Trying to Understand – CX31993 + 97220 Made My HD 490 Pro Noticeably Brighter

[Broccoli]

I recently got my first pair of decent headphones (HD 490 Pro), so consider me new to the hobby. For the first few months, I was connecting them to the front panel headphone port on my desktop PC, which has an ASUS motherboard. (The board features a typical SupremeFX setup with dual op-amps. The front panel is powered by the OPA1688 op-amp and the S1220A codec.)

Some people claim that as long as the motherboard output is loud enough and there is no hiss or noise, then motherboard audio is perfectly adequate and no further gear is necessary. However, I’ve since learned that the situation may not be so straightforward. Motherboard audio can exhibit frequency response roll-off, and high output impedance can alter the frequency response of headphones in undesirable ways. Based on this, the general consensus seems to be that at the very least, a DAC/amp dongle is recommended.

I then turned to Crinacle’s dongle reviews, where he lists recommendations across various price points. In the tier just above the Apple dongle, he recommends the CX31993 + 97220 combo. I ended up buying a no-name dongle with that chipset from AliExpress. To my surprise, the dongle made a clear and immediate difference. It made my HD 490 Pro sound noticeably brighter to the point of sibilance, requiring a little EQ in the treble range.

At this point, you probably see where I’m going... I’m trying to understand the audible difference the dongle made. Should I assume that the CX31993 + 97220 combo has a flat frequency response, and that my motherboard was coloring the sound by softening the treble? Or could there be something off about the implementation of my particular dongle?

Lastly, a more general question: should I go all in and get a dedicated desktop DAC/amp (something like the Topping DX1), or is my $8 dongle realistically sufficient for the HD 490 Pro?

Thanks in advance!

 

[DVDdoug]

Your headphones have a nominal impedance of 130 Ohms, and the higher it is the less the frequency response is affected by the headphone amp's output impedance. But it also depends on how much the nominal headphone impedance varies over the frequency range. I wouldn't expect impedance to be a problem but that's just my expectation and I don't know the specs...

Amir's Review shows a broad impedance bump in the mid-bass range and a slight slope-up above 10kHz. If the headphone's output impedance is too high you'll get frequency response variations that follow the impedance curve... A bump in the bass and another boost at the highest frequencies. But without knowing the output impedance we can't know the amplitude of those variations.

Another possibility - Windows has some settings that can alter the sound and usually they are usually specific to the particular device so they could be different for a USB device and the internal soundcard. And some soundcard & motherboard manufacturers add their own additional sound control utility (EQ, etc.).

 

[raindance]

Plus the chipset used has almost no impact on the sound, it is all about how well the circuit is implemented, period. Excessive brightness usually is an indicator of excessive harmonic distortion, or some silly OS-related "sound enhancement", or output impedance of the amp not playing nice with the headphones.

 

[Broccoli]

@DVDdoug @raindance

Thank you both so much for your comments! I am new to this, and your responses have given me much to look up.

So, after googling a bit it looks like the output impedance of USB dongles typically is very low, 1 Ω and less... I guess we therefore can rule out the impedance of the USB dongle itself as a factor... Probably should not result in audible differences, right? If the dongle is coloring things, it must be implementation, harmonic distortion, etc, as you pointed out.

I dual boot Linux and Windows, and the added brightness persists across operative systems, so I think we probably can rule out OS related quirks. (For the record: On Windows I don't use any additional audio drivers for the motherboard, nor for the USB dongle. On Linux I use Pipewire as my sound server and EasyEffects for applying EQ.)

This leaves the final issue. Apparently motherboards / front headphone jacks, can have obscenely high output impedance? While I don't think there are any measurements for my specific motherboard, is it fair to say that high motherboard output impedance (in combination with the impedance of the headphone itself) theoretically might result in a darker sound than the dongle? In other words, the dongle isn't exaggerating the treble, my motherboard was muting it?

 

[raindance]

Without measuring you'll never know... Do you have any other source to compare?

At the price point I'm not sure it's worth the hassle to measure output impedance. Many motherboard outputs are pretty bad though so there's a chance the dongle is honest...

My 2c worth of opinion - I'm not sure I would use a USB dongle for 120 ohm headphones. For 32 ohm ones, sure, but I think you're maybe running into power limitations because the dongle can't swing enough voltage for that impedance. That could make it sound bright for sure.

 

[oleg87]

The HD490 is not super sensitive to output impedance, but it's certainly possible your motherboard's headphone outputs are high enough to affect tonality - I've seen measured values in the range of 300ohms (though it seems unlikely to be this high on a newer board that makes a point of "premium" audio), 75ohms is a value that seems to come up a lot as well. I don't know what magnitude of perceived difference we're talking about here - I think you could likely attribute about a 1 dB'ish (but pretty wide-band) bass "boost" to the motherboard.

 

[kemmler3D]

Apparently motherboards / front headphone jacks, can have obscenely high output impedance? While I don't think there are any measurements for my specific motherboard, is it fair to say that high motherboard output impedance (in combination with the impedance of the headphone itself) theoretically might result in a darker sound than the dongle?

Without measurements of the specific motherboard it's hard to say, but I think it's plausible. The worst headphone output I've ever heard (the only extremely obvious distortion I've ever heard from a headphone amp) was from PC-integrated audio.

A good next step would be trying out the same tracks on the same headphones but a different known-to-be-high-performance DAC. If a friend has one, or you can try one in a hifi shop, you get another data point.

Alternately: Is it possible you're listening louder with the new dongle? Louder = more perceived emphasis on treble/bass, and the dongle would likely be producing higher output at a given OS volume setting.

 

[Broccoli]

Thanks for the feedback, guys. I realize I am asking a difficult question that is hard to quantify without measurements, so thanks for bearing with me.

@kemmler3D That is a very good suggestion, unfortunately I live in a rural area, so I don't have access to any audio stores nearby. (I am aware of the connection between volume and perceived music quality, and when A/B'ing I make sure to listen with comparable volume, so I don't think this is the issue.)

The HD490 is not super sensitive to output impedance, but it's certainly possible your motherboard's headphone outputs are high enough to affect tonality - I've seen measured values in the range of 300ohms (though it seems unlikely to be this high on a newer board that makes a point of "premium" audio), 75ohms is a value that seems to come up a lot as well. I don't know what magnitude of perceived difference we're talking about here - I think you could likely attribute about a 1 dB'ish (but pretty wide-band) bass "boost" to the motherboard.

Seems plausible, and 1 dB would be audible I guess. Earlier in this thread @DVDdoug linked to a graph with the HD 490's impedance curve, and while the peak of the curve is around 80-90 hz, the entire area up to 1khz is somewhat elevated. There is also a little impedance boost starting at 10k hz and above, but I am thinking that the perceived brightness I hear perhaps is a combination of factors: So, if the impedance of my motherboard is an issue, the loss of the additional boost in the bass, combined with a little more sparkle up top, might in combination make me perceive the dongle as brighter. So it's not necessarily just a treble boost I hear, but also the loss of a little warmth?

I guess I at one point could throw money at the issue and see if there's a difference. But if I'm getting another DAC/Amp, then I'd like to take my time and look into the alternatives before making an investment, so don't expect an immediate update to this thread. I will update the thread with my findings, but expect it to take time. Might not be this month, nor the next.

At the price point I'm not sure it's worth the hassle to measure output impedance. Many motherboard outputs are pretty bad though so there's a chance the dongle is honest...

My 2c worth of opinion - I'm not sure I would use a USB dongle for 120 ohm headphones. For 32 ohm ones, sure, but I think you're maybe running into power limitations because the dongle can't swing enough voltage for that impedance. That could make it sound bright for sure.

Fair enough, that another reasonable possibility. In terms of combination DAC/Amps, what would you personally consider a bare minimum for a 130 Ω headphone?

 

[raindance]

Bare minimum is something that can swing some voltage, usually not USB powered. Schitt might have something. I had an amp from Channel Islands Audio, no longer selling products unfortunately - but you could find a used one - that I thought was decent for higher impedance phones. Generally you're looking at amps that can push a decent amount of power, like 0.5 watts or so rather than 50 milliwatts but the key is that they can swing adequate voltage to produce sufficient power at 120 ohms or whatever (ohms law).