Hello all,
 
I have the Waves DiGiGrid IOS and use it with Sonar Platinum.  Since build v9r11 of their driver, I have not been able to load plugins in StudioRack and have them run from the IOS' server.  However, they will work fine when loaded locally.  The meters won't function and audio output either doesn't work or is not affected by the plugin when the plugins are running from the IOS' server.  StudioRack seems to work fine in other channels / tracks.  This bug has been reported to Waves, but I was curious if anything on Cakewalk's side is affecting it.  I understand Cakewalk may have a fix coming soon, so I wanted to verify this.  I was able to reproduce this tonight with the latest build of Sonar Platinum.
 
Your input is appreciated.

Best,
Stephen Salazar

Has anyone from Cakewalk looked into this?
 
Best,
Stephen Salazar

Although I have not tested this throughly, anyone out there having this problem should try upgrading to the March 23rd release of Waves installs.  They seem to have corrected the issue, as I'm able to run my Waves plugins via SoundGrid using busses again.
 
Best,
Stephen Salazar

Hey Stephen,
In an effort to reduce latency and avoid dropouts, I've been looking into the DiGiGrid systems, namely the Waves Impact to run my Waves plugs.  
Other than the above listed foible are you satisfied with the way Sonar is compatible with it?  Has it helped to free up CPU and Ram processing?  Noticeable latency improvements?
Cheers,
Mark Moore

Hi Mark,
 
It looks like I jumped the gun.  I saw that audio was being passed through when the SoundGrid plugins were activated on the IOS server from busses...I even heard that it was the plugin-processed sound, so I was under the impression it was working.  However, I started to notice glitching when I tested today, as if it was dropping out and had a jittery, static-like quality to the sound.  With all of my Waves plugins in StudioRack inserts configured to load from the SoundGrid server, I saved a new copy of my project file.  I reopened the new copy and noticed the problem was still present where audio would not pass through the StudioRack plugins in busses (no audio / no meter activity).  Therefore, I'll be writing Waves / DiGiGrid again for an update on when it will be fixed.  Please note that it's fine to run StudioRack plugins for tracks...the problem just exists in the busses.  This did NOT happen until v9r11 of the driver.  I'm hoping they can get the driver and firmware issue resolved soon (especially considering I'm paying for the WUP).
 
My project is set at an 88.2 K sampling rate.  When I don't have the StudioRack plugins switched to the IOS server in busses, I'm able to mix with the following settings in Soundgrid Studio:
WSG Processing:  112 Samples / 1.3 ms
Soundgrid Driver Buffer (Network) 208
Soundgrid Driver Buffer (ASIO) 256
 
One thing to consider is you'll have to experiment with the buffer settings that work for you.  When I had all of the StudioRack plugins running from the IOS, I couldn't get audio to playback without glitching unless I raised my network buffer to the maximum setting (1088) and ASIO at 512.  Of course, this is with the problematic bus plugins running from the IOS, so I'm curious how that will be once the problem is fixed.  Although the StudioRacks in the busses won't produce audio or visuals on the meters unless I run them locally from my computer, I'm still able to load quite a few plugins in the tracks with StudioRacks.
 
Performance:
There's a little delay getting the audio engine active.  You've probably seen it in the demo videos posted on Waves / DiGiGrid websites.  Also, opening more than one StudioRack at a time may cause a dropout in Sonar, but this requires experimenting with the buffer settings.  Once you have the right settings, the engine stays pretty stable and the dropouts become less of a problem.  Dropouts usually don't occur during playback.  I don't have all of my Waves (or other third-party plugins) loaded in StudioRacks.  I've outlined this below to give you an example of what I'm doing with the hardware I use.  The IOS still has voices and DSP to work with, and I'm still learning how to use it efficiently.
 
Project Files:
10 stereo tracks
8 mono tracks
Drum parts are currently frozen from Superior Drummer 2 (which works with virtually no latency when tracking from my electronic kit).
 
Plugins:
* Bass drum (StudioRack Stereo on IOS with Scheps 73, Kramer Tape, and RCompressor) + TS-64 Transient Shaper
* Snare Top and Bottom sent to a bus (StudioRack Stereo in the bus currently running locally with Kramer Drums...perhaps unnecessarily in the StudioRack due to the bus issue) + TS-64 Transient Shaper
* Hi-Hats (JJP-Cymb & Perc...no StudioRack)
* Toms (StudioRack Stereo on IOS with CLA Drums, Aphex Aural Exciter, and Q1)
* Overheads (JJP-Cymb & Perc, DeEsser, and Q4...no StudioRack)
* Drum Room (Scheps 73, DeEsser, and Q4 - no StudioRack)
* Bass (StudioRack Stereo on IOS with C6-SC (sidechained with bass drums and snares), CLA Bass, REQ 4, dbx-160, and MaxxBass) + Devil-Loc Deluxe
* Guitars (Manny Marroquin EQ on 2 tracks)
* Guitar Solo (Manny Marroquin EQ, RCompressor, H-Delay, DeEsser, Aphex Aural Exciter, and Vitamin...no StudioRack)
* All drums sent to parallel compression bus (StudioRack Stereo currently running locally with H-Comp)
* Lead vocals to StudioRack Stereo in BUS currently running locally with Butch Vig Vocals, CLA Vocals, RVox, DeEsser, and Aphex Aural Exciter)
* Harmony vocals to StudioRack Stereo in BUS currently running locally with Butch Vig Vocals, Manny Marroquin Reverb, DeEsser, and Aphex Aural Exciter) + Eventide H3000 Factory
* Guitars sent to bus for parallel compression (CLA-76 and OneKnob Phatter...no StudioRack)
* L2 temporarily on the Master (no StudioRack)
 
I realize I can move some of those plugins in the tracks to StudioRack, but I'm doing a balancing act with performance and trying not to overload the IOS.  I may eventually decided to use the StudioRacks exclusively for tracks to get around not being able to use them in busses while still making good use of the IOS.  My CPU meter in Sonar is barely peaking, so I know I can get away with opening more plugins that run off my CPU.
 
One thing about the DiGiGrid IOS...it's super flexible, routing-wise.  Also, I can load plugins for tracking with no noticeable latency, which is great for when you want to hear a little reverb on your voice or other instruments.  There are many ways to accomplish this, too, including loading those plugins from eMotion ST on the IOS server, which is nice.  It took a while to wrap my head around the routing options of the IOS, but once I figured it out, I realized I could configure it exactly the way I wanted.  It's a great sounding and useful product.  It just has a few issues with Sonar at the moment.
 
System:
ASUS X99-A with gigabit Ethernet
ASUS ThunderboltEX II (installed, but not currently in use)
Intel Core i7-5960X @ 3.0 GHz (Haswell 8-core)
Corsair Vengeance LPX RAM (32 GB)
Samsung XP941 512 GB SSD system drive
2 Western Digital 500 GB Velociraptors for audio and sample drives
XFX TSK Multipli X54 display card (AMD Radeon HD 5450)
Corsair Obsidian 550D with Corsair Pro Series 860 Watt PS (AX860i)
Noctua NH-D15 Premium CPU Cooler
Windows 10 Pro 64-bit
 
I figured you wanted some details, so this might give you an idea.
 
Best,
Stephen Salazar
 
 
 
 

Hi Mark,
 
So I decided to move all Waves plugins that were in the tracks on my project to StudioRack Stereo inserts.  I had no glitches during playback of my mix (even when switching between the different StudioRack inserts).  Additionally, I still have plenty of headroom left on the IOS with the same buffer sizes mentioned previously.

Voices used:  18 of 64
DSP used:  48 to 62%
 
Best,
Stephen Salazar

Hey Stephen,
Thanks for the detailed response.
It appears you would have a very CPU intensive workflow if not for the IOS.  I'm vacillating on the purchase simply for the fact that I use a lot of non-SoundGrid plugs (ie., SoundToys, Slate, Kush Clariphonic and Pusher, etc.)
However, I love using the Waves SSL channels and CLA compressors on practically every channel and busses, and I run a lot of busses/auxes.  My template is:
Master
Sub
Music
Vocals
FX
Room
Drums
Toms
Toms Verb
Kick/snare crush
NYC compression
Percussion/Loops
Bass
Bass crush
El Guitars
El guitars Spring verb
Acoustic gtrs
Keys
Aural Exciter
Slap
Vox Echo
Vox Spring
Each with SSL or ProChannel running on them.
 
Along with that my sessions usually run about 36-42 tracks with 14 channels of drums, 3 channels of bass (DI, amp and distortion for bite), lots of guitars, keys, vocals, (lead and double tracks) bgvs, percussion and loops.
As you can see I need all the CPU and memory I can get and started looking into the SoundGrid system to help with that.
By the time I'm finishing a demo Sonar is limping pretty badly and I hate rendering because I'm often asked to recall a mix I did weeks earlier.
But, creativity is borne from limitations I guess.
Good chatting.
Mark.
 
 

You might find that getting a heavier hitting computer would be as or more useful than getting a digigrid system, and maybe cheaper too. AMD processors just are not that good at high end processing these days, sadly. A high end AMD chip generally competes with an i3 or maybe i5 when you talk heavy workloads. So you could look at getting a Haswell-E series 6 or 8 core chip. Those, particularly overclocked, are beasts with tons of power for CPU hungry programs. Not an inexpensive proposition, but then neither is a digigrid setup and a faster host CPU will run everything.
 
Also allows you to get more memory, if you wish, a Haswell-E on an X99 board supports 128GB of RAM currently.
 
Just worth looking at before you pull the trigger on a digigrid system. If you aren't interested in rolling your own you can talk to Jim Rosenberry on this forum, selling custom DAW systems is his profession. From him you'd be looking at somewhere in the realm of $2100 to get an OC'd 6 core Haswell-E system with 32GB of RAM, with prices going up from there as you wanted more hardware.

On the new i7-6700 skylake processor, a resource hog like Ozone 7 barely registers at all. I've yet to find a situation that has maxed it out yet (at least not on purpose).

As a simple example of a Haswell-E I loaded up a project I have with 12 virtual instrument tracks in it. Those are BFD3 for one track and a split of Kontakt and Synthmaster for the others. All the synths were set to the highest quality resampling, Kontakt sample size and complexity varied but I had a few instances of Kirk Hunter Concert Strings 2 and 3 loaded, which are pretty big and 3 in particular has some intense scripting. About 9GB of memory used. Nothing bounced, all VIs playing in realtime.
 
I then tossed on a bunch of superfluous effects on the channels to load things up a bit. 3 effects per channel. Most channels had QL Spaces (convolution reverb) on them already, to that I added a selection of EQs I have like LP-64, Q-10, and Passive EQ and compressors like V-Comp, H-Comp, Ultrachannel, and VC67A. This was just haphazard, I didn't actually configure them, just tossed them in for load. The master channel had a couple of hitters on it, MSpectraldynamics and IK Stealth Limiter. Also console emulation was turned on for all channels.
 
Total load in task manager shows about 23-25% at all times. This is with an i7-5930K running at 4GHz.
 
So nothing super scientific, but gives you some idea of the load handling you can expect out of a Haswell-E and if it might be sufficient for your uses. You could roughly triple that load and expect it not to have any real issues.