SET vs OTL

Is it possible that the reason SET tends to sound harmonically richer than a push-pull OTL amp is that the latter has a phase splitter that causes information to be lost?

After all, when we speak of 'SET vs OTL', what we mean is single-ended transformer-coupled vs. push-pull direct-coupled.

Ralph, I like your amps and have a lot of respect for you & your company, but to say that the <1% of 2nd-order harmonic added by a SET at moderate volumes on appropriate speakers, when studies have shown up to three times that amount is not audible, just doesn't make sense to me.

A good SET just gives me more 'chills of realism' than an OTL push-pull amp.

Mind you, I'm no musician and don't have perfect pitch.

I meant to say "but to say that the <1% of 2nd-order harmonic added by a SET at moderate volumes on appropriate speakers *is what causes this harmonic richness*, when studies have shown up to three times that amount is not audible, just doesn't make sense to me."

Atmasphere,
Paul's last question is one I share - does the phase splitter itself ever create
distortion issues (i.e. something like 'hysteresis') when the signal is re-
merged?

Second, I have always had questions about 'parallel' SET circuits, and
wondered whether or not one could, with the right tube, do a single-ended
OTL circuit which was 'accordion amplifier' circuit style (and if done right, use
more than two tubes). I expect this will be revealed to be a stupid question as
I am surely overlooking something obvious, but... I'll ask anyway... :^)

Has anyone ever used/made/heard a singled-ended 'accordion amplifier'
circuit? I read about this a while ago (TubeCAD Journal) but have never seen
anyone talk about one, though my gut is that this exists under another name
and I do not know about it.

 
When driven from a balanced source, I thought the dual differential Atma-Sphere circuit required no phase splitter?
 

Atmasphere,

regarding your point on fakeness and removing harmonic distortion... I understand that in an OTL / push pull amp, some of the harmonics will cancel-out showing lowered (measured) harmonic distortion. But if it is the even order harmonics that are being cancelled then the odd-ordered ones are left to dominate, the sound will be much worse/less natural even when measured harmonic distortion. Is this not an issue with OTLs?

As for SETs, the problem is (correct me pls) that even order dynamics can be accentuated by the tube's interaction with the transformer, making the sound more lush/"pleasing" but less natural/neutral.

Regarding my speakers they are Asa monitor pros (http://www.asa.fr/productsPages/gd_monitor_pro.php),
The monitors have a very flat impedence curve and are 8 ohms/ 89db.
They come with a pair of active subs that sort of make up on frequencies below 40hz (cut off point) for where my small OTL amp was lacking.
My listening room is not big and I usually listen to jazz.

Limited budgets mean that one usually buys used equipment from overseas, so it is often not possible to audition... (very hard to find used GM20 Graafs, but did see a Navison Audio OTL on Agon recently). Hence my question on whether a good SET (eg the Cary 300SE) would sound more natural than an OTL amp of similar price/quality.