Project Polaris

Ian is 100% correct.

The filter goes between the source and the amp and it creates the 'opposite' frequency curve as the headphone has.
So if the headphone has a +3dB hump of a certain width than the filter has a frequency response that shows a -3dB dip with the same width as the hump.
Add both the +3dB of the headphone and -3dB of the filter and you end up with 0dB so compensated.

As ALL headphones wobble around and have peaks and/or dips of various dB's all headphones thus need their own 'opposite' curve to become 'flat'.
This is what the modules do.. they define the opposite curve.

Of course you can come close by using EQ and pull up and down various sliders but it will never be as exact as the 'exact' curve.
A filter can come closer with good headphones.

The Ember and O2 (as well as all other amps out there) all have equally flat frequency responses.
What Ian says is correct... you often read amp A has boosted bass or suppressed mids or whatever.
It is complete bollocks they are all equally flat (if well designed)

They can still sound different for 2 reasons.
1: output resistance interacting with headphone impedance creating a frequency dependent voltage divider (forget about the damping factor nonsense)
2: Added harmonics.

See it like this...
A pure sinewave of 70Hz has an amplitude of lets say 0dB and lets assume that is 1V
A pure sine wave has no harmonics.
Now we have a tube amp that puts out the same 70Hz tone also at 1V
It ADDS harmonics so the first harmonic = 140Hz and is at -40dB which is a factor 100 smaller in amplitude so 0.01V
The total output voltage is thus 1.01V and consists of 70Hz + a small signal of 140Hz
When you will be looking at the FR plot of the non distorting amp (lets call this the O2) than it will show 0dB at 70Hz (1V = 0dB).
When you will be looking at the FR plot of the distorting amp (lets call this the Ember) than it will show 0dB at 70Hz (1.01V = +0.08dB) BUT a plot is created referenced to 1kHz and 0dB.

at 1kHz the amps do the same and also at the other frequencies so from 20Hz to 20kHz both amps will show the same.
with the same input signal (a pure sinewave) the O2 will always give 0dB and the Ember always +0.08dB so both are equally 'flat'
When we make a plot of the Ember and want 1kHz as 0dB than the input signal is lowered by -0.08dB so at 1kHz it shows 0dB and also at all the other frequencies.

BUT... here is the kicker.
The O2 will only let you hear the 70Hz and the Ember will let you hear the 70Hz + a MUCH softer 140Hz but you'll hear it non the less.
Just not as 2 separate tones but as one tone with a more natural 'warmth' to it.
When the 70Hz and 140Hz were not related and would be 'free running' you would hear 2 tones.

So both amps are equally flat but because the added harmonic signal is VERY small in amplitude it won't SHOW on the plot (well as a marginal raise in amplitude) but your ears will hear it and measurement equipment that shows THD will show the signal and an FFT plot will also show it.

Now when you have 'corrected' the anomalies of the FR of the headphone it will be tonally correct but the Ember will still add the harmonics (tube goodness).

Of course it should be noted that when a 70Hz (or another frequency) pure tone is played through a headphone that headphone itself will also add quite a few harmonics. The amount of harmonics is frequency dependent as well so will differ from headphone to headpone.

The HD650 has exceptionally low distortion b.t.w. and once FR corrected it will thus sound 'better' than some others.
The 'added tube goodness' will be faithfully reproduced by the headphone and is thus audible.

You explain things SO clearly Frans.

Your description kind of reminds me of additive synthesis as in the old Yamaha DX synths, except the frequencies are added above the fundamental which creates a very harmonic laden and possibly 'jangly' sound that is so prevalent in the 80s.

Wow!

I do actually understand what you're saying there. I sometimes feel that a lot of so-called experts barely know any more than I do myself. Not so with you Frans. It's awesome that you're prepared to share that kind of knowledge with folk like me. Many wouldn't take the time. Kudos to you my friend.

And yes, MASH. That was it. I think it might have been Philips who used the bitstream tag. I was the first to get a CD player among my friends. I got it on my 21st birthday, lol. A Denon DCD-1100 I think. It was awesome at the time. My local Woolworths had 3 CD's. Brothers in Arms, by Dire Straits, Zoolook by Jean-michel Jarre & a rendition of Tchaikovsky's 1812 overture and Marché Slàve with REAL CANNONS!

I bought all three.

Gordon.

As do I (on Facebook I mean)

Yeah mate, it kinda puts us islanders to shame.

So there should be no audible difference between solid state amps if they are design correctly and have low output impedance? What about the speed and the other amp characteristics besides FR, should they affect what we are hearing?

The FR of most amps nowadays is pretty flat. I find it very difficult to differentiate between amps and when I read descriptions of how an amp change the sound drastically, I always suspect something is amiss. Generally, I find many overstate the differences massively.

Solid state amps are amazingly close to each other. Most react very fast as well. You'd need a quick ear to hear any differences. We just don't hear as fast as we think. Op amps are way faster than us.

Generally, in an amp, I listen out for noise issues, volume pot travel and how it relates to volume, connections being good, and most important .... Enough power to supply proper transients and dynamics on recordings that actually contain that kind of information. Unfortunately, too many recordings have been processed to death so that the dynamic range is squashed to pulp.

On some reviews, the reviewers write what they have listened to. If you check the recordings, you often find that they have been compressed terribly, so when the reviewer goes on about massive dynamics, then you know that something is really amiss.

I'd say that most modern ss amps are pretty similar and the last fraction of quality comes with huge expense. Getting distortion lowered, noise level or noise floor minimal, but more important, supplying a decent amount of power comes at quite a cost and unfortunately, many people might well not be able to tell the difference in spite of what they write.

In fact, I doubt whether they really do hear those differences.

As Frans eloquently put it....... Complete bollocks!!!!