Hi Stuart,
bassdude wrote:
Shroomz, can you give an example of why the alignment is so critical. I'm just trying to understand.
Well there's several reasons, but the main one is the cruncher of a question,
"how much sample deviation between channels is acceptable". Well, none in a stereo device as far as I'm concerned. None in a multichannel device like a mixer either if I can build it in a way which reliably loads time after time, from day to day without the random phase alignment issues caused by a device being dynamically assigned accross dsps. I think that a stereo or multichannel device should be sample aligned in the sense that what you put in is aligned in the same way when it comes out. That means that I want to be able to feed all channels the same test tone, phase invert what comes out of any of it's channels & completely cancel that signal out with one of the other channels' outputs.
Quote:
For instance when I send drum tracks to the vp8, the different mic positions will introduce delays anyway so this alignment is really no concern to me at all. And 2-3cm percieved difference? I don't know anyone that could detect that with their ears!
In no expert on drum micing, but I'm quite sure that micing drums or cabs from very specific positions is what gives you a certain sound, specifically when close-micing amp cabs where position, direction & distance are crucial.
I could quite easily build a device like the VP-8 or new VM mixer I'm working on which isn't reliable. It wouldn't be dsp locked, it would be dynamically assigned to dsps by SFP. One problem with doing that would be that if you've got the perfect sound from your mic positioning, I couldn't 'guarantee' that it would sound the same every time you load the VP-8 or mixer. I could certainly *guarantee* that the non-dsp locked VP-8 or mixer wouldn't be sample aligned every time it gets dyamically assigned.
Consider that someone wants to process multiple 100hz sine waves. I know for a fact that most of us here could audibly hear a 1 or 2 sample difference in channel alignment manifested in audible phase problems
Quote:
The vp8 is 8 in 8 out indivdual isn't it? So why the concern that each channel must be perfectly aligned together. A 2-3cm difference would be about 0.1-0.2 ms which is nothing.
I'd rather a device like the VP-8 or a mixer was reliable in the sense that it performs EXACTLY the same (to the very sample) each time it loads. I'm not sure If I'd put my name to it otherwise.
Btw, I made some design adjustments last night & the mixer I'm building will now only require that you have 2 x 4 dsp cards min' instead of what I told Voidar above.