Vengeance Producer Suite - AVENGER - 1.8.5 the main thread

[Jac459]

Too bad I didn't know that before purchasing it. Burned myself a couple of times with software that doesn't work well on Apple.. I guess I need to do some research before buying next time.. Anyway thanks for sharing

I think you need to put that in context. VPS avenger works very very well with Apple silicon.
I had a few crash on a very specific use case. Alart from that it is working great.

[sqigls]

well said, i think in the day and age of misinformation you should be a tad more specific.
I've personally never had a crash with Avenger, which is saying something.
2012 macbook pro, mojave; 2021 macbook pro monterey, cubase 9,11,12.
never had a crash report instigated by Avenger on either machine.

[m-ac]

I'm not talking about crashes, I'm referring to CPU issues. I've gone above and beyond as far as reporting these issues, showing videos etc, as well as testing umpteen scenarios - apparently all for nothing. Maybe it's a logic + avenger thing. I've tested on 6 different macs now, 5 versions of OSX, and 6 different processors (intel and silicon). The thing just isn't made for mac, it's developed on pc then compiled for OSX after the fact.

I've got my fingers crossed for V2, because Ive spent a fkn fortune on this thing with expansions. Yeah I know...why, if I was having issues. I dont know, maybe I'm an optimist.

[Hanz Meyzer]

The performance is quite poor, on an Apple M1 pro, too, and also happens on Windows. It is caused by a lot of CPU spikes which the plugin produces, you can see it in Bitwig's cpu graph. No other plugin does this, so I agree with @m-ac.

I don't think V2 will be different here, since this problem seems to be ignored since 2016.

[zvenx]

The performance is quite poor, on an Apple M1 pro, too, and also happens on Windows. It is caused by a lot of CPU spikes which the plugin produces, you can see it in Bitwig's cpu graph. No other plugin does this, so I agree with @m-ac.

I don't think V2 will be different here, since this problem seems to be ignored since 2016.

Plus 1.
rsp

[hurricaneaudiolab]

A plain, bare-bones saw oscillator with all effects disabled shouldn't take this much CPU, man. Wish Avenger supported multicore like DIVA.

[twal]

I don't how long I can keep using Avenger, I have a one Osc patch with 2 voices and chorder going along and I'm hitting over 100% cpu when playing a handful of keys. Granted, my computer is a cheap one but... I either have to wait to upgrade or get Avenger later?

[Jac459]

I don't how long I can keep using Avenger, I have a one Osc patch with 2 voices and chorder going along and I'm hitting over 100% cpu when playing a handful of keys. Granted, my computer is a cheap one but... I either have to wait to upgrade or get Avenger later?

That's really really surprising... I have an m2 pro, so agreed it is supposed to be very performant but 1 Osc with 30 voices (6voivesx5 instances) is like 1 or 2% cpu.

[Hanz Meyzer]

Cpu spikes are not visible on a normal daw cpu meter: you really need to use bitwigs graph view to get a clue. When a single core reaches 100%, audio will drop out, no matter how many cores your cpu has… thats why it is common practice in plugin development to avoid ANY kind of cpu spikes. The spikes by avenger are very short but also extremely high. A common cpu meter will only show the average over a bunch of audio buffers. Some daws like renoise are more clever and show some kind of peak usage, but again, this is already averaged. A lot daws even show the sum of all cores, which is not helpful at all in this scenario, as explained above. Use bitwig to measure the cpu usage of a plugin. Click the cpu display to open the graph.

And a proper way to display cpu usage in a daw is to show the maximum peak of all cores divided by the max possible amount of all cores. So you see the max peak of single core. This turns out to be more complicated with new performance/energy saving cores. Yet, since the cores in a daw are assigned to the plugins dependent on the processing graph of the daw (on which track is the plugin, usually one track or even one group = one core), so not assigned dynamically on-the-fly, you might understand now, why already a single core at 100% causes a drop out. Maybe some day a clever mind will come up with a better way to process a daw graph, but switching cores for a process also is very costly.

[Jac459]

Cpu spikes are not visible on a normal daw cpu meter: you really need to use bitwigs graph view to get a clue. When a single core reaches 100%, audio will drop out, no matter how many cores your cpu has… thats why it is common practice in plugin development to avoid ANY kind of cpu spikes. The spikes by avenger are very short but also extremely high. A common cpu meter will only show the average over a bunch of audio buffers. Some daws like renoise are more clever and show some kind of peak usage, but again, this is already averaged. A lot daws even show the sum of all cores, which is not helpful at all in this scenario, as explained above. Use bitwig to measure the cpu usage of a plugin. Click the cpu display to open the graph.

That's exactly what I am doing sir .

I do have cpu Spikes but it is only on full songs and I can't say is it is avenger.

I also feel it is only seems I moved to the beta V5.

[Hanz Meyzer]

Ok try this: put the audio buffer to 256 samples. Load a single avenger instance and browse thru the factory library, while playing a bunch of notes. I can see drastic spikes here and there, even on my m1 pro, might depend on the preset, but not at all on the number of oscs. Now compare that with zebra 2 or hive.

[Jac459]

Ok try this: put the audio buffer to 256 samples. Load a single avenger instance and browse thru the factory library, while playing a bunch of notes. I can see drastic spikes here and there, even on my m1 pro, might depend on the preset, but not at all on the number of oscs. Now compare that with zebra 2 or hive.

Ok, I'll do later today. I think I'll compare with Falcon too. I just have Hive, not zebra.

Also Hive is CLAP so cpu management is supposed to be better.

[Trancit]

nevermind...

[Chipi]

The performance is quite poor, on an Apple M1 pro, too, and also happens on Windows. It is caused by a lot of CPU spikes which the plugin produces, you can see it in Bitwig's cpu graph. No other plugin does this, so I agree with @m-ac.

I don't think V2 will be different here, since this problem seems to be ignored since 2016.

Windows 11 64bit, NO problems here. Uses less cpu than many other popular synths.

[Hanz Meyzer]

Windows 11 64bit, NO problems here. Uses less cpu than many other popular synths.

Seems it was already to hard for you to properly read what I wrote above. Avenger behaves poorly regarding performance on any kind of system, because of cpu spikes. And it seems that you can't see spikes in your DAW.

@Jac459 For example try "AR charting atmo chord arp 1/2" from factory/arps. Spikes a lot here, even only using ONE note and 3-5 oscs at once. And my machine has a M1pro which is way faster than most intel machines, also the memory bandwidth is a ton faster. Using 48kHz, 256samples.

I think something is very fishy with the voice allocation code, since the spikes always seem to happen on a note start, so it is more obvious while using an arp. Also the sampler osc doe not seem to be the fastest out there, either. Compare this with any other commercial plugin, you won't find any spikes, for good technical reasons.