• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Do You Need an External DAC/Headphone Amplifier?

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,393
Location
Seattle Area
A question is raised from time to time whether one needs an external DAC or headphone amplifier for desktop computer use. I thought I address that with a set of measurements in this article. There will be more to come because unfortunately, there is no one representative computer sound performance.

For this run, I will be testing my everyday laptop which is my main workhorse for a ton of work from listening tests to running my Audio Precision analyzer. It is an HP Z series Laptop that I bought 2-3 years ago at some $2,200. It has a Core i7 processor and is designed for "workstation/corporate" work as opposed to consumer.

As a way of reference I thought I compare its headphone output performance to that of Topping NX4 DSD portable headphone amplifier and DAC. See its review here: https://audiosciencereview.com/foru...sd-dac-and-portable-headphone-amplifier.3507/

The NX4 DSD retails for $160.

Measurements
Let's start with our usual dashboard view of the HP Laptop:
HP Laptop Z Series Dashboard Measurement.png


Starting from the right, we see pretty decent performance with nothing out of ordinary jumping out in the FFT spectrum. Distortion is 0.005% which amounts to 87 dB SINAD (signal above distortion and noise). Ranking that among recently reviewed products we get this:

1534568968189.png


It actually beats the $2,400 Schiit Yggdrasil DAC. :)

The Topping NX4 DSD pulls way ahead here with a SINAD of 104 dB.

Output voltage is a problem though with an anemic level of just 0.6 volts RMS. We like to see 2 volts for a DAC let alone a headphone amp.

As to add insult to injury, it has a pretty high output impedance of 17 ohms, causing its output to drop considerably with lower impedance loads:

1534569234364.png


The Topping NX4 DSD runs circles around it with 2 volts unloaded output and impedance of just 0.9 ohms.

Let's compare power versus distortion+noise at 300 ohm:
HP Laptop Z Series Power vs Distortion at 300 ohm.png


Power output is just 1 milliwatt compared to 59 for Topping NX4 DSD. The "good" news is that it doesn't get distorted and stays competitive with Topping until it runs out of juice.

Let's look at the same at 33 ohm:

HP Laptop Z Series Power vs Distortion at 33 ohm.png


As usual, when we reduce the load impedance the output power goes up. Here, the HP laptop produces 4 milliwatts compared to 144 milliwatts for Topping NX4 DSD.

Conclusions
From pure noise and distortion, the HP Z series laptop does pretty well. There are no warts and the output is pretty clean. The main issue is just lack of drive power. There simply is not enough juice to drive anything other than IEMs to high listening levels.

Even with IEMs such as my Etymotic ER4 SR, I just have enough power. If the source material has low level, sometimes I can't get enough volume out of it.

Of note, I have performed hundreds of critical listening tests using this laptop and my Etymotic IEMs. So the distortion and noise in general is pretty good so the need here is an external headphone amplifier, more than a DAC. But then again, the drive level is pretty low here so might as well go for a combo DAC+headphone amplifier like Topping NX4 DSD or its many competitors.

EDIT: measurements of my desktop motherboard posted here: https://www.audiosciencereview.com/...ernal-dac-headphone-amplifier.4116/post-96588

As always, questions, comments, corrections, etc. are welcome.

-----
If you like this review, please consider donating funds for these types of hardware purchases using Patreon (https://www.patreon.com/audiosciencereview), or upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
 
Last edited:

Blumlein 88

Grand Contributor
Forum Donor
Joined
Feb 23, 2016
Messages
20,524
Likes
37,057
I've tested a handful of laptops. That is more or less what I've seen with most. Decent SINAD, low distortion and low output levels. If you use IEMs you think it has plenty good enough sound. If you are in quiet surroundings non-IEM use seems okay. But hook up something better and the abundant power makes for a difference worth having usually.

Now I haven't tested any gaming laptops, those might fare better. I've had mid-level consumer laptops or higher priced machines for business use.
 

JJB70

Major Contributor
Forum Donor
Joined
Aug 17, 2018
Messages
2,905
Likes
6,148
Location
Singapore
I think this is one of the elephants in the room of audio equipment, that many computers, tablets and smart phones actually have pretty good DACs and audio output and that if the device delivers sufficient volume with your preferred headphones then in many cases you really don't need an external DAC or amp. Obviously if people discern a difference using an external amp/DAC then go for it, and I do think there is a pleasure in ownership but I think for many users the onboard sound is perfectly adequate. I used to carry a Pioneer DAP around, and it struck me that unless I was really listening intensely to identify nuances (which I don't do when commuting, staying over in hotel rooms, if listening to music in the office while working etc) then the sound from my Xperia smart phone was to all intents and purposes as good and decided it would be easier to just carry a single device. The other elephant in the room looking at the measurements is that for some parameters an HP lap top out performs certain audiophile type products.
 

Zilfallion

Member
Joined
Feb 16, 2018
Messages
82
Likes
65
I'm pretty sure my moderately old mid-low tier ASUS laptop would do horrible in these measurements. It sounds OKAY for my moderately high impedance HD 6XX, and even provides a fair amount of power, but it's got a super high noise floor when used with IEMs, and I seek to avoid using it's headphone jack as much as possible because of that. Oh, and it has grounding issues when plugged in I get loud mains hum.

Now, that being said, I'd love to see how my desktop's motherboard fares[Though you'll never get me to part with it for that long. My desktop is too important to me]. Should take a multimeter and a dummy load and test output voltage and stuff sometime, but based on rough sound level testing I've messed about with and sensitivity, it's got at least a ~1 V RMS output that drives my HD 6XX fairly well(I don't listen at the levels you seem to in some of your listening segments, but my Motherboard gets the HD 6XX far louder than I'd ever care to have on my head), and a relatively low noise floor that's hard to hear for my somewhat sensitive IEMs. Considerably more powerful than my Galaxy S5 phone.

If you ever get into motherboard testing(Difficult because you'd need a CPU as well, and some amount of DDR4, though you could probably get set up cheaper than some of the DACs that pass through here and share the DDR4 between AMD or Intel test benches if you had two), I'd love to see testing of some of the ALC 1220 implementations on the past 2 generations of boards(Intel 200/300, AMD 300/400 series boards.) In particular from ASUS and Gigabyte. ASUS probably being much more popular from what I've seen on the Headphone reddit, and they have their "Supreme FX" modification of the Realtek ALC 1220, but my board is a Gigabyte. Subjective impressions from most people with ALC 1220 boards that have tried it has been fairly good.
 

Umlautica

Member
Joined
Jul 25, 2018
Messages
18
Likes
40
The lack of power was made clear in the conclusions but the high output impedance seems to have been skipped over. Any reason for this?

The 17 ohm output impedance is high enough that a reactive and low impedance load will effectively change the magnitude response. This doesn't show up in the purely resistive load tests but I'm guessing that it would be noticeable with some IEMs.

EDIT: also, thanks for taking the time to measure the less-interesting gear. Not many do so it's nice to see.
 
Last edited:

bcall

Member
Joined
Jul 7, 2018
Messages
5
Likes
5
Quick question, was the laptop solely on battery power for these tests?
 

Grave

Senior Member
Joined
Jun 30, 2018
Messages
382
Likes
204
lol better than schiit :facepalm:

The integrated sound I have tried seems to have audible problems though, so I would not use it.
 

mindbomb

Active Member
Joined
Nov 11, 2017
Messages
284
Likes
175
@amirm
How is the jitter on the laptop? I am concerned that there will be a schiit modi-like situation with noise shifting with load, since it shares the board with other components.

also, thd+n vs frequency for the dac, to make sure 1khz isn't an outlier.
 
Last edited:

JustIntonation

Senior Member
Joined
Jun 20, 2018
Messages
480
Likes
293
I have a HP Probook. Second Probook I own. I've always felt both Probooks have the best onboard audio output quality I've ever heard out of a laptop.
Also had three IBM/Lenovo laptops, two quite old one very new. And their line out soundquality is just horrible.
I would recomend some testing of the frequency response though as I do feel my probook has a little bump in the bass.
 

bennetng

Major Contributor
Joined
Nov 15, 2017
Messages
1,634
Likes
1,692

garbulky

Major Contributor
Joined
Feb 14, 2018
Messages
1,510
Likes
827
"Measures better than the Ygdrassil". Having heard the Gugnir MB and many onboard sound devices. I think that's all I need to know about using measurements as a predictor for listening.
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
"Measures better than the Ygdrassil". Having heard the Gugnir MB and many onboard sound devices. I think that's all I need to know about using measurements as a predictor for listening.
I think this is the wrong forum for you.

You repeatedly rail against the facts measurements reveal, so I suspect you will find more comfort in an audiophile forum where your dogma will find companionship.

What it tells us is you use sighted and uncontrolled listening methods. Methods that draw incorrect conclusions due to bias, expectation and non identical conditions.
 
Last edited:
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,393
Location
Seattle Area
"Measures better than the Ygdrassil". Having heard the Gugnir MB and many onboard sound devices. I think that's all I need to know about using measurements as a predictor for listening.
You don't know what you have heard. You know what you have perceived. You have to isolate your listening tests to only sound.

Remember, with this very laptop, I have passed many double blind tests that audiophiles would not dream of taking let alone passing.

You have to develop some ability to perform listening tests properly or forever you will be living in the Matrix. Did you take this test for example? https://www.audiosciencereview.com/...ear-the-lossy-codec-a-public-blind-test.4090/
 

garbulky

Major Contributor
Joined
Feb 14, 2018
Messages
1,510
Likes
827
You don't know what you have heard. You know what you have perceived. You have to isolate your listening tests to only sound.
To avoid circular arguments, I'll let this one go :) . I'll agree that I know only what I have perceived. - While apparently the measurements appear to know even less of what I perceive. :D (Sorry, had to slip that one in there! :D )

Remember, with this very laptop, I have passed many double blind tests that audiophiles would not dream of taking let alone passing.

You have to develop some ability to perform listening tests properly or forever you will be living in the Matrix. Did you take this test for example? https://www.audiosciencereview.com/...ear-the-lossy-codec-a-public-blind-test.4090/
I like that test. I don't think I've done that one. But I have performed (non blind) tests to try to identify between mp3 and wav. - Above 160 kbps at least
The answer it's very hard to distinguish between the two. I don't think I could reliably differentiate in a test like that especially because I couldn't differentiate in a sighted A/B comparison
 
Top Bottom