The other potential explanation I’ve found relates to distortion profiles, and is here:
I think this Schiit is breeding when I’m not looking …
This just needs a pair of Sys, to allow easy swapping between sources and amps, an Eitr to bypass USB, and maybe a Mani and Jill for turntable/needledrop duties …
(Color differences between components are due to artificial vs. natural light combined with the position of the camera).
I did one more experiment to get a perspective on dynamic range compression. This time, I drove the Magni 3 from a Hifiberry DAC+ fed from Roon. Again I used a 50 Hz sine test file.
This time, I reduced the levels digitally in Roon and then measured the corresponding output value from the Magni 3. I ran 3 experiments:
Low Gain, Medium Volume (around 12 o’clock)
Low Gain, High Volume (Around 4 o’clock)
High Gain, High Volume (around 4 o’clock)
The relationship of input to output stays linear for the first 30 dB in low gain, low volume, and the first 50 dB otherwise. In both cases, the non-linearity corresponds to very low voltage readings that may be at the limit of my multimeter’s precision, so I wouldn’t read to much into it.
When looking at voltage, one would expect it to decrease by a power of 10 for every 20 dB, and that is in fact what we see.
From this, I would conclude that in my experiment, I’m not seeing any dynamic range compression.
Here’s a link to the raw data.
Dang that is a shrine to redundancy in an awesome way!
Since I don’t actually listen with a voltmeter strapped to my ears but rather headphones like the LCD2C, I decided to pull out my E.A.R.S. and measure the SPL from the headphones relative to digital attenuation in Roon. Playing full range pink noise and setting 0 dBFS at 100 dBA SPL, we get the following:
Until you get close to 50 dBA SPL, the slope of the line is a perfect -1. I was in a noisy environment, and it turns out that my noise floor is right around 50 dBA SPL, so that accounts for the poor behavior after that point.
I interpret this to mean that there’s no dynamic range compression here either, neither due to the potentially compressing amp nor the headphones which some people (like myself) have called macrodynamically somewhat restrained vs a dynamic driver like the HD58X.
Raw data is available in the last two tabs of this spreadsheet.
Macrodynamics, what is it and how can it vary?
Great work Percy. I don’t profess to understand all the in’s and out’s of what’s been discussed but it’s great discussion and I am really impressed at your experiments. Each day’s a learning day..
I have to agree with those who have reported about the problems with the Magni 3. Before I bought the Magni 3 I already owned a Schiit Lyr 2, Magni 2 Uber, Modi Multibit, Loki and Wyrd.
The Magni 2 Uber was not a very good product from any point of view. The only way it could be used was if you used it with headphones with a rolled off top end. The high frequencies were harsh and anything I tried with that amp turned out to be a very short listen.
The Lyr 2 is another amp entirely. It costs 4.5x as much as the Magni 3 and is worth every penny. I love the sound of a hybrid Tube driver/Mosfet output type amp. The Lyr 2 sounds like whatever tube you try. Its output is totally transparent and allows the tube roller a platform to experiment with whatever 6DJ8 type tube you want to try.
Currently I am using NOS Philips 7DJ8 tubes in my Lyr 2 and couldn’t be happier with the sound. As stated before, the Lyr 2 is discontinued but I have read that the new Lyr 3 is actually an improvement over the Lyr 2. I do not know how the sound could be improved but maybe one day I will try one. I am in no hurry.
But back to the Magni 3…
When the Magni 3 was released it got lots of positive press. There were many people who gave it glowing remarks, so I decided to try one. I placed the new Magni 3 into my Schiit audio stack (Wyrd, Modi Multibit, Loki, Lyr 2) and pulled out the Lyr 2.
The initial sound was harder and less involving. I reasoned that the amp was new so I gave it two months to burnin. After two months the initial sound softened a bit but it still had a hard high frequency character that made long term listening impossible.
During that two month period I tried several different headphones. I tried my older Sennheiser Momentums, Sennheiser HD 580, Grado SR 60, HiFi Man HE-400S and my Focal Clears. If the headphone had a rolled off high end, the Magni 3 was fine but when coupled to the HE-400S or the Focal Clear, the Magni 3 made the sound difficult to listen to for any length of time.
I did an A/B test between the Magni 3 and my Lyr 2. I realize that this isn’t a fair test given the difference in price but the Magni 3 was supposed to be some great giant killer. It is not.
The difference between the Magni 3 and the Lyr 2 is immediately noticable. The Magni 3 imparted a hard high frequency character that I have referred to multiple times and the Lyr 2 showed an immediate improvement of that problem. When I swapped between the two amps the sound was like an A vs B. Both amps had a wide bandwidth but the Lyr 2’s bandwidth was rendered without any harshness. It was suddenly possible to listen for long periods of time again when I put the Lyr 2 back in my Schiit Audio stack.
I wanted to like the little Magni 3 and gave it every chance to shine. It could not. My Magni 3 is now sitting it its original box but I will not resell it. I do not want to rip off any other audiophile. The Magni 3 is much better than its predecessor, the Magni 2, but it still sound too harsh in the high frequencies, whatever the technical explanation.
If you want a good headphone amp, I cannot suggest the Magni 3. It is powerful and can transform a weak and anemic output into a really robust sound. Unfortunately it comes with a harsh high frequency character that makes listening for long periods of time difficult.
When you listen to the Magni 3 you may not notice the harsh high frequency issue I wrote about. The problem shows itself as the inability to listen for long periods of time. If you A/B the Magni 3 with any high quality headphone amp you will be able to verify what I wrote about. The difference is obvious.
It is a shame. Like I wrote above, I wanted to like the Magni 3. What’s not to like? It is a very inexpensive headphone amp that has enough power to drive any headphone that you want to use. All of that for $100USD? Yup it is a pretty good product but when you compare it to a good headphone amp you will be able to tell what the difference is.
The bottom line is, do not try to drive a good, wide bandwidth, set of headphones with the Magni 3. If you use cheaper headphones, with a rolled off high end, then your should be OK with the Magni 3. It is great for the money but be realistic about its abilities.
I think this is a good example of hype being bad for everyone … customer/listener, reviewer and manufacturer. If people are coming off reviews and impressions with the expectation that an, entry-level, $99 solid-state amp is going to perform the same, or sound the same or better, as a $450 tube-hybrid amp, with 4x the power, then it is would either have to be one of an almost impossibly rare class of products (sometimes called “giant killers”) or the reviewer is doing themselves and everyone else a disservice.
This is one reason why I advocate using reviews as a means to narrow down things to an audition list, and then choose based on one’s own listening from there. And if no audition is possible, then I would generally take that item off the list.
Back to the Magni 3 specifically …
From the listening I’ve done here so far, which includes comparing the Magni 3 to the Vali 2, a Lyr 3, and a Jotunheim (just from the Schiit line-up), the Massdrop CTH and LCX, and then compared to my genuinely high-end stuff, like my SPL Phonitor X and Woo WA234 MK2 MONO, using headphones ranging from HD6XX up to the Focal Utopia, LCD-4 and Abyss AB-1266 Phi CC, it seems to be about what I would expect.
Which is to say, it’s a nicely performing, higher-than-average-power, $99 amp. I like it better than some of it’s immediate competition. Quite a bit better in some cases. Some of that is based on sound, some based on it not having a stupid layout. But mostly the sound.
And I can’t say I find it is actually “harsh” in absolute terms when taken purely on its own, even with the highest end headphones I know of - but it is certainly not as smooth or composed as, say, Lyr 3 in those cases (which isn’t really surprising).
Personally I’d go with the Vali 2 over the Magni 3.
But I wouldn’t be trying to pair either with high-end headphones for my own uses. The headphones may be the most impactful part of the chain, but I’ve always found a balanced approach to system building yields a much better end result than focusing too hard in one area.
The highest end amp with which I have experience is the Ember II, so my perspective is much more limited than yours. The one thing I’d add to what you’re saying is that synergy definitely seems to be a thing. I loved the DT 1990 with the Ember II at its 35 ohm output impedance, as do I love the RN QT2. Yet, I prefer the LCD2C from the Magni 3, perhaps because it’s solid state(?). The HD58X sounds best to me from the Topping NX4, which is also solid-state but based on IC’s rather than a discrete design and battery powered rather than mains powered.
So, long story short, it’s not just a question of price, but also of matching the amp’s design (solid state vs tube, discrete vs IC, low vs high output impedance, etc.) with the specific headphones you’re using.
In so far as you’re willing to use EQ, things get even more interesting, since you’re now trying to match amp with headphones and EQ profile, the last of which can (and should) be adjusted to work with the other components.
Synergy is an under-appreciated facet of system building in my opinion.
It can be the difference between “not for me” and “amazing”. Some things just do not work well together. And some things work unexpectedly well together - and become quite a bit more than the some of their parts.
In some cases that’s quite predictable - e.g. pairing low-impedance dynamic headphones with OTL tube-amps - or trying to drive an HE-6 off a cell-phone - neither case is going to turn out well.
Having spent a good while comparing Modi 3 and Modi MB and Magni 3 and Vali 2, and every combination there in, I was ready for something else …
Most interestingly for me, here, is that I have owned ALL variants of the Bifrost. The last time I had one around it was surprisingly difficult (though reliably possible) to distinguish between it and the Modi MB. That isn’t really the case anymore … it is much easier to tell the current unit apart from the current Modi MB. More about that when I get proper reviews of all these units posted, however!
I really like my Lyr 3. I have a few tubes that synergize with it quite well, if you want to borrow a few for testing.
I’d be interested to hear your thoughts on Bifrost - I’ve been expecting an update to it for over a year now, and it never seems to come.
We need a totem pole of all Schiit equipment on top of each other. The ultimate aluminum monolith of sound haha.
Which tubes might I ask? I have a bugle boy in my Vali right now which I am really enjoying!
I have a Westinghouse that is excellent, and a few others for flavor.
I was telling @Torq he could call it the leaning tower of Schiit lol… I’m looking forward to it’s final form… hopefully it doesn’t transform into a golem of Schiit…ok I’m all Schiit out
WOW Bifrost! I have a very old Bifrost here that’s waiting for the right time for THE update.
Schiit is doing very good stuff at decent prices but sometimes I’d like a slightly nice build qulity for their entry level stuff. The askew plastic volume pot of my Fulla 2 always bothers me. Still great item for the price.
This is a brand new Bifrost unit … as in it arrived new, on loan for review, about two weeks ago.
As mentioned above, the last Bifrost Multibit unit I had (and heard) was not that much different to the Modi Multibit. You could tell the difference, if you knew what to listen for, and paid really close attention, but even then the differences were very small.
This is a different kettle of fish - even compared to a new (acquired at the same time) Modi Multibit. They’re much easier to tell apart, and the differences favor the Bifrost Multibit. It’s closer to the original Gungnir Multibit than it is the last Bifrost MB or current Modi MB.
Is this another unannounced analog upgrade / B serial # / silent V2 firmware release kind of thing? (Note: Those things don’t bother me, it’s just hard to keep track of what’s what is all.)
I’ve been using the Mimby for my office setup, waiting for a Bimby design update before I pulled the trigger. I’d like to have matching chassis w/ the Lyr 3, but as you said - they didn’t sound different enough to justify the upgrade.
Hard to say, excepting that the serial number on this unit does start with a B. Which may or may not mean anything … given how long ago I last had a Bifrost on hand, let alone one that had just come from the factory.
All I can say for sure is that it has clearly discernible audible improvements over a Modi MB that was sourced at the same time.
It’s a shame there was so much ridiculous whiny-self-entitled-bitching the last time there was a slipstreamed revision (around the Yggdrasil Analog 2 upgrade announcement)- otherwise things would probably just be announced. The only ones that really know what’s going there are Schiit themselves and it’d be remiss of me to speculate.