The article also shows the need for substantial subsonic filtering.
I agree with that.
The FM effect he describes made me think about warp being mostly mechanically vertical and electrically L-R or Side.
This goes back to an observation I made earlier...
On 'phones it seemed that simply steering warp to mono improved clarity because it was eliminating the "out-of-head" pressure differences.
The warp energy was there - it was just moved to a different place.
Maybe there was some FM, but on phones the FM component shouldn't be that audible because the displacements are small.
Holman's example however was with speakers having larger cone displacement.
On speakers, steering warp to mono improved clarity but I wasn't hearing out-of-head pressure differences either way.
Something about it just sounded better despite the warp still being there.
L-R warp reproduced on speakers would seem to have double the cone displacement - or double the FM deviation - than mono warp because the cones are moving in opposite directions.
I'm going to listen to this again to understand why moving the subsonic to mono sounds clearer than having it in Side.
For whatever reason simply steering warp to mono makes a big difference.
Is not the best solution though.
Substantial sub-sonic filtering of L/R seems better.
Even better than that IMHO is aggressively HP filtering Side since the warp energy seems to be concentrated in L-R and the musical energy - due to elliptic EQ and vertical crossover - isn't concentrated there.
By HP filtering Side (vertical) and leaving Mid (lateral) untouched the effects of the filter would seem to be less.
This is what our member Eric proposed some time ago.
The more I look at it the more it makes even more sense.
Finally, what the article doesn't look at is the effects of ticks & pops which often give transients well above Fig 1
Yeah, they sure do.
10 to 15 dB or more.
The amount of headroom required when recording flat is already a lot.
Add a click on top of that and it's easily capable of hitting the guardrails.
How clipped clicks affect de-clicking is something I haven't tried but on my flat test transfers I was conservative to not to clip them.
I was glad I found this article.
I didn't save all my old back issues of Audio but I must have kept it for this one article.