TOPIC:

Hi I am considering the Flex for a studio monitor setup. Can anyone tell me the round trip latency in these situations


a) Analog in -> DSP (no Dirac) ->  Analog out
b) Analog in -> DSP (with Dirac) ->  Analog out

Thanks.

 

Hello,
3,52ms for Analog in -> DSP (no Dirac) ->  Analog out

thomasjacquot wrote:

Hello,
3,52ms for Analog in -> DSP (no Dirac) ->  Analog out

Is that something you've been able to measure yourself? Also, is this on a unit without Dirac functionality (as distinct from it being available but switched off)?

Thanks.

yes I measured it 2 days ago. it is a unit without dirac. just the minidsp plugin.

thomasjacquot wrote:

yes I measured it 2 days ago. it is a unit without dirac. just the minidsp plugin.

Thanks. I asked about the Dirac situation since I believe a unit being Dirac capable may increase the latency even when Dirac is switched off. I believe this is the case with the SHD series for example, although I've not been able to test this myself.

Please give me the answer

thomasjacquot wrote:

Please give me the answer

Errr, to what question?

A unit that has Dirac capability will have a bunch of latency......regardless of whether Dirac is enabled.
I have measured my SHD.  The minimum latency is 25mS.
www.minidsp.com/forum/minidsp-for-newbie...rocessing-time#57559

Dave.

With no fir , it’s low, very low, low enough you shouldn’t have any sync issues, like under 1ms low, more like a small fraction of a ms, like a few nanoseconds maybe….

with fir, you’ll max out at 10.33ms approximately with late centering or 2048t…
most likely with a 1024t filter you’ll be 5.33ms 

 

denver8me4dinner wrote:

With no fir , it’s low, very low, low enough you shouldn’t have any sync issues, like under 1ms low, more like a small fraction of a ms, like a few nanoseconds maybe….
 

I suspect that is incorrect and may mislead people, based on the measured example above and the performance of the likely very similar 2x4 HD stated here:

www.minidsp.com/forum/minidsp-for-newbie...e-product-line#38684

Unless you have some accurately measured data of you own to back your claim up?

What part is incorrect 

I’ve been using 2x4hd for years and the flex seem the same to me 

Used REW and made the flex the input and output….. ram an impulse and it was a decimal point in ms in IR offset….. 

so…….. now sure I follow you 

denver8me4dinner wrote:

What part is incorrect 

I’ve been using 2x4hd for years and the flex seem the same to me 

Used REW and made the flex the input and output….. ram an impulse and it was a decimal point in ms in IR offset….. 

so…….. now sure I follow you 

I think you followed me just fine. I referred to two examples of data that suggested what you said was incorrect but opened it up to you to provide data to support your claim.

I think I'd need to have a play with the USB loop-back in REW just to get an idea what this would pick up. It won't represent the normal use-case but I can see it might pick up the largest latency component.

The latency of the 2x4HD is 3.1mS.

I don't know what........."With no fir , it’s low, very low, low enough you shouldn’t have any sync issues, like under 1ms low, more like a small fraction of a ms, like a few nanoseconds maybe…."  means, but it sounds like complete nonsense to me.
More actual testing and less speculative rhetoric would be appropriate.

And, as I said, any unit that is Dirac capable......whether Dirac is enabled or not.....will have much greater latency.

Dave.

denver8me4dinner wrote:

Used REW and made the flex the input and output….. ram an impulse and it was a decimal point in ms in IR offset….. 
 

What did you use as a timing reference? Having a quick play with this approach using my SHD I can't see a way to try to measure latency this way, since either an 'acoustic' or loopback timing reference will have the same latency as the actual 'measurement'.

Right now I'm going to stick with miniDSP's official answer of 3.1 ms being right.

Oh, and it's 3.1 ms NOT 3.1 mS. The latter means milli-Siemens not milli-seconds.

Oh for crying out loud.  Are you in some sort of anal mode today?
If we're really going to get technical about it even "ms" is not correct.  The proper abbreviation would actually be "msec".

BTW, I'm not any sort of "official" source for miniDSP.  Just an experienced user who's actually measured the latency of many miniDSP products.

Also, you don't need a timing reference to perform this measurement.  You do it differentially..

Dave.