What aspects of a tube amp/preamp contribute to "holographic" effect on soundstage?


Can someone explain to me is there any part of a tube amp or preamp that contributes to a more holographic image than another otherwise similar tube amp/preamp?

For example, listening to my PrimaLuna Evo 300 integrated - moving between triode and ultra linear mode. I have read that the ultra linear mode is more "solid state" sounding and has more linearity through the frequency range as opposed to the smooth finesse in the mid range of the triode mode.

However to my ears, using KT 150 power tubes in this unit, in the ultra linear mode I seem to hear greater depth of the layering, what I would call "holographic." The cymbals, piano, bass, vocals, etc all seem to occupy a more defined space front to back in the soundstage than in triode mode. Triode mode in this integrated seems to have a flatter presentation across the left to right soundstage - not as deep or into the room as ultra linear.

Perhaps this is unique to PrimaLuna since there is higher distortion (and higher gain) in the ultra linear mode it is presenting as more "holographic." 

When I owned a Luxman MQ88uSE, which I believe used a triode mode with KT88s, I remember there was MORE of a holographic effect - or a "deeper" soundstage than the Evo 300 in either triode or ultra linear mode. So it is hard to square a direct comparison between the two as to what would affect the deeper/more holographic sound of one amp over another.

Is there any known aspect of tube amp topology that can directly be related to being more holographic versus less holographic? Like KT88s will always be more holographic, or triode mode will always be less holographic, etc?

Thanks!
33c1efbf 2142 46a3 a4bb c4042a2db044Ag insider logo xs@2xjsqt
There are three things I can think of:
1) linearity- this is why triodes and ultralinear are more 3D, because they are more linear
2) bandwidth- phase shift can mess with your perception of the soundstage. In tube amps the only way to prevent this is by having enough bandwidth so that the audio passband is unaffected. So 10KHz and below is where the soundstage information exists; to prevent phase shift in this region you need 100KHz bandwidth. This is easier to get with triodes and ultralinear due to the reduced output impedance.
3) Noise is an issue. This can be caused by ground loops within the circuit, resulting in a noise floor that is made up of harmonic and inharmonic information masquerading as noise. So how the audio circuit is grounded can affect its ability to reproduce low level detail, which is important for the soundstage and depth in particular.

I should point something out though. Just because you have a 'triode' mode and 'ultralinear' mode does not say that the amp is doing either of those proper justice. This is because triodes usually have a lower output impedance, so the optimal load for them is different. They also have less gain, which will affect how much feedback is actually occurring. Bias points are different too. In a nutshell, unless the designer really did their homework, compromises will likely overshadow the results. So if I were you I would be hesitant about drawing conclusions from such a small sample size.


If a particular amp sounds good to you, that is what is important.


Thanks @atmasphere this all makes sense.

I guess I hear on different recordings that the different settings sound better. Really record-dependent. Triode sounds better on a flatter and more balanced/produced record like a lot of the 80s/90s audiophile pop/rock (Dire Straits, Toto, etc) and ultralinear seems to sound better on sparser jazz arrangements like Bill Frisell’s Valentine or Dominic Miller’s Absinthe.


everything... the whole design and its execution

not to mention the unit’s interaction with other gear in the chain (most notably the speakers)

and most importantly... the placement of the speakers and room it is playing in
@jjss49 I understand where you are coming from but I am talking about all things being equal - the speaker placement, the system, etc - just swapping out one tube amp for another and leaving everything else the same, one amp has a more holographic image than the other - so what is it about the differences in those amps that has that effect.
99% of soundstage is recording dependent! Recordings, analog or digital, vary widely in the spatial aspect one can hear! 
@jsqt 

it is still the entire design and execution of the whole amp and how it interacts with the speaker load... across different amps, you cannot isolate one component or part or aspect versus another... everything works together to produce the sound

in a few amps that are designed to run different types of power tubes (like primalunas, quicksilvers etc) the user has an opportunity to swap tubes, or switch UL/triode to hear a difference, but how do you compare the vastly different sound of a primaluna running kt88's to an audio research vt100-2 running same tubes with different power supply, driver circuit, bias circuit, transformers, transformer taps, internal wiring, rca vs balanced connections, chassis damping?  it all matters

sorry for the inconvenient answer but this i believe is the reality