Meters

[bayswater]

I usually have a meter (nothing else) on the Master fader strip. I noticed in the recently added window that shows CPU load for plugins, that just about all the meters I have use a lot of CPU. Why would that be? Doesn't seem that they have to do that much with the signal. Certainly no more than a compressor, and logically, less. bx_meter is a real hog.

It does matter if I move them to somewhere other than the Master strip, it not something to do with that part of the signal path.

The only meter I have that doesn't do this is the PDP VU, and even that uses more than most other plugins.

Any suggestions for useful alternatives?

[mikehalloran]

Make certain you have the latest

Brainworx bx_meter - Changelog

Version 1.17.1 (Oct 9, 2023)
- General: fixed CPU spikes on Mac OS

https://www.plugin-alliance.com/en/prod ... meter.html

Many bx_ plugs were updated for this issue. In my case, it was bx_limiter True Peak

More info here:

viewtopic.php?t=73762

[bayswater]

Make certain you have the latest

Brainworx bx_meter - Changelog

Version 1.17.1 (Oct 9, 2023)
- General: fixed CPU spikes on Mac OS

https://www.plugin-alliance.com/en/prod ... meter.html

Many bx_ plugs were updated for this issue. In my case, it was bx_limiter True Peak

More info here:

viewtopic.php?t=73762

That looks promising. Thanks.

[JSmith1234567]

I had that problem too, and ended up figuring out a "meter bus".

I was getting latency and some other stuff, all for just a meter!

I put a bus on one of my master 2-bus sends that sends to a bus that just has a meter inserted on it.

Then I use this (free!) meter;

https://www.refusesoftware.com/simplemeter

I did an EXHAUSTIVE search for meters off and on for quite awhile, one of those weird Covid obsession rabbit-holes, trying to find the least-cpu intensive and least annoying, resizable meter I could find, trying really everything out there, and this was the winner by far.

I wanted a meter that also didn't keep flashing all of the au-courant LUFs/inter-channel-modunation/quauntum-solar-flares etc., just green until you hit red, so it doesn't keep distracting your eye all day unless it's getting too hot constantly.

This is the same guy that makes the terrific "Lowender", which is fabulous...let me replace my hardware DBX 120-X when i went in-the-box.

One thing about the "meter-bus".

For some reason you can put a plugin on a track in DP, but it won't actually work unless it has a physical output assignment.

I have not found a way to have a blank or artificial output.

You can send to another bus, but nothing will work unless somewhere down the line there is a physical output.

There is probably an easy way to do this, I just haven't figured it out, and this works.

So I assign that meter-bus to an output I'm not using for anything important on my audio interface, and turn the output-fader on the bus down to null.

Or zero? Is "null" zero? Whatever!

[JSmith1234567]

I do wish Motu would create a simple linear peak-meter plugin instead of calibration, and also add meters to custom-consoles.

And PLEASE give custom-consoles a facelift!

It's a tremendous tool, but not a great GUI.

And it could use some additions and new features.

I read about "environments"; and such in Logic.

No idea what they do, but it sounds good?

[bayswater]

Yes I know about the latency issues. I’m seeing high CPU usage — not peaks, just a high sustained CPU usage level — in an empty idling project.

On getting a plugin to see a signal on a channel with no assigned output, yes there has to be an output for a signal path to exist. But you can install some virtual driver like BlackHole, and use that. BlackHole itself doesn’t have to go anywhere.

[bayswater]

Make certain you have the latest

Brainworx bx_meter - Changelog

Version 1.17.1 (Oct 9, 2023)
- General: fixed CPU spikes on Mac OS

https://www.plugin-alliance.com/en/prod ... meter.html

Many bx_ plugs were updated for this issue. In my case, it was bx_limiter True Peak

More info here:

viewtopic.php?t=73762

That looks promising. Thanks.

That made no difference. On a large template with low buffer size, a few tracks playing and loads of plugins and VIs, the overall CPU usage is at about 50% with bx-meter and about 25% without it. PSP VU makes no noticeable difference.

[HCMarkus]

Waves WLM Plus after any mastering plugins. All I need.

[mikehalloran]

...

That made no difference. On a large template with low buffer size, a few tracks playing and loads of plugins and VIs, the overall CPU usage is at about 50% with bx-meter and about 25% without it. PSP VU makes no noticeable difference.

Hmmm,,,,

You don't say where you are getting those numbers:
Here?

58.png (67.48 KiB) Viewed 581 times

These percentages are all divided by the number of cores to get the real CPU hit. It's not unusual for them to be much higher than 100. Most plug-ins are single-core however.

On my system. bx_meter shows 21% with and 14% without which is normal for anything. The Real Time meter dances a little but never jumps over 25% either.

For kicks, I fired up iZotope Insight 2 and it showed 58% in the Activity Monitor (above) and the Real Time bar was showing the same.


If the window at the bottom is showing those numbers, that's not good. This is my CPU Load with that 58% from Insight 2.

CPU load.png (32.73 KiB) Viewed 581 times

With DP, I'm looking at the Real Time Performance window. When it's hitting red, I hear distortion. RTP does roughly correspond to single core performance but it's a much more accurate indicator of what's going on under the hood. You can enable/disable it in a project but I have it set to always display and showed how in that other thread.

To keep the attachments at three, I made a composite image with Insight 2 on top and bx_meter 1.71 on the bottom in the same project with no other changes made.

Real time bx_meter vs Insight 2.png (72.27 KiB) Viewed 581 times

[bayswater]

I'm just talking about the real time meter in DP. In an idle template with bx_meter on the Master Track, whether in play mode or not, the meter wavers around the middle of the band (aka 50%). Without bx_meter, it sits at about ¼ of the way in from the left (or 25%). Presumably it is using a lot less CPU when measured across cores in Activity Monitor, but as you say, problems arise with the real time meter in DP hits the red.

I can load of lots more audio, MIDI, VIs, plugins etc, and the real time meter doesn't change much. I'm wondering why bx_meter is showing such a big effect. What's it doing? All it has to do is watch the signal level and display it. Same as other meters that show no noticeable effect on the RT meter.

[James Steele]

You should be able to go into a mode in DP’s Audio Performance window that will show you the CPU hit of each plug-in.

[mikehalloran]

I'm just talking about the real time meter in DP. In an idle template with bx_meter on the Master Track, whether in play mode or not, the meter wavers around the middle of the band (aka 50%). Without bx_meter, it sits at about ¼ of the way in from the left (or 25%). Presumably it is using a lot less CPU when measured across cores in Activity Monitor, but as you say, problems arise with the real time meter in DP hits the red.

I can load of lots more audio, MIDI, VIs, plugins etc, and the real time meter doesn't change much. I'm wondering why bx_meter is showing such a big effect. What's it doing? All it has to do is watch the signal level and display it. Same as other meters that show no noticeable effect on the RT meter.

That is puzzling. If you were using v. 1.70, I would expect that but v. 1.71 fixed it for me.

[bayswater]

You should be able to go into a mode in DP’s Audio Performance window that will show you the CPU hit of each plug-in.

I could but its not that useful. If I look at the plugin performance window there’s a long list of items. Most show very little CPU use and bx_meter shows much more than all the others.

I’l check the version again. It was the one the installer provided so I have to assume it’s the most recent.

I really raised this because it was a surprise., and I thought I misunderstood what a meter does. I don’t need to use bx_meter. There are lots of alternatives.

[James Steele]

Most show very little CPU use and bx_meter shows much more than all the others.

Well there you go! Then I'd say you're right. bx_meter is using much more CPU than the others. If I have a choice between two utility type plugs that perform the same function and one his hitting the CPU harder than the other, I'll go with the one with the lighter CPU hit. That's one of the reasons I think I still have Ozone 9 Elements on my Mac. It had a lighter CPU hit than the full version of Ozone 9. I have Ozone 11 now, but again, sometimes use the one with the lighter hit.

[James Steele]

Oh... and I just put bx_meter on my main output and even the Apple Silicon native version (I'm using 1.17.1) is a CPU pig on my Mac Studio and it appears to not be very stead, but spikes. Attached is screenshot of performance meter. The AVE% is acceptable, but the problem is that MAX%. 31.44%!!!! Seems like it's unstable and spiking and even at times uses more CPU than Ozone 11. Something must be poorly coded in that plug I would guess. I'm using the AU. I could try the VST3 and see if that's better but I doubt it.

Maybe my version is old. Actually i guess not. Just checked and 1.17.1 is latest version.

bx_meter.jpg (42.31 KiB) Viewed 479 times