Is this possible?

Avene wrote on 5/3/2003, 9:24 AM
Lets say if I route a track to a bus out, that being a separate out on my audio interface. That signal then goes into an external fx unit with it's output coming back into Vegas through a spare channel on my audio interface which has been record enabled with the input monitor on. Now if I was to send this channel to another bus out and to a second fx unit and back into Vegas again assigned to another record enabled channel with the input monitor on, would I end up with any kind of latency? As far as I can work out, the latency would be the same for all record enabled channels, and those with audio already recorded. Am I correct here?

Basically I want to get more creative with my sounds by sending them through multiple fx units at once, and having the final effected sound coming back into Vegas where it can easily be recorded to a new track. I think I'll need more auido ins and outs of course, along with some more external fx units.

Comments

wazer wrote on 5/3/2003, 10:57 AM
Hi Avene!

i use RME HDSP w/Multiface. 8in/8out analog, SPDIF i/o, ADAT i/o. BufferSize/Latency : 512 samples - by DEFAULT, w/ASIO driver.
VEGAS 4.0's "automatically detect and offset for hardware recording latency" checkbox always checked.
My mixer is a Mackie 32-8. i don't use patchbays.

Now here are my first combination test results for your qouestion.
...

i started with a NEW track, with a whole song on it. 24 bit session, 24bit recorded wave, 44100 Hz.

i set this track (TRK 01) to MASTER bus (ANALOG out to the Mackie 32-8's track 1-2), and sent to BUS A (aux, 0.0 dB) too, which was a DIGITAL OUT to my Lexicon MPX-1.

i've created a NEW track (TRK 02), armed fo REC, (INPUT MONITOR ON, INPUT came from MPX-1's DIGITAL OUT, FX Mstr Mix 0%)

i set this track (TRK 02)to MASTER bus again, so i had the doubled signal on the MASTER bus.
At this moment i've HEARD some latency - didn't recorded.

i sent (aux) this TRK 02 to BUS B, which was an another ANALOG out pair to my Mackie mixer through the RME outs. I set this incoming signal to bus 3-4 on the Mackie (L/R mix on Mackie was OFF, so this means a simple NON-DELAYING ANALOG insert to another ANALOG tracks on the mixer) to get into my DRAWMER 1961 EQ.

I've created a NEW track (TRK 03), armed fo REC, (INPUT MONITOR ON, INPUT came from Mackie 32-8's bus 7-8 - to RME input 7-8 -, becouse the EQ was sent to this pair of bus on the Mackie, but not to L/R mix again.

So i had THREE tracks in my VEGAS 4.0, one was played, two was RECORDED.

i've HEARD D-O-U-B-L-E-D latencies/flams. (i mean for eg. a kik was played back THREE times insead of ONE loud sound!!)

BUT!

After a short period of recording, i ZOOMED in to see what i've got.
Arm for RECs / OFF.

MAAAN!

TRK 02 (MPX1 rec) was RECORDED and played back ALMOST IN SYNC (26-28 samples later) with the original TRK 01 - BUT PHASE REVERSED!

The TRK 03 (1961 EQ rec) was RECORDED and played back around 1094 samples LATER, and PHASE REVERSED again - becouse of the MPX-1)
The 1094 samples reminds me the DOUBLE 512 samples latency - but maybe i'm wrong.

It IS a very creative idea from you, but in my test doesn't passed..

What i do is always REC and REC and REC everything (incl. 100% wet FX tracks), and "S"+trim, and "S"+trim, and "S"+trim... Sometimes (OK, often)i use the PhaseReverse option...


Thanks for the testing idea!

w!

And sorry for my broken English if i wasn't understandable. I've tried to be..



pwppch wrote on 5/3/2003, 3:00 PM
Why not just use DX effects non-destructively or print the FX to a recorded dry track?

Every out -in - out - in pass will introduce additional latency on the final input path.

There is a buffering of data going out.
There is a latency on the input by at least one buffer of the output.

What this means is that for each buffer you play out, the input you will record will arrive 1 buffer later to the hardware, which then has to wait one more buffer before it can be mixed into the output which will then be output and delayed by another buffer before it can be recorded back in again.

Basically, latencies are added in your scenario.


So, the simplest scenario - with no FX in the analog chain:

Using ASIO drivers with a buffer frame size of say, 4 ms and an input offset latency of 1 buffer (4 ms). Then you would have 4 + 4 + 4 + 4 = 16 ms of latency introduced. Some ASIO drivers will actually introduce additional latency on input as it depends on the hardware/driver. From the track perspective, the alignemnt should be with in a few samples for the first track, depending on the accuracy of the ASIO drivers, but there will be a delay in the overall final signal. The second recorded track would have an offset of at least 8 ms.

What you will hear using software monitoring is a double flam, depending on the latency/buffer sizes used by the ASIO driver. If you are also monitoring your seperate outputs, you may get an additional flam.

Peter

Avene wrote on 5/3/2003, 6:47 PM
Hello Wazer, many thanks for giving that a try. I've been trying to experiment doing this myself, but unfortunately for some reason I can't get any audio out of my C-Port interface on outputs 3,4,5,6 etc. The C-Port's mixers show the levels, but the fx units the outs are connected to don't receive any signal.

Oh well, at least it was worth a try.. Thanks again.
Avene wrote on 5/3/2003, 7:17 PM
Hi Peter. Thank you for the detailed explanation here. I normally would use internal DX effects, but I'm still very attached to a couple of my outboard units like the Ensoniq DP/2. It's got a nice vocoder. And I just thought that I could run a vocal from one track with another signal as a modulator from another track, then send that off to another outboard unit such as a compressor etc.

So my best option would be to record each external effect to a dry track, or print the fx as you said? So essentially it would be no different than recording a vocalist or live part? I realised enabling the input monitor introduces a delay, although I was hoping that delay would be the same throughout. So yes, I understand why that wouldn't work now with the output buffers and all. Just one question here, why is the input monitor on each new track set to 'on' by default?

Also, apparantly Logic has these input channels where external audio signals from an audio interface can be mixed alongside recorded tracks. They all play in sync and can be mixed down together in real time. So basically this is where I got the idea from, and thought that maybe something similar could be achieved with Vegas. There wouldn't happen to be any plans for such a feature?
pwppch wrote on 5/4/2003, 8:36 PM
>>Just one question here, why is the input monitor on each new track set to 'on' by default?
<<

Just a choice we made. The average user wants this behavior. You can change it and save it as part of the default track prefs if you'd like.

>>Also, apparantly Logic has these input channels where external audio signals from an audio interface can be mixed alongside recorded tracks. They all play in sync and can be mixed down together in real time. So basically this is where I got the idea from, and thought that maybe something similar could be achieved with Vegas. There wouldn't happen to be any plans for such a feature?
<

As I remember it, this is done using hardware mixing in Logic, not software. ASIO- and I believe EASI - has a mechanism to permit the host software to control the input channels so it is "in the mix". Basically the audio input bus (return from the external FX unit) is mixed in the sound card vs having to be mixed in the software. Buffering delays cannot be made to disappear, no matter how you do it. You must buffer the data out and then wait one buffer before you can get it back. This effect is addative. There is no such thing as zero latency when it comes do PC audio unless you are doing it in the hardware - and even then there will be some latency.

The idea of mixing external sources from the Vegas Mixer page has been discussed. This can be done either in software or in hardware using ASIO. Still this is different than doing the output-input-output-input sceneario that you described.

Peter

Avene wrote on 5/4/2003, 9:31 PM
Cheers, thanks once again for the insight.