Schiit Magnius Balanced Headphone Amp Review

[maverickronin]

And a remote (can't believe they charge extra for it btw), a digital touchscreen color display, a mobile app, additional XLR and RCA inputs, mono XLR out. Are there any other headphone amps with all of those bells and whistles?

A touch screen on a simple headamp/preamp is just silly. All it should really do is show the attenuation. Switching inputs would be much easier with dedicated buttons instead, especially when there are only 4.

they probably could have included a Modius board and still been less than 20% of HPA 4 price.

I don't doubt it would be possible to undercut Benchmark by skimping on the bling (see above...) but the Freya S was already 20% of the HPA4's price and only had half the steps of HPA4.

 

[Racheski]

A touch screen on a simple headamp/preamp is just silly. All it should really do is show the attenuation. Switching inputs would be much easier with dedicated buttons instead, especially when there are only 4.

But...but...it’s like all digital man...don’t be so analog. Yes I agree it’s a bit silly; seriously though is there anything else out there with similar features?

 

[maverickronin]

But...but...it’s like all digital man...don’t be so analog. Yes I agree it’s a bit silly; seriously though is there anything else out there with similar features?

Maybe the ADI-2 Pro if you only need 1 analog input?

I think that's as close it gets if you need that many inputs and a screen. The ADI-2's screen does actually do stuff though.

 

[o7_brother]

On the 300 ohm power graph, what are the power values for the THX amp and the Topping A90? It's hard to tell from the graph and they aren't written anywhere I can find.

 

[Jimbob54]

On the 300 ohm power graph, what are the power values for the THX amp and the Topping A90? It's hard to tell from the graph and they aren't written anywhere I can find.

Your best bet is to go to the reviews for those 2 amps- the graph here gets messy and a lot cuts off. Look at the cross hair readings bottom right corner on the 300 ohm graph on both reviews. They are both more than ample at 300 ohms for anything you will throw at them btw- especially with 4v balanced in and balanced out.

 

[o7_brother]

Your best bet is to go to the reviews for those 2 amps- the graph here gets messy and a lot cuts off. Look at the cross hair readings bottom right corner on the 300 ohm graph on both reviews. They are both more than ample at 300 ohms for anything you will throw at them btw- especially with 4v balanced in and balanced out.

I did, that info is not in those reviews.

 

[Jimbob54]

I did, that info is not in those reviews.

Nothing easy here !

Thats single ended out 196mw for the a90

The 789 is around 150mw - see the Monoprice 887 review- it shows it compared to 789 (basically the same performance too)

No idea what balanced in, balanced out is- some brain will chip in I'm sure

Bottom line- a90 "beats" the THX amps- they all deliver more clean power than the vast majority of users will ever need.

 

[solderdude]

I did, that info is not in those reviews.

From the plots (and calculated from the plots) in 300 Ohm:

789 SE: 200mW
A90 SE: 200mW
Magnius SE: 250mW

789 bal: 0.8W
A90 bal: 0.8W
Magnius bal: 1W

All fully comparable powerwise

 

[o7_brother]

789 bal: 0.8W
A90 bal: 0.8W
Magnius bal: 1W

But this very reviews says 762 mW for the Magnius, so that can't be right?

From the 300 ohm graph in this review, the order from least to most powerful is clearly 789 -> Magnius -> A90. Looks to me like like roughly 500, 800 (762), and 900 mW respectively, just from looking at the clipping points.

 

[solderdude]

But this very reviews says 762 mW for the Magnius, so that can't be right?

762mW is with no distortion at all, at the knee. See how the grey line (Magnius) rises to 1W ?
specs. acc. to Schiit:
Maximum Power, 300 ohms: 1000mW RMS per channel
But yes with distortion up to 1%
specs acc to drop for 789:
Output power: 800 mW, 300 ohms, <1% THD (watts per channel)

So it is a matter of choosing the maximum distortion opposite output power.

800mW to 1W is just 1dB difference.
For all practical intends these are equally powerful.
the difference between 762mW and 800mW is 0.2dB.

 

[JohnYang1997]

762mW is with no distortion at all, at the knee. See how the grey line (Magnius) rises to 1W ?
specs. acc. to Schiit:
Maximum Power, 300 ohms: 1000mW RMS per channel
But yes with distortion up to 1%
specs acc to drop for 789:
Output power: 800 mW, 300 ohms, <1% THD (watts per channel)

So it is a matter of choosing the maximum distortion opposite output power.

800mW to 1W is just 1dB difference.
For all practical intends these are equally powerful.
the difference between 762mW and 800mW is 0.2dB.

A90 is definitely 1W for 300Ohm with BAL out BTW.

 

[solderdude]

A90 is definitely 1W for 300Ohm with BAL out BTW.

Not doubting the specsheet nor you, the info was based on Amir's plots which showed 196mW for SE 300 Ohm which translates to almost 0.8W balanced.
I have no idea why it clips so hard and even slightly folds back where the specsheet says 250mW in SE.

Anyway... 0.8W or 1W is immaterial in practice. There are far more important aspects than the last dB in power levels one will likely never reach.

 

[JohnYang1997]

Not doubting the specsheet nor you, the info was based on Amir's plots which showed 196mW for SE 300 Ohm which translates to almost 0.8W balanced.
I have no idea why it clips so hard and even slightly folds back where the specsheet says 250mW in SE.

Anyway... 0.8W or 1W is immaterial in practice. There are far more important aspects than the last dB in power levels one will likely never reach.

Amir's measurement already shown 900mW+ power in the same graph. SE was a little off which should be 250mW. It's important as it's made in a comparison.

 

[solderdude]

I was going of the published plot in post #289.
Any idea why it folds back at 195mW to about 150mW ?
Seems unlikely the A90 will do this.

 

[JohnYang1997]

I was going of the published plot in post #289.
Any idea why it folds back at 195mW to about 150mW ?
Seems unlikely the A90 will do this.

The behavior (not the loss of power) is caused by the protection circuit in TPA6120A2. As we know the loop gain on A90 is exceptionally high it clips very hard.
I still don't know why it's 195mW instead of 250mW. But in post #1
https://www.audiosciencereview.com/...lanced-headphone-amp-review.15252/post-481766
It's 900mW+ with BAL out.

 

[BAMCIS]

The HPA 4 is stupidly expensive, but it's stepped relay volume control kicks any pot to the curb and makes a much bigger difference to usability and audibility than a few dB of already inaudible SINAD.

It's pretty cool that Shiit got 125 db SNR from such a simple design at such a super low price. I wonder if Magnius 'sweet spot' sounds as good or better than my Freya S and if the stepped relay volume in the Freya S is on par with the one in the HPA4 / LA4. I might get the Magnius just to have a single knob to control volume on my subs and try it on my main system as a pre amp. Would you say Magnius should sound better than any other pre amp that uses a pot, since none are anywhere near 125 db SNR?

 

[wyup]

I am curious to know if similar performace difference applies to line outs (RCA vs XLR). Say, if I had a Modius/Magnius stack connected by XLR, would that performance penalty still apply using Magnius RCA outs to connect to an amp?

 

[solderdude]

Yes, the 'penalty' however is at inaudible levels when using music in both cases so while technically measurable there are no sonic consequences.
It is only a very low amplitude 2H distortion of which many folks even claim it is even beneficial (tubey).

I really don't get the fuss about this.

 

[Zensō]

Yes, the 'penalty' however is at inaudible levels when using music in both cases so while technically measurable there are no sonic consequences.
It is only a very low amplitude 2H distortion of which many folks even claim it is even beneficial (tubey).

I really don't get the fuss about this.

I agree, though I think this comment in the review is why people are concerned:

“Hmmm. This is no good. We have headphone amplifiers clocking at SINAD (sum of noise and distortion) at 120 dB and here we are short of that by good bit at 94.”

 

[ElNino]

It is only a very low amplitude 2H distortion of which many folks even claim it is even beneficial (tubey).

I really don't get the fuss about this.

The problem isn't the relatively low level harmonic distortion... it's the IMD, which is orders of magnitude worse. The harmonic distortion results for SE are misleading.