Conrad,
Sorry for the delay in answering, I was not at home yesterday.
Conrad_PA5Y wrote:...On 2m in particular I have a very noisy blower and need downward expansion to keep background noise down to an acceptable level. It is important that you can hear this I feel. Otherwise any breathing as the gate opens prematurely really does sound awful!
So many great strides have been made in the available audio processing features of PowerSDR. It is now second only to external rack processing or full-on digital audio workstation software. Unfortunately, this does not include the gate/expander. Perhaps someday there will be a gate that has adjustable attack/hold/release times. The expansion ratio is adjustable, that is the gate "percentage" setting. Put that at 100% and it is a pure gate. I've found that 90% works best for me, although I don't actually use the gate, I still perform my noise gating in external DAW software. However, I have moved all of my other audio processing into PowerSDR. If and when the gate in PowerSDR gets more sophisticated, I feel confident I can abandon the DAW entirely at that time.
Would I be correct in assuming that the lowest latency would be available by driving the line input directly from a mixer and listening via VAC with an ASIO driver?
Actually, no. The lowest latency is obtained using a high quality, zero latency audio interface with ASIO support for both transmit and receive. Doing this effectively shortens the audio paths, which are otherwise quite lengthy and torturous (refer to
this thread). However, care must be taken to optimize the ASIO driver and buffer settings, and of course the VAC settings, to obtain the lowest possible latency. With a UMC202HD audio interface I could normally do about 10mS better than the "standard" audio path, and although I haven't measured it, I know I am doing even better with my near-zero latency Presonus Studio interface. Bryan, W4WMT, gets some really outstanding performance out of his professional PCIE digital audio card, but that thing costs a mint!
I have both a Behringer UMC202DD and a Focusrite Scarlet 2i2 available, the Behringer achieves around 9.6ms round trip latency with my (soon to be improved) system.
Experience has shown that the Behringer is a superior performer compared to the Focusrite. It has a faster and more stable driver that is easier to adjust.
I have no concept at this stage just how much latency the audio processing will add.
This is where you are going to become disappointed in the latency of the MON function. First, the total latency of the leveler, EQ, CFC compressors and post-EQ create is substantial, measured in many ten's of milliseconds. Then add to that about another 30mS of latency through the low latency transmit filter. The result, even with the most optimized VAC/ASIO setup, is not commensurate with real-time monitoring. This is why you are much better off creating a WAV file recording of your unprocessed audio coming in from the microphone/sound interface (radio or VAC) and then playing that into PowerSDR as a loop using the built-in facilities available via the Wave menu when making adjustments. Rob, W1AEX shows how he does that in his excellent Youtube videos-- see
this thread.
I am also dismayed to find that I can only have one ASIO based VAC running.
Someday we may see a new VAC output that carries the same audio stream that goes to the radio CODEC. Preliminary work has been done on this and perhaps it will get into a future version of PowerSDR.
73,
Scott