MuLab 7.1.30

Official support for: mutools.com
Post Reply New Topic
RELATED
PRODUCTS

Post

A test version of MuLab for OSX 64 bit is here:

http://www.mutools.com/galaxa/

This is a serious rewrite of the Mac specific code and has these advantages:

* 64 bit version = present and future ready. (previous MuLab for Mac was 32 bit only and in a dead-end street)
* A bunch of problems with VST plugin editors should be fixed now. (Carbon based editors are completely dropped)
* MuLab gets closer to iPad etc. No concrete plans yet, but the option is available.

Please post feedback as much as possible about this brand new Mac version. There might still be issues that need to be fixed but Mac OSX is a very volatile platform and so i need feedback from different systems. Also about using VST plugins. Please let me know. Thx!

Post

windows version ? Please

Post

MuLab for Windows update coming soonish, lets first focus on the new Mac version.

Post

:party:

Feedback forthcoming!

:party:
I started on Logic 5 with a PowerBook G4 550Mhz. I now have a MacBook Air M1 and it's ~165x faster! So, why is my music not proportionally better? :(

Post

Installed and have found the following:

1. All my plugs verified except Tassman4 64-bit VST. Times out on the verify, so I have no error log to share. Rather than recompiling it in 64-bit, they apparently wrote their own crappy 32->64 bit bridge! :dog:

WaveShell still fails verification. That's to be expected due to their funky special shell... :borg:

Only lost 4 plugs that are 32-bit only. Still trying to decide if it's the end of the world. One is Luxonix' LFX-1310 multi effect unit, which I will definitely miss.

Gonna try out JBridge on them to see how they do.

2. CPU usage is ~10% higher on the 64-bit version. Even just loading a New project and hitting play, the new version uses about 5% more CPU than the 32-bit version. I'm not sure what the expected result is supposed to be.

3. Load balancing between the two CPU cores is still goofy. Can only use 50% of my CPU before the sound starts glitching.
I started on Logic 5 with a PowerBook G4 550Mhz. I now have a MacBook Air M1 and it's ~165x faster! So, why is my music not proportionally better? :(

Post

Quick note: JBridge seems to be working great--one window, no GUI issues. I can play Slayer2 through the Luxonix wah-wah and it sounds awesome! :band:
I started on Logic 5 with a PowerBook G4 550Mhz. I now have a MacBook Air M1 and it's ~165x faster! So, why is my music not proportionally better? :(

Post

syntonica wrote:Installed and have found the following:
1. All my plugs verified except Tassman4 64-bit VST. Times out on the verify, so I have no error log to share. Rather than recompiling it in 64-bit, they apparently wrote their own crappy 32->64 bit bridge! :dog:
WaveShell still fails verification. That's to be expected due to their funky special shell... :borg:
Only lost 4 plugs that are 32-bit only.
Ok sounds like all is ok on MuLab's side then, right?
syntonica wrote:2. CPU usage is ~10% higher on the 64-bit version. Even just loading a New project and hitting play, the new version uses about 5% more CPU than the 32-bit version. I'm not sure what the expected result is supposed to be.
One moment i also had an impression that the cpu usage is a bit higher. Nothing changed in the MU audio code though. I'll have a look at it.
3. Load balancing between the two CPU cores is still goofy. Can only use 50% of my CPU before the sound starts glitching.
What type of processor do you have? How many real physical cores? (not 'logical' or 'hyperthreaded' cores) What setting do you have for MULAB -> Audio Setup -> Num Audio Processor Threads ?

Post

Mulab seems fine, other than the CPU usage issue, although on shutdown, I'm setting messages like "104 errors logged." I've peeked inside the logs and nothing too gruesome to report. Majority are VST errors. :D

My CPU is an i5 with 2 cores. Processor threads is set to Auto (2).
I started on Logic 5 with a PowerBook G4 550Mhz. I now have a MacBook Air M1 and it's ~165x faster! So, why is my music not proportionally better? :(

Post

syntonica wrote:Mulab seems fine, other than the CPU usage issue, although on shutdown, I'm setting messages like "104 errors logged." I've peeked inside the logs and nothing too gruesome to report. Majority are VST errors. :D
Note that with this new version all log output from MuLab is also visible in the Mac Console. Please copy-paste these errors in an email to me. I want to have a look at that. Thx.
My CPU is an i5 with 2 cores. Processor threads is set to Auto (2).
If your i5 has 2 cores then MuLab's "Auto" setting should have a (1) after it, not (2). Can you please double-check what that setting is and report back. Note that if it's set to Auto (1) and you would look at the CPU meter in your system then it's very probable that the reported maximum is ~50% cause the audio engine is only using 1 core out of 2 = 50%. Try setting it to 2 just to check the difference. Note however that it's not recommended to use all cores for audio processing, keep at least one free for other tasks, otherwise chance raises audio gets interrupted by tasks with a higher priority. But it always depends on the system and you can test the "2" setting on your system ie. using both cores for audio processing.

Post

I made a detailed comparison of the cpu usage of M7.0.47 (32 bit), M7.1.7 (32 bit)(unpublished) and M7.1.7 (64 bit). Of course all 3 times with the exact same audio setup settings on the same computer. I opened the Dream Of The Piano demo and started playing it from 18.1 and looked at the average cpu meter value in MuLab (click once on the cpu meter = "Avg value"):

M7.0.47 32 bit = ~30%
M7.1.7 32 bit = ~30%
M7.1.7 64 bit = ~31%

So practically no difference, maybe a tiny little bit more for the 64 bit version.
Then i also used an internal cpu consumer module to make a similar comparison and the results were the same: almost no difference (cpu load was 67% in this second test))

Can you please also make a detailed analysis of the "10% more cpu" situation you're encountering. I'd like to repeat that case so to research it.

Post

mutools wrote:MuLab for Windows update coming soonish, lets first focus on the new Mac version.
Good to know ,Thanks Jo.
mutools wrote:* MuLab gets closer to iPad etc. No concrete plans yet, but the option is available.
I am only on windows and very rarely on Mac but this seems an attractive option 8)

Post

Hi Jo,

I've attached a screenshot of what I see re: CPU and Audio settings.
Screen Shot 2016-09-20 at 8.36.07 AM.png
I tried 1 and MuLab choked at 25%. 2 chokes at 50% so I tried 4 and my project ran at about 60% of CPU with no glitches and the MenuMeters show the CPU load is balancing evenly across all 4 threads.

(I thought I had messed with all this before and found the Auto setting worked best, but this time, I am using a project which I purposely overloaded to use both cores.)

This happens on both 32- and 64-bit versions.
You do not have the required permissions to view the files attached to this post.
I started on Logic 5 with a PowerBook G4 550Mhz. I now have a MacBook Air M1 and it's ~165x faster! So, why is my music not proportionally better? :(

Post

Okay, re: CPU usage between 32- and 64-bit versions, MuLab is reporting the same percentage, more or less. However, Activity Monitor is reporting more usage when running 64-bit. These are taken in the last bar of the Krafty Groove demo.

32-bit:
32.png
64-bit:
64.png
Let me know if there's any other diagnostics you wish to see.
You do not have the required permissions to view the files attached to this post.
I started on Logic 5 with a PowerBook G4 550Mhz. I now have a MacBook Air M1 and it's ~165x faster! So, why is my music not proportionally better? :(

Post

syntonica wrote:Hi Jo,
I've attached a screenshot of what I see re: CPU and Audio settings.
I tried 1 and MuLab choked at 25%. 2 chokes at 50% so I tried 4 and my project ran at about 60% of CPU with no glitches and the MenuMeters show the CPU load is balancing evenly across all 4 threads.
(I thought I had messed with all this before and found the Auto setting worked best, but this time, I am using a project which I purposely overloaded to use both cores.)
This happens on both 32- and 64-bit versions.
So your cpu seems to have 4 cores, that's why the Auto = 2, cause the Auto setting uses half of the cores. It really depends from system to system what is the best setting. If a setting of 4 works fine for you, then leave it at 4.

By the way, how is the new MuLab 7.1 64 bit for OSX going? Do you think it's ready to replace MuLab 7.0?

Post

syntonica wrote:Okay, re: CPU usage between 32- and 64-bit versions, MuLab is reporting the same percentage, more or less. However, Activity Monitor is reporting more usage when running 64-bit. These are taken in the last bar of the Krafty Groove demo.
32-bit: 32.png
64-bit:64.png
Let me know if there's any other diagnostics you wish to see.
Are you sure you have the exact same audio setup for MuLab 32 bit and MuLab 64 bit, especially same audio output and same Num Processor Threads? If yes, then ok i'm not sure why OSX shows such a higher number for the 64 bit version, i assume it's because of the 64 bit. Oh, thinking loud, it may be also because M7.0 32 bit used OSX' Carbon sub-system while MuLab 64 bit uses OSX' Cocoa sub-system which i can imagine to use more CPU and RAM cause Cocoa is a much more dynamic system. This should not affect your audio too much cause the audio runs at a higher priority than the UI level, but the figures you see there in the Mac's Activity Monitor are the total CPU amount of an app, thus in this case UI activity + audio activity. That's why it's a less relevant figure than MuLab's CPU meter, which purely measures the audio processing. And as you say that MuLab's CPU meter shows more or less the same values (same here) then i think all is fine. Right? Please double-check it from the practical side: Making music!

Post Reply

Return to “MuTools”