Schiit Audio Kara F Preamplifier, Gungnir 2 DAC, and Wotan Power Amplifier
- REVIEW
- by Steven Stone
- Jun 10, 2025

Back in the 60s, most stereos were bought and sold as complete systems. My first, purchased from NYC’s Sam Goody’s in 1967, consisted of a Garrard turntable, H. H. Scott receiver, and a pair of EPI 100 loudspeakers for the combo price of $479. The Schiit Audio GigaStack’s approach is similar, except you still have to add a source component and speakers. The GigaStack trio of components consists of the Schiit Gungnir 2 DAC, Kara F or Kara F+ preamplifier, and the Wotan power amplifier. For the review stack I chose the Kara F rather than the F+.
Using components from one manufacturer offers several advantages over mix and match. First, since that was how the components were conceived, designed, and tested, chances of a mismatch in terms of input and output impedances are eliminated. Second, if you do experience any difficulties, it’s far easier for a manufacturer to solve the problem since they are intimately familiar with the optimal way to configure all the pieces in the signal chain.
The Kara F Preamplifier
The $799 Kara F is a Kara with the addition of the Forkbeard module. From the outside, the Kara appears to be basically a Freya S with headphone capabilities. But the Schiit site claims that “the Kara kills the Freya S from all outputs, in all modes.” Well, okay, then. The technical reasons for the performance improvements include higher rail voltages, boosted from ±17V to ±32V, along with the new 48V power supply to support the new higher voltages.
Operationally the Kara F is quite similar to the Freya S. (I’m more than a little bit familiar with the Freya S, since I own two of them.) Inputs and outputs on the Kara F include two balanced and three single-ended inputs as well as one pair of balanced outputs along with two single-ended outputs. The Kara F supports three operational modes—passive, active gain 1, and active gain 5. In my system, I rarely changed from passive to active modes for additional gain. The 4-volt maximum signal level available from my reference DAC’s balanced XLR output was so robust that in my usual loud listening level the Kara F was still between 15 and 20dB down from unity gain in passive mode.
Levels control on the Kara F, like the Freya S, consists of a 128 stepped-relay attenuator accessible via remote control. I like this stepped volume a lot. Not only is tracking between channels at all volumes consistent, but it offers an aural cue that the volume change has occurred with audible clicks. While some might find the relay’s clicks while changing levels distracting, I appreciate having a sonic confirmation that the remote control’s level commands were successfully received.
The Kara F includes one single-ended ¼” headphone output. According to Schiit Audio its maximum output is 1 watt RMS into 32–300 ohms. It supported my hardest-to-drive headphones easily. With both the Beyer-Dynamic DT990 600-ohm and the Dan Clark Stealth, I never needed to raise the volume above –20dB in passive mode. Switching to a pair of sensitive in-ears, the Empire Ears Zeus, I had to turn the levels down to –60dB. If I turned up the volume to full, without any signal I could hear a faint hiss, but at normal listening levels the Empire Zeus was as quiet as a grave.
The F in Kara F stands for Forkbeard, which is Schiit’s first foray into iOS apps. It offers all the controls available on the Kara’s physical remote. It also duplicates the controls available for the Gungnir 2 DAC and the controls on the front panel of the Wotan power amplifier. For me the primary advantage the app has over the physical remote was that it supplied numbered dB level readings, so I could know exactly what the gain level setting was instead of guessing. The Forkbeard app even offers haptic control, so when you change the volume, it vibrates by way of feedback. I wish, however, that the app’s volume control increments were more precise. Single-step adjustment of volume is available via Forkbeard—just tap the left side of the dial to decrease volume, and the right side to increase volume. The Gigastack and Forkbeard combo also enable Visual Volume, which shows where the system will exceed its limits, in the form of yellow-and red-colored bars on the volume dial.
Since I have been using the physical Schiit remote for several years, it took me a while to wean myself away from it to the app. One particular ergonomic quirk of the Karra F hastened my migration—the Kara F remote control’s eye has a limited angle of acceptance below its own horizontal plane. When I held the remote at waist level from my primary seating position, the Kara F usually did not respond. I had to raise the remote to shoulder height to achieve a consistent reaction. Yeah, boohoo, but this was never a problem with the Freya S due to the different placement of the IR eye. With the IOS app remote control, no line-of-sight is required.
One feature that I found essential on the Freya S was retained on the Kara F—the silent switching between inputs. Unlike many preamplifiers which produce a click or momentary pause, the Kara F produces no sound whatsoever when switching inputs, which makes it ideal for matched-level instantaneous A/B switching between two source components.
During my listening sessions with the Kara F in the system, I could hear nothing to alert me that its presence was in any way altering the sound passing through it. I heard the same level of musical information without noise or distortion that I was accustomed to hearing though the Freya S. Does the Kara F sound audibly superior to the Freya S? To my ears they both produce stellar sonics…but the added functionality of headphone connections and the Forkbeard app certainly makes the Kara F a more complete component than the Freya S.
The Gungnir 2 DAC
Most DACs in the $1599 Gungnir 2’s price range utilize either a second-party Delta-Sigma DAC chip or an R2R DAC as their primary digital decoding device. The Gungnir 2 doesn’t go the AKM, ESS, or Crystal DAC chip route. Instead, the Gungnir 2 relies of what they call a “multi-bit” design called “Multiform,” utilizing a Texas Instruments DAC8812C x 4 with digital filtering performed by a SHARC DSP processor. Unlike many DACS, the Gungnir 2 allows for upgradability due to its replaceable DAC/analog card. Also, firmware updates can come into the Gungnir 2 via the Forkbeard dongle.
Unlike every other manufacturer’s DAC I’m familiar with, the Gungnir 2 has two USB input options. One is the Unison 384, which goes up to 384/32, while the other is for sources up to 192/32. I tried both and didn’t hear any major sonic differences between them. I suspect most users will employ the Unison 384 input, so that is what I used during subsequent listening sessions. I have made quite a few recordings using DSD, so I was disappointed to find that the Gungnir 2 does not support DSD (or MQA, for that matter). If your own library is heavily populated by DSD and MQA, and you don’t wish or have the ability to transcode into PCM (Roon does this automatically if configured to do so, which is how I addressed the issue), the Gungnir 2 would probably not be high on your list of future DACs.
Operating the Gungnir 2 via the Forkbeard app was a pleasure. It even displays the bit rate of your current music file. You can turn on and off the non-over-sampling feature, invert the phase, change the DAC’s digital input, and mute its output from the app.
I have to admit that I hate doing matched-level A/B comparisons between DACs. The testing is time-consuming, and the results are often frustrating. I mentioned earlier that the Kara F is ideal for A/B comparisons of DACs due to its silent input switching. I gave that feature a thorough workout, comparing the sound of the Gungnir 2 with my current reference DAC, the Gustard A-26. Their prices are within $100 of each other, and the Gustard has been my primary reference DAC for the last three months.
Using Roon’s ganged feature, I could send the exact same feed to both DACs simultaneously. Each DAC was connected to my network from the same network switch via continuous runs of Ethernet cable. I employ 75 feet of CAT 8 to the Gustard, which goes directly into its built-in Ethernet port. The Gungnir 2 was connected to a Raspberry Pi4B via one meter of Audience AU24 USB cable. The Raspberry Pi was configured in Diet-PI as a Roon endpoint, and the PI was connected to the network via 60 feet of CAT 6 back to the switch. That’s the two slightly different signal chains. While I would prefer if they had been identical, the lack of a network/streaming input connection on the Gungnir 2 made that impossible.
Before I could do this listening test, I had to critically match the two DACs output levels. After accomplishing that (not easy since often the level differences between two DACS are not 0.5dB apart, while most DAC volume controls are in 0.5 increments of a dB) After that, I finally began listening. My hope, when I do a matched-level A/B comparison is that I will hear routinely noticeable sonic differences. I hate to admit that hearing any differences between the two DACS after repeated tests using my own recordings and reference commercial tracks was, for me, impossible.
At my usual listening levels during the test, the Kara F’s volume level was set around –20dB. Add that to the S/N figures of either of these two DACs, which are both better than –120dB, and you have a noise floor that is nearly –140dB down! For me, and most humans I know, that’s in the inaudible range. Both DACs produced identical soundstages and were ferences. I hate to admit that hearing any differences between the two DACS after repeated tests using my own recordings and reference commercial tracks was, for me, impossible.
At my usual listening levels during the test, the Kara F’s volume level was set around –20dB. Add that to the S/N figures of either of these two DACs, which are both better than –120dB, and you have a noise floor that is nearly –140dB down! For me, and most humans I know, that’s in the inaudible range. Both DACs produced identical soundstages and were equally adept at precise imaging. Bass extension and tonality between the two were also indistinguishable. When I controlled the switch-over, I kept listening for even the most subtle sonic tells that would alert me to one DAC’s presence over the other. Near the end of my testing, I had my wife make the change on the Kara F from DAC A to DAC B, while I attempted to determine blind when the switch had been made, and if a switch had been made…I got it wrong every time.
The Wotan Power Amplifier
The $1999 Wotan power amplifier ranks as the heaviest component I’ve had to install in my system since I refurbished my pair of JL Audio f112 subwoofers. At 54 pounds, it’s a beast. Its power capabilities are worthy of its moniker, with 200 watts RMS into 8 ohms, doubling to 400 watts RMS into 4 ohms. It is a dual-mono differential design that Schiit calls its High Power Nexus™ circuit. This circuit has universal feedback that can be turned off, changing its gain from 26dB to 32dB and affecting its damping factors. With feedback the damping is greater than 50 into 8 ohms, with no feedback the damping factor is reduced to greater than 30 into 8 ohms.
The Wotan has not one, but two internal fans. When the Wotan is turned on the fans spin briefly. During my listening sessions they did not come on again, which isn’t surprising given that my Spatial Audio X-2 loudspeakers are sensitive and an easy impedance load. The Wotan is not completely silent, however. I noticed a low-level continuous hum from inside the chassis. The level was so low that at my listening position I couldn’t detect it, but it was always present. At the speaker’s drivers, the Wotan was absolutely silent with no hiss or hum even when the Kara F’s volume in the passive mode was at maximum and my ear about an inch from the drivers. For those who abhor noise, I suspect it would be very hard to find a system that has less noise at full output than the Schiit Audio Kara F in passive mode/Wotan combination.
For those who wish to experience a zero-feedback amplifier design, Schiit Audio has obliged with a button on the front chassis (and in the Forkbeard app) that lets you turn feedback on and off. I tried it and almost immediately went back to the universal feedback mode. Without feedback, the Wotan was too wild and wooly for my tastes. I immediately noticed the lack of speaker driver control when feedback was eliminated. I suspect there are some systems that might benefit from the no feedback setting, but mine was not one of them.
GigaStack Combination Setup
Setting up the GigaStack was easy and straight forward. First, I removed one of the two DACs connected via XLR to my Freya S preamplifier and replaced it with the Gungnir 2. After some listening, I replaced the Freya S with the Kara F. Finally, I removed one of my stacked amplifiers, the Orchard Audio Starkrimson Ultra, and then placed the rather substantial Wotan power amplifier on top of my reference Pass X-150.8, and placed my other current reference amps, the FOSI V3 mono amplifiers with Sparkos op-amps and the GAN power supply upgrade, on top of the Wotan. This arrangement made it possible to switch between power amplifiers in about a minute.
I realize that when most audiophiles change components they are doing so because they want or expect a change or improvement in their system’s sonic capabilities. I’m rather different. When I put a new component into my signal chain, first I listen for how it may have degraded, rather than improved, the system’s capabilities. Why? Because my room and source material have been refined and developed over many years to the point where the default sonics of my system are, for me, basically correct. And while there is always room for the possibility of incremental refinements, if I notice no audible change in the system sonic capabilities, I consider that a good thing.
Together, the GigaStack performed impeccably. Once I acclimated to the Forkbeard app, I preferred it to the supplied physical remotes. Not having to be concerned with the directionality of an RF remote is more than a little bit liberating.
The GigaStack Sound
The Schiit Audio GigaStack “house sound” is everything you would expect from a high-performance audio component, and in my system sounds like virtually nothing at all. By this I mean that every recording played through the system sounded like the recording was supposed to sound. On old mono tracks, the soundstage width was a pinpoint in space, while with my live concert recordings the soundstage had all the three-dimensionality that a carefully placed M/S microphone array can deliver. In terms of low-level detail, I had no complaints. On one of my reference field recordings, made outdoors during a Rockygrass Academy workshop, of Chris Eldridge and Bryan Sutton playing “Church Street Blues,” it was easy to listen into the mix. I could not only hear the two guitars clearly, but also the sound of the river 50 feet away, as well as the occasional tapping from the mandolin-making workshop 150 feet away. On my recording of the Boulder Philharmonic performing the Brahms Requiem, the chorus was above and behind the orchestra. Through the GigaStack system I could easily hear that the chorus was located above and behind. I could also hear the conductor Theodore Kuchar’s occasional exhalations.
Kuchar’s additional breath emphasis reminds me of an exchange between the Boulder Philharmonic’s conductor, Ozzy Lehnert, and J. Gordon Holt. After listening to our recording of the orchestra Lehnert asked, “Can you make it sound more forgiving?” Holt responded, “Only God can forgive.”
Summary
I’ll admit to never having had much use for blingy-looking audio gear. In my personal world view, gold is of greater value inside a component rather than on knobs, buttons, or ¼”-thick faceplates. So, it should be no surprise that I find the looks of the Kara F and Gungnir 2 to be fine and dandy. Their shared basic chassis design has a unique visual appeal that’s far superior to a plain aluminum project box, without tipping over into shiny metallic excess.
At the beginning of the review, I mentioned my first system from Sam Goody’s. It’s price, when adjusted for inflation to 2025 dollars, comes to $4593.35, which is only slightly more than the GigaStack’s current $4247 price tag.
Comparing that stereo system’s capabilities with the GigaStack is like comparing my first car, a 1970 Saab 96, to a current-generation Subaru hybrid. The GigaStack delivers a level of sonic value that far eclipses what could be had for the same money in my youth. Granted, all electronic devices have gotten less expensive, but high-performance audio has benefitted even more due to the convergence of parts quality, refinement of design, and economies of scale dictated by demand. Schiit is a prime example of high-performance audio sonics, packaged in a straight-forward manner and priced for the value-conscious music lover. So, what’s not to like?
Specs & Pricing
Kara F solid-state preamplifier
Analog inputs: Two balanced XLR, three unbalanced RCA
Outputs: One balanced XLR and two unbalanced RCA
Dimensions: 16″ x 8″ x 2″
Weight: 12 lbs.
Price: $799
Gungnir 2 DAC
Inputs: Coaxial RCA SPDIF, optical SPDIF, USB
Formats supported: PCM up to 384/32
Output: balanced and unbalanced fixed output
Dimensions: 16″ by 8″ by 2″
Weight: 12 lbs.
Price: $1599
Wotan solid-state power amplifier
Output power: 200Wpc into 8 ohms, 400Wpc into 4 ohms
Inputs: Balanced XLR and single-ended RCA
Input impedance: 47k ohms
Dimensions: 16″ x 13″ x 3.875″
Weight: 54 lbs.
Price: $1999
Complete system price: $4247
