Contribute
Register

Logic Pro and the i9-9900K- A Test

Joined
Aug 29, 2012
Messages
56
Motherboard
Gigabyte Z390 Aorus Pro
CPU
i9-9900K
Graphics
RX 580
Mac
iMac, MacBook Pro
Classic Mac
eMac, Power Mac
Hello all!

I'm creating this post in the hopes that it will provide a bit of reference for those who are unsure about the power of processors today; because I was unsure before I bought my components. My problem, while searching for components, was that I could not find any real-world test results as to how effective todays Intel processors would be with Logic (specifically) and what I would actually need to power my projects. So about 3 weeks ago I got fed up with running around in circles, bit the bullet and built my new Hack:

Gigabyte Aorus Pro (non-wifi/bt)
i9-9900K
Sapphire RX580 Pulse 8GB
Vengeance 32G 3000MHz
Samsung EVO 970 250GB
Noctua DH-15
EVGA P2 750W
Fractal Design R6

After getting it all working nicely, and a 7 hour download later, I decided to see what this build could actually do and if I would be disappointed.

I have not overclocked anything. Everything is running stock speeds. My soundcard is an Allen & Heath Xone DB4 (no, I'm not a dj). I have my powerbar plugged into a watt meter that tells me the total power consumption of my build (including monitor). Total power consumption for this build at idle is average 68 Watts and 25C.

Starting a new project in Logic, I created 8 "Software Instrument" tracks each with 8 bars of empty midi. The channels were populated with Alchemy, Ultrabeat and Sculpture. I then chose 8 different (kinda) random patches and threw a schwack of random midi into the regions. I then chose between 3-5 audio effects for each channel trying not to use the same effects twice (with some exceptions within EQ and Reverb). I then added 3-5 "Automation Parameters" per track. I finally copied those 8 tracks until my processor would overload, deleting 1-2 tracks at a time until the song would stay stable during loop-play.

Well, in order to stay stable with these parameters, my max track count is:

64 tracks @ 96kHz and 1024 samples; averaged 64C and 244W consumption
152 tracks @ 44.1kHz and 1024 samples; averaged 67C and 270W consumption


Logic Test.jpg


I understand that this test is quite subjective. Everyone has a different style and components, and every musical experiment is just as different requiring different needs. I'm not a professional but my needs are heavy on soft synths, effects and automation of which a meaty processor is quite capable of handling. That being said this test was for me, not so much a disappointment (are you kidding?), but rather a realization that I could have saved myself $200 and went with the i7-9700K or even the 8700K. None of my projects have ever been over 24 tracks of solid, non-stop audio. More tracks doesn't necessarily make a better song. It's like getting 4000MHz ram. Paying top top dollar for performance with the rest being a massive plastic belt-buckle, spray-painted gold (comes with cape and hero pose). People may say I'm "future-proofing". Well heck, my 6 year old 3770K (non-oc'ed) was still able to pull it's weight (it didn't really like my 20+ track experiments though, but there are other reasons why I bought a new comp now). "Future-proofing" has never worked for me. I say this to myself as much as I say it to you: know your style from what you have actually made in the past and buy accordingly.

Cheers!

BTW: It sounded like crap. Although the "Progressive House kit" track on solo made me giggle it was so silly. That poor drummer.
 
Last edited:
Joined
Aug 22, 2017
Messages
333
Motherboard
ASUS Rog Maximus IX Hero
CPU
i7-7700K
Graphics
RX 580
Mac
iMac
Mobile Phone
iOS
Hello all!

I'm creating this post in the hopes that it will provide a bit of reference for those who are unsure about the power of processors today; because I was unsure before I bought my components. My problem, while searching for components, was that I could not find any real-world test results as to how effective todays Intel processors would be with Logic (specifically) and what I would actually need to power my projects. So about 3 weeks ago I got fed up with running around in circles, bit the bullet and built my new Hack:

Gigabyte Aorus Pro (non-wifi/bt)
i9-9900K
Sapphire RX580 Pulse 8GB
Vengeance 32G 3000MHz
Samsung EVO 970 250GB
Noctua DH-15
EVGA P2 750W
Fractal Design R6

After getting it all working nicely, and a 7 hour download later, I decided to see what this build could actually do and if I would be disappointed.

I have not overclocked anything. Everything is running stock speeds. My soundcard is an Allen & Heath Xone DB4 (no, I'm not a dj). I have my powerbar plugged into a watt meter that tells me the total power consumption of my build (including monitor). Total power consumption for this build at idle is average 68 Watts and 25C.

Starting a new project in Logic, I created 8 "Software Instrument" tracks each with 8 bars of empty midi. The channels were populated with Alchemy, Ultrabeat and Sculpture. I then chose 8 different (kinda) random patches and threw a schwack of random midi into the regions. I then chose between 3-5 audio effects for each channel trying not to use the same effects twice (with some exceptions within EQ and Reverb). I then added 3-5 "Automation Parameters" per track. I finally copied those 8 tracks until my processor would overload, deleting 1-2 tracks at a time until the song would stay stable during loop-play.

Well, in order to stay stable with these parameters, my max track count is:

64 tracks @ 96kHz and 1024 samples; averaged 64C and 244W consumption
152 tracks @ 44.1kHz and 1024 samples; averaged 67C and 270W consumption


View attachment 404085

I understand that this test is quite subjective. Everyone has a different style and components, and every musical experiment is just as different requiring different needs. I'm not a professional but my needs are heavy on soft synths, effects and automation of which a meaty processor is quite capable of handling. That being said this test was for me, not so much a disappointment (are you kidding?), but rather a realization that I could have saved myself $200 and went with the i7-9700K or even the 8700K. None of my projects have ever been over 24 tracks of solid, non-stop audio. More tracks doesn't necessarily make a better song. It's like getting 4000MHz ram. Paying top top dollar for performance with the rest being a massive plastic belt-buckle, spray-painted gold (comes with cape and hero pose). People may say I'm "future-proofing". Well heck, my 6 year old 3770K (non-oc'ed) was still able to pull it's weight (it didn't really like my 20+ track experiments though, but there are other reasons why I bought a new comp now). "Future-proofing" has never worked for me. I say this to myself as much as I say it to you: know your style from what you have actually made in the past and buy accordingly.

Cheers!

BTW: It sounded like crap. Although the "Progressive House kit" track on solo made me giggle it was so silly. That poor drummer.
Hey @Sparkmac thanks for your post. I started a thread aimed at optimizing our hacks for audio production as I'm finding out that despite the solid Geekbench 4 test results, Logic doesn't seem to be taking full advantage of its beastly power. At least not to the level a real iMac would.

Also, are you using standard Logic instruments and effects in this project and would you be willing to upload it to test here?
 
Joined
Aug 29, 2012
Messages
56
Motherboard
Gigabyte Z390 Aorus Pro
CPU
i9-9900K
Graphics
RX 580
Mac
iMac, MacBook Pro
Classic Mac
eMac, Power Mac
Hi @luckyal

Ya know... I never thought of making this into a benchmark that other people could test on their machines. I just considered it to be a one-off because the focus of this test was to see if I even needed a 9900k, which of course I realize now that I didn't and that when the time comes for the next hack I will do a rethunk. Aaaaaaaannnnnnddddddddd because I treated this as a one-off, I threw out the original file. :clap: But if you like I can create a new one that is similar, do all the benckmarking again and upload my results and the file. Though after reading your thread, I'm not sure how much more my test will unveil on your hack since my test is not very complex at all, but I guess we can find out. Gimmie a bit 'o time and I'll throw something together.
 
Joined
Aug 29, 2012
Messages
56
Motherboard
Gigabyte Z390 Aorus Pro
CPU
i9-9900K
Graphics
RX 580
Mac
iMac, MacBook Pro
Classic Mac
eMac, Power Mac
Here ya go @luckyal.

During "recreation", I realized that I missed something very important: I could test all the way down to 2 cores. So I did. And I'll probably never do it again. A few things to note: this test was run on Mojave 10.14.4 using system def 18,3. I didn't use my soundcard (xone: DB4). There are so many different soundcards out there that I thought I would remove it from the equation and just use the sound off the mobo. Also, the "Average Watts" is measured without the monitor; computer watts only. I did this to be a bit more accurate (I didn't know it but my 6 year old 27" Asus monitor is eating 20 watts.).

Have fun testing this Logic file. Only Apple plugins are used, nothing external. If you've downloaded all the Logic expansion packs you should be fine. No audio files were used so the hd is not worked at all. This is a test just for the processor.

Cheers!

BTW: Make sure you turn the volume down to start. It's... well.... hmmmm......

408318
 

Attachments

Last edited:
Joined
Aug 22, 2017
Messages
333
Motherboard
ASUS Rog Maximus IX Hero
CPU
i7-7700K
Graphics
RX 580
Mac
iMac
Mobile Phone
iOS
Here ya go @luckyal.

During "recreation", I realized that I missed something very important: I could test all the way down to 2 cores. So I did. And I'll probably never do it again. A few things to note: this test was run on Mojave 10.14.4 using system def 18,3. I didn't use my soundcard (xone: DB4). There are so many different soundcards out there that I thought I would remove it from the equation and just use the sound off the mobo. Also, the "Average Watts" is measured without the monitor; computer watts only. I did this to be a bit more accurate (I didn't know it but my 6 year old 27" Asus monitor is eating 20 watts.).

Have fun testing this Logic file. Only Apple plugins are used, nothing external. If you've downloaded all the Logic expansion packs you should be fine. No audio files were used so the hd is not worked at all. This is a test just for the processor.

Cheers!

BTW: Make sure you turn the volume down to start. It's... well.... hmmmm......

View attachment 408318
Thanks I'll check this out when I get home
 
Joined
Aug 22, 2017
Messages
333
Motherboard
ASUS Rog Maximus IX Hero
CPU
i7-7700K
Graphics
RX 580
Mac
iMac
Mobile Phone
iOS
I was able to pull 84 tracks at 96Khz but only 121 tracks at 44.1. I'm not sure I can recreate this test faithfully unless I know the exact count of each of the types of tracks you had. For example if you have 24 Alchemy tracks and only 19 Ultrabeat tracks, that's going to make a huge difference if the two were reversed because Alchemy is far more intensive on the system than Ultrabeat.
 
Joined
Aug 29, 2012
Messages
56
Motherboard
Gigabyte Z390 Aorus Pro
CPU
i9-9900K
Graphics
RX 580
Mac
iMac, MacBook Pro
Classic Mac
eMac, Power Mac
The original is 6 tracks Alchemy, 1 Sculpture, and 1 Ultrabeat. All I did was copy all 8 tracks then subtracted one at a time until the loop stayed stable. So, for instance: "16 Cores @1024 Samples @ 96kHz" is 60 tracks for me: 8 tracks copied 8 times is 64 tracks, minus 4 is 60. So: 6x8=48 Alchemy, 1x8=8 Sculpture, 1x8=8 Ultrabeat, minus the last 4 tracks which would be 2 Alchemy, 1 Sculpture and 1 Ultrabeat (as per how the tracks are stacked), leaving me with 46 Alchemy, 7 Sculpture and 7 Ultrabeat. Click on each of the tracks and you can see exactly what they are. Are you using all 16 cores? Cause your 84 tracks at 96kHz is 22 tracks better then my best score! 121 tracks @ 44.1 seems slightly low but only in comparison to my count.

Edit: "I'm not sure I can recreate this test faithfully unless..." After thinking about it, this is a bit of an odd statment since it's just a simple copy and paste. In the "track header" region (not the midi region), click the first track. Then hold the "shift" key and click the eighth track header. All eight track headers with their midi regions should be selected. Then hold down the "option" key and click drag all eight track headers down until they clear the the first eight. Release. Repeat. That's as faithful as you can get. :)
 
Last edited:
Joined
Aug 29, 2012
Messages
56
Motherboard
Gigabyte Z390 Aorus Pro
CPU
i9-9900K
Graphics
RX 580
Mac
iMac, MacBook Pro
Classic Mac
eMac, Power Mac
I just quickly remade the "16 Cores @ 1024 Samples @ 96kHz" test and took a screenshot of it because I didn't do it before like I should have. Oh, and something else I didn't mention was that I did not overclock. I'm running stock speeds.


408348
 
Joined
Aug 25, 2012
Messages
721
Motherboard
Gigabyte Z390 Designare
CPU
i9 9900k
Graphics
RX480
Mac
MacBook Pro
Classic Mac
Performa, Power Mac, PowerBook
Mobile Phone
iOS
24 tracks on my 6700k... Hyperthreading doesn't seem to work at all. Core utilization 40%. 340% total CPU load. That's bad.
Cubase goes up to 70% core utilization and nearly 700% total CPU load. Something is wrong on my system.
 
Top