latency free monitoring?

edmann · Post by **edmann** » Sun Oct 23, 2005 9:39 am

Greetings Everyone

I have a SCOPE card, and am wondering if there is any provision for "latency-free monitoring", which I presume would bypass software (?) during recording.

I need to live-record 8 channels and hear it all realtime as it is being performed...w/o an external mixer.

I admit to knowing nothing about this as related to using this card - which BTW I love

All input welcome, thanks in advance

Ed Mann

astroman · Post by **astroman** » Sun Oct 23, 2005 9:48 am

yeah, sometimes one doesn't see the forest because of the trees...

the Scope mixer's out IS latency free and has always been, including FX
most other manufacturers only provide a dry bypass for this, but we get the real story

cheers, Tom

edmann · Post by **edmann** » Sun Oct 23, 2005 9:55 am

wow..I see. Do you know if this includes A/D sent in thru ADAT litepipe?

at0m · Post by **at0m** » Sun Oct 23, 2005 10:07 am

Yes, all IO are latency free - or almost, there's a sample or 2 off - inherent to the ad/da conversion. It works the same on any digital mixing console, dsp's being hosted in a dedicated console or on a PCI card.

edmann · Post by **edmann** » Sun Oct 23, 2005 10:09 am

thank you thank you!

cheers

Ed

Immanuel · Post by **Immanuel** » Mon Oct 31, 2005 1:17 am

There is more something like 2ms seconds of latency in both the ad and the da converter. This is due to oversampling. Old converters where much faster, but speed came at a price in sonic quality.

astroman · Post by **astroman** » Mon Oct 31, 2005 2:34 am

could you elaborate, please ?
oversampling isn't interpolating over a certain period, but processing at a higher clockrate.
The number of units may increase, but the timeslices decrease, so at the ends it's probably on par.

I only measured the time to send a Scope signal to a digital mixer via adat and then back into scope, which gave me 20 samples delay, equivalent to 5 samples per adat pass.
dunno how to measure conversion speed

cheers, tom

Immanuel · Post by **Immanuel** » Mon Oct 31, 2005 3:16 am

There was a thread at REP about it once in the Dan Lavry moderated section. I could be wrong about the reason, but I remember it as oversampling being the one.

You can measure the delay in your entire converter chain by sending a signal from Scope out thru analog (or your external converters analog) then back thru one or the other analog and record it simultatiously with the original signal within Scope. Check the waveform and count how meny samples the doubleconverted signal is behind.

edmann · Post by **edmann** » Mon Oct 31, 2005 5:45 am

does delay correspond to sampling rate?

alfonso · Post by **alfonso** » Mon Oct 31, 2005 6:21 am

On 2005-10-31 05:45, edmann wrote:
does delay correspond to sampling rate?

Well, if it's measured in samples, it definetly should.
I'm not an expert though, but I'd think so.

Post by **hubird** » Mon Oct 31, 2005 6:37 am

samplerate 44.1 = 44100 samples per second.
one millisecond = 44,1 sample.
f.i. latency of 5 ms = 220,5 samples.

at0m · Post by **at0m** » Mon Oct 31, 2005 7:30 am

Look, someone did all the calculations and measurements (allright, it was v1 hardware) and posted the results on http://www.planetz.com/Pulsar/PerformanceLatency.html !

If it may consolidate you, the delay is inherent to any format conversion, at least caused by the buffers of the translator. It will occur in any hardware device, be it a large Tascam ADAT console or a pulsar daughterboard.

Enjoy!

Immanuel · Post by **Immanuel** » Mon Oct 31, 2005 2:42 pm

Converter latency is related to sampling frequency - yes. Higher frequency = lower latency.