JUST POSTED➡M4 Mac Mini: Insane Value for Music Production, But… | M4 vs M4 Pro vs M3 Pro vs M1 Pro th-cam.com/video/sUcIO18W3oE/w-d-xo.html 🙋FAQ🙋 *Q: Will you test the base M4 Mac mini? / Will you compare the M4 chip to the M4 Pro chip? A: Yes to both! These tests take time (as mentioned below) and I'm working on it! - *Q: Why isn’t [Insert DAW] included?* A: It takes an enormous amount of effort and time to create these testing videos, because I do everything myself from coming up with the testing methodologies, doing the actual testing many times to ensure repeatability, analyzing the test data, to writing the script, creating motion graphics, and filming and editing the video. I do this outside of a day job, and I want to help as many people as possible, so I chose these 7 DAWs simply because they are the most popular ones according to the statistics from Google Trends and TH-cam Trends. That said, testing LUNA, Bitwig, Digital Performer etc is certainly under consideration! - *Q: Where is the CPU monitor from?* A: Launch Activity Monitor, and hit CMD + 3 on your keyboard. - *Q: Why are you using the built in speakers?* A: Because it’s a variable that is found across all Macs. All the plugins used in this video also have a free trial. This allows you to recreate this test, or similar tests on your own. Also, changing the audio device won’t affect how a DAW uses the CPU cores. - *Q: It’s the devs’ intention to not have their DAW use efficiency cores.* A: I don’t care whether it’s intentional or not. At the end of the day, I just wanted to show that if a DAW doesn’t fully use the efficiency cores, you just get *less* performance compared to a DAW that does on the same chip, so you should keep your DAW of choice in mind when choosing an Apple Silicon chip. - *Q: Using efficiency cores can cause system issues, hangs etc.* A: In my experience, when ALL cores are fully used to the max, the system is laggy, but that’s to be expected. It doesn’t really cause any real issues, though. - *Q: No one uses that many tracks.* A: Please watch the video. I have explained that it’s not about the actual number of the track count. The track counts are just there to highlight the performance differences. If I don’t try to overload the DAWs, how could I measure the differences? Also, have you seen Jacob Collier’s sessions? Some people do have hundred of tracks in their DAW project. - *Q: Efficiency cores are too slow to make a difference.* A: That’s simply wrong. They might be slower than the performance cores, but they do give you a significant performance boost when utilized, as my tests showed. Many users have also shared experiences that corroborate my test results and conclusions. - *Q: In the real world, no one maxes out the CPU like that.* A: What I want to show is a simple fact: if your DAW can fully utilize both the efficiency cores and the performance cores on Apple Silicon chips, you can save some money by buying an older chip, or a chip with less overall cores. Conversely, if your DAW can’t, you will need to look at the number of performance cores rather than the total core count when deciding which Apple Silicon chips to buy. - *Q: Your tests are comparing apples to oranges. You should compare chips with the same performance cores/efficiency cores rather than the total core count.* A: I intentionally try to match the total core count because, for most people, that’s what they look at when they are looking into buying/upgrading a Mac. In fact, on Apple’s website where you configure your Mac, it doesn’t even tell you the performance core to efficiency core distribution; it only tells you the total core count. 10 vs 10 vs 12 vs 12 just seems a lot more straightforward than 10 vs 12 vs 14 vs 16 to me. - *Q: Are the fans loud when the CPU is maxed out?* A: Not for 16-inch MacBook Pros, even on High Power Mode, IMO. However, the 14-inch M4 Pro MacBook Pro’s fans do get really loud on High Power Mode. I plan to make a dedicated video to discuss this! - *Q: You are using different buffer sizes. Studio One's Dropout Protection was set to High instead of Maximum* A: It would be correct that the actual, technical buffers for the mix engine used is not all the same. However, it was deliberate decision I made to reduce *PERCEIVED* variables. What I mean is that when you set the Dropout Protection to Maximum in S1, it actually displays 2048 as the process buffer right underneath it, and this will 100% confuse many people and they would point out I “did not use 1024 buffer for S1.” I posted a video addressing this and showing test results with Dropout Protection set to Maximum: th-cam.com/video/HSGmveHv0mY/w-d-xo.html Most importantly, the main goal of this video is to show that some DAWs' CPU overload ceiling is higher than others. In other words, on the same hardware, you could do more with some DAWs than others. This is a difference between being able to just barely play back a huge mixing project smoothly, and just not being able to play it at all. - *Q: Did Logic Pro 11.1 update change anything?* A: No. I made a community post about it, if you are interested! - DAW VERSIONS TESTED: Logic Pro 11.0.1 REAPER 7.27 Cubase 14 Pro 14.0.5 Ableton Live 12 12.1.1 FL Studio 24.1.2 Pro Tools 24.6.0 Studio One 7 7.0.0.102470
You should have indicated that the processor variations are inconsistent. i.e All are not binned models or all are not unbinned models, but rather a mismatch of binned and unbinned models across the chip generations.
@@wisdomyaw03 Thanks for the feedback, but actually, the consistency I chose is the total core count, and not “binned vs unbinned.” 12 vs 12 vs 10 vs 10 is simply WAY easier to compare than 14 vs 12 vs 12 vs 10. To me, the latter just makes everything hard to compare. The fact that a chip is binned does NOT affect the performance of the rest of the cores-it's not like they operate on a much slower clock speed just because 2 cores are “disabled.” To me, this way of comparing just makes it a lot easier for everyone to understand. You are welcome to disagree, but I personally prefer this way :)
Interesting - I’m wondering, I’ve made the experience a few times already that a project wouldn’t playback anymore because of cpu overload when I was listening back through the internal speakers but the project still played back fine without cpu issues when connected to my interface. So I always thought it does make a difference (since the interface is an external sound processing card isn’t it?). Are you sure it doesn’t affect the cpu? Any idea what else can cause this experience? Thank you so much for your fantastic videos!
@@JamesZhan I appreciate the rationale behind your chip selection-which isn’t of much concern to me as much as your disclosure of these variations in order to alleviate any confusion people might have. I happen know these chip variations very well so I’m able to extrapolate the results and make sense of it. One thing I’ll say about your reason for this chip selection is that, the number of cores-hence trying to match them in terms of core count -is not necessarily important since each generation sometimes offer improvements through increasing the number of cores. If you take the base models(binned) for example, they all start at the same price. Likewise, the fully spec’d unbinned models. So, I think comparing on the basis of price would be more appropriate rather than the number of cores, so that an upgrade would be clearer for a viewer what improvements they are getting. That’s just my opinion. Cheers!
@@seńor_t007I am still using the Mac Pro 2015 with logic 😂 still works like a charm. I am in a semi professional field. So I am a student in music production but I am also working for a company doing music and sound design. And like I said my MacBook ist still doing its thing
I really Really REALLY appreciate the communication style you utilize. Moderate tone, effective conveyance of information and complete lack of sensationalism. TH-cam seems to be overtaken with people desperate for eyeballs and have become some of the most obnoxious, over the top barkers of their message. Thank you!
Glad you noticed! One of the key things I always strike for on this channel is that I present evidence to support my conclusions. I'm pretty brand-agnostic, actually, but if I appear to really like a product, I always try to show the why. I'm not totally unbiased or infallible, but at least I show people how I get to my conclusions. I also try to be very transparent when I'm NOT super familiar with the kind of product I'm reviewing (e.g. my Keychron Q1 Max review video), so people know it's more of a “this thing is cool and I like it” video and less of a in-depth review. It also helps that I don't rely on this channel for my living-it's my hobby and so there's no incentive for me to shill for any company :)
I agree with you 100% Rick. It takes me back to days when most people were honest and focussed on the evidence and truth instead of the current mess of clickbait greed and "truth doesn't matter" global movement, as we all witness in the increasing number of populist governments. Our dad's used to say, "When I was your age, the world was a very different place!". Now we're saying it to our kids! :) Keep up the great work James! The truth is gratifying and will eventually enlighten the masses... :)
@@JamesZhan Hi James. I have a M1 Pro 32GB Ram. I was thinking of trading in my MBP for a M4 MBP 24GB Ram. Would 24GB be enough? I use a lot of plugins and external SSD for drum samples and instruments.
Thanks - You provide an incredibly useful service to the music industry. I am about to totally revamp my music set up, and your videos cut through so much confusion. You are appreciated.
@@JamesZhan Thank you - I would be delighted. I have always used Reason with Windows - I am looking forward to seeing how Reason does with the M4 - I may just jump to Reaper and have done with it. Either way, very well done in creating actual useful content, not just opinion. I will continue to follow and recommend your channel.
@@JamesZhan Believe it or not, I read in your community replies when the video would be up and I was refreshing my browser on your channel at 12pm EST 🤣 But that's just because I am totally obsessed with getting info about the new models because this will be my first own MB. I've been writing music on an M1 Air 8gb which my music college gave me for the time of my studies. But that is coming to an end and I want to produce my Bachelor's Project on this thing. The M1 Air was actually fine while I was just using Logic stock plugins. But since I started using Neural DSP, Kontakt Libraries and Toontrack products it is reaching its limits. Especially RAM I think. Will probably get the M4 Pro 16" and the base 24gb will be great for now but my brain is yelling: "BUT WHAT IF IN 2 OR 3 YEARS YOU WANT TO USE EVEN MORE LIBRARIES AND WE RUN OUT OF RAM AGAIN AAAAAAAAH!" The 48gb upgrade is 460€ though which is insane. A video on RAM is what I crave now 😄
I never have a response to like and subscribe requests from the channel creator, but you're pretty much the only channel I've found that makes this specific niche content that I am looking for. Thank you for the informative videos!
Thank you for doing such thorough tests like this. Very much appreciated. I really hope you are able to do a Mac Mini M4 vs M4 Pro test/shootout as well. Thanks!
@@JamesZhan please consider and compare the fan noise between Mac mini m4 and Mac mini m4 pro. Also throw some light if the fan noise in Mac mini m4 pro occurs only when cpu is maxed out or its all time thing ? Please help us out 🙏
Thanks! I would like to have an test with Mac Mini M4 and M4 Pro comparison. Until where Mac Mini M4 could go in fact. But Great job here. You are the best.
I rarely comment videos on youtube, but man you're actually the only one who made a kind of video I was looking for in weeks. Thank you very much! This was very helpful coming from an M1 Max I had doubts about upgrading or not. Subscribed!
30+ years experienced Software Engineer here. Great video - most of your conclusions seem pretty right to me. And now my 2c Higher core utilisation for the same track count is probably a bad thing (Logic @ 10:20 ). It's fair to expect Logic to use more cores but forcing it to use more cores may simply cause "busy work". Splitting up and sharing work that can just as easily be done on fewer cores will potentially even reduce performance, although that's not evident in your test. A kid carrying a balloon cannot be sped up by adding more kids to the task. The funky patterns in Studio One at 13:30 are a classic. Part of the app is sharing work between core 8 and 10 without properly splitting the workload. They are taking turns at doing the job, when one is busy, the other is idle. This is possibly slower than just having one core do the whole thing but it may also be done deliberately by the OS to help with heat dissipation. You see this happen at 14:00 where the work is clearly offloaded from one core to another twice, briefly, and when this happens the audible popping is due to momentary delays in "handover" between them. Most musicians understand that crackle can happen when, say, one or more buffers worth of work miss a deadline and this can also happen because of another application running. Having a bit of un-utilised capacity or shutting down other apps should reduce this. This happens when the workload spikes - complex algorithms sometimes have very peaky workloads and good developers plan for this, not just the typical situation. It doesn't matter if the average workload is easy, for audio it really matters that the spikes do not cause glitches so developers might be forced to adopt a conservative policy here. In my experience it's most likely the reason efficiency cores have not been fully used by some apps is just because the developers haven't yet done the extra work to get that last juice out of it. Maybe they're slacking, maybe it's rocket science, maybe it's misguided management - I doubt it's a conspiracy. It's also possible that Apple knows and even advises developers to avoid full utilisation because regularly running a machine (especially a laptop) at full capacity is more likely to overheat it and this reduces the life of the machine. Machines built to run flat out (e.g. some servers) cost a lot more money because of this. Apple may also be making assumption about typical usage - sometimes even musicians want to balance resources with other apps they run at the same time. Having a setting for this seems ideal to me. It's common for developers to have a policy of not using all the cores because the computer can become unresponsive for other tasks. If there are 8 cores, the programmers may be deliberately using 7 of them so there's a spare to do any other concurrent work. Power users everywhere want to be able to override this and settings for doing so are common. Also, obviously this stuff is way complex but I would say upgrading a machine could still give a performance improvement for many users even though CPU cores are not being maxxed out. The best way is to somehow try the machine out in a realistic situation and judge for yourself. Apple used to offer no-questions-asked return and refund within 14 days.
Wow this is incredible. Thank you so much for taking the time to share this knowledge with us! I'm glad to hear that you found my conclusions sound. I'd gone over them many, many times before I published the video, but getting a software engineer's approval is extra reassuring that I'm not putting out misinformation (I'm always worried about doing that). RE: the funky pattern in S1. I'm guessing this is an S1 issue, since it only happens in S1? If so, do you think it's an error/bug in the software? Or is it the VST? I think, as a user, I would LIKE to have the option to use the efficiency cores to process audio if needed. Some of these Apple Silicon chips have a 50/50 split between e cores and p cores, so that other half can really make a difference between being able to mix a song and not. I like that REAPER let you choose how many cores you want it to use. With Logic, they weirdly give you the choice, but only for performance cores. This is also critical for people with a limited budget. Let's say someone is only able to afford the base M2 MBA, with 4 performance cores and 4 efficiency cores. REAPER will allow this person to do more on the laptop compared to Logic. And you are right; ideally, people should actually test the computer themsevles. I'm only here to provide a reference point. There are so many different use cases in music production, that no number of testing can truly reflect an individual's use case.
@@JamesZhan About the funky pattern. It's not likely to be considered a bug per se. My guess is *part* of the software that's running (probably S1 itself but hard to be 100%) has not been (re)written for the multi-core world we live in now. When programming you have to write very different code to enable multiple cores in parallel. It's almost always simpler to write it "single threaded" although these days this simply is not acceptable for the reason you demonstrate. An app is made of countless components and thsi might just be one component that was active during your example test. It could be the VST, the DAW or many other things. It might only happen when certain stars align or it might always be there. Even developers who work on the software may take a week to figure out what causes it so I would only be guessing. It does certainly relate to multi-core utilisation though, and this is notoriously hard to do well. Ultimately it's much more important how much work the machine can reliably do than what the graphs say of course.
@@chromosundrift As someone who have written a few plugins myself, this is very accurate. Dealing with multicore in any DAW where it's a mountain of code is daunting job.
Apparently as it is multicore processing with e cores and p cores mixed can cause dropouts, due to inherent challenges and a poor apple implementation (and documentation it seems). You're right that nobody wants full load on the cpu for audio, crackles will come before that (that's why ableton 'cpu' meter does not show actual cpu usage but the time ratio it takes to process the audio (100% is the full length of the buffer). Also ableton now prevents the OS from switching the load to E cores by creating a dummy load when the load is too light (they used to recommend a small buffer to force the cpu to switch). That tells me Apple needs to work on their real time audio implementation if devs have to use such a crude workaround.
Also obviously you want to leave some room for the GUI rendering (and everything else including the OS that's running). I suppose they can use the e cores for that.
Surprised to see that the Ableton Team isn’t taking full advantage of the processing power. Hope that changes in the future. Thanks for all your efforts on this James!
I think people asked Ableton that and the official response was that they had no plans to make use of the efficiency cores. Something about efficiency cores being only meant for light tasks (I'm just pulling this off the top of my head so I could be totally be misquoting, though)
Thanks! Wish i could give more, but i need to buy the RIGHT gear now that I've watched this. You really are saving people money and time in a clear and concise way. Just wish you had Bitwig in your DAW lineup, but that's fine LOL
Wow, thanks so much for the tip! No timeline yet, but let's just say I see all the Bitwig, LUNA, and Reason comments; just too many DAWs, too many tests, and too little time 😅
+1 on adding Bitwig Studio - I'd REALLY like to know where it fits in the lineup for relative performance w/ the DAWs you've tested. EXCELLENT WORK! Liked and Subscribed, James. 🙏🏻
James Zhan,I would honestly be very confused and pretty much lost if it wasn’t for this channel. You have a great way to take a complicated subject and turn it into something even I can really understand, and I appreciate that more than you’ll ever know. Thank you!
I guess there's a reason why my day job is producing educational videos, haha! I'm glad you found it easy to understand, as I do really try my best to parse all the tech info to achieve that goal :) I appreciate your kind words!
As a music producer myself my mixes in Ableton are usually anywhere from 45 to 70 tracks total with anywhere from 2 to 7 VSTs per track. You can imagine the level of taxation that puts on my M1 Pro but with the workload being distributed to my Apollo Twin X it’s still able to handle all that so far without any issues so it might still not be worth the upgrade to the M4 pro for me unless it saves me a lot of wav/mp3 export time then I might consider it. It would be nice to see export time differences as well between M1 to M4 CPUs! By the way it’s been really hard to find a video like this. Other cpu benchmarks I’ve seen related to music production simply stack multiple wav file tracks without using any VSTs. That is a useless test. I really appreciate both the level of effort and detail you’ve put into this video.
Thank you for your kind words! How's the core activity looks like when you mix in Ableton with that number of tracks and VSTs? Does Ableton only use the performance cores? (you will need to make sure there are no other heavy apps running for an accurate reading, though) I've confirmed this many times on my end but it's always nice to see it confirmed by other people!
@@JamesZhanso I just ran a couple tests in Ableton both with my Apollo sound interface on and off. In both cases the efficiency cores 1,2 were showing about 50-75% workload the entire time (which is totally different from what you get) with performance cores 3,4,5,6 working the hardest. I found it interesting that all the cores were working just as hard whether my Apollo was on or off… but I guess it makes sense considering only the UAD plugins get disabled when the Apollo is off.
@@Killyang That's really interesting! Thanks for sharing. If you stop and start the playback every like 4-5 seconds, do you see the efficiency cores' activity graphs fall and rise correspondingly? I've found that, sometimes, if there's several processes running in the background, when I make a DAW that can't use efficiency cores to play a huge project, macOS would “shift” all the other processes to the efficiency cores, making them look like they are processing audio in the DAW, when they are just handling all the other processes as the DAW is using up all the performance cores. This is why I often reboot the computer and quit as many apps as I can in the Activity monitor before running these tests.
@@JamesZhan agreed. I made sure to do a full reboot and only open Ableton. When the music stops the green graph on all the cores stays about half full and when the music continues the graph elevates maybe another 20% gradually on all the core but the difference is more pronounced on the performance cores. The uptick in the efficiency cores is there but not as obvious. The graph only goes all the way down (black) once i quit Ableton. Hope that helps!
This is insanely useful for me. Thank you for putting this all together. Have been having issues with Studio One for some time now and this video fully convinced me I should move to Cubase.
Apart from this top notch content, i would request you to add on one more thing and that is Unified memory usage (for a better understanding of ram usage to swap memory usage ratio )
Drew at UA said that Luna does use the efficiency cores as well as the performance cores. In my testing I have found that to be correct. Perhaps James could include Luna in the next test to confirm this.
It's amazing how much work you put in! All the graphics, the many tests. Respect, you are by far the greatest DAW tester for Apple computers on TH-cam.
10:37 I also think Apple does this on purpose (the way Logic Pro doesn’t directly utilize the E cores like you can get the other DAWs to). I would absolutely love for an employee / software engineer or even a spokesperson to respond to this statement. I would be super curious to find out how they would respond… like maybe there is some less obvious reason to do that…, which I think would be important for ALL of us to know. Again… I love what you are doing here.
When you get weird HARD pops and a CPU dropout that is called an "X-run" in the Unix/Linux/BSD world. It use to be a word everyone knew then it stopped being a problem when Apple switched to intel from PowerPC. But it seems Studio One is having an X-run Issue. X-runs happen when a core allocation jumps from one core to another to alleviate pressure for too much code coming down the pipeline for preemptive multitasking. I remember this happening all the time on the old Power Macs G5s since they had 2 PPC Dual Core CPUs and plugins didn't know where to send the execution correctly so it would jump from core to core causing WAY worse Xruns than the M4 does the drops would last for almost a full second sometimes. But those are called Xruns. Great Job as always James!! You truly are a treasure to the audio community!
Interesting. I am about to make the jump to S1 on Solicon (5 Pro) from Intel and this is a little discouraging to be honest. Might start exploring Cubase. However, I don't do massive mixing, - I am more focused on tracking, so multi-channel simultaneous recording (20+ channels) is the performance angle I will be looking at.
@@trebleboost7 If you are coming from any intel machine even a XEON mac pro you will see a huge improvement even with a base M4 Mac mini. I am a track engineer and that is the bulk of my work since I have a space big enough to track drums and console. The 2 things you can do to improve tracking performance is a using a Thunderbolt Interface and stay away from USB interfaces. This will lower latency and actually lower CPU usage since USB needs to talk to the CPU for every byte is writes to a drive, Thunderbolt doesn't have this problem. Thunderbolt actually runs from the PCI bus. The difference is staggering. Example in Reaper using the same interface an Antelope audio Orion Studio Synergy core which has both USB and Thunderbolt with a 48K sampling rate and a 128 buffer. On USB the latency is 8.4 milliseconds. Thunderbolt its 2.2 milliseconds. That doesn't sound like a lot until you have 50 track up with plugins and you need to record something else. On Thunderbolt I can set my buffer to 128 and never change it through the entire tracking and mixing process. On USB I have to raise it once I have about 10 to 15 tracks with plugins on them. Thats just the nature of the 2 ports. But as far as the M4 Pro you will see a huge improvement over any Intel Mac.
Very informative. My M1 Max Studio with Logic works like a charm but I might upgrade to the next one for video. Or wait another year. These chips are glorious.
James, thank you so much for all the effort you put into these videos, both on the research and testing side, and on the clarity and detail of the presentation.
Thank you for the content! Is it strange that I see more efficiency cores used than performance cores when mixing projects in DP11? Showing about 37% Ecores and 20% Pcores. Mac mini M4pro 24/512.
@JamesZhan from your video I was really blown away by how powerful the base m4 pro is! I was debating the max but pro is the way I feel! Ill be upgrading from m1 pro which is already a beast but does start to fall apart when I mix/master. In terms of spec i am thinking 1 or 2tb for storage as I currently have a 1tb and have about 100gb left. The one thing on my mind is the nano texture display (yet to see this myself in person) and then my other thought is debating whether to upgrade to 48gb unified mem. What are your thoughts on more unified mem? Do you think it will affect heavy plugin usage?
This benchmark and analysis has got to be the best one I've seen so far. Very thorough, very mindful and very self conscious of your own "limitation" (for lack of a better term). I'm definitely subscribing this high quality content!!
James, you're so precise, and I really love how you get to the truth to make the best buying decision for audio production. I look forward to what you think about the mac mini computers and which is the best one to buy. Subscribed and thank you!
Once again, excellent work, James! I really hope your channel grows even more, because your content is truly high quality work and you deserve to reap the rewards! Definitely my nr1 favorite Music/Prod/Tech channel! All the best
Thank you James for this fantastic test video! Everything was beautifully explained and illustrated. Looking forward to the next one about fan noise and thermals.
JUST POSTED➡M4 Mac Mini: Insane Value for Music Production, But… | M4 vs M4 Pro vs M3 Pro vs M1 Pro th-cam.com/video/sUcIO18W3oE/w-d-xo.html
🙋FAQ🙋
*Q: Will you test the base M4 Mac mini? / Will you compare the M4 chip to the M4 Pro chip?
A: Yes to both! These tests take time (as mentioned below) and I'm working on it!
-
*Q: Why isn’t [Insert DAW] included?*
A: It takes an enormous amount of effort and time to create these testing videos, because I do everything myself from coming up with the testing methodologies, doing the actual testing many times to ensure repeatability, analyzing the test data, to writing the script, creating motion graphics, and filming and editing the video. I do this outside of a day job, and I want to help as many people as possible, so I chose these 7 DAWs simply because they are the most popular ones according to the statistics from Google Trends and TH-cam Trends. That said, testing LUNA, Bitwig, Digital Performer etc is certainly under consideration!
-
*Q: Where is the CPU monitor from?*
A: Launch Activity Monitor, and hit CMD + 3 on your keyboard.
-
*Q: Why are you using the built in speakers?*
A: Because it’s a variable that is found across all Macs. All the plugins used in this video also have a free trial. This allows you to recreate this test, or similar tests on your own. Also, changing the audio device won’t affect how a DAW uses the CPU cores.
-
*Q: It’s the devs’ intention to not have their DAW use efficiency cores.*
A: I don’t care whether it’s intentional or not. At the end of the day, I just wanted to show that if a DAW doesn’t fully use the efficiency cores, you just get *less* performance compared to a DAW that does on the same chip, so you should keep your DAW of choice in mind when choosing an Apple Silicon chip.
-
*Q: Using efficiency cores can cause system issues, hangs etc.*
A: In my experience, when ALL cores are fully used to the max, the system is laggy, but that’s to be expected. It doesn’t really cause any real issues, though.
-
*Q: No one uses that many tracks.*
A: Please watch the video. I have explained that it’s not about the actual number of the track count. The track counts are just there to highlight the performance differences. If I don’t try to overload the DAWs, how could I measure the differences? Also, have you seen Jacob Collier’s sessions? Some people do have hundred of tracks in their DAW project.
-
*Q: Efficiency cores are too slow to make a difference.*
A: That’s simply wrong. They might be slower than the performance cores, but they do give you a significant performance boost when utilized, as my tests showed. Many users have also shared experiences that corroborate my test results and conclusions.
-
*Q: In the real world, no one maxes out the CPU like that.*
A: What I want to show is a simple fact: if your DAW can fully utilize both the efficiency cores and the performance cores on Apple Silicon chips, you can save some money by buying an older chip, or a chip with less overall cores. Conversely, if your DAW can’t, you will need to look at the number of performance cores rather than the total core count when deciding which Apple Silicon chips to buy.
-
*Q: Your tests are comparing apples to oranges. You should compare chips with the same performance cores/efficiency cores rather than the total core count.*
A: I intentionally try to match the total core count because, for most people, that’s what they look at when they are looking into buying/upgrading a Mac. In fact, on Apple’s website where you configure your Mac, it doesn’t even tell you the performance core to efficiency core distribution; it only tells you the total core count. 10 vs 10 vs 12 vs 12 just seems a lot more straightforward than 10 vs 12 vs 14 vs 16 to me.
-
*Q: Are the fans loud when the CPU is maxed out?*
A: Not for 16-inch MacBook Pros, even on High Power Mode, IMO. However, the 14-inch M4 Pro MacBook Pro’s fans do get really loud on High Power Mode. I plan to make a dedicated video to discuss this!
-
*Q: You are using different buffer sizes. Studio One's Dropout Protection was set to High instead of Maximum*
A: It would be correct that the actual, technical buffers for the mix engine used is not all the same. However, it was deliberate decision I made to reduce *PERCEIVED* variables. What I mean is that when you set the Dropout Protection to Maximum in S1, it actually displays 2048 as the process buffer right underneath it, and this will 100% confuse many people and they would point out I “did not use 1024 buffer for S1.”
I posted a video addressing this and showing test results with Dropout Protection set to Maximum: th-cam.com/video/HSGmveHv0mY/w-d-xo.html
Most importantly, the main goal of this video is to show that some DAWs' CPU overload ceiling is higher than others. In other words, on the same hardware, you could do more with some DAWs than others. This is a difference between being able to just barely play back a huge mixing project smoothly, and just not being able to play it at all.
-
*Q: Did Logic Pro 11.1 update change anything?*
A: No. I made a community post about it, if you are interested!
-
DAW VERSIONS TESTED:
Logic Pro 11.0.1
REAPER 7.27
Cubase 14 Pro 14.0.5
Ableton Live 12 12.1.1
FL Studio 24.1.2
Pro Tools 24.6.0
Studio One 7 7.0.0.102470
You should have indicated that the processor variations are inconsistent. i.e All are not binned models or all are not unbinned models, but rather a mismatch of binned and unbinned models across the chip generations.
@@wisdomyaw03 Thanks for the feedback, but actually, the consistency I chose is the total core count, and not “binned vs unbinned.” 12 vs 12 vs 10 vs 10 is simply WAY easier to compare than 14 vs 12 vs 12 vs 10. To me, the latter just makes everything hard to compare.
The fact that a chip is binned does NOT affect the performance of the rest of the cores-it's not like they operate on a much slower clock speed just because 2 cores are “disabled.”
To me, this way of comparing just makes it a lot easier for everyone to understand. You are welcome to disagree, but I personally prefer this way :)
Interesting - I’m wondering, I’ve made the experience a few times already that a project wouldn’t playback anymore because of cpu overload when I was listening back through the internal speakers but the project still played back fine without cpu issues when connected to my interface. So I always thought it does make a difference (since the interface is an external sound processing card isn’t it?). Are you sure it doesn’t affect the cpu?
Any idea what else can cause this experience?
Thank you so much for your fantastic videos!
Quick question: is there a difference between the 24Go Ram and the 48? Thank you!
@@JamesZhan I appreciate the rationale behind your chip selection-which isn’t of much concern to me as much as your disclosure of these variations in order to alleviate any confusion people might have. I happen know these chip variations very well so I’m able to extrapolate the results and make sense of it.
One thing I’ll say about your reason for this chip selection is that, the number of cores-hence trying to match them in terms of core count -is not necessarily important since each generation sometimes offer improvements through increasing the number of cores. If you take the base models(binned) for example, they all start at the same price. Likewise, the fully spec’d unbinned models. So, I think comparing on the basis of price would be more appropriate rather than the number of cores, so that an upgrade would be clearer for a viewer what improvements they are getting. That’s just my opinion. Cheers!
The best content creator for testing new mac and apple products related to music production, mixing, mastering and more!
hands down.
Really appreciate the kind words! Glad you like the videos :)
The only Engineer and Content Creator out there doing this.❤
Absolutely true. James is GOAT
Mr. Zhan is the truth!
“Fight the fear of missing out”. To me the most important sentence in this video.
I've heard so many musicians thriving on the OG M1 Mac mini!
@@dalil260175 yeah, great quote for sure!
I’m still on the intel right before the M1😅
@@seńor_t0072018 intel mac mini here but think Im going m2 studio ultra or max. don’t need tb 5
@@seńor_t007I am still using the Mac Pro 2015 with logic 😂 still works like a charm. I am in a semi professional field. So I am a student in music production but I am also working for a company doing music and sound design. And like I said my MacBook ist still doing its thing
I really Really REALLY appreciate the communication style you utilize. Moderate tone, effective conveyance of information and complete lack of sensationalism. TH-cam seems to be overtaken with people desperate for eyeballs and have become some of the most obnoxious, over the top barkers of their message. Thank you!
Glad you noticed! One of the key things I always strike for on this channel is that I present evidence to support my conclusions. I'm pretty brand-agnostic, actually, but if I appear to really like a product, I always try to show the why. I'm not totally unbiased or infallible, but at least I show people how I get to my conclusions. I also try to be very transparent when I'm NOT super familiar with the kind of product I'm reviewing (e.g. my Keychron Q1 Max review video), so people know it's more of a “this thing is cool and I like it” video and less of a in-depth review.
It also helps that I don't rely on this channel for my living-it's my hobby and so there's no incentive for me to shill for any company :)
I agree with you 100% Rick. It takes me back to days when most people were honest and focussed on the evidence and truth instead of the current mess of clickbait greed and "truth doesn't matter" global movement, as we all witness in the increasing number of populist governments. Our dad's used to say, "When I was your age, the world was a very different place!". Now we're saying it to our kids! :)
Keep up the great work James! The truth is gratifying and will eventually enlighten the masses... :)
Agree! We Appreciate your neutral objectivity 👍
I second this, very well put!
Best Mac Audio benchmark channel ever, thank you James!
Thank you for your kind words! I’m humbled to hear that
@@JamesZhan Hi James. I have a M1 Pro 32GB Ram. I was thinking of trading in my MBP for a M4 MBP 24GB Ram. Would 24GB be enough? I use a lot of plugins and external SSD for drum samples and instruments.
Thanks - You provide an incredibly useful service to the music industry. I am about to totally revamp my music set up, and your videos cut through so much confusion. You are appreciated.
Wow thank you so much for the tip! I'm glad my videos are helpful. And good luck with the revamp! I would love to hear how it goes
@@JamesZhan Thank you - I would be delighted. I have always used Reason with Windows - I am looking forward to seeing how Reason does with the M4 - I may just jump to Reaper and have done with it. Either way, very well done in creating actual useful content, not just opinion. I will continue to follow and recommend your channel.
Yeah but this guy should be testing Gig Performer!
You rule, dude. Thank you for doing this!
Wow, thank you so much for the super thanks! I'm glad you're finding these videos helpful.
When i will earn i will too contribute
Thanks for providing such meaningful content for free.❤
This series has turned into one of my most anticipated videos of the year now. Thanks for all of the effort you put into these videos!
It's absolutely bonkers to hear that someone was looking forward to my video 😅 thank you so much for your support!
@@JamesZhan Believe it or not, I read in your community replies when the video would be up and I was refreshing my browser on your channel at 12pm EST 🤣 But that's just because I am totally obsessed with getting info about the new models because this will be my first own MB. I've been writing music on an M1 Air 8gb which my music college gave me for the time of my studies. But that is coming to an end and I want to produce my Bachelor's Project on this thing. The M1 Air was actually fine while I was just using Logic stock plugins. But since I started using Neural DSP, Kontakt Libraries and Toontrack products it is reaching its limits. Especially RAM I think. Will probably get the M4 Pro 16" and the base 24gb will be great for now but my brain is yelling: "BUT WHAT IF IN 2 OR 3 YEARS YOU WANT TO USE EVEN MORE LIBRARIES AND WE RUN OUT OF RAM AGAIN AAAAAAAAH!" The 48gb upgrade is 460€ though which is insane. A video on RAM is what I crave now 😄
Thanks for the great video!
If you can somehow run the fan noise and thermal throttling on a Mac mini m4 pro 14 cores that would be awesome 😎
+1 That exact machine is in my cross hairs.
The bus chain test is such a nice test. This is exactly what I (and i guess other audio producers as well) needed to know. Thanks !
I never have a response to like and subscribe requests from the channel creator, but you're pretty much the only channel I've found that makes this specific niche content that I am looking for. Thank you for the informative videos!
Wow, thank you for your kind words!! I really appreciate that.
This is the best comparison video on the whole TH-cam!
The best guy on TH-cam for MacBook and musicians
Thanks! Great stuff. A lot of work, for sure.
Great job as always!
Youre truly amazing fr… thank you again man. Nobody else is doing anything like this
Wow, thanks so much for the tip! It really helps with funding purchases to make videos like this.
Thanks James, for putting all that tedious analysis in for our benefit! Much appreciated.
You've done a really massive job! Thanks!
please add BITWIG STUDIO !
I tested Ableton and Bitwig.Bitwig's performance is comparable to Ableton.
Can't wait for M4 mini BASE test results on these daws! I like the clear style of presentation.
M1 base model 2020 here, it’s still super good, no upgrade needed
Thank you James for all the tests, it’s very appreciated!
Awesome content!
Danke!🙃
Thank you for doing such thorough tests like this. Very much appreciated. I really hope you are able to do a Mac Mini M4 vs M4 Pro test/shootout as well. Thanks!
Already working on it :)
@@JamesZhan please consider and compare the fan noise between Mac mini m4 and Mac mini m4 pro. Also throw some light if the fan noise in Mac mini m4 pro occurs only when cpu is maxed out or its all time thing ? Please help us out 🙏
Thanks! I would like to have an test with Mac Mini M4 and M4 Pro comparison. Until where Mac Mini M4 could go in fact. But Great job here. You are the best.
Thanks so much for the tip! I will be working on that :)
This is the most useful Tech Review on Apple silicon Computers for Audio Production users, thank you James, you are amazing!
I rarely comment videos on youtube, but man you're actually the only one who made a kind of video I was looking for in weeks. Thank you very much! This was very helpful coming from an M1 Max I had doubts about upgrading or not.
Subscribed!
Thanks so much for your kind words! I'm glad you found the video helpful
Really good extensive testing. This really helps the whole audio community!💎
I subbed for the insane detail and calm, objective demeanor. This is highly professional.
Same 👍
30+ years experienced Software Engineer here. Great video - most of your conclusions seem pretty right to me. And now my 2c
Higher core utilisation for the same track count is probably a bad thing (Logic @ 10:20 ). It's fair to expect Logic to use more cores but forcing it to use more cores may simply cause "busy work". Splitting up and sharing work that can just as easily be done on fewer cores will potentially even reduce performance, although that's not evident in your test. A kid carrying a balloon cannot be sped up by adding more kids to the task.
The funky patterns in Studio One at 13:30 are a classic. Part of the app is sharing work between core 8 and 10 without properly splitting the workload. They are taking turns at doing the job, when one is busy, the other is idle. This is possibly slower than just having one core do the whole thing but it may also be done deliberately by the OS to help with heat dissipation. You see this happen at 14:00 where the work is clearly offloaded from one core to another twice, briefly, and when this happens the audible popping is due to momentary delays in "handover" between them.
Most musicians understand that crackle can happen when, say, one or more buffers worth of work miss a deadline and this can also happen because of another application running. Having a bit of un-utilised capacity or shutting down other apps should reduce this. This happens when the workload spikes - complex algorithms sometimes have very peaky workloads and good developers plan for this, not just the typical situation. It doesn't matter if the average workload is easy, for audio it really matters that the spikes do not cause glitches so developers might be forced to adopt a conservative policy here.
In my experience it's most likely the reason efficiency cores have not been fully used by some apps is just because the developers haven't yet done the extra work to get that last juice out of it. Maybe they're slacking, maybe it's rocket science, maybe it's misguided management - I doubt it's a conspiracy.
It's also possible that Apple knows and even advises developers to avoid full utilisation because regularly running a machine (especially a laptop) at full capacity is more likely to overheat it and this reduces the life of the machine. Machines built to run flat out (e.g. some servers) cost a lot more money because of this. Apple may also be making assumption about typical usage - sometimes even musicians want to balance resources with other apps they run at the same time. Having a setting for this seems ideal to me.
It's common for developers to have a policy of not using all the cores because the computer can become unresponsive for other tasks. If there are 8 cores, the programmers may be deliberately using 7 of them so there's a spare to do any other concurrent work. Power users everywhere want to be able to override this and settings for doing so are common.
Also, obviously this stuff is way complex but I would say upgrading a machine could still give a performance improvement for many users even though CPU cores are not being maxxed out. The best way is to somehow try the machine out in a realistic situation and judge for yourself. Apple used to offer no-questions-asked return and refund within 14 days.
Wow this is incredible. Thank you so much for taking the time to share this knowledge with us!
I'm glad to hear that you found my conclusions sound. I'd gone over them many, many times before I published the video, but getting a software engineer's approval is extra reassuring that I'm not putting out misinformation (I'm always worried about doing that).
RE: the funky pattern in S1. I'm guessing this is an S1 issue, since it only happens in S1? If so, do you think it's an error/bug in the software? Or is it the VST?
I think, as a user, I would LIKE to have the option to use the efficiency cores to process audio if needed. Some of these Apple Silicon chips have a 50/50 split between e cores and p cores, so that other half can really make a difference between being able to mix a song and not. I like that REAPER let you choose how many cores you want it to use. With Logic, they weirdly give you the choice, but only for performance cores.
This is also critical for people with a limited budget. Let's say someone is only able to afford the base M2 MBA, with 4 performance cores and 4 efficiency cores. REAPER will allow this person to do more on the laptop compared to Logic.
And you are right; ideally, people should actually test the computer themsevles. I'm only here to provide a reference point. There are so many different use cases in music production, that no number of testing can truly reflect an individual's use case.
@@JamesZhan About the funky pattern. It's not likely to be considered a bug per se. My guess is *part* of the software that's running (probably S1 itself but hard to be 100%) has not been (re)written for the multi-core world we live in now. When programming you have to write very different code to enable multiple cores in parallel. It's almost always simpler to write it "single threaded" although these days this simply is not acceptable for the reason you demonstrate. An app is made of countless components and thsi might just be one component that was active during your example test.
It could be the VST, the DAW or many other things. It might only happen when certain stars align or it might always be there. Even developers who work on the software may take a week to figure out what causes it so I would only be guessing. It does certainly relate to multi-core utilisation though, and this is notoriously hard to do well.
Ultimately it's much more important how much work the machine can reliably do than what the graphs say of course.
@@chromosundrift As someone who have written a few plugins myself, this is very accurate. Dealing with multicore in any DAW where it's a mountain of code is daunting job.
Apparently as it is multicore processing with e cores and p cores mixed can cause dropouts, due to inherent challenges and a poor apple implementation (and documentation it seems). You're right that nobody wants full load on the cpu for audio, crackles will come before that (that's why ableton 'cpu' meter does not show actual cpu usage but the time ratio it takes to process the audio (100% is the full length of the buffer). Also ableton now prevents the OS from switching the load to E cores by creating a dummy load when the load is too light (they used to recommend a small buffer to force the cpu to switch). That tells me Apple needs to work on their real time audio implementation if devs have to use such a crude workaround.
Also obviously you want to leave some room for the GUI rendering (and everything else including the OS that's running). I suppose they can use the e cores for that.
this the video I’ve been waiting for from the creator who’s the best at this, thanks for doing this so fast 🙏
Surprised to see that the Ableton Team isn’t taking full advantage of the processing power. Hope that changes in the future. Thanks for all your efforts on this James!
I think people asked Ableton that and the official response was that they had no plans to make use of the efficiency cores. Something about efficiency cores being only meant for light tasks (I'm just pulling this off the top of my head so I could be totally be misquoting, though)
They can't make it reliable enough for now (see my other posts) but they say they're working on it.
Great comparison! Can you please also do Bitwig? It’s an amazing DAW! :)
Very very helpful. Jame's style is so easy to listen to and very thorough comparison.
These videos are the most helpful I have found. Thank you. Please continue doing the test that mixes through buses.
Thanks! Wish i could give more, but i need to buy the RIGHT gear now that I've watched this. You really are saving people money and time in a clear and concise way. Just wish you had Bitwig in your DAW lineup, but that's fine LOL
Wow, thanks so much for the tip! No timeline yet, but let's just say I see all the Bitwig, LUNA, and Reason comments; just too many DAWs, too many tests, and too little time 😅
+1 on adding Bitwig Studio - I'd REALLY like to know where it fits in the lineup for relative performance w/ the DAWs you've tested. EXCELLENT WORK! Liked and Subscribed, James. 🙏🏻
You had me after watching the first 1:16 !!! This is what I’ve been looking for
Thanks running these tests!!!!!
Glad to help!
I was WAITING on your review of this! Thank you!
James Zhan,I would honestly be very confused and pretty much lost if it wasn’t for this channel. You have a great way to take a complicated subject and turn it into something even I can really understand, and I appreciate that more than you’ll ever know. Thank you!
I guess there's a reason why my day job is producing educational videos, haha! I'm glad you found it easy to understand, as I do really try my best to parse all the tech info to achieve that goal :) I appreciate your kind words!
I can't imagine how much work went into this. Ty James!
Thanks for the benchmark. My M4 pro is ordered 🎉
This was really well done: good DAW coverage, systematic, thorough. Thanks for the hard work!
Thanks so much for the tip!
my dude.. you are amazing. thank you for such great content.
Thanks for this benchmark vid 👍Exactly what I needed to see, upgrading soon from M1 Pro with Cubase 😎
Glad it was helpful!
Nobody comes close to the quality and credibility of your videos! Great job 👏👏👏👏
Thanks a lot, I'm lucky to be a Cubase user!
50k$ wow. he deserves it
@@antgeeze7129 vietnamese dong, not dollar
Love that Reaper is featured first.
Your time spent in breaking down performance among the various chips and DAW's is greatly appreciated!
As a music producer myself my mixes in Ableton are usually anywhere from 45 to 70 tracks total with anywhere from 2 to 7 VSTs per track. You can imagine the level of taxation that puts on my M1 Pro but with the workload being distributed to my Apollo Twin X it’s still able to handle all that so far without any issues so it might still not be worth the upgrade to the M4 pro for me unless it saves me a lot of wav/mp3 export time then I might consider it. It would be nice to see export time differences as well between M1 to M4 CPUs!
By the way it’s been really hard to find a video like this. Other cpu benchmarks I’ve seen related to music production simply stack multiple wav file tracks without using any VSTs. That is a useless test. I really appreciate both the level of effort and detail you’ve put into this video.
Thank you for your kind words! How's the core activity looks like when you mix in Ableton with that number of tracks and VSTs? Does Ableton only use the performance cores? (you will need to make sure there are no other heavy apps running for an accurate reading, though)
I've confirmed this many times on my end but it's always nice to see it confirmed by other people!
@@JamesZhanso I just ran a couple tests in Ableton both with my Apollo sound interface on and off. In both cases the efficiency cores 1,2 were showing about 50-75% workload the entire time (which is totally different from what you get) with performance cores 3,4,5,6 working the hardest. I found it interesting that all the cores were working just as hard whether my Apollo was on or off… but I guess it makes sense considering only the UAD plugins get disabled when the Apollo is off.
@@Killyang That's really interesting! Thanks for sharing. If you stop and start the playback every like 4-5 seconds, do you see the efficiency cores' activity graphs fall and rise correspondingly?
I've found that, sometimes, if there's several processes running in the background, when I make a DAW that can't use efficiency cores to play a huge project, macOS would “shift” all the other processes to the efficiency cores, making them look like they are processing audio in the DAW, when they are just handling all the other processes as the DAW is using up all the performance cores.
This is why I often reboot the computer and quit as many apps as I can in the Activity monitor before running these tests.
@@JamesZhan agreed. I made sure to do a full reboot and only open Ableton. When the music stops the green graph on all the cores stays about half full and when the music continues the graph elevates maybe another 20% gradually on all the core but the difference is more pronounced on the performance cores. The uptick in the efficiency cores is there but not as obvious. The graph only goes all the way down (black) once i quit Ableton. Hope that helps!
Man you’re kickass for this! Subscribed!
I love the fact that Cubase is well programed to use Apple Silicon, this was the reason why i ditched Studio One and even Logic Pro.
This is insanely useful for me. Thank you for putting this all together. Have been having issues with Studio One for some time now and this video fully convinced me I should move to Cubase.
Apart from this top notch content, i would request you to add on one more thing and that is Unified memory usage (for a better understanding of ram usage to swap memory usage ratio )
Great testing.
Great video man. Any chance of you testing out UAD Luna DAW as well?
Drew at UA said that Luna does use the efficiency cores as well as the performance cores. In my testing I have found that to be correct. Perhaps James could include Luna in the next test to confirm this.
@ Thank you.
The video we all waited for. Your information is so clear and well tested, showed with all the visuals and stuff, your a genius. thank you
It's amazing how much work you put in! All the graphics, the many tests. Respect, you are by far the greatest DAW tester for Apple computers on TH-cam.
Really appreciate your kind words! Happy to hear you liked my work.
you're more in-depth than the average youtuber. I'm subbing
10:37 I also think Apple does this on purpose (the way Logic Pro doesn’t directly utilize the E cores like you can get the other DAWs to). I would absolutely love for an employee / software engineer or even a spokesperson to respond to this statement. I would be super curious to find out how they would respond… like maybe there is some less obvious reason to do that…, which I think would be important for ALL of us to know.
Again… I love what you are doing here.
Hands down the best guide for music producers that want to upgrade their Mac setup!! Amazing 🙌🏼
Alright, here we go! C'mon Ableton, don't disappoint.
OH well😵💫🧐😩~~~ I’M considering CUBASE for my money’s worth~~between that or the other two at the highest performance
I thought move to Ableon was my everything-now I’ve learned it’s about the core of the DAW not just the cores of the MAC
Live uses for LIVE~~been ditched logic and that beach ball ~~~terrible~shame on apple
@@antgeeze7129 don't worry, you will do fine with ableton 😄
@@antgeeze7129 Even despite the results, there's just simply no DAW like Ableton for me.
This was a great breakdown and I loved how you mentioned the time stamp guide at the beginning of the video
When you get weird HARD pops and a CPU dropout that is called an "X-run" in the Unix/Linux/BSD world. It use to be a word everyone knew then it stopped being a problem when Apple switched to intel from PowerPC. But it seems Studio One is having an X-run Issue. X-runs happen when a core allocation jumps from one core to another to alleviate pressure for too much code coming down the pipeline for preemptive multitasking. I remember this happening all the time on the old Power Macs G5s since they had 2 PPC Dual Core CPUs and plugins didn't know where to send the execution correctly so it would jump from core to core causing WAY worse Xruns than the M4 does the drops would last for almost a full second sometimes. But those are called Xruns. Great Job as always James!! You truly are a treasure to the audio community!
Interesting. I am about to make the jump to S1 on Solicon (5 Pro) from Intel and this is a little discouraging to be honest. Might start exploring Cubase. However, I don't do massive mixing, - I am more focused on tracking, so multi-channel simultaneous recording (20+ channels) is the performance angle I will be looking at.
@@trebleboost7 If you are coming from any intel machine even a XEON mac pro you will see a huge improvement even with a base M4 Mac mini. I am a track engineer and that is the bulk of my work since I have a space big enough to track drums and console. The 2 things you can do to improve tracking performance is a using a Thunderbolt Interface and stay away from USB interfaces. This will lower latency and actually lower CPU usage since USB needs to talk to the CPU for every byte is writes to a drive, Thunderbolt doesn't have this problem. Thunderbolt actually runs from the PCI bus. The difference is staggering. Example in Reaper using the same interface an Antelope audio Orion Studio Synergy core which has both USB and Thunderbolt with a 48K sampling rate and a 128 buffer. On USB the latency is 8.4 milliseconds. Thunderbolt its 2.2 milliseconds. That doesn't sound like a lot until you have 50 track up with plugins and you need to record something else. On Thunderbolt I can set my buffer to 128 and never change it through the entire tracking and mixing process. On USB I have to raise it once I have about 10 to 15 tracks with plugins on them. Thats just the nature of the 2 ports. But as far as the M4 Pro you will see a huge improvement over any Intel Mac.
Thanks so much for sharing this information! Super interesting stuff.
This is the most comprehensive information on these machines when it comes to music production...awesome content and information...💯👍
I'm still on an intel i5 mac pro, lol... the upgrade to M4 (this or next month) is gonna literally change my life😂😂
Be ready to prepare to reinstall /revalidate a lot of your plug ins. However it seems worth it……so far
Doesn't get better than this. Thank you James for your incredible content.
I absolutely love your tests man - thanks for doing them ❤
Thank you so much for your kind words, Dom!☺️
Thanks
Very informative. My M1 Max Studio with Logic works like a charm but I might upgrade to the next one for video. Or wait another year. These chips are glorious.
Yeah video is a completely different ball game. You would have to look at CPU cores, GPU cores, AND the media engines.
Thanks! Can you do LUNA? So, M4 Max has no use for audio? Thanks!
James, thank you so much for all the effort you put into these videos, both on the research and testing side, and on the clarity and detail of the presentation.
best test out there... Thanks bro.
Bloody brilliant stuff, you are my guild star in this world, most appreciated your efforts are
Thank you for the content! Is it strange that I see more efficiency cores used than performance cores when mixing projects in DP11? Showing about 37% Ecores and 20% Pcores. Mac mini M4pro 24/512.
Thanks for this great and really informative video ❤
It is unbelievable how much effort you put into those tests, especially the mixing scenario. Thank you!
Was waiting for this video before I made my decision ! You a legend James !!!
Thanks for your kind words! What are you going to buy?
@JamesZhan from your video I was really blown away by how powerful the base m4 pro is! I was debating the max but pro is the way I feel! Ill be upgrading from m1 pro which is already a beast but does start to fall apart when I mix/master.
In terms of spec i am thinking 1 or 2tb for storage as I currently have a 1tb and have about 100gb left. The one thing on my mind is the nano texture display (yet to see this myself in person) and then my other thought is debating whether to upgrade to 48gb unified mem. What are your thoughts on more unified mem? Do you think it will affect heavy plugin usage?
Thanks!
Wow, thanks so much for the tip!!
Thx for this very transparent comparison. Good job
very nice content, respect
AMAZING VID JAMES! Learned so much! Waiting for the same vid on the M4 chip compared to this M4 Pro!
The best comparison!
as always thank you so much for taking the time to make this video!
Huge work ✨
Appreciate you doing this bro!
appreciate you taking what is a bit of a niche subject but so helpful
Thank you for the effort put into making these videos James! Great analysis.
KING ZHAN totally killin it⚡⚡
Wow, incredible work with this video. Thank you!
This benchmark and analysis has got to be the best one I've seen so far. Very thorough, very mindful and very self conscious of your own "limitation" (for lack of a better term). I'm definitely subscribing this high quality content!!
Thank you for the kind words, I appreciate it!
I said it in the past and i say it again, your videos are THE BEST about Macbook Pro's to choose for a music production use. Keep the hard work!
HARD WORK 🔥 thanx bro
James, you're so precise, and I really love how you get to the truth to make the best buying decision for audio production. I look forward to what you think about the mac mini computers and which is the best one to buy. Subscribed and thank you!
King James, phenomenal testing and insights as always!! Thank you!!
Once again, excellent work, James! I really hope your channel grows even more, because your content is truly high quality work and you deserve to reap the rewards! Definitely my nr1 favorite Music/Prod/Tech channel!
All the best
Am new to your channel -- very professional. BTW, what video filter are you using?
Thanks so much for the tip! Not exactly sure what you mean by filter? If you are asking about lens filter, I film with a Black Pro-Mist filter on.
@@JamesZhan Interesting bloom filter. Makes you look almost like you were AI generated.
@@ikoino Haha really? I thought the difference was very minimal. It's only 1/8 diffusion
This is such a useful video thank you so much
Thank you James! Excellent comparison and detailed analysis. We appreciate all that you do!
Thank you James for this fantastic test video! Everything was beautifully explained and illustrated. Looking forward to the next one about fan noise and thermals.