While I wait for my Tier 3 Dwarf to arrive I have been dreaming about what I would do if I was designing a DSP audio appliance from scratch.
The most obvious application for me is real-time processing for low-latency tracking to a DAW.
One of my biggest headaches is having to overdub guitar through an amp simulator into an already-busy session with lots of tracks and effects.
Tracking guitar in this scenario often requires disabling reverbs or freezing tracks or bouncing the session down and overdubbing to that because otherwise the CPU is overburdened and simply cannot handle tracking cleanly at low buffer sizes.
Having an external DSP box to run the amp sim(s) would shift the burden of real-time processing away from the CPU and allow it to focus on running the session.
The Dwarf is almost the ideal box for this but not quite because it lacks digital IO. Analog IO inevitably means more latency added at the AD/DA stages. Also it has only two outputs which means that if I want to record a stereo rig I cannot simultaneously record the clean guitar signal.
I’d love the next version of the Dwarf to have ADAT/SPDIF lightpipe and/or AVB. In my mind this would be the ultimate standalone audio DSP for tracking.
Well, as an aside, I didn’t realize that the Dwarf could function as a USB audio interface. (Can it?)
But this is irrelevant. The real reason to use digital audio is that many of us in the studio have organized our workflows around a primary audio interface which is usually an integral part of our setups. Using the USB audio on the Dwarf would disrupt this workflow and defeat the purpose of having a high-end interface in the first place.
It’s entirely unreasonable to expect someone entrenched in his/her workflow to interrupt a session to switch from, say, an RME UFX II with it’s exceptionally reliable and performant drivers to some generic interface with class-compliant drivers. Especially if the session relies heavily on the hardware and software routing capabilities of the UFX II.
Using the USB audio on the Dwarf would defeat the purpose of having a high-end interface. The solution is to route digital audio out of the interface to an external device and back in again, latency-free.
Thanks but I was thinking about using CC eth port to send digital audio over UDP or other realtime protocols. The idea is to distribute signal processing stuff over multiple devices when SPDIF it’s not available.
Ah I see, but this would require to override however these pins are currently configured. I’m not sure if you can get suitable transfer rates out of this either.
But sounds like an interesting “hack” none the less.