QueenZone discussion #1459056

aristide1 wrote: Another return of the Golders Green comet? Smart of you to avoid the word "quadraphonic" this time. Although "center channel extraction filter" suggest you still have delusions about a quadraphonic source. In fact you've done some M/S equalization. The mid channel - the mono component of a stereo source - it's a derivative not a real center channel as you like to believe.

GinjaNinja wrote:

aristide1 wrote: Another return of the Golders Green comet? Smart of you to avoid the word "quadraphonic" this time. Although "center channel extraction filter" suggest you still have delusions about a quadraphonic source. In fact you've done some M/S equalization. The mid channel - the mono component of a stereo source - it's a derivative not a real center channel as you like to believe.

I don't claim that the recording is quadraphonic. It was suggested by another QueenZoner and though the consensus later became that it wasn't, the SQ decoding process still gave some interesting results. "Center Channel Extractor" is literally the name of the filter I used, go pick a fight with Adobe. Indeed, true isolation of the mid channel is impossible. I called it a "crude, mostly-vocals" track as that's exactly what it sounded like.

Mrmarioanonym wrote: aristide1, if you can do it better than ginjaninja, then do it, otherwise just shut up. you're always rude to people. it's like you're a little spoiled child.

aristide1 wrote: If you are aware the recording is not quadraphonic and the center channel is not real, then why all these decodings and extractions, instead of a classic equalizer (with M/S capabilities if you need)? You may obtain better results, or at least you can do something on your own, not only observing "interesting results" which are the unpredictable side effect of a failed decoding algorithm.

GinjaNinja wrote:

aristide1 wrote: If you are aware the recording is not quadraphonic and the center channel is not real, then why all these decodings and extractions, instead of a classic equalizer (with M/S capabilities if you need)? You may obtain better results, or at least you can do something on your own, not only observing "interesting results" which are the unpredictable side effect of a failed decoding algorithm.

First of all, by "all these" are you referring to my single share from over five years ago and an attempted improvement at one track? The "extractions" are nothing to do with quad decoding, as I'm sure you know, similar results can be achieved with any stereo track with centre-panned vocals. I didn't make an attempt at equalisation as the vocals and guitar occupy a similar range of frequencies, boosting both wouldn't really have helped to bring the vocals out. I also didn't want to risk ruining the balance of the sound with any extreme frequency adjustments. Fine, let's call the interesting results a product of phase cancellation, as that's all I really did. I've actually tried going back to Stark's remaster and using the center channel extractor (you can relax, no quad decoding involved with this one), the results are pretty much the same.

aristide1 wrote:

GinjaNinja wrote:

aristide1 wrote: If you are aware the recording is not quadraphonic and the center channel is not real, then why all these decodings and extractions, instead of a classic equalizer (with M/S capabilities if you need)? You may obtain better results, or at least you can do something on your own, not only observing "interesting results" which are the unpredictable side effect of a failed decoding algorithm.

First of all, by "all these" are you referring to my single share from over five years ago and an attempted improvement at one track? The "extractions" are nothing to do with quad decoding, as I'm sure you know, similar results can be achieved with any stereo track with centre-panned vocals. I didn't make an attempt at equalisation as the vocals and guitar occupy a similar range of frequencies, boosting both wouldn't really have helped to bring the vocals out. I also didn't want to risk ruining the balance of the sound with any extreme frequency adjustments. Fine, let's call the interesting results a product of phase cancellation, as that's all I really did. I've actually tried going back to Stark's remaster and using the center channel extractor (you can relax, no quad decoding involved with this one), the results are pretty much the same.

It's not a correct approach to reject equalization just because there is some overlapping of different instruments. If you look at vocal and guitar from multitracks, Killer Queen for example, you will notice the energy has a different distribution, although the range is more or less the same. I've taken the spectrum of the vocal track, compared it with the target track, Killer Queen from Seattle 77, obtained a transfer profile, then used it to produce the best approximation of the "vocal channel" for the whole recording, with the purpose of boosting. It is the same thing you want to do, but has nothing to do with center channel or panning. The point is the overlapping effect was minimum, and you should not worry too much about it. You may try the same thing here by comparing vocal spectrum from a mostly vocal part with the same part from a sbd recording (Rainbow maybe), then apply the resulting profile to the current track. It's almost automatic using a MeldaProduction module or equivalent software, and is quite fun to play with spectral profiles.