A technical support community for Apple Logic Pro users.

  • 1
  • 2
  • 3
  • 4
  • 5
  • 9
 
TTOZ
Topic Author
Posts: 102
Joined: Wed Mar 04, 2009 6:58 am

New Logic Multitrack Benchmark Test

Sat Aug 25, 2018 3:57 am

well this is it:

NewLogicBenchmarkTest.zip
(7 MiB) Downloaded 1447 times


I really think I have created the perfect new benchmark test.. and props to Evan, didn't realise how long it would take to get the balance right.

On my 2015 macbook pro 2.8hghz quad, i get 37 tracks with turbo disabled or 38 with turbo enabled.. Their really is something to this turbo boost switch software.. difference is.. CPU is at 74 degrees instead of 95!
Turbo boost enabled will initially get 45 or even 50 tracks for about 8 or so bars but then overheats and drops the cpu frequency.. the macbook pro just can't maintain it. And then, "audio overload" message.

I will inform the logic test topic as well as other forums. Space Designer just took up too much ram for so many tracks so i replaced with platinum verb, which WILL still load just fine for everyone even though it's "hidden" in Logic 10.4

Details of test:

44K
Tracks on load : 128 instrument tracks each with sculpture, EQ, multipressor, Chorus, Auto Filter and Platinum Verb, playing pleasant 4 note chord

Tracks with their midi enabled by default: 50 (i came to this figure for people with new hex core macbook, hex core mac pro users and so on, it's very easy to enable more just like with Evan, mute tool will load with project so just click the midi clips to enable or disable one by one!)

Logic total ram use from Fresh Launch and project load : 1.1 GB

Project Length: 32 bars looping at 130 BPM (just what my default is.. if you don't like this let me know and i'll change it)

Audio track selected on Load so you can just play away and all cores will be distributed correctly. You literally don't need to do anything other than mute and unmute clips on a track by track basis.

My idea is that it has to successfully play the 32 bars AND make the loop point without the overload message appearing, for it to be considered a valid result.

My results:

Crystal well 2.8ghz macbook Pro i7 Quad, 16GB ram, 1TB SSD, Nvidia 750M

Internal macbook audio. Buffer: 128, Process Buffer Range: Medium.

Turbo Disabled: 37 tracks

Turbo Enabled: 38 tracks

This is obviously a much more intense test and much more suitable for modern machines.

If a machine out there really CAN play the whole included 128 tracks, it's very simple, just highlight the last track and press command D to duplicate it and copy the midi clip over till you reach your maximum tracks.

Cpu load is spread perfectly across the test, real cpu is low 80's % per core.. Logic is hammering away...

I'll test my imac pro this week, but I am guessing it will be 100 tracks...

I am happy to make it 255 tracks default if that's what you all prefer, and anything else you want changed, just let me know,

so we can get it to be the new default test.. it makes so much more sense with computers being as powerful as they are today.

Cheers

Ps have posted this at gearslutz as well so will take any input from both forums on board and notify both accordingly

PPS it was suggested i use chromaverb and sculpture hi res mode, which of course was my first thought, but i wanted to maintain backwards compatibility with older logic versions. Now someone has told me that 10.4 songs can only be opened in 10.4.. is this true? Can anyone confirm?

I no longer have access to any previous X version.

If it's true, the test is still relevant, as someone kind that has kept an older version around could possibly transfer it and upload it for any 10.x user.

And regardless, as has been shown with first i9 result below, the test is very intense as is.
Last edited by TTOZ on Sat Aug 25, 2018 9:45 am, edited 1 time in total.
iMac Pro 8 core 3.2ghz, 64Gb Ram, 4x TBolt SSD, CalDigit TS3+ Dock, OS 10.14.6, Logic 10.4.7/PT2019.10, Motu Microlite x2, Tons of midi synths, Apollo 8x2.
 
User avatar
Ploki
Posts: 4175
Joined: Tue Mar 04, 2008 2:59 am
Location: Slovenia
Contact:

Re: New Logic Multitrack Benchmark Test

Sat Aug 25, 2018 6:12 am

i really want someone to do this on the new i9 macbook.

my macbook can pull off 29 tracks stable, 30 it doesnt chew through the cycle.

CPU is not maxed out... and one core is not even used at all.
Screen Shot 2018-08-25 at 15.09.38.png
Screen Shot 2018-08-25 at 15.09.38.png (168.83 KiB) Viewed 7978 times
2018 Mini i7 | 2018 13" Quad i5 | Mini: Mojave 10.14.6 / MBP: Catalina 10.15.1 | Logic 10.5
RME FireFace 400 / 800 | ROLI Seaboard RISE 25 | ROLI Blocks
 
TTOZ
Topic Author
Posts: 102
Joined: Wed Mar 04, 2009 6:58 am

Re: New Logic Multitrack Benchmark Test

Sat Aug 25, 2018 9:39 am

Ploki wrote:
i really want someone to do this on the new i9 macbook.

my macbook can pull off 29 tracks stable, 30 it doesnt chew through the cycle.

CPU is not maxed out... and one core is not even used at all.
Screen Shot 2018-08-25 at 15.09.38.png


Are you sure you are at 44.1 khz?

I will tell you, in istat, i have 8 threads, and Logic is using 650% here.. (max possible is 800%).
With you and the i9, the theoretical max is 1200%. I presumed logic would get close to 1000%

30 tracks is way too low.. No way that makes sense..

Ok, here are the questions to get to the bottom of it

1) Have you applied Apple's thermal patch for the new macbooks?
2) Running at 44k?
3) what is your buffer and process buffer set at?
4) In Logic preferences, are all 12 threads enabled? Does logic performance meter show 12 bars?

There really is a problem with overheating and the new macbooks, unfortunately the macbook chassis can not even come close to dealing with the heat of the i9.. that's why after all the reviews, it's pretty much unanimous advice to just get the base cpu with top ram and graphics.

The i9 will dip under base clock with full load, even with the thermal patch.

My last advice is to try turbo boost switcher which is FREE
https://www.rugarciap.com/turbo-boost-s ... -for-os-x/

This will keep your i9 at 2.9ghz

It is very likely it will stop ALL overheating.. Considering when my turbo boost is disabled i get 37 tracks, I presume you will get 50 with 6 cores at 2.9ghz.

Please give it a go..

as i showed above, because of heat and fluctuating clock speed, having turbo enabled for multi core tasks that use all cores in a macbook is a pointless waste of heat and shorter processor lifespan.

My cpu has not gone above 80 degrees ever since i disabled turbo no matter what I am doing.

I also now undervolt it 0.80 mv, and it never goes above 75.

Interestingly, if i don't undervolt it, and turbo is enabled, mhx gets throttled at about 94 degrees.

with undervolting, it allows the temp to go to 98 and maintain a higher turbo.. that is way too hot for comfort for me, so I have more or less permanently disabled turbo.

Volta doesn't work on all macbooks but the watts power limiting feature should work as well as turbo enable/disable (yes, volta also allows control over turbo BUT it requires turning off a portion of SIPS and is only a one week trial. Cheap app though).

By the way the one core not being used at all is normal for Logic X.. i have never seen the last core used unless I arm a track to be "live".. It reserves it for that and various tasks.

:)

You will notice if you click on a sculpture track and start playback again, suddenly the last core is used :)

I think this is all to do with Logic's new "multi thread for record arm and playback tracks" option.

Ok, so i changed Logic's setting to just multi thread for playback tracks.

My test results are identical.. however, now all the "live" tracks are stacked onto one core..

I pressed the R button on three of the sculpture tracks and started playback.. as soon as i hit a midi key you see all three tracks go into live mode, and logic's final core almost full. So.. there you go:)
iMac Pro 8 core 3.2ghz, 64Gb Ram, 4x TBolt SSD, CalDigit TS3+ Dock, OS 10.14.6, Logic 10.4.7/PT2019.10, Motu Microlite x2, Tons of midi synths, Apollo 8x2.
 
User avatar
Ploki
Posts: 4175
Joined: Tue Mar 04, 2008 2:59 am
Location: Slovenia
Contact:

Re: New Logic Multitrack Benchmark Test

Sat Aug 25, 2018 11:18 am

TTOZ wrote:
Ploki wrote:
i really want someone to do this on the new i9 macbook.

my macbook can pull off 29 tracks stable, 30 it doesnt chew through the cycle.

CPU is not maxed out... and one core is not even used at all.
Screen Shot 2018-08-25 at 15.09.38.png


Are you sure you are at 44.1 khz?

I will tell you, in istat, i have 8 threads, and Logic is using 650% here.. (max possible is 800%).
With you and the i9, the theoretical max is 1200%. I presumed logic would get close to 1000%

30 tracks is way too low.. No way that makes sense..

Ok, here are the questions to get to the bottom of it

1) Have you applied Apple's thermal patch for the new macbooks?
2) Running at 44k?
3) what is your buffer and process buffer set at?
4) In Logic preferences, are all 12 threads enabled? Does logic performance meter show 12 bars?


No i meant i want someone to test the i9 because I'm opting for one.
I tested on 2012 2,7GHz quad.
buffer is 1024, and all 8 threads are enabled. Also using 32bit mixer (i wanted to see if it affects CPU on my last project and forgot to change it back)
I have multithreads enabled only on playback threads
2018 Mini i7 | 2018 13" Quad i5 | Mini: Mojave 10.14.6 / MBP: Catalina 10.15.1 | Logic 10.5
RME FireFace 400 / 800 | ROLI Seaboard RISE 25 | ROLI Blocks
 
TTOZ
Topic Author
Posts: 102
Joined: Wed Mar 04, 2009 6:58 am

Re: New Logic Multitrack Benchmark Test

Sat Aug 25, 2018 12:40 pm

Ploki wrote:
TTOZ wrote:

Are you sure you are at 44.1 khz?

I will tell you, in istat, i have 8 threads, and Logic is using 650% here.. (max possible is 800%).
With you and the i9, the theoretical max is 1200%. I presumed logic would get close to 1000%

30 tracks is way too low.. No way that makes sense..

Ok, here are the questions to get to the bottom of it

1) Have you applied Apple's thermal patch for the new macbooks?
2) Running at 44k?
3) what is your buffer and process buffer set at?
4) In Logic preferences, are all 12 threads enabled? Does logic performance meter show 12 bars?


No i meant i want someone to test the i9 because I'm opting for one.
I tested on 2012 2,7GHz quad.
buffer is 1024, and all 8 threads are enabled. Also using 32bit mixer (i wanted to see if it affects CPU on my last project and forgot to change it back)
I have multithreads enabled only on playback threads


Ok, then I am pretty happy with my result for 2014 (2015 build) 2.8 quad macbook.

Buffer is 128 but it makes no difference as an audio track is selected anyway.. only the process buffer size matters in this case...

I am using 64 bit mixer.. so that sound about right.. my max "consistent" result is 38 so that's just over 20% more than yours.. architecture boost from sandy/ivy bridge, clock for clock, was the big one, which shows the difference here.. all the other changes have been less impressive since then, regarding clock for clock performance.. it's been more about die shrinks, temperatures and the big one, core count.

Of course, i see you have an RME, so that could come into play too...

I bet you if you set buffer at 128 your result doesn't change, as long as any of the sculpture tracks aren't selected.

PS, i generally leave my process buffer at small, as it's good enough, and access music say to do that for virus TI.. so...
iMac Pro 8 core 3.2ghz, 64Gb Ram, 4x TBolt SSD, CalDigit TS3+ Dock, OS 10.14.6, Logic 10.4.7/PT2019.10, Motu Microlite x2, Tons of midi synths, Apollo 8x2.
 
User avatar
Ploki
Posts: 4175
Joined: Tue Mar 04, 2008 2:59 am
Location: Slovenia
Contact:

Re: New Logic Multitrack Benchmark Test

Sat Aug 25, 2018 12:48 pm

changed processing buffer to Small and buffer size to 128, mixer to 64bit.
and got to 33 tracks stable. wierd
2018 Mini i7 | 2018 13" Quad i5 | Mini: Mojave 10.14.6 / MBP: Catalina 10.15.1 | Logic 10.5
RME FireFace 400 / 800 | ROLI Seaboard RISE 25 | ROLI Blocks
 
TTOZ
Topic Author
Posts: 102
Joined: Wed Mar 04, 2009 6:58 am

Re: New Logic Multitrack Benchmark Test

Sat Aug 25, 2018 1:11 pm

Ploki wrote:
changed processing buffer to Small and buffer size to 128, mixer to 64bit.
and got to 33 tracks stable. wierd

well it actually makes more sense.. cause i was always led to believe haswell was a 10% clock for clock increase.. and i have 100 mhz extra.. so that actually makes much more sense as I am 15% higher.

Hope to hear from others trying this!

Also can anyone with an older LPX version confirm whether it loads or not?

Cheers
iMac Pro 8 core 3.2ghz, 64Gb Ram, 4x TBolt SSD, CalDigit TS3+ Dock, OS 10.14.6, Logic 10.4.7/PT2019.10, Motu Microlite x2, Tons of midi synths, Apollo 8x2.
 
TTOZ
Topic Author
Posts: 102
Joined: Wed Mar 04, 2009 6:58 am

Re: New Logic Multitrack Benchmark Test

Sat Aug 25, 2018 8:47 pm

ok, good news, it load in 10.2.4 :)

So anyone on older logic X, feel free to try.
iMac Pro 8 core 3.2ghz, 64Gb Ram, 4x TBolt SSD, CalDigit TS3+ Dock, OS 10.14.6, Logic 10.4.7/PT2019.10, Motu Microlite x2, Tons of midi synths, Apollo 8x2.
 
JakobP
Posts: 1411
Joined: Sun Aug 17, 2014 1:49 pm
Location: Stockholm

Re: New Logic Multitrack Benchmark Test

Sun Aug 26, 2018 3:46 am

Opens fine in 10.0.7 on 10.12.6 :) My mbp 2.4 i7 quad ivy bridge could handle 37 tracks with LPX at around 530% cpu, built-in audio at 128 samples i/o buffer. Wouldn't take the cycle all the way with 38 tracks and around 550%.
Logic Pro 10.5.1 • MBP 15" Early 2013 • OS X 10.14.6 • 2.4 GHz Intel Core i7 • 8GB RAM • 750GB SSD • Mac Mini Late 2012 • OS X 10.14.6 • 2.5 GHz Intel Core i5 • 4GB RAM • 500GB HD • Focusrite Saffire 6 USB • Presonus Inspire 1394 • Roland UM-One mkII • Behringer FCB1010 • Gibson ES335 • cheap midi keyboard
 
JakobP
Posts: 1411
Joined: Sun Aug 17, 2014 1:49 pm
Location: Stockholm

Re: New Logic Multitrack Benchmark Test

Sun Aug 26, 2018 4:45 pm

Also tried my 2.5 i5 mac mini, it could handle 14 tracks after a couple of overloads.Same OS and LPX versions.
Logic Pro 10.5.1 • MBP 15" Early 2013 • OS X 10.14.6 • 2.4 GHz Intel Core i7 • 8GB RAM • 750GB SSD • Mac Mini Late 2012 • OS X 10.14.6 • 2.5 GHz Intel Core i5 • 4GB RAM • 500GB HD • Focusrite Saffire 6 USB • Presonus Inspire 1394 • Roland UM-One mkII • Behringer FCB1010 • Gibson ES335 • cheap midi keyboard
 
JakobP
Posts: 1411
Joined: Sun Aug 17, 2014 1:49 pm
Location: Stockholm

Re: New Logic Multitrack Benchmark Test

Mon Aug 27, 2018 1:38 am

Ploki wrote:
...my macbook can pull off 29 tracks stable, 30 it doesnt chew through the cycle.

CPU is not maxed out... and one core is not even used at all.
Screen Shot 2018-08-25 at 15.09.38.png

Looking at your screenshot, it seems your mbp is using the intel graphics instead of the descrete gpu, I wonder if that could reduce your track count ? Here, the mbp switches to descrete gpu when Logic starts. Both your machine and the OP:s have considerably higher geekbench scores than mine, and I expected mine to perform worse, I now suspect the overhead of more advanced versions of OS and LPX could be the reason for my relatively high track count? Experts ? :)
Logic Pro 10.5.1 • MBP 15" Early 2013 • OS X 10.14.6 • 2.4 GHz Intel Core i7 • 8GB RAM • 750GB SSD • Mac Mini Late 2012 • OS X 10.14.6 • 2.5 GHz Intel Core i5 • 4GB RAM • 500GB HD • Focusrite Saffire 6 USB • Presonus Inspire 1394 • Roland UM-One mkII • Behringer FCB1010 • Gibson ES335 • cheap midi keyboard
 
User avatar
Ploki
Posts: 4175
Joined: Tue Mar 04, 2008 2:59 am
Location: Slovenia
Contact:

Re: New Logic Multitrack Benchmark Test

Mon Aug 27, 2018 2:47 am

JakobP wrote:
Ploki wrote:
...my macbook can pull off 29 tracks stable, 30 it doesnt chew through the cycle.

CPU is not maxed out... and one core is not even used at all.
Screen Shot 2018-08-25 at 15.09.38.png

Looking at your screenshot, it seems your mbp is using the intel graphics instead of the descrete gpu, I wonder if that could reduce your track count ? Here, the mbp switches to descrete gpu when Logic starts. Both your machine and the OP:s have considerably higher geekbench scores than mine, and I expected mine to perform worse, I now suspect the overhead of more advanced versions of OS and LPX could be the reason for my relatively high track count? Experts ? :)


i suspect my thermal paste is all but dried out on this machine so it cannot maintain turbo very long.
With internal gpu i get 33, with discrete a track less, probably because it puts more stress on the cooling system.

Might also be the OS
2018 Mini i7 | 2018 13" Quad i5 | Mini: Mojave 10.14.6 / MBP: Catalina 10.15.1 | Logic 10.5
RME FireFace 400 / 800 | ROLI Seaboard RISE 25 | ROLI Blocks
 
User avatar
Ploki
Posts: 4175
Joined: Tue Mar 04, 2008 2:59 am
Location: Slovenia
Contact:

Re: New Logic Multitrack Benchmark Test

Fri Sep 21, 2018 4:20 pm

oh wow, the i9 is rocking 67.

this is pretty impressive gotta say... :) roughly twice as much as my old i7, which coincides with geekbench benchmarks.

edit: make that 70.
2018 Mini i7 | 2018 13" Quad i5 | Mini: Mojave 10.14.6 / MBP: Catalina 10.15.1 | Logic 10.5
RME FireFace 400 / 800 | ROLI Seaboard RISE 25 | ROLI Blocks
 
User avatar
Ploki
Posts: 4175
Joined: Tue Mar 04, 2008 2:59 am
Location: Slovenia
Contact:

Re: New Logic Multitrack Benchmark Test

Thu Oct 04, 2018 9:31 am

Anyone has 2.2 or 2.6 2018 to test?
2018 Mini i7 | 2018 13" Quad i5 | Mini: Mojave 10.14.6 / MBP: Catalina 10.15.1 | Logic 10.5
RME FireFace 400 / 800 | ROLI Seaboard RISE 25 | ROLI Blocks
 
User avatar
skijumptoes
Posts: 1882
Joined: Wed Jul 15, 2009 3:09 pm

Re: New Logic Multitrack Benchmark Test

Thu Oct 04, 2018 11:33 am

Tried one of my old 2008 Mac Pro (2x Xeon 2.8ghz Quad) 18GB DDR, and it got to 41 tracks. I'm gonna cry the day i have to get rid of them, still plenty enough power there for me and what i do.

OSX 10.13.6 LPX 10.4.2

Crazy thing is, i can run at 40 and use the internet etc. while it's playing, but once you add those extra tracks in it kills it lol. Seems like a good test also as i got no pops/crackles - it just bottles out as soon as i hit 42 yet 41 will run and run. Well done!

Last core isn't being fully utilised though.
 
User avatar
Ploki
Posts: 4175
Joined: Tue Mar 04, 2008 2:59 am
Location: Slovenia
Contact:

Re: New Logic Multitrack Benchmark Test

Thu Oct 04, 2018 2:12 pm

10 years, and a super thin laptop packs nearly twice as much punch as a tower used to. wow.

i used to have one of those dual quads 2008... it was one fine machine. so much fun taking it apart.
2018 Mini i7 | 2018 13" Quad i5 | Mini: Mojave 10.14.6 / MBP: Catalina 10.15.1 | Logic 10.5
RME FireFace 400 / 800 | ROLI Seaboard RISE 25 | ROLI Blocks
 
TTOZ
Topic Author
Posts: 102
Joined: Wed Mar 04, 2009 6:58 am

Re: New Logic Multitrack Benchmark Test

Sun Oct 07, 2018 10:19 pm

it's just the architecture of the chips are so much better now, clock for clock.

I remember when I did full productions with VI's on the 2 core 2.93GHZ (windows) in 2008 ish, then went to an imac 3.33ghz dual core sometime in 2010 i think it was, no hyperthread, and i was wowed with power LOL. The thing is that as chips have gotten more powerful with multiple cores, VI's have come out that use an entire core for one instance, so we've never really gained that much power in the audio world. Every time processors are improved, it seems analog modelled plugins are as well, and the cycle goes on. The other thing I have noticed is that OSes are memory and cpu hogs now too.. For audio, if a modern snow leopard existed, it would do everything i could possibly need for my DAWs, and would run so quickly on a current CPU.. High Sierra does not feel all that fast even on my imac pro, and resizing retina apps causes skips and jumps.. My current imac pro feels slower navigating logic and the OS (probably due to 5k screen) than my 2011 27" imac with i7-2600 processor, SL 10.6.8 and Logic 9.1.8. That felt like lightning all the time.
iMac Pro 8 core 3.2ghz, 64Gb Ram, 4x TBolt SSD, CalDigit TS3+ Dock, OS 10.14.6, Logic 10.4.7/PT2019.10, Motu Microlite x2, Tons of midi synths, Apollo 8x2.
 
User avatar
skijumptoes
Posts: 1882
Joined: Wed Jul 15, 2009 3:09 pm

Re: New Logic Multitrack Benchmark Test

Mon Oct 08, 2018 4:31 am

Yeah definitely agree on the snappiness of apps and OS's nowadays, it's weird really, the performance from a user point of view seems to dip quite early in when you put a machine under load... But then the machine just keeps going and going after that.

Must admit, i thought DSP or utilising the GPU would be a big thing now for audio users, but it still (As far as i'm aware for us Logic/Mac users?) comes purely down to CPU.

Logic X has always felt a little bit like Logic 9 with extra bolted on though, certainly not the rewrite that Final Cut appeared to be. But i would love to see it updated to maybe run it's interface more in a hardware sense via GPU instructions which is separate from the CPU task at hand. Not sure how do-able that is though, but the screenset changes recently are nice to see.

Also, in terms of 'snapiness' and UI performance, the amount of stuff going on in the background nowadays is ridiculous, i had setup a database for my clients to login and see where i am with stuff, which i host on this machine. As a precaution i placed a firewall on here too and boy was that an eye opener when i started opening different apps and plugins up, makes you paranoid when a plugin is contacting an amazon server (Why?!). I almost regretted installing it, cause i'm happy not knowing all that's going on or else i get OCD about it! :)

While the firewall is waiting for you to yes/no the connection, the plugin or app will just sit like a dummy 'I wanna talk to someone, please say yes!', reminds me reading about a problem facebook had with their website plugin, it went offline a year or two back, causing thousands of sites to fail as they couldn't retrieve the little facebook 'follow' button and i don't think there was any timeout in there! So.. it's maybe possible all this has an affect on the experience somewhere?

So much relies on so much nowadays, and i think that's true on an OS level too, and much of the 'planned' obsolescence that Apple get tarred with perhaps? i.e. we know they can progress stuff fast, but is it necessary when all you're doing is changing the window, if the view out of it is the same?

I guess i'll be a better judge of that when moving to a newer machine.
 
User avatar
KC
Posts: 347
Joined: Thu Jul 12, 2007 2:27 pm
Location: West Michigan
Contact:

Re: New Logic Multitrack Benchmark Test

Mon Oct 08, 2018 5:06 am

I ran this on my 2012 MP - got up to 92 tracks before an error. 128 i/o buffer Medium Process Buffer
Kayle Clements
composer

clementunes.com | music
2012 MacPro 12 core 3.06ghz • 10.13.6 (High Sierra) • 64GB • Metric Halo 2882 • Logic X (10.4.8)
 
User avatar
Ploki
Posts: 4175
Joined: Tue Mar 04, 2008 2:59 am
Location: Slovenia
Contact:

Re: New Logic Multitrack Benchmark Test

Mon Oct 08, 2018 9:21 am

wow, nice!
2018 Mini i7 | 2018 13" Quad i5 | Mini: Mojave 10.14.6 / MBP: Catalina 10.15.1 | Logic 10.5
RME FireFace 400 / 800 | ROLI Seaboard RISE 25 | ROLI Blocks
  • 1
  • 2
  • 3
  • 4
  • 5
  • 9