Well, you could, instead of putting the delay on an Insert on the vocal track, use a Send on the vocal track, routed to a stereo Group, with the delay inserted on that, and then insert the Cubase Mix6to2 plugin on the Group, where you can invert the phase of just the Right channel, if you wish.
On the other hand, I’d be interested to hear how mono-compatible such a configuration would be (same with your original scenario ).
Ah, the old Mix6to2! More than just a mono button workaround for those who don’t use the Control Room! (Years ago you helped me to do that, I’ve still got that preset in there somewhere, I believe!) I had no idea/recollection it did phase reversal too, leave it to you to remember that!
On the one vocal track I did these two wideners on:
The pitch shift and differential delay collapsed to mono not great, but certainly better than a simple Haas delay.
The 2nd one collapsed to mono pretty well, to be honest. I guess the phase reversal protects against comb-filtering to a nice degree. FWIW, it’s been described as doing well when collapsed to mono, and I didn’t find anything to contradict that.
I also applied a LPF to both of the panned tracks, so maybe that helped the one that was simply delayed/panned hard Left to not comb-filter too much with the source track panned down the center.
3) It makes me want to go back to the panned pitch-shifting/differential delay method, and reverse the phase of one side and see if that works.
Not so much for mono issues, but comparing the two methods (again, I’ve only done this on one track):
The 1st method (i.e., the pitch shift one) had a definitely noticeable, but very slight (to my ears) instability in the stereo image … it seemed like it shifted around a little bit. It was so slight though that I could only hear it by isolating the track, as soon as I put some music to it, it wasn’t noticeable at all.
Also, the pitch shift method had a little bit less comb-filtering in stereo than the Ping-Pong phase reversal method, but that difference went away pretty much when I applied the hi cut filters to the panned Ping and Pong.
I was going to suggest Parrotspain’s approach, but he beat me to it.
But the thing I couldn’t get past was the overall concept.
Switching the polarity of the return is only going to change the comb filter cancellation/doubling at the frequencies that coincide with that length of delay.
In other words, the polarity switch might have a strong and predictable effect at say, .3 ms delay, since it’s nearly in phase with the original signal. But 30 ms and 60 ms, for example, is so far removed from the original waveform that the cancellation/doubling frequencies would be totally random. 60 ms might affect say, 200hz and its harmonics in a noticeable way, but 65 ms would comb filter an entirely different set of frequencies.
Rather than setting up some weird routing to see if you like the comb filter effect when you flip the polarity,
it seems even more simple to just increase/decrease the delay time by millisecond increments until you get a sound you like. There’s no magic way to eliminate the comb filtering, it’ll always be there at some set of harmonics, so just pick the ones you like.
BTW, this is very similar to the technique of delaying the various mics on a drum kit to avoid phase cancellation.
On a drum kit though, some of the critical resonances are stable, so that by sacrificing some frequencies, you can regain others. But it’s a real revelation to listen to the flangey effects that multi-micing creates, as you experiment with various delays.
You can waste many hours looking for the right combinations before realizing that it was just as good before you changed anything! (Don’t ask me how I know this)
And registered, thanks for that really thought provoking discussion. I came across the idea of the Ping-Pong 30-60 msec delays with polarity reversal in an old SOS article (May 2012, Geoff Smith, Creating and Using Custom Delay Effects, audio example 11g).
As what you say seems to make so much sense, I can only guess that the phase reversal isn’t so much for tonal control, but more for a widening effect (since the example was in a section of how to widen vocals). But to be honest, I can’t figure out why phase reversal of the 60 msec Pong would make it wider.
For what it’s worth, the 30-60/phase reversal Ping Pong on the only vocal track I tried it on a) did do a nice job of widening, and b) collapsed to mono a little better than some other methods I tried. When I get back home, I’ll just flip the polarity back and forth to check to see how much of a contribution that had.
Of course we can never know another’s pain completely … but I feel like I used up years of my life doing that with early reflections just this week. All I could do was make the overall tone and comb-filtering sound different by playing with delay times/EQ/etc. … in the end I could never convince myself I had found a way to make it sound better than just picking some reasonable delay based on virtual room size and moving on. Man, I could have done a lot of other things with those hours!
You wouldn’t happen to know if the SHEPPi plug-in .dll is 32 or 64 bit?
I’ve tried it in Cubase 6.5.5 64 bit, and Cubase crashed after a couple of minutes with this plug-in on the mix buss! …seemed to work fine, but when I tried to switch the presets…Cubase stopped responding so I had to force shutdown Cubase!
It sounded cool while it lasted though!
Oh yea… who ever thought that GUI background was a good idea needs a nap!
Can’t be positive, but I think I read it was 32 bit, so maybe a wrapper would help. But I like what it did a lot w/ just the “home-made” single 30msec/phase-inverted Pong. I gather it’s the phase inversion that widens it and makes it more mono-compatible, but I can’t confirm.
I remember playing with short period delays to create the illusion of extra stereo width many years ago on mixes.
The idea is that you generate a sound, let’s say a snare drum and pan it fully to the right speaker, then take a short period echo (no feedback) and pan it to the left speaker. When the delay is set around 15 -25 ms at a suitable level. The human ear/brain detects this as a room reflection rather than a sound source and re-calculates the position of the original snare sound a little further to the right ie. beyond the loudspeaker. The delay is not then interpreted as a separate sound.
If you then set up another similar delay panned right for sounds coming from the left, you can repeat the effect in reverse You should end up with a psycho-acoustic increase in width of your soundscape on your chosen sounds. This can be set up as FX sends and applied to taste.
You may need to adjust the delay times for minimum phase cancellation for reasonable mono compatibility as others have commented.
In recent decades, I think we have all become accustomed to sound sources that have been chorused or pre-processed into lush stereo soundfields and maybe these older methods of mixing for width have been put aside…