Cool. Thank you both for taking the time to share your best practices. Cheers!

Hi - 

Has anyone found which is the best option overall when using VEPro as a server on a separate machine to your DAW?

A. Many instances (tabs) of VEPro with say just a single Kontakt plugin instance inside and just the main return to the DAW (maybe 120 VEPro instance?)?

or,

B. Fewer instances of VEPro with multiple (up to 16) Kontakt plugin instances inside each with 16 stereo returns back to the DAW?

I've been doing B for some years but hear that some people put a single instrument on each VEPro instance(???).

Thanks!

Originally Posted by: casalena

Just an update - I built the same exact template with VE 6 and it loads more than 70% faster.  Even in standlaone.  Loading plugins, even moving buses or channels around takes forever.  Engaging 32 outputs in cubase took over a minute, with VE6 took 10 seconds.  It's pretty bad.

That would be in line with the earler automation lanes post. I will be going back to VEPro 6 for my next project.

Hi - 

VEPro servers on the network do not show up in the plugin for selection in my network.

For VEPro I use a dedicated point-to-point ethernet connection with no switch and fixed IP addresses for the two machines. There is no DHCP on that 'network'. All visible if I instead use my 'internet' network with DHCP instead. 

Should I still expect to see the instances even with just a point-to-point cable?

Both DAW machine (OSX) and VEPro machine (Win10) have two ethernet cards, one for 'internet' and one for VEPro. On the VEPro machine the 'internet' card is disabled when not downloading stuff.

Thanks.

From another user:

"The issue is they added a ton of automation lanes to VE Pro that means cubase auto creates all these auto lanes per return lane from an instance. So say you have 130 return stereo lanes, each one in the Cubase mixer now creates 20 auto lanes too. So now you have the Cubase mixer trying to create 2,860 mixer lanes every time it tries to open a session instead of 130. And there's no way around this with the VST3 standard -- if the plugin presents auto lanes they will be created in the DAW mixer."

Out of interest, is anyone having these issues _not_ using Cubase/Nuendo?

Originally Posted by: magnumpraw

I have experienced the same issue on Cubase 10/Mojave (Main DAW) and W10 (slave comps)

My template has my 6 slave VEPro instaces saved in a Freeze state (this is for various workflow reasons).

When I unfreeze an instance, VEP6 only takes a couple seconds to unfreeze and connect to the slave comp.

VEP7 takes over 5 mins!!! It is also much slower when refreezing that instance in a project.

I had to just revert to VEP6 on all my computers for now. Hoping this gets resolved.

Hi magnumpraw - 

It's a drag huh. Did you have any issues moving back to VEP6 with projects saved in 7?

Thanks!

Thanks for the reply.

Hmm - I wonder if it's a Windows defender thing maybe scanning remote calls if that's even a thing. Or maybe a paging or caching thing - don't know. Strange to see the slave machine CPU drop to just OS level during that 'hang' period.

So frustrating. If I find a solution I'll post it. 

Did you ever get this fixed? Was there some configuration that sorted it?

I currently have the same issue and more:

+ Long close time that appears like a hang (same as you) for many minutes and longer than load time;

+ A long wait after Cubase/Nuendo has connected to all the VEP instances. Watching the CPU monitor on the Windows VEP slave machine it goes up as each instance is connected and then once all are connected, and while in Cubase there is a progress spinner, the CPU on the slave goes down to 1 or 2% as if VEP has closed, stays there for 4 or 5 minutes then quickly goes up to about 30-40% and Cubase completes and I can start work. This is the same for coupled or decoupled - no difference. 

+ And also a super slow gui in VEP when connected to DAW.

Thanks!

Also probably worth observing...

The UI is only sluggish when the DAW is connected to the instances. When the preserved instances are disconnected the UI is as fast as normal.

Anyone got any clues please? Something is bringing VEP to a crawl and I can't figure out how to fix it. Thanks!

Hi - 

Does anyone know what might be going on here, please...?

With multiple instance tabs running and a bunch of VI's (mainly Kontakt), drop-downs across VEP (e.g changing output channels on mixer tracks, or changing Kontakt midi channels for instance) become very slow, taking a few seconds to respond and update when you click or move.

Things of note:
+ When the server project is loaded without being connected to DAW it is not slow to respond. It is only sluggish when there is an active connection to the DAW.
+ In other Windows apps, it is not slow when clicking around and doing stuff
+ CPU utilisation is around 35% (41% when playing)
+ Disk and network utilisation low
+ I've tried putting default thread count in VEP to 1, 2, and 8. No Difference. I've tried setting Kontakt to multi-core and note multi-core. No difference. 
+ No audio drop-outs are happening. It's just insanely slow and frustrating when trying to do any config. 
+ I am working via Remote Desktop Connection from Mac to PC. Use Hardware Acceleration for graphics is turned on. And in Device Manager on the PC it appears to be using the NVIDIA graphics card, not a virtual one for RDC.

Any ideas, please?

Thanks!

Graham

Windows 10 Pro, Intel Xeon 4 core (8 virtual), 256GB RAM, all SSDs, dedicated gigabit ethernet to DAW machine. Nuendo 10.

Hi - 

Does anyone have any suggestions for the best way to migrate big templates from using the "Kontakt 5" plugin to the new Kontakt 6 "Kontakt" plugin, please?

Surely there's a way of not having to recreate the template?

Many thanks,

Graham

Hi - 

Does anyone know how to fix this, please...?

As I add more VEPro instances/tabs with 2 instances of Kontakt in each one, the MIDI channel dropdown gets progressively slower. With 19 VEPro instances and 2 Kontakt instances in each, selecting any Kontakt instrument MIDI channel dropdown takes up to 3 seconds for the channel numbers to appear.

I've tried:

+ Increasing the default threrad count per instance in VEPro

+ Turning on and off the fast GUI switching VEPro setting

+ Turning off and on the HiDPI checkbox in VEPro settings

+ Removing MIR from the eLicenser so it doesn't load it with VEPro startup

+ Decreasing the default thread count to 1 and turning on 8 cores in Kontakt settings instead

+ Turning off all unnecessary Windows services and features

+ Changing performance to prefer Background Service instead of Programs, and vice versa

+ Turning off Superfetch

+ Disabling Turbo Boost

+ Disabling the Paging File

None of these made any difference

CPU is no more than 22% utilised. I have more than 100GB of RAM still free, all drives are SSDs. Audio works fine, it's just the GUI dropdowns are becoming crazily slow.

It's as if Kontakt is sharing a single resource across all instances and not load-balancing it across cores, or there is some weird bottleneck somewhere that is not being managed appropriately????

Any ideas?

VEPro Machine: Win 10 Pro, Xeon 4 core, 256GB, SSDs, Quadra K620 GPU.

Thanks!

Oh yes, yes! +1

Hi - 

It used to do it, and now it doesn't for some reason... Is there a way to get VEPro to automatically maximize the window when a new instance is created? It's a pain having to manually maximize every time. My machines are all Windows 8.1.

Thanks!

Graham