Hiby FC3 - Summit-Fi DongleDAC according to the Df-metric

Equipment / Electronics / Gear / Accessories Portable DAC/AMPs & Dongles
1

This little dongle really deserves its own thread if the guys at HypeTheSonics are even conditionally right about it: based on their (or SoundExpert.org’s) measurements and resulting Df-metric (difference in dB between the output signal from the device over a 32-ohm load and the mathematically calculated perfect analog equivalent of the input signals), it might just be the best fidelity/price and fidelity/size DAC-amp on the market.

Based on the above info and hoping to hear the magic for myself, I got the 2022 version this year, with the more elegant “slot” indicator LED that replaces the previous “blob” they had on:

HibyFC3
HibyFC31200×2100 218 KB

First impressions as posted on AliExpress:

Good inline DAC/amp, better sound than my phone’s jack, pushes more power for more bass amount and impact, better low-mids for fuller voices. Doesn’t perform as well as Oppo HA-2 for either bass or low vocals, but that one is much larger. Seems a bit brighter than the HA-2 but I have to do a more serious comparison to be sure. I tried it plugged into the iFi Micro iDSD amp vs. the full Micro by itself and I can’t tell the difference.

Initially I thought it wasn’t working but it turns out it was delivered with internal volume saved to level 0 or very low, so I had to turn it up first from its volume rocker before I could hear anything. :slight_smile:
Connected to a Windows 10 and the auto installed driver didn’t see it as a device with microphone signal, only as “speakers”. I will investigate more later.
The only regret I have with this purchase is that some of my money went to supporting the British marketing format MQA which is not lossless and has no right to call itself master quality.

(I probably just needed the driver from the Hiby website.)

This Thursday I also took some time during a break in a quiet spot at work to A/B it against the FiiO BTR3, listening through the KSC75X.

If I plug the BTR3 into the phone and use it inline, it’s about half as loud as the FC3, so after I cleared that up I just lowered the FC3’s internal volume to 16/30 for the rest of the test, so I could keep the phone’s volume constant and be able to switch faster.

For most music I couldn’t really detect any stand-out difference except maybe for that nagging feeling like I had with the Oppo HA-2 that the FC3 was overall brighter. So after a number of songs not finding any big difference - and knowing what I knew about the BTR3 running off battery from my JM6 Pro comparison - I switched to the more powerful Bluetooth&battery mode of the BTR3 and readjusted the volume. Not long after that, I found the difference I was afraid of while listening to Lisa Hannigan’s “Splishy Splashy”: similar to my early comparisons to other beefier units, the Hiby FC3 immediately lost to the BTR3 on dynamics and realism and stage depth, to the point I’d say it took out 2/3 or more of the joy of listening to that song.

There’s something really strange going on with this design - I don’t understand how it can at the same time push more absolute volume than the BTR3 at max while sounding way less dynamic at equal loudness. I guess it’s just not good at servicing sudden spikes in power demand / doesn’t have great slew rate or whatever that’s called, when powering headphones directly. Because curiously it sounded indistinguishable from the iFi Micro iDSD’s DAC when plugged directly into the Micro’s amp. So I have to conclude they just haven’t designed this thing as a good “final stage” to hook directly to anything except the most weaksauce headphones, and the only way it might retain that Df-metric based title of champion of fidelity/price is if it’s used as a DAC and plugged into some other beefy amp (hopefully one clean enough to not trash all that fidelity).

On that notion, I’m planning to compare it to the BTR3, Oppo HA-2 and iFi Micro used as DACs into the E/90X energizer of the Koss ESP/95X, since that’s now my best transducer and that system is where I will want to be using my absolute best DAC going forward. (But… I might also give a chance to that high-gain mode that HypeTheSonics are talking about being available in the FC3, maybe that’s what fixes the dynamics. Though it’s pretty darn inconvenient to activate since it requires plugging in something that at least initially looks like a 150-ohm+ load on both channels. I’d have to plug in both impedance adapters I have + the KSC75 for a total of 30+75+50, and keep my fingers crossed.)

Other FC3 reviews I’ve seen around:

https://andyaudiovault.wordpress.com/donglemadness/hiby-fc3/ (I scoffed at Andy’s review at first, but after getting the device and listening to it I ended up agreeing with him: the FC3 is lean and un-authoritative with a lot of transducers, it’s just not good at delivering raw power, however summit-fi it might be when it does achieve synergy)

https://medium.com/bedrock-reviews/hiby-fc3-review-42761b6da958 (found the link here in HFG; only has literally 1 thing to say about the sound, but I guess it’s better than nothing)

1 Like
2

Done. Went in for the A/B, A/C and A/D comparisons today and… the FC3 is brighter and more detailed than all of these (with the Oppo and the iFi used via their line-outs). The BTR3 seemed to come closest but it was also the 3rd one evaluated, so it might have been fatigue that had me not discerning as much of a difference, or it’s simply that it’s a newer design that the other two and just reveals detail better.

Now if I didn’t have any measurements to look at I’d primarily suspect that the FC3 is abnormally bright and that’s the only reason it appears more detailed and that’s why it “disagrees” with 3 other quite respectable DACs, but… in this case I know of a little Df-metric that disagrees, that says the FC3 produces something so similar to the analog equivalent of the input signal that it should be practically indistinguishable to human hearing (if driving a 32-ohm load with no special sensitivity problems). So since I subjectively enjoy hearing that extra treble and I also saw Serge Smirnoff having a nice discussion about his Df-metric over at ASR and finding substantial agreement with other people’s attempts at constructing a singular metric of audio fidelity… I have to go for door #2 in this case: I want to say the FC3 is actually showing fidelity and representing the exactly correct amount of treble and detail, and all of the other 3 in my line-up are veiled to various degrees (possibly because the manufacturers thought most people don’t really want to hear all that crispness up top, revealing defects in mastering/recording etc., and also some people want a more ‘analog’ sound and those should be kept around as customers as well etc.).

So with these results, right now my conclusion is the FC3 is my best DAC for plugging into another amp, and this is what I will prefer to use with my electrostatic system going forward (but really with everything, as long as I can plug it into some other amp stage I will try to have it in every chain lol).

3

It’s a shame that it’s so weak. An AMP is almost mandatory to bring anything out of it.

1 Like
4

Yeah, if I was into sensitive IEMs I might be able to use the FC3 as-is, but with my types of headphones… nah. Even if I said OK, it’s a tiny dongle, it’s only relevant for outdoorsy vacation-y listening and therefore it only needs to power the KSC75, it still can’t do it alone, it’s no replacement for the BTR3.

If they could somehow make an “FC3 mk2” with the same clean output but more power, and not 3x the price, they could just wipe the whole market niche around it, but so far they don’t seem to have found the magic formula for that. The FC4 is more powerful but far less clean, even on BAL-out, the FC5 is BAL-only, still not as clean (0.0015 vs 0.00025 THD+N) and has no microphone-in (not a real replacement for the headphone jack, which is the first job of any dongle), and the FC6 is way bigger and more expensive and wastes space with unnecessary fluff like a screen and a menu and selectable digital filters bla bla. :roll_eyes:

5

Welp, I can confirm the high-gain mode exists - the FC3 just has to see a really high impedance on its audio jack when you first plug in its USB. For example if I hook it up to some amp’s line-in, plug in the USB, then unplug the amp and put in my KSC75s all of a sudden the bass is filled in and has punch. Seems about what I’d expect for a shift from 1.3 to 1.9-ish Vrms, even though the absolute volume increase isn’t more than 3 dB.

What didn’t work out for me was getting it to switch using the above series of adapters + KSC75, maybe because the total is not quite high enough past 150 ohms, maybe because of that “smart” DUNU adapter that doesn’t show the same impedance in both directions (not a simple series resistor).

Just ordered another 150-ohm series-resistor adapter for $4.3 - I think with this one I will be able to take the FC3 portable and trick it to go into high-gain mode and drive my KSCs properly.

6

Questyle M15 is always the dongle that makes people think desktop amps aren’t needed. It outputs over 22mW on the balanced connection, which is insane. I think the unbalanced power is about twice what the base Schitt outputs.

There are plenty of reviews on it. I don’t own one, but if will probably be my next amp.

7

Ehh, what I see is people have trouble specifying exact power limits for the M15 because of its current-amplification design. Those 70-ish mW specs at 32 ohms can’t be the whole story or it wouldn’t be any kind of replacement for desktop units, it wouldn’t even surpass the 5x cheaper FC3 which can do 100 mW @32.

But anyway, when I said summit-fi, I did mean fidelity, not power, so for the massive price difference I don’t think the M15 is a better fidelity option, not even close.

8

They aren’t having that much trouble, given this is on the product specification page

“Output Power： 3.5mm:RL=300Ω,Po=11.97mW, Vout(Max)=1.895Vrms,THD+N=0.00045% 4.4mm:RL=300Ω,Po=22.60mW, Vout(Max)=2.624Vrms,THD+N=0.00057%”

source - https://questyleshop.com/products/m15

The interesting part is that both HBB and DMS rave about the fidelity of the M15. So much so, that HBB didn’t even mention the power. Seriously go look up their reviews on YouTube.

9

That’s not the point. I wasn’t implying the M15 sounds bad, just that it doesn’t justify the extra 200 USD once you’ve achieved transparency at 50-60 USD, unless you really really want that additional power-in-a-small-package that’s specific to the Questyle.

10

Not anymore.
JCally AP20, which was mentioned some time ago, turns out to be pretty, darn powerful… And it costs $60.

obraz
obraz720×975 96.9 KB

I own it and it’s the dongle I pick up the most, when I need one.

2 Likes
11

I highly doubt that they are technically equivalent, in transparency, sound quality, or power.

Under that logic, just buy the Apple dongle, they are $8. Otherwise you are just spending 7x the money. Note the lack of power is the best argument against the Apple dongle.

When you talk about gear in a thread labelled summit-fi, I thought we were talking about getting the best gear, not the cheapest acceptable gear.

12

I have thought about that model ( all three headphone jacks is a nice feature ). Sometimes the Cirrus Logic DAC chips can add a tinny sound to my ears, how is the AP20?

13

Sure, I’ve seen high power specs from JCally and was eyeing an AP90 myself there for a while, though it doesn’t quite reach the 320 mW I’d like to put into my HE-400i if I start taking it on vacations, but…

Distortion specs. Chinese companies are only mandated by law to ensure 10%(!!) THD+N or better at the max power they promise to the consumer. And though some of them want to build a solid reputation and do tell you what you need to know about their THD+N at different power levels, you always have to be careful about what information is being withheld and preferably look for independent measurements to be sure. JCally don’t seem to be among the ones most concerned with the signal quality specs, unfortunately, so it’s hard to tell what fidelity you’re really getting with all of that power.

14

Nope. Summit-fi = summit fidelity, no defects introduced in the signal while converting and reproducing it. Power matching I see as an independent issue, meaning you can have different summit-fi sources at different power levels or transducer types/impedances. Otherwise we’d have to declare just one type of source as “summit-fi”, i.e. whatever can play transparently through all transducers, and then there would be very little to discuss except the iFi Micro varieties and maybe a couple of other things similar to those.

15

Yeah, true that.

What exactly do you mean by “a tinny sound”? Tinnitus? Some high-pitched noises?
I can’t say I’ve heard anything like it.

16

Is 22mW a high or a low value? If anything I believe people would complain about on-paper power specification of M15.
FYI, I used to have one, it didn’t fail to drive anything I threw at it including 500Ohm low sensitivity buds nor HD6xx, though I did not have any low impedance low sensitivity planar at hand when I was using that. But the case is above statement applies to Qudelix 5k as well or basically any modern source I tried.

1 Like
17

Do we really have to do this? I already told you the problem is at low impedances. Their official specs @32ohms paint the M15 as a submediocre device that can’t even work as a line source of 2Vrms. And then you have that insanity of the low-gain mode producing slightly more power than the high-gain @32. But reviewers say in practice there seems to be much more (clean?) power available, so there’s something problematic there, maybe the traditional measurement setups don’t play well with this kind of amplifier design. I consider it currently an open question what the real max. power of the M15 is into 32 ohms.

18

M15 has as much power as it states in specs, or maybe more true would be to say as much power as it measures for example in here - the raw measurement is not a problem

The advantage of it is that current mode amps can have by design much better control on the slew of the signal (lets call it a speed of signal level) than voltage ones. It means that if a voltage mode amplifier may have problems with transient peaks of the music (“dulling them out”) the CMAs should be able to handle those swings better.
I never heard anything dulled out by too weak of a power available by amp as long as I had some (20dB?) headroom over what is calculated in online calculators, but there are people I trust that claim that some stuff - like Yinman 600Ohm buds sound really dull on weaker sources even though by calculation they should be easily driven by them and recommend hundred of mW sources.

1 Like
19

Prof. Wolf didn’t go into speculations on why that might be the case, but he also mentions at the end of the article that it’s mind-boggling how little power the M15 seems to have @32, after blowing most dongles out of the water @300:

Anyway, since some reviews talk about it sounding “organic” or “musical” I think it’s not for me (just like the AP20 because its CS43131 also has some reviews like that), I much prefer the supposedly-harsh, supposedly-too-digital sound of Sabre DACs - that’s top fidelity for me, nothing veiled, nothing rounded out or rolled off, just truth. If your music’s harsh it should sound harsh, if you’re playing harsh music too loud for too long, your DAC should help create fatigue, and that means it’s doing its job right. :slight_smile:

20

Like the sound is coming out of a tin can, or has a thin, metallic sound.

Lol, I liked the description a bunch. Most of the dongle dacs seem to use either the Sabre ES92xx or the Cirrus Logic CS43xxx. I tend to view the Sabre as the better one, but they aren’t all harsh.