Phase-Invert Difference with EQ+

[spwilson27]

Has anyone noticed this oddity about EQ+ with regards to it's phase-inverted signal?

1. Create a phase inverted copy of a source sawtooth by creating two fx layers, and on one of the layers use the built-in tool device to invert both the left and right channels.
2. Insert EQ+ in one of the channels.

Even with no filters added, ED+ will cause a phase shift and distort the signal enough that the phases between the layers no longer perfectly match. This doesn't happen with pro-Q3.

What does this mean EQ+ is doing - even without any active filters?

[spwilson27]

The resulting merged signal does look like comb-filtered in the spectrogram. Perhaps EQ+ is adding some un-compensated for delay?

Edit:
This could actually be a pretty sever bug in implementation if so, it would me that simply by adding an empty eq onto a channel that you are introducing comb-filtering.

Edit 2:
Using the track delay device I moved the EQ+ ahead ~.11ms and now the comb filtering almost perfectly cancels out. It looks like EQ+ is in-fact adding some amount of delay?

Edit 3:
Bouncing a track with 100+ EQ+'s delayed the track and modified the timbre. Maybe there is some sort of all-pass or DC Offset removing filter active?

Edit 4:
Reached out to support to check if they knew about this:

yes, the EQ+ introduces a few samples delay. As it is not a phase linear EQ in general, it is not intended to be used in parallel tasks.

Seems odd that it introduces delay without allowing for the daw to compensate like it does for other plugins, especially when the EQ doesn't have any filters, but guess it's their decision.

[ReleaseCandidate]

Yes, this is a known problem: viewtopic.php?t=561412

It should be documented that EQ+ does not compensate latency.

[_leras]

Yes, this is a known problem: viewtopic.php?t=561412

It should be documented that EQ+ does not compensate latency.

This.