thank you
M1 Max with audiogridder is crushing right now
You deserve 1M+ subscribers
Haha thanks! Maybe someday lol
Logic is also calling their drum midi library, “AI”. Toontrack has been doing it for over 10 years
Could you run a test with the Ableton Live 12 and the newest Bitwig?
I can't believe they never fixed this for LP11. Interesting that on the CPU meter within logic, it is showing 12 processing threads, but on activity monitor it is obvious the performance cores are the ones doing the heavy lifting while the e-cores are not doing much to assist at all.
I look forward to your videos. Thank you!
Thanks for the great analysis and clear explanation.
Time for a studio one 7 vs Ableton 12 vs new FL Studio version, reaper & co.
Bro, you make great videos
Great work. Thanks. I'm mainly a Cubase 13 user but dabble in Logic Pro 11.
another great videos of yours. actually Logic 11 runs smooth on my Macbook Pro M3. I've never seen a system overload pop up, as it did all the time with my Macbook Air M1
I'm glad you said this. I'm thinking of diving in to Mackbook M3 Pro. What CPU and Ram do you have. How many tracks to you typically record and play on?? Thanking you in advance.
@@Darrin.Crawford I had a Macbook Air M1 from 2021 and switched it for the 14 inch Macbook Pro M3 with 18 gbRam. Try to get it with a student discount (which I did, I just found someone who was a student and they gave me 200 us$ off at the apple store) I only do instrumentals with 1 vocal track, however, I noticed that whenever I had to switch on low latency on the Macbook Air, now I can run tons of plugins almost withotu any latency due to the smaller buffer size which can handle more plugins at once. Usually I have tons of saturation, compression, delay, reverb plugins of all kind, all running through busses with sends and returns, with parallel side compression, parallel comp bus, etc... tons of routing and still the Macbook Pro M3 can handle everything. Never heard the Fan even! never! it has. Even though I did lots of research because the M3 models have less perfomance cores than the maxed out M1 or M2 models, there is a video on YT showing the differences in regards of perfomance cores, that's all you need to care about. You could also get 8 cores of which 4 are perfomance cores but 32 Gig Ram however I once read that heavy plugin processing and live recording benefits more from more perfomance cores rather than more Gig. So the 18 gig ram are completely fine. I hope I didn't confuse you too much, it's easy to get lost, just get the macbook Pro M3 model with 18 gb ram and you're good for the next 5 years at least. It's an investment for sure!
🧢 boy stop it
This is by design for consistency. The E-cores are significantly weaker than the P-cores and are reserved for non-audio application/OS processing. Every time playback is started in Logic it redistributes the tracks across the cores to balance out processing, not the individual plug-ins on them. Since all P-cores are equally powerful it can freely do so without having to worry about moving a very plugin heavy track to a weaker E-core and causing an overload.
Agreed. I feel like the content creator's premise is just flawed. Using the efficiency cores for realtime audio would be a bad idea.
@@EricGoetzMusicbut third party daws like reaper are able to use e-cores for audio processing...
it's true, but the efficiency cores are actually quite powerful. And as meislit9217 points out, other DAWs can and do use all cores. Perhaps it would require too much of a Logic rewrite, and the performance as it is, is enough. Perhaps Apple knows something we don't, and if ALL cores are maxed out, the extra heat makes the clock speed drop a llittle, and the benefits are negliable. No idea. But it's interesting for sure.
Maybe it is by design, but it results in less performance compared to other DAWs that are make use of the efficiency cores
@@EricGoetzMusic Why exactly would it be a bad idea? If almost half of my computer's chip is e-cores (e.g. 11-core M3 Pro has 5 e-cores and 6 p-cores), I would absolutely want to have the option for them to be all maxed out when I'm doing intensive tasks.
My tests showed that you can get significantly more performance from a chip when a DAW can fully use the e-cores.
I would love to see your analysis of Ableton Live 12 on this topic.
Thank you for taking the time to conduct these tests and for sharing these results. What core settings do you recommend for an M1 Mac Studio Ultra and M1 Mac Pro?
Hey may I ask what is the purpose of the Efficiency Cores? Thank you!
Thanks for your video and time! It does help us Mac Music producers! Keep up the great work !!
I have waited fot this video...thank you
Key factors to note here - these are laptops. The efficiency cores are there to optimize for laptops. It is rumored that going forward with M4, the desktop versions like the Mac Studio are going to have fewer efficiency cores, and more performance and GPU cores. So that will change these equations significantly. Should they adjust Logic to use efficiency cores more, even if that’s not the what they’re intended for? I’m not sure. For the DAWs that use the efficiency cores more, does that impact battery life and/or thermals?
Thank you for these videos - very clearly explained and deeply tested. I'm about to upgrade from intel to silicon (Logic and Ableton user) and very interesting that M1 Max with 64gb ram might be the best option and considerably cheaper than an M3.
Hi james. Do you think new CPUs (Qualcom Snapdragon X Elite and Intel Lunar Lake) that eliminates hyperthreads, taking as a priority single cores instead of Multi-cores, would present better results for DAWs performance in terms of more channels with more plugins?
I just opened a "real world" session randomly and looked at the CPU usage and it basically used a lot more of the efficiency cores than your example, so I'm not sure your selection of plugin doesn't matter.
Can u make logic settings video pleeeeeeeeeeeese
thank you keep doing what you do!
If you set laptop to low powermode the effiency cores star to work with the other cores
My guess about chromaglow is that it is heavily using the neural engine, which is improved/bigger on the M2 chip, and that’s why you can get more tracks than on M1.
04:20 I'd be interested in 10c M1 Pro/Max vs 12c M2 Pro/Max as they both have 8 performance cores while the M2 Pro/Max being clocked higher
Thank you for the tests, and they are valid -- but wondering if you've tried to record enable an audio track with a few plugins, and few plugins on the master buss at the same time. There is some strangeness in how Logic has always handled this - it's very easy to get an overload because Logic forces those processes onto one core only. It's weird and not sure why this is the case, and I doubt that has changed in LP11 - but it's a very interesting and important thing to consider when considering which Mac to buy or when to upgrade. This is why I'm waiting to upgrade my M1 -- because single core performance could make the biggest difference when it comes to recording tracks while playing back a lot of existing tracks (and while keeping the buffer low to decrease the latency of live input tracks). thanks again for your videos...
Oh here we go! I can’t wait for the test on the hopefully coming new Mac Studio with an M3 or M4 chip. I’m running a mid 2014 MacBook Pro and been waiting for a year for this new Studio.
Mac Studio M1 base model is cheaper and likely better given the performance core count is higher. Not every daw and plug-in are utilising the e cores.
Like others have said, when you upgrade the m1 you can keep it and run audiogridder.
I’m looking to upgrade from late 2013, what do you recommend for music production also for mixing in pro tools with lots of plugins ?
When they 1st released Logic Pro 11, there was talk that certain things wouldn't work on an Intel chip because of it's lack of a Neural Processor. I wonder if Chroma Glow is utilizing that?
Many thanks to you James for reviewing these detailes, it's a shame Apple doesn't care about such optimisation, when Reaper (much smaller software producer) did.
Hey is the key takeaway the same for ableton 11? Does Ableton mostly run/use performance cores similar to Logic?
very useful info, thank you!
I'm glad you found it helpful!
This is absolutely ridiculous! (Not your test or results, but Apple's decision to have its own software fail to take advantage of its own computer's efficiency cores!) If Apple itself will not generate software to run on efficiency cores, why would any other developer do so? And why sell us computers that are "more powerful" only to find out they actually are not unless the software is developed in a very specific way?
Well, the thing is many other developers really *do* make their software utilise the efficiency cores. It's miserable that Apple doesn't, but they're not necessarily the best programmers for their own hardware.
Hello James, great work as always! Can we have the wallpaper on the background?? Cheers!
Test the M3 MacBook Air or Pro???? I’m still looking for a test that compares the base M3 chips with Reaper.
You can take a look at my M3 Pro chip testing video! There's no point testing every single chip (nor do I have the financial ability to do so, haha), because the core utilization will be the same.
@@JamesZhan so I know that the M3 Pro has less performance cores and more efficiency cores but that doesn’t seem to affect Reaper as much (the DAW I use) as it does other DAWS.
@@nerdpunk8922 reaper can use all cores. And all M pro chips still has the same amount of p-core or more compared to the M3 in air. The m3 non pro/non max model in the entry level mac pro is the same as in the air, it comes with a card reader and a hdmi port, but only support 1 external monitor, while the air supports 2 if the internal screen is off and you have one display adapter per usb-port. But any M3 pro/max, or m2 pro/max will outperform m3 air with reaper, but at a higher cost. This test is not intended as a real world test, so you can expect the track count to be much lower in real usage, so it doesn't tell you what you can expect, so it isn't a real life performance test, thus there it would not tell you exactly what the real life performance difference would be, so it can't really guide you in terms of determining if the price difference is too high.
Logic appears not to even distribute among the performance cores well once you make track stacks and start to sidechain. It seems to overload a single core in that case. Make two relatively cpu heavy track stacks and try side-chain one over the other (very real world) and Logic will error with a single core spiked.
I would love to know how this test would go with the M1 Ultra v. Max. Or M2 Ultra.
Never mind, I see there is a vid on this already. Great reviews!
Ableton 12 video?
Really appreciate this analysis. "Efficiency cores" seem kind of useless, if they're only being used to offload low-power tasks then having a whole core sitting mostly idle basically all of the time seems like a waste of resources.
Very helpful. Thanks, mate!
Although it's kind of disappointing that Logic Pro cannot benefit from using all cores I still don't see this as a bigger problem, I've worked entire production and mixing sessions on a simple M1 chip and haven't had any problems yet, Logic is just so well optimized that it can handle (almost) everything that you throw at it.
Any grounds for class action for these abysmal findings?
Did Avid update the use of cores in protools ??
Ableton 12 plsss 🙏🏻
I was hoping for this test, just to be sure they had not addressed that issue. I saw no information suggesting that they had.
But I still wonder if avid really has. They have had some buggy releases, so it would be interesting to see a comparison, to see if they are showing the same to be expected perprmance boost, in your test scenario. I have seen requests for ableton 12 and latest bitwig, but if they have not announced support for e-cores, I dont think they have implemented it. Apple could have, with logic, without making an announcement, but that is apple, you mever know what they announce and what is just implemented without a mention.
Very helpful, thanks!
Thank you! Which DAWs can utilise the M1 Max to it's fullest potential?
Excellent video!
thanks for the video, James.
disappointing, yes..Apple’s decision(s). But not very surprising..from them. i’m sure they’ve got thermal and & sustainable load performance reasons..for choosing this approach…but i am also thinking they will be glad one will need to buy the more expensive machine, in order to get the performance Logic ‘deserves’ ..
happy holidays ;)
thanks again
Does anyone know if iPad processors work the same way? Specifically regarding audio production on iPad.
Do you know how the plugins on one channel are distributed to the cores? I sometimes get an overload signal with what does not seem like many tracks and I wonder why. Is it possible that one channel with many plugins is causing it all by itself?
Also - do you (or anyone here) know how UADx plugins compare to the average plugin use of processing power? They seem to be quite heavy on the CPU.
Also - THANK YOU FOR YOUR VIDEO!!
Could anybody explain or tell the reason, why the Logic Pro goes fluently on MacBook 2019 I9 on Catalina, and does not go without a lag on BigSur and more recent macOS... many thanks in advance 🙏
How does eleven do with a 2020 intel chip on a iMac ?
Nice vid, James. One thing I noticed about the Chromaglow screen capture: there is significantly more economy core usage on the M1 than in the other M1 clips, or vs the M2 chromaglow clip. Could that indicate an update or something was happening in the background on the M1 during the chromaglow trials, giving the unexpected result?
thx!!!
Thanks for the video.
Pointing out that a benchmark is not realistic is silly, that's not it's purpose and it's near impossible to make a realistic 'project test'. keep it up James, you got the best channel on youtube on this topic!
What does it mean? Should I buy the mac Mini with a lot of ram and thats it ?
May I asked where can I check the performance chips on each chip ?
Hi James, I’m in a little bit of a pickle and cant decide between a refurbished M1 Max MacBook pro with 64G Ram and 4TB SSD and a new M3 Max with 36 GB Ram and 1TB SSD. Both are in the same price range. I mainly want to use it for film scoring and I work on Logic. What would you recommend?
Hi I am a new music production student and in the exact same boat. Leaving my comment here in case anyone replies.
For film scoring, I'm assuming you will need to use a lot of sample libraries in your projects? If that's the case, I would get M1 Max with 64GB of RAM because that's what you will need if you use tons of sample libraries.
I replied!
@@JamesZhan Hi James thank you so much for replying. I will be using logic or Ableton until I find my preferred daw. And I know my course will involve Pro tools. I will be making a range of music and curating soundscapes for short films. Would you recommend the same 2021 M1 MacBook Pro?
I am a little bit worried about buying the older Mac for it to be unusable in a year. But I don't want to pay more money for a worse performing Chip.
Thank you so much for your reply James! Basically Ram is is the key here, rather than a newer chip and more cores. I also like that the older mac has a lot of SSD. Mac charges such ridiculous prices for upgrading your storage
What about the M3 max with 16 cores?
could you do a video about producing music on Logic Pro, using different iPads like Pro, Air etc?
Please make a benchmark for intel ultra 7 vs m3 air
Have you tried leaving the processing threads at Automatic? Read somewhere that this made Logic use the efficiency cores but I don’t have a silicon Mac to test yet
May I ask what screen you are using ? I see one in the background!
Thx for testing!!!
What about Avid Protools?Would be nice to get Infos!
Can you test DAW on Mac Studio's?
🙏🏼
Great video! So bottom line, what is the best choice of Mac if I use Logic and also want to do 4k video editing on a single machine?
Once you throw video editing into the equation, it's a lot simpler: just get the latest chip. The M2 and M3 chips all have more powerful media engines that would be useful for video.
can anyone explain why logic 11 doesn’t have 2048 buffer?
HI HOW FAR CAN I GET ON AN M1 8GB WITH REAPER USING SOFT SYNTHS LIKE ARTURIA WITH ABOUT 5 PLUGINS ON EACH TRACK PRO Q3 , UAD 1176 ECT - I NEED LOW LATENCY FOR OVERDUBBING SOFT SYNTHS AND KEMPER
So what does it mean when my 2020 13" MBP with quad core i7 can do the same stress test you did and running up to 146 tracks using Logic 11? Used a mono gtr track with Saturn 2 set to high quality. I even added the Townhouse Buss Compressor and Ozone 10 EQ on the master buss to boot. Are we saying I have the same performance as a Apple Silicon chip? I did let it run for about 10 mins and the fans were definitely spinning hard but it kept going. Obviously mileage will vary since I usually have a mix of plugins and some that might use more CPU.
I'd imagine that the efficiency cores aren't really meant to handle heavy workloads like this and so the incremental amount of performance gain they'd add might not be worthwhile for the system scheduler to make use of them. It does look like a bunch of system stuff is getting offloaded onto them, at least, otherwise the efficiency cores would be reading 0%.
Also I'd expect that most plugins (especially third-party ones) would be single-threaded, which would explain the ChromaGlow discrepancy. I think ChromaGlow would also be using the neural cores on Silicon, which is also pretty uncommon for non-Apple things to make use of.
In my other Mac testing videos featuring DAWs that can fully make use of the e-cores, the performance gain you get from fully using e-cores is NOT "incremental." Testing data showed that it's actually quite significant-enough for users of those DAWs to be able to get a "lesser" chip while getting MORE performance out of their money compared to a more expensive chip.
@@JamesZhan Ah, that's interesting. I'll have to watch some of your other videos on the subject.
1024 is fine for editing audio, but you're gonna want 128 if you're recording, especially with heavier-cpu plugins like U-He Diva.
As a Logic user with an M1 Max chip, I tried to reproduce your test. I loaded a normal pop production into Logic. Nothing particularly many plugins. The efficiency cores kick in first. Almost exclusively. So most of the work runs through these cores. Then I reproduced your test. I got 370 mono tracks with Saturn and Chroma Verb. Then the performance cores are at 100% but the efficiency cores are maybe at 30%. You would expect it to be the other way around. I don't know why that is. But the tests with maximum load do not reflect reality. I'm a bit confused.
I think it's good that you're doing the tests. Thanks. I'm interested to see how the M4 Max will perform. I wish you continued success with your videos.
Regarding the ChromaGlow tests: Could it be that the better performance of the Neural Engine of the newer processors is responsible for the results?
That's possible!
Logic just dropped its update 11.1 curious if data show's its optimized.
what about single core performance? pretty sure m3 pro should have much shorter bounce times, opening heavy plugins, etc. It's weird that you only have one metric to measure 'performance' as if amount of tracks running at once is the only metric for measuring performance / user experience
can you do one on Studio One 7???
It’s coming in my M4 Mac review
The biggest bottleneck for Apple M chips is ram and not the CPU. This is especially true for virtual orchestra templates for film scoring where its typical to employ hundreds of sample tracks which are memory and SSD intensive. CPU is more important for things like video editing where speed of rendering and transcoding video uses the CPU to its max.
*RAM size, not RAM Speed.
It is an issue on 512gb models because that ram is slow as hell, but other RAMS are fast.
Still if the ram size is low, it will use caching, but that happens on every machine.
The issue is that Reverb and VST Synth weight a lot. Ableton tried to fix that with max for live. But still...
Obviously they’re doing this on purpose, it’s their own software. Are they doing this to keep these cores available for background tasks? Or to extend battery life? They should allow the user to enable use of all cores.
we just opened 2 different sessions on M1 Max and M2 Max Mac Studios with Logic 11.0.1. and in both cases all CPU Cores are used ... so is the problem fixed? why are there different results on that? thats strange
Hmm interesting. "Used" as in fully maxed out? Because I'm currently doing more Mac testing in preparation for the M4 Pro review video and the latest Logic 11 is still unable to maximize the efficiency cores.
@@JamesZhan on the m2 it was a 150 track song with tons of plugins, so only little room left. On M1 max the session was smaller
Hey! Very nice video. One question... my iMac 27" (2019) is maxing out with bigger projects (Ableton). It's an Intel i5 (6x 3.0 GHz) with 64GB RAM. The RAM is not the problem here. I think 32GB would actually be more than enough. I think it's more about the CPU itself...
If I want to upgrade (and if I want to get some portability) which is better?
1. MacBook Air M2 16/512
2. MacBook Pro M2 Pro 16/512
I'm using very CPU intense plugins from iZotope (Ozone 11 etc...), Kontakt, Fabfilter etc... is the M2 enough or should I go with a M2 pro?
Thanks in advance...