mixing DSP and Native plugins: good guide?

[ThosSounds]

Hi all,

Anyone know of a good guide for mixing DSP and Native insert plugins on audio tracks? Opinions/rules of the road, do's and dont's, etc..? Searching for a thread or such but not finding one, at least not from the last few years?

Thanks!

[WildHoney]

Originally Posted by ThosSounds

Hi all,

Anyone know of a good guide for mixing DSP and Native insert plugins on audio tracks? Opinions/rules of the road, do's and dont's, etc..? Searching for a thread or such but not finding one, at least not from the last few years?

Thanks!

My understanding (from the TDM days) is that you will save voices by using DSP plugins on auxes and native plugins on audio tracks. Further, once you're on an aux, each time you go from a DSP to a native plugin, you'll be using up a voice (or two). Similarly, on an audio track, each time to go from native to dsp, you'll be using voices. As far as I can tell, there's never really a reason to use DSP on an audio track unless you're needing to track through the plugin (guitar amp models are my primary example -- I always use the DSP version to avoid latency while tracking). Not claiming to be an expert on this issue, but have been managing it in sessions for a long time with these principles.

[JFreak]

Voices are a non-issue with HDX, a little bit has changed since TDM days.

But as a rule of thumb you don't want to go Native-DSP-Native-DSP-Native-DSP-Native but instead first try to survive with full DSP mixer; and once you need to go native, you just set all the plugins in a track Native if possible. You only get the latency hit once and save DSP for the tracks that can be left full DSP.

[JFreak]

Originally Posted by WildHoney

My understanding (from the TDM days) is that you will save voices by using DSP plugins on auxes and native plugins on audio tracks.

Ahem, there are two things that you apparently mix with each others.

First of all, in the early TDM days the only plugin you actually could instantiate on an aux track was a TDM plugin. RTAS was not an option. Then came HTDM (host-tdm, basically native processing) to go around that, and if memory serves me right it was PT7 when we were allowed to use RTAS on an aux track. So that's probably why you think it's better to use DSP plugins on aux tracks.

Another thing is voices. What is a voice? It's an input point of TDM mixer. Every time signal goes through TDM it needs a voice. So if you have an audio track, what happens first is signal gets recorded to disk and then routed to TDM mixer (1 voice used) and without any plugin processing it goes to output and that's it. TDM-only processing doesn't eat voices, because the signal is inside TDM mixer all the time. But things change once you insert a native plugin. Then the signal goes out of the TDM mixer to host CPU for processing and then gets back to TDM mixer (1 voice used). If you now send that signal to output, you have used 2 voices. But if you have TDM-RTAS-TDM-RTAS-TDM plugin chain, your signal needs to go out of the TDM mixer another time and get back in once again so it needs third voice for it, etcetera.

My understanding is that HDX has a different formula for voice calculations so what I just wrote only applies to TDM.

[dwaynedelario]

From my experience wiith HDX, less delay compensation will be required if you place Native plugins before DSP plugins on audio tracks. On aux tracks with assigned inputs (that are routed / in use), plugin induced delay seems to be the same regardless of order.

So where applicable, put DSP after Native on audio tracks if you must mix and match.