HIFIMAN EF400 Desktop R2R DAC and Headphone Amplifier Review 7

HIFIMAN EF400 Desktop R2R DAC and Headphone Amplifier Review

Value & Conclusion »

Setup, Customization, and Performance

Desktop Setup


The HIFIMAN EF400 is extremely simple to setup given there is only a single USB input to handle, and keep in mind that you can't plug in both of the USB input connections at the same time since there is no input switcher here. They are just there to give you two different options to choose from, be it Type-B or Type-C. Then connect the AC power cable and turn on the device to see one of four operating modes lit up on the front. In my case it was set to High Gain/NOS out of the box and you can easily rotate the left knob to choose the other three functions. There is a plastic protective film over the black section on the front that you may want to peel off at this point but I left it as-is to minimize scratches to the glossy panel. Using the knob to go between the different states will also result in momentary audible clicks but there should not be any persistent clicking/hissing otherwise. I saw reports of some customers having said issues and those seem to have being resolved simply by turning the EF400 off and on again or actually necessitating an RMA. It does seem to be an isolated issue and I had no problems with this unit.


Using the HIFIMAN EF400 with Windows 10 or 11 doesn't require any additional drivers, with Windows recognizing the device automatically. We see that the default firmware provides PCM playback options up to 32-bit, 192 kHz and, from what I can see anyway, there does not seem to be native DSD playback as opposed to DSD being first converted to PCM. This puts the playback support of the EF400 behind less expensive offerings from the likes of FiiO, TOPPING, S.M.S.L etc that use a combination of an XMOS USB interface and a third-party DAC from the likes of ESS/AKM/Cirrus Logic. This is also where the OS and NOS modes available here come in to where NOS would basically just pass on the exact playback rate chosen to the headphones outputs, say, without any oversampling in place and only analog filters in the way. OS mode doesn't seem to be really worth it on this R2R DAC topology to begin with, but you are welcome to try it out since it has the potential to allow for vocals to be more forward facing as a result of slightly more cramped mids compared to the NOS mode. That said, we do see the integrated USB decoder still allows for all your favorite players to recognize the device in WASAPI (Windows Audio Session API) mode accordingly. There should be equivalent drivers pre-configured in newer macOS builds too, but I am not familiar enough to speak on this front.


If you have a device that doesn't accept UAC 2.0, or even if you simply want to try other options, then there are audio drivers available from HIFIMAN. As of the time I write this review, they need to be procured by emailing HIFIMAN customer support rather than simply finding it on this page where I would normally expect it to be. I did contact HIFIMAN about this and was provided this Google Drive link to the installer that you may want to use too, and they did say they will inform the relevant people about hosting the drivers on the service page directly henceforth. Note that the installer file mentions the use of SXW USB audio drivers from 2021, and installation is trivial as seen above. If anything, there are no options to customize the install location, desktop/start menu shortcuts etc!


What it then does is create a new output device in Windows although there is no USB audio control panel as with the likes of those devices using XMOS drivers. The playback options are unchanged in Windows, although you do get ASIO (Audio Stream Input/Output) mode for use in your media player of choice. I demonstrated above with JRiver again and set the audio player to the same output. At this point, you can implement further tweaks, but those are player dependent. In practice I would not really bother with these drivers unless your client device fails to see the EF400 or requires drivers for whatever reason.

Audio Performance


The HIFIMAN EF400 is a weird thing, to say the least. It's the only source I've tested to date that has a different sound signature from the different headphones outputs even with the exact same set of headphones! I had noticed a lot of inconsistent reports from customers as well as other reviewers to where it took for someone to put the EF400 through objective testing to see what's going on. In particular, note the output impedance of the four ports on the front that are all measured in the few ohms as opposed to being well under 1 Ω for other sources we've tested to date. The 3.5 mm single-ended output was the biggest culprit by far to where I really can't recommend the use of IEMs with the EF400 here, and even the 6.35 mm and 4.4 mm ports are not ideal with lower impedance earphones unless they have a linear impedance curve such as the Symphonium Helios, the 64 Audio Duo, and the Custom Art FIBAE 7 Unlimited. Everything else will experience a warmer sound owing to either a dip in the sub-bass response or a deliberate increase in the mid-bass and lower mids or both simultaneously. The same can be said about lower impedance dynamic driver headphones, but these are less common on the market. What you will instead see are lower impedance planar headphones and here things are a bit different again.

Seen above is the HIFIMAN HE1000se and Susvara tested on the EF400. The former is a 35 Ω, 96 dB/mW set and the latter a 60 Ω, 83 dB/mW set making them quite different in how easy, or not, they are to drive. These are also planar sets which are more likely to have a linear impedance curve in that even if the output impedance of the source is higher than everyone's favorite rule of thumb—no more than 1/8th the rated impedance of the headphones—the entire frequency response curve should still be evenly affected and not have one region change relative to others. I still found the HE1000se to sound better off the balanced outputs despite them having a higher output impedance than the 6.35 mm single-ended output that seemingly cramped up the soundstage and also affected the trailing ends of tones. Using the more demanding Susvara also showed that the high gain mode, even when volume matched to the low gain mode, fared similarly better. This could be the result of the Class AB current amplification circuitry that is playing more to its strengths at the highest Vrm outputs. The EF400 is rated to put out up to 10.7 Vrms at 36 Ω (4.4 W per channel max at slighly worse THD+N) so it will certainly get the Susvara loud enough even though personally I've had it sound better off other amplifiers such as the HIFIMAN EF1000 and the Feliks Audio Envy.

The other use case with the EF400 would be higher impedance dynamic driver headphones but here too I found the R2R DAC make them arguably even warmer than I'd like. Open-back dynamic driver sets already don't have the best bass extension so I suppose it's a personal preference if you wish to make this more pronounced. In my opinion, the EF400 worked best with open-back planar magnetic headphones that leaned towards the neutral-bright sound signature in the first place. The HIFIMAN Arya was a match made in heaven to where I have to urge Arya owners to consider purchasing the EF400! The Ananda Stealth worked well too and the Edition XS less so—this is where you would start to notice the EF400 becomes more like a warm smoothening filter that does blunt the otherwise clean and analytical sound that the Arya has more to spare, if you will. This made for a highly enjoyable, non-fatiguing sound with pretty much any music genre I threw at it. Acoustic jazz and classical music were especially wonderful, and male vocals also came off quite engaging. The MOONDROP VENUS was another good pairing with the EF400, and the Audeze LCD-XC was the only closed-back set that I could justify as being an overall positive experience. The vast majority of other planars tend to be warm and/or dark to where the added warmth and blunting of the EF400 is more a negative for me.
Next Page »Value & Conclusion
View as single page
Aug 22nd, 2024 01:38 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts