Topic: Fixing an off-center recording

[nulldogmas]

Ultimate Vocal Remover seems to be a bit better than iZotope RX in splitting vocal stems, and at least has multiple processing modes to try. Maybe give that a shot first to split off the vocals, then use iZotope to split the remaining instrumental stem to isolate the piano?

https://ultimatevocalremover.com/

[WiFiJeff]

Thanks for the suggestions.  I am beginning to hope that this is possible to fix, since I stumbled on two easy improvements.  The first was to play the recording on front stereo speakers instead of headphones, that immediately got the piano out of my left ear and closer to a realistic sound stage.  Then in Rx I tried a plugin called BinauralDecoder.  Before I realized it was intended to make a binaural recording from a first order ambisonic feed, I listened to what it did to my recording without loading any parameters, the piano sounded centered and great, the voices moved from center and right over to the left but overall the effect was nicer.  Only two concerts between now and month-end, so I'll play around and if I get a better result I'll try to figure out what I did to get it!

[robgronotte]

I've been saying this in several different threads for awhile now - the free program Ultimate Vocal Remover 5 is great for splitting vocals from music.
It does occasionally have trouble when some of the instruments are in the vocal range but I don't think this would happen much with only a piano for instrumentation.

[WiFiJeff]

I've been saying this in several different threads for awhile now - the free program Ultimate Vocal Remover 5 is great for splitting vocals from music.
It does occasionally have trouble when some of the instruments are in the vocal range but I don't think this would happen much with only a piano for instrumentation.

Thanks.  I have just downloaded this and will install it this weekend and see what it can do.  The earlier suggestions of fine-tuning the vocal stem separation won't fly.  I am working with a two and a half hour opera, not a five minute "song."  The Rx cleanup, which I am used to doing, took 8-10 hours of close listening.  It is also pretty clear when a cough removal is successful without hurting the music or having to re-examine the whole five acts.   

Meanwhile, I have an acceptable (to me, at least) fix.  After trying the Binaural Decoder with a phase-inverted right track to get the voices back over to the right more, I guessed that maybe the Decoder was looking for ambisonic WXYZ tracks and tried to use a mix of my left and right as the first (L) "mono-ish" input, the result of this was not so good.  But then I noticed that a straight mix of left and right as my left track, with the right track unchanged, gave a pretty good result.  I then messed with the RX phase plugin (a new toy, play with it!), I found that a 60 degree shift for the new left track seemed good (I am sure this is not optimal, I find it hard to compare results as what is better in one section of the music might not be so elsewhere).  After almost 20 years, I also have no memory of the staging so I can't be sure I am actually getting the voices where they should be on the stage, however the result has a stereo "feel" and is listenable for long stretches, so I am pretty happy.

Jeff

[Gutbucket]

That reminds me I need to check out Ultimate Vocal Remover.

Binaural Decoder may in part be doing something similar to the Mid/Side stuff, Ambisonics being formally extended Mid/Side essentially.  Your mention of the 60degree phase shift sounding right reminded me of something I read in the Q/A section of the Voxengo PHA-979 phase rotation plugin, which may or may not be relevant in this case-

"Q. PHA-979’s description says it can be used “to setup headphone monitoring so that it closer resembles sound stage produced by stereo speaker monitoring”.  That sounds interesting, but I don’t see how this functionality is implemented? A. A corresponding preset is available.  PHA-979 only models speaker phase positioning – 60 degrees phase difference between speakers.  It is not about time difference – the speakers are located at equal distances from the listener, so there is no delay between arrivals of signals from both speakers.  At the same time, phase difference between speakers is constant for all frequencies, it is equal to 60 degrees and it changes perception of the stereo field greatly."

[hoserama]

A friend recently sent me a file of a nice piano recital he taped from the extreme right side of the house, with the sound quite definitely coming from the left side of the headphones.  The average volume for the left and right track is quite close, in fact the right channel measures marginally louder.  Clearly there must be timing issues here as well accounting for the perceived placement of the piano off to the left (which it was).  Is there any suggested delay that might correct this, has anyone successfully tried this?

Don't think this has come up, but S1 Imager from Waves would be perfect for this. I've used it to center off-balance recordings multiple times.

[Chilly Brioschi]

iZotope moved Ian Stewart's Article:

https://www.izotope.com/en/learn/5-ways-to-adjust-phase-after-recording.htmlI

A detailed, but worthy read

[goodcooker]

That's a great article^

If some of the terms being used applied to stereo field and phase relationships aren't completely clear (and there's a lot of conflicting info and terminology out there) this article condenses some of the fundamental and often misunderstood ideas and makes them clear in plain layman's language.