Latency comparison: 8000 versus eclipse

[thehorse]

Can anybody provide exact figures on how the h8000 compares to the eclipse in terms of latency.  I am primarily interested in the difference in a purely digital setup (ie. spdif, adat).  It seems a lot of folks use the eclipse in a live rig, which suggests the ADA total time must be reasonable.

I am contemplating move to good digital effects, which my RME UC handles very well, but I have many years of zero latency analogue experience.

I know the 2016 in and out is less than 2ms, for example.  I also know that the factor stomps are pretty fast.

thanks.

[nickrose]

We don't publish exact figures for this, but as a rough guess, with analog I/O the latency is defined by the A/D/A converters, and is probably in the region of a few mS. With digital I/O the latency is determined by internal buffering and (depending on the effect) is probably between 4 and 20 samples, or from about 80uS to 400uS at 48k sampling.

[thehorse]

Thanks!  Indeed, I use my RME UC for all my ADA and it has both spdif and adat, so I am most interested in how many samples to get in, modify, and get out at 48k.  From the sounds of it, there is very little measurable latency in terms of digital I/O and algorithm processing (assuming that complicated pitch shifting likely requires several ms).

One follow up: if you were to run a stereo signal into the H8000 and eclipse (no ada, just digital I/o) and use both of their respective DSPs in series using the same effects (eg. dsp A: modfactor alg; dsp B: timefactor alg), would there be any difference in total digital latency (I/O and processing) between the eclipse and H8000?

cheers.

[nickrose]

thehorse:

there is very little measurable latency in terms of digital I/O and algorithm processing (assuming that complicated pitch shifting likely requires several ms).

True, except that pitch shifting usually has a minimum of about 20mS (it has to buffer a few complete cycles of the audio).

thehorse:

would there be any difference in total digital latency (I/O and processing) between the eclipse and H8000?

There should not be a big difference, but they are different units so will not be exactly the same. Note that any given effect may introduce delays which will appear as latency, and even when the units are running the "same" effect, there may be differences.

[Author]

Posts