Schiit Audio



What an incredible sounding little device! Plenty of power for my Hifiman 400i’s and great build quality. There really is a difference with the tubes, I wasn’t sure with it only being a hybrid, and being SO cheap, but I thought ay worst, back it will go for an exchange for the Valhalla. My only complaint is, while all metal contraction, it’s still small and lightweight. A heavy cord will move it on the shelf. But I won’t exchange it for a “bigger” unit!!
Highly Recommended and help support a couple of great semi retired guys!!


You know I’ve always had that complaint with my Schiit stack but never heard that come up in any reviews. It is an issue especially since the tops of the units are slippery and the feet they provide don’t help much. Have to be very careful when plugging and unplugging things.


It probably doesn’t come up in reviews because it’s common to all small audio components. When the device often weighs less than the cables it is connected by, you pretty much have to put your hand on it when making/breaking connections or it’s going to move.

Now, if you want to reduce the potential for the things (be they Schiit, JDS Labs, Topping etc.), to slide around when just sitting using them normally then, other than using a more compliant cable, sticking an appropriately sized (cut with scissors/craft knife) piece of neoprene/rubber gasket sheet under each of the units will keep it pretty firmly planted.

You can either remove the feet, or use sheet thick enough to clear them, and things will stay put much more reliably.


I’m always too concerned with thermals to put large rubber sheets under the units, but maybe I’m too paranoid!


Regarding Schiit…I’m personally not a fan…

Criticism of the Magni 3 widespread across many forums. Following a good deal of research, the main issue appears to be aggressive clamping of the maximum input voltage. This compresses the sound–i.e., recreates the Loudness Wars with any source and degrades quality. The output improves if you turn down the input volume and stick to non-demanding headphones. But then, why bother with a headphone amp at all?


I only have a (now-discontinued) Lyr 2 amp from Schiit, but it does exactly what I wanted it to: provide a warm tubey sound, and provide enough power for even my hardest-to-drive headphones. I’ve had zero trouble with it as well. It gets plenty warm, but that’s kind of part of the territory with tubes in the first place. I just keep it well ventilated.

I guess I could complain about the money I’ve spent on NOS tubes, but that’s really a little more my fault than Schiit’s.


Are you referring to what’s discussed here. I’m not sure I get it, and I don’t think I’ve heard anything like this with my Magni 3 (fed either from a Hifiberry or an LG V20 at max volume).


Yes, I am. Compression is the sense of ‘flatness’ in recordings, and is very obvious too me. Specifically, songs that should go LOUD at certain points instead go f-l-a-t. No increase in sound, just the same as before. To test this, find an old 1980s CD with (for example) The Police. Compare it to a remastered year 2000 compilation CD from The Police. In the first case you’ll hear the audio peaks, in the second you hear loudness all the time.

[I used The Police as an example because I happen to have a 1980s CD and circa 2000 CD from them.]


I’m familiar with dynamic range compression, I just don’t hear my Magni 3 doing it nor do I understand the mechanism by which it would do it.


Aggressive protection circuitry for line-level input sources.


I think I’ll need to bust out my multimeter and take some measurements. From reading that thread, it seems like the Magni 3’s potentiometer sits between the input and the gain stage. The pot shouldn’t do any compression, just attenuation.


Input attenuation doesn’t cause dynamic range compression.

What’s happening is that your source’s output level is too high … and when you hit higher peaks in the music you’re going over the acceptable input level and it’s getting clamped.

Absent a fault with that Magni 3, the fix is easy … turn down the source volume. If you can’t, then it has a very abnormally high output level. In which case the only fix is a different amplifier that a) doesn’t attenuate the input and b) has sufficiently high input headroom that it doesn’t also wind up clipping the input stage.

This most certainly does NOT happen with “any source”. In fact outside studio/pro-level outputs I can’t see it happening at all unless you’re either amping an amp, or you’ve got variable level output and the source is turned up too loud. It definitely isn’t happening with any of the myriad sources I have sitting here.

It’s perfectly fine not to like the thing, or the company, but claims that it happens with all sources, which is demonstrably false, help no one.


For my own edification, what does this mean in electrical terms? Is the amp clipping peaks, or is something else going on?


I searched for any viable explanation for why the Magni 3 (and only the Magni 3) made everything I threw at it sound bad. One explanation is poor QC for my unit. But, similar critiques are widespread.


What source are you using?

Among the people I know that are typically using amplifiers (Magni 3 or otherwise) at that end of the market, a staggeringly high proportion of them are feeding them out of something that has variable level output …

It’s entirely possible that the unique internal topology of the Magni 3 means it is more aggressive in it’s input protection. That just means it’s not the amp to use with your source(s).

I have one sitting here that isn’t acting the way you describe fed out of any number of standard DACs. Though I can certainly make it act funny if I hook it up to my RME ADI-2 and crank up the output level on those.


I’d have to reverse engineer the circuit to say specifically. But from what I’ve seen described by Schiit in regard to this, there is input protection that applied at a certain input level - and when that occurs the input level gets attenuated (rather than clipping).

That’s a safe approach … putting a solid-state amplifier into hard clipping is a really effective way to destroy a speaker or headphone connected to it. Much more common to kill a transducer that way than by feeding it too much clean power.


I agree with your explanation. It’s an entry level product and must handle a wide range of use cases. And, I may well have had a bad unit, but I returned mine many months ago – no more testing possible.


All good.

I’m still curious as to the source(s) you used with it though. If I have one (or more) of them, I could test it myself directly.

I’ve certainly seen similar comments to yours elsewhere. But in all cases I’ve been able to talk to the person(s) involved, it’s turned out either to be too high an input level or a faulty unit. Both of which are issues that are readily remedied - even when the remedy is a simple “wrong amp for that job”.


[EDIT] I completely revised this post because I initially wired my test load incorrectly, with the load in parallel to where I was measuring voltage instead of in series. I’ve now corrected and re-measured. This approach also didn’t involve burning any resistors :slight_smile:

Disclaimer - I’m not an electrical engineer, nor do I play one on television, nor am I even a competent electrical hobbyist. Don’t try this at home, you might damage your equipment and/or cause a fire.

Since it seemed fun, I ran a little test :slight_smile:

First, I constructed a fake 22 ohm load using a screw-type 3.5mm plug and a 22 ohm resistor.

Then I played a 50 Hz sine wave test tone at -3 dBFS using Room EQ Wizard (nothing special about the program, but it allows me to generate test tones).

I then used a true-RMS multimeter to measure the output voltage between the left channel and and the resistor connected to ground, from a variety of devices:

Macbook Air at Max Volume - 1.33 Vrms

Macbook Air -> Magni 3 at Max Volume, Low Gain - 2.601 Vrms
Macbook Air -> Magni 3 at Max Volume, High Gain - 8.63 Vrms

Macbook Air -> Topping NX4 at Max Volume, Low Gain - 1.33 Vrms
Macbook Air -> Topping NX4 at Max Volume, Low Gain, Bass Boost On - 2.33 Vrms
Macbook Air -> Topping NX4 at Max Volume, High Gain - 3.60 Vrms

Macbook Air -> Topping NX4 at 2 Vrms -> Magni 3 at Max Volume, Low Gain - 3.97 Vrms
Macbook Air -> Topping NX4 at 2 Vrms -> Magni 3 at Max Volume, High Gain - 12.53 Vrms

Macbook Air -> Topping NX4 at 3.39 Vrms -> Magni 3 at Max Volume, Low Gain - 8.26 Vrms
Macbook Air -> Topping NX4 at 3.39 Vrms -> Magni 3 at Max Volume, High Gain - 15.13 Vrms

Then I calculated the gain for all of the amplifier scenarios scenarios by dividing the output voltage by the input voltage:

Macbook Air -> Magni 3 at Max Volume, Low Gain - 1.96
Macbook Air -> Magni 3 at Max Volume, High Gain - 6.49

Macbook Air -> Topping NX4 at Max Volume, Low Gain - 1
Macbook Air -> Topping NX4 at Max Volume, High Gain - 2.71

Macbook Air -> Topping NX4 at 2 Vrms -> Magni 3 at Max Volume, Low Gain - 1.99
Macbook Air -> Topping NX4 at 2 Vrms -> Magni 3 at Max Volume, High Gain - 6.27

Macbook Air -> Topping NX4 at 3.39 Vrms -> Magni 3 at Max Volume, Low Gain - 2.43
Macbook Air -> Topping NX4 at 3.39 Vrms -> Magni 3 at Max Volume, High Gain - 4.46

So on low gain, the Magni 3’s gain does not go down with input level, however on high gain, it does seem to go down a bit with input level, especially once you get past 2 Vrms in.

P.S. It was also interesting to learn that the NX4’s low gain is in fact unity (1) and that the bass boost at 50 Hz is nothing to sneeze at.



I’m pretty sure I’ve seen reference from Schiit that input levels up to 7v RMS are fine. Though the more I think about it, I also think I’ve seen them say that it isn’t actually input attenuation that’s causing the issue, but an interesting artifact of what happens if/when the unit does clip it’s output.

Obviously the best people to ask are Schiit themselves - who are generally very responsive to emails.