Soundgrid Studio Latency

Hi all,
Just wondering if anyone can enlighten me on a problem I am having regarding Soundgrid Studio?

My setup seems to be working well, however, I have an untenable about of latency when using StudioRack and tracking in both LogicProX 10.7 and ProTools Ultimate 2021.10. (Using the SoundGrid Studio core audio Driver)

My setup is a MacPro running Big Sur 11.6, A Soundgrid DLI running an AVID OMNI in SGP mode.
Soundgrid Studio is set to Latency Optimized.
I’ve tried this setup on various occasions over the last few years and I just can’t seem to get the advertised results with my DAWS.

I am used to and very familiar with RME UFX’s (I have a couple of those) and the Totalmix software for its direct monitoring, which provides no perceivable latency at all.

Or, am I just approaching this all wrong? I can obviously setup a channel in SGStudio and put the same plugs on it and it works by monitoring in there, however, I was under the impression, I could just do it with the StudioRack plugin (toggling between states in Protools for example using the Peripherals/Midi/HUI method, which works), but there’s a heap of latency.

Any one got any ideas??

Just thought I’d follow up…
I think I’ve got it. Still lots I don’t quite understand, but it was something in my patching.

I’m using the Soundgrid Driver in Logic, but from there, my setup is a bit unusual.

I wanted to combine my two main setups in to one.
ie I have 8 channels of ADAT running to and from my UFXII to the OMNI (all set to ADAT I/O) then the DLI in SGP mode.

There’s no latency at all. Amazing. Just not sure what I did to fix it.

However, I’m still trying to work out the m-s StudioRack plugin and how it can be reflected properly in SoundgridStudio.

…and now, after it all working so well… I cannot get audio to pass in Soundgrid Studio’s “STUDIORACK” page in the mixer, when recording. Works fine in playback. Aaaargh!!!

Aaaah, I finally got it !!! I wasn’t matching the input source to the StudioRack source.

I was just under the impression that you had to assign any slot, as long as it was different to others.
For example, if want to record 4 takes using the same input but differing StudioRacks, I have to make sure that either the previous ones are set to “None”, or set to playback only.

What absolutely blew my mind was that in the PATCH, I can even assign inputs from my RME UFXII directly to spare CoreAudio Inputs and use the direct monitoring through plugins, with what seems like only another ms of latency. It’s really fast!!!

Bit of a wow moment fo me. So just though I’d post to see if it may help others, as there’s not a lot of in depth info around about all of this.

Your suggestion is very helpful to us. Thanks for your precious post.