Jump to content

IK Multimedia: High CPU on latest AT5 and Lurssen


Jacques Boileau

Recommended Posts

I have done some not very scientific but still valid CPU usage tests on these two plugins. I have a good but somewhat aging PC (i7 4th generation) and unfortunately can't afford having new versions of plugins that grab much more CPU than the previous version. So, for me and for you if you need plugins to be easy on your CPU, we can choose which version we want to use. Depends if things in the release notes of the latest version are a requirement for you usage or not.

Amplitube 5 - v5.0.3 and v5.1.0

For this test I created a project in Cakewalk using Basic.cwt. I then added AT5 to the audio track and left it at the default preset. I didn't play, just left it idle. I took my reading from the Cakewalk performance module.

  • v5.0.3: Audio Processing 1.3 Engine Load 10-15%
  • v5.1.0: Audio Processing 2.5 Engine Load 20-30%

Lurssen Mastering Console - v1.1.1 and v1.0.3

For this test I have used LMC standalone because that is the way I master my stuff. I used Windows Task Manager to determine the CPU usage of the two versions while playin a song in LMC. Of course these fluctuate, but it is easy to see that there has been quite an increase in CPU usage and if you are using LMC inside Cakewalk on a big project, it may make it non usable in that scenario. It certainly did for me.

  • v1.1.1: 6% - 7%
  • v1.0.3: 1.5% - 2.5%
Edited by Jacques Boileau
  • Like 2
Link to comment
Share on other sites

34 minutes ago, Jacques Boileau said:

I have done some not very scientific but still valid CPU usage tests on these two plugins. I have a good but somewhat aging PC (i7 4th generation) and unfortunately can't afford having new versions of plugins that grab much more CPU than the previous version. So, for me and for you if you need plugins to be easy on your CPU, we can choose which version we want to use. Depends if things in the release notes of the latest version are a requirement for you usage or not.

Amplitube 5 - v5.0.3 and v5.1.0

AmpliTube 5 has been updated to 5.1.1.

Link to comment
Share on other sites

2 hours ago, Jacques Boileau said:

I have done some not very scientific but still valid CPU usage tests on these two plugins. I have a good but somewhat aging PC (i7 4th generation) and unfortunately can't afford having new versions of plugins that grab much more CPU than the previous version. So, for me and for you if you need plugins to be easy on your CPU, we can choose which version we want to use. Depends if things in the release notes of the latest version are a requirement for you usage or not.

Amplitube 5 - v5.0.3 and v5.1.0

For this test I created a project in Cakewalk using Basic.cwt. I then added AT5 to the audio track and left it at the default preset. I didn't play, just left it idle. I took my reading from the Cakewalk performance module.

  • v5.0.3: Audio Processing 1.3 Engine Load 10-15%
  • v5.1.0: Audio Processing 2.5 Engine Load 20-30%

Lurssen Mastering Console - v1.1.1 and v1.0.3

For this test I have used LMC standalone because that is the way I master my stuff. I used Windows Task Manager to determine the CPU usage of the two versions while playin a song in LMC. Of course these fluctuate, but it is easy to see that there has been quite an increase in CPU usage and if you are using LMC inside Cakewalk on a big project, it may make it non usable in that scenario. It certainly did for me.

  • v1.1.1: 6% - 7%
  • v1.0.3: 1.5% - 2.5%

Just out of curiosity I tried your Amplitube 5 tests with my i5, 9th gen PC.  Here are some numbers that may help establish a benchmark with newer hardware. I do see an increased (doubled) load with the latest version.

Amplitube 5 - (tested with VST3)

  • v5.0.3 Audio processing 0.8%; Engine load 5-8%
  • v5.1.1 Audio processing 1.8; Engine load 10-15%

 

  • Thanks 1
Link to comment
Share on other sites

37 minutes ago, abacab said:

Just out of curiosity I tried your Amplitube 5 tests with my i5, 9th gen PC.  Here are some numbers that may help establish a benchmark with newer hardware. I do see an increased (doubled) load with the latest version.

Amplitube 5 - (tested with VST3)

  • v5.0.3 Audio processing 0.8%; Engine load 5-8%
  • v5.1.1 Audio processing 1.8; Engine load 10-15%

 

With a multi layered guitar song with a lots of doubled guitars and details on multiple tracks, double the load makes a heck of a difference!

For what is added between 5.0.3 and 5.1.1, it is hard to understand what made it such a jump.

Although I understand the new VIR technology must take some part of the new CPU load in AT5, but AT5.0.3 is double AT4. So AT5.1.1 is four times as CPU hungry as AT4!

Sometimes I wish we could disable VIR, as it is the case in AT5 CS, for lesser important guitar parts to free up some CPU. Just a thought.

  • Like 1
Link to comment
Share on other sites

20 minutes ago, Jacques Boileau said:

With a multi layered guitar song with a lots of doubled guitars and details on multiple tracks, double the load makes a heck of a difference!

For what is added between 5.0.3 and 5.1.1, it is hard to understand what made it such a jump.

Although I understand the new VIR technology must take some part of the new CPU load in AT5, but AT5.0.3 is double AT4. So AT5.1.1 is four times as CPU hungry as AT4!

Sometimes I wish we could disable VIR, as it is the case in AT5 CS, for lesser important guitar parts to free up some CPU. Just a thought.

I bought into SampleTank 4 MAX as a pre-release  deal. I was still running my gen 3 at that time. The jump in CPU load in the initial v4 release compared to SampleTank 3, made it almost unusable for me.

Last year I upgraded to my current i5-9600K, 6 cores locked at 4.3GHz, with turbo to 4.6GHz. Got a little more breathing room now, and it handles most things I throw at it now without the cooling fans revving up.

I get the impression that the devs at IK are building cool new features using the latest hardware, with little regard for optimizing code, so that previous gen hardware can take advantage of them. 😢

Link to comment
Share on other sites

28 minutes ago, abacab said:

Last year I upgraded to my current i5-9600K, 6 cores locked at 4.3GHz, with turbo to 4.6GHz. Got a little more breathing room now, and it handles most things I throw at it now without the cooling fans revving up.

My i9-10900K 10-core processor O.C. to 4.9GHz is not (yet) choking on me.

Edited by Bapu
Link to comment
Share on other sites

14 minutes ago, abacab said:

Well that one should be good for a couple more years!!! :)

All kidding aside, I'm hoping it will last 5 more years, then I'll probably only need one more upgrade to last the rest of my life because by that time I'll own ever plugin known to man and aliens alike and most likely will only WUP after that (just to get in Fleer's craw).

  • Haha 3
Link to comment
Share on other sites

  • 3 weeks later...

Hi Jacques, I saw your post on the other thread. I see you have a "i7 4th generation ". This sounds similar to my old DAW setup. I was running an older i7 Mac Mini. I started to notice that certain newer VST products on the market such as Amplitube 5 and Massive X were really taxing the system and I realized there was little I could do anymore to keep the system running faster. I already had upgraded the hard drives to SSDs. But an older gen i7 is an oler gen i7.

In you case, I would suggest just stick with Amplitube 4 (I believe you get a copy of that if you have AT 5 and most of the collections show up in that) or go with another guitar sim. There are tons of other that are quite good and nowhere near the resource hogs of Amplitube. Try Klevgrand STARK.. I am still blown away by that one.

re: Amplitube 5 .. I can't explain the reason for the resource usage other that to assume that all the new fancy graphics in the interface and the sound engine is built on some sort of framework that just uses more CPU overall. As much as a Amplitube user and fan I am , I find I flip back and forth between 4 and 5.  4 sounds totally different and in some cases I prefer using 4 instead of 5 when using some amps and effects.

good luck.

Link to comment
Share on other sites

24 minutes ago, telecode 101 said:

Hi Jacques, I saw your post on the other thread. I see you have a "i7 4th generation ". This sounds similar to my old DAW setup. I was running an older i7 Mac Mini. I started to notice that certain newer VST products on the market such as Amplitube 5 and Massive X were really taxing the system and I realized there was little I could do anymore to keep the system running faster. I already had upgraded the hard drives to SSDs. But an older gen i7 is an oler gen i7.

In you case, I would suggest just stick with Amplitube 4 (I believe you get a copy of that if you have AT 5 and most of the collections show up in that) or go with another guitar sim. There are tons of other that are quite good and nowhere near the resource hogs of Amplitube. Try Klevgrand STARK.. I am still blown away by that one.

re: Amplitube 5 .. I can't explain the reason for the resource usage other that to assume that all the new fancy graphics in the interface and the sound engine is built on some sort of framework that just uses more CPU overall. As much as a Amplitube user and fan I am , I find I flip back and forth between 4 and 5.  4 sounds totally different and in some cases I prefer using 4 instead of 5 when using some amps and effects.

good luck.

Thanks for the input! I know my PC is aging fast, but I will live with it for a while. I own AT4 and still have it installed. It is 1/4 the cpu load than the current version of AT5, so yes it's a good solution. The other solution is to freeze tracks when I am done tweaking them. But for now, I am still using the versions prior to the cpu increase. For LMC that is quite fine. For AT5, I have the drawaback of not being able to use X-Gear, so I will upgrade AT5 at the end of the GB since I intend to get these.

I just think it's in IK's best interest for them to check if there is something that can be optimised or not. The more efficient the plugins are the less support calls they get and the more satisfied their customers are. Not a waste of time. Remember that noting really fancy was added to these two products between the version that saw the increase in cpu. The fancy graphics and great sound was already there!

I am not dissatisfied, I just feel something can probably be done about it and felt giving my input and findings was important.

Edited by Jacques Boileau
  • Like 2
Link to comment
Share on other sites

The quetion is are we able to choose install version (older versions) had the same issue with roland cloud , i hit the support , no way to get old isntaller , crazy , i copy and paste old dll from laptop to desktop and it worked , lesson learned and i knew it !! i tend to become to much confident versus my old dayz where even comp was plugged to internet !!

Link to comment
Share on other sites

1 hour ago, Zo said:

The quetion is are we able to choose install version (older versions)

Yes, no problem! The older install versions are available in "My Products" at the IK website.

Just click on the "past releases" button to access them.

  • Like 2
Link to comment
Share on other sites

11 minutes ago, abacab said:

Yes, no problem! The older install versions are available in "My Products" at the IK website.

Just click on the "past releases" button to access them.

Thks , you rock !!!!

  • Like 1
Link to comment
Share on other sites

5 hours ago, Zo said:

@Jacques Boileau be sur to check if in both cases the HD engine is in the same setting ......

Very good point Zo. I did make sure in my test that it was the same configuration, but I have redone the tests with and without HD to see what difference it makes and if it is an avenue to save some cpu. From the numbers below (average I mentally made since the numbers vary) you can see a small gain if HD is disabled, but not as much as I would have expected.

Without HD

1.0.3: 2%
1.1.1: 6.5%

With HD

1.0.3: 3%
1.1.1: 7.2%

I did not take readings, but with DDM off you have a very small difference, but nothing worth not using the feature.

Edited by Jacques Boileau
Link to comment
Share on other sites

6 hours ago, Jacques Boileau said:

Very good point Zo. I did make sure in my test that it was the same configuration, but I have redone the tests with and without HD to see what difference it makes and if it is an avenue to save some cpu. From the numbers below (average I mentally made since the numbers vary) you can see a small gain if HD is disabled, but not as much as I would have expected.

Without HD

1.0.3: 2%
1.1.1: 6.5%

With HD

1.0.3: 3%
1.1.1: 7.2%

I did not take readings, but with DDM off you have a very small difference, but nothing worth not using the feature.

Really. Strange , are you cheking asio meter for values or real windows values ? 

Link to comment
Share on other sites

Please sign in to comment

You will be able to leave a comment after signing in



Sign In Now
×
×
  • Create New...