There's a puzzling invention. Hey, I guess it works if you have a quad setup, but your needle (?) is only mono-compatible...?
It would only be "needed" for quad discs played on mono radio in order to somehow preserve the original mix.
Jackpot, I’ve found the patent. Highlights from a quick read while my lad watches Bluey: Both channels are phase shifted by half the total shift. So for a 90° setting, one channel is shifted -45° and the other +45° I still wasn’t clear which channel was being shifted in which direction, but it’s possible the devil is in the extensive details. The shifting is only done between 30-30000Hz. Anyone else spot anything relevant? Might dig out my unaltered WAV files of Wheels of Fire this weekend and try a 45 degree rotation in each channel.
Couldn't wait, and had a quick tinker with White Room. I think +45 on the left channel and -45 on the right channel is the sweet spot, at least for that track - it removes the seasick effect just like rotating the right channel 90 degrees, but it feels like the vocals are a bit more centered. This also doesn't push the right channel up into clipping territory, resulting in two pretty balanced looking channels. I forgot how bad the original album sounded, I've only listened to fixed versions since I wrote the first post!
Checking this with the opening cut (Rain in My Heart) on the Sinatra record, and the vocal is definitely more centered; also, when I OOPS the track, while the vocals stay, it seems to be in phase (little to no reverb)? Do you notice this with the Cream tracks, @Alexlotl ? (Bluey, from the one episode I caught in the doctor's office, seems really a fun and charming show.)
I noticed the OOPSing didn't completely eliminate the vocal, but did reduce it a fair bit. I checked with my older 90 degree right channel rotation recordings and it was roughly the same. The general effects on the vocals are very similar to the old +90, but the soundstage feels better with the +/- 45. Bruce's vocals always feel slightly right-of-center on this album anyway, even in the tracks without CSG. Anyway, re-did all the affected tracks with the +/- 45 settings and giving it a first listen on speakers now. Bluey is a blessing and a curse; it's wonderful kids telly, and is very smart about gilding the philosophical pill, but on the flipside no real-life parent is able to live up to the standards set by those damn dogs! Amusingly, there's actually an episode about that in Series 2.
On Wheels of Fire the alignment of the CSG'd tracks is slightly off, and if memory serves the balance is slightly off on the entire album. Those will both prevent the vocals from completely dropping out when OOPS'd (although both can be fixed).
At least with Cream, you know some tracks are not CSG -- with Frank that info isn't available -- I might play around with my favorite track, #3 - Little Green Apples, as on some of my listens to the -90 degree right test, I notice a more marked difference after the first two tracks, likely due to sparser, more folk arrangements compared to orchestral pop.
It's a damn shame the WoF multis burned in the Atlantic fire. This album would definitely be a contender in the Albums that need a remix poll.
I finally got around to trying this method. The first track was "I've Gotta Be Me" from Steve's Sammy Davis Jr. "Greatest Hits" CD. Both the original LP and this track have CSG. I applied PhaseBug 90 degrees (sorry, forget whether it was left or right), and the voice is now centered and sounds right. There is a non-CSG version of the track on the Reprise CD "I've Gotta Be Me - The Best Of Sammy Davis Jr. On Reprise." That track and my newly-repaired track sound pretty much the same. Works for me!
Do try it again with the complementary 45 degree shifts I describe above, I think it gives a better result. I’ll update the original post when I get time.
Another note about 45-45. The above track sounds better with L +45, but the Sergio Mendes "Fool On The Hill" sounds better to me with R +45. Am I missing something?
Maybe the machine was hooked up backwards on one of them? It's supposed to change the phase as much as 120°, so - as much of a pain as it is - just experiment and find out what sounds best for each song.
I think this could be automated to find the best setting, based on what the original CSG processor was capable of doing I imagine a program which would, using a 15 second or so sample, step through degrees (which word I won't type out) of changing phase on one channel of -120, -105, -90, -75, etc. in 15 degree increments ending with +75, +90, +105, +120 (I think these 15 degree steps would cover all possibilities of the original unit) - and internally, and externally play, the sum to mono to find which setting gives the loudest mono. Let us say it automatically finds that -90 is the loudest. It then would indicate to apply +45 to one channel and -45 to the other, and done.
Yeah, as much of a pain as it is, you can save settings, so do +45/-45, +50/-50, +60/-60, and the inverse of each of those, and just toggle between them and decide which sounds best. Automating it would save the effort but would take longer. Besides, if you land on something really good before you test all of them, you won't need to do any further tests!
Or I could set up a bunch of samples or even full tracks, go have lunch, and the automated program would do it all to every full track............ Say you have a compilation master, or CD as described in this thread, 14 tracks, where you think you hear CSG on tracks 2,7,and 13, but you don't know what setting of CSG because each could have used different settings. Just feed all 14 tracks in and go have lunch. "zero" is one of the settings it would try, so tracks without CSG would find the zero setting as best and make no change to that track. Come back from lunch and all 14 would be processed correctly. There might be been a little ( +/- 30 original setting) CSG on tracks 9 and 10 that you didn't notice. The program would correct those too. This first shot is a brute force method - the program can be optimized to jump around in the list of settings to be tried, to narrow in faster on the best undo-CSG setting for a track. Then correct every track for absolute phase.
I just re-did my Sinatra Cycles CD with the 45 degree shifts, and it was definitely an improvement over my previous attempt using one channel with a 90 degree shift. Nice work and thanks!
Yes. There are a whole lot of different degree variations in the fourteen claims, some of which appear to use different offsets in each of three separate frequency ranges. However, the thing that was most apparent to me is that the main text and the supporting images describe a +60 -60 configuration, total difference of 120. There is no attempt to specify left or right - just Channel A and Channel B. Going to try out 60-60.
On both these tracks, I like R +60 L -60 better than my previous 45-45 versions. Mendes' piano moves full left, while the vocals are centered.
^ This. 90° one way or the other seems to be the norm. My one experience with a 120° track was a CSG'ed US single of "A Song Of Joy" by Miguel Rios on A&M.
I've just received my DCC gold Wheels Of Fire, and to celebrate, here's our host's mastering of "White Room" with the CSG removed. This song finally sounds the way it deserves to!
