View Single Post
  #15  
Old 06-10-2021, 12:04 AM
Matt Hepworth's Avatar
Matt Hepworth Matt Hepworth is offline
Member
 
Join Date: Dec 2013
Location: Phoenix or SLC
Posts: 557
Default Re: Question regarding both HDX and Carbon

Quote:
Originally Posted by digiot View Post
Not following what you are saying. The only time latency is an issue on native systems is when live tracking. With the Hybrid engine that is taken care of.
VI’s have a much lower latency then live inputs. You don’t need to run VI’s at super low buffers (64 or lower)to not perceive latency.
RME and Quantum are solid on low buffers but still not as solid as HDX/Carbon.
As for UAD, if you are fine with rerouting everything through their 3rd party mixer just so you can track with no latency then cudos to you.Their workflow is horrible not even counting their lackluster DSP. I love their plug-ins but unless you have tons of their DSP chips they are not usable in a full mix.
Native is subject to buffer, protocol, converter delay, and driver. VIs are only affected by output latency, true. At 256 buffer that's about 9ms with Carbon and about 6ms with HDX. An Apollo WITHOUT LLM enabled (monitoring straight through PT and not using Console) is 6ms output latency at 256. Quantum is 5ms.

I agree Carbon and HDX aren't terrible for VIs, but if that's the focus they should be avoided, since native systems will provide superior latency for less investment.

However, HDX minimum buffer at 48kHz is 64 still, IIRC, while native interfaces are 32.
Taking all the above into consideration, the *absolute minimum* output latency is 4ms for Carbon, 2.5ms for HDX, 1.5ms for Apollo, and sub 1ms for Quantum.

Keep in mind this whole thread is about *native* plugins for input monitor paths with Carbon/HDX vs Native interfaces. Native is built for this and performs better. If we're talking audio input monitoring then DSP would come into play and Carbon (and, to a lesser extent HDX) would prove far superior to native RTL.

Last edited by Matt Hepworth; 06-10-2021 at 12:23 AM.
Reply With Quote