You'll need to disclose the Ram of each Mac here. Other wise these tests are pointless and yes Logic performes better with more Ram versus swap memory.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Who wants to run the Logic tracks benchmark on their M2/M3?
- Thread starter eicca
- Start date
- Sort by reaction score
was that M3 test done with a 1024 buffer? Definitely want to redo that test at a much lower buffer size. Apple Silicon is weird that way. I get horrible performance in Ableton when my buffer is any higher than 256 (M1 Pro). I usually work at 128 and don't think ive ever noticed a major hiccup. My M3 Pro just came in a couple of hours ago so I'll be stress testing it.
Hi, I went to a shop today and made a Logic Pro X Test2 on 3 M3 Machines. All 18MB Mem, AudioConfig as recomented 1024, All prozessors on. Here are my results. M3 Pro 12 Core 16" = 180 Tracks, M3 Pro10Core 14" = 147 Tracks, M3 14" = 123 Tracks. Can somebody make a test with M2pro MAX .. My M1 8 core /16GB = 114 Tracks I use Logic since Notator so I know how to check ..
...Best
...Best14" MacBook Pro with the 12 Core M3 Pro CPU, 18 GB RAM, 1TB HD. I get the system overload at 172 tracks.
I also ran this with the Sonoma 14.1.1 update that just updated last night, although not sure if that would make a difference.
So 20 more tracks than the 12 core M3 Pro with 18GB RAM. A but surprised it isn’t more.
Exactly, strange, wasn't it? I was surprised too. I still have to check what exactly is different. Plus it was a lot cheaper. So I did it right. I'll be fine with the 192 tracks for the next few years....
Those bar-graphs are definitely not drawn correctly to scale.
194 is about 60% of 328, yet its graph is well below 50% of 328. Just eyeballing it, it looks more like 1/3.
Also 118 is about 60% of 194, but again, its graph is well below 50%
Finally, 118 is around 36% of 328, i.e slightly more than 1/3, yet its graph is well under 1/4 of 328's graph.
Here's a simple text-only representation, using X's in a mono-spaced font
Code:
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX 33 X's
XXXXXXXXXXXXXXXXXXX 19 X's
XXXXXXXXXXXX 12 X'sIt's not about the graph, because if you cut out the common part, it's OK.
It's about the result. Is it possible that the MBP 16" M3 Max achieved 328 tracks?
Last edited:
I guess we're gonna have to find someone to reproduce that before we will know for sure.
Not quite sure what is meant by 'tracks' ? Are we talking audio tracks, VSTi synth tracks, sample player tracks ?
I'd think there is a difference between running multiple tracks of DIVA vs audio tracks with Amplitube and or multiple fx vs Kontakt 7. What sample and bit rate is used ? Without this information it's hard to judge what this all means.
I'd think there is a difference between running multiple tracks of DIVA vs audio tracks with Amplitube and or multiple fx vs Kontakt 7. What sample and bit rate is used ? Without this information it's hard to judge what this all means.
Cutting out the common part is deceptive. The whole point of a bar graph is visual comparison of size. If you can't do that with the visible bars, then what's the point?
The bar sizes are what matter, and visual games like this just undermine credibility to me.
“Tracks” means tracks within a specific project file that’s downloaded from the web page of this benchmark’s creators.
Logic Pro M1 (Pro & Max) Benchmarks - Music-Prod.com
Here is the Logic Pro X Benchmark project file to use. Just download it, run it and see how many tracks you can get to play.
music-prod.com
This particular project is set up to be hugely inefficient. A base M1 chip hits system overload when running about 90 tracks of that test IIRC.
But when tested with a project set up more like a real-world example, the base M1 handled over 100 synthesizers, something like 600 audio tracks and a total of over 1000 plugins running before it hit system overload.
I don’t remember which YouTube video that was from but it was two British dudes and their production studio back when the M1 Mac Mini first launched.
M1 Ultra supposedly did 311, and it sounds like the M3 Max is on par with the M2 Ultra performance-wise. So 328 for M3 Max sounds totally realistic. And incredible.
Thanks Eicca. I believe I prefer more real-world examples that mirror my usage than generalized tests that may not reflect what my actual usage would be. Music production is rather diverse in that some use mostly audio tracks where others VSTi synths and others sample based synths, all getting very different results from their chosen SOC.
Yes, I recall that YouTube video with the two British guys. Their videos were very interesting.
M2 Max base Studio 209 Tracks with 32 samples. 204 Tracks with 1024 samples. So silicon really runs better on low latency
