Review and Measurements of JDSLabs ODAC Rev. B Compared to Fiio E10K

[amirm]

This is a concurrent review of two DACs, the famous "O2 DAC" by Nwavguy from our neck of the woods and the E10K Olympus 2 from Fiio. They both use the TI PCM5102 DAC chip so makes for an interesting comparison.

Retail price for the JDSLabs version of the O2 DAC is $149 through Amazon: https://www.amazon.com/JDSLABS-ODAC...?srs=7812820011&ie=UTF8&qid=1510959113&sr=8-4

The Fiio E10k Olympus 2 is half the price at $73: https://www.amazon.com/FiiO-E10K-Headphone-Amplifier-Black/dp/B00LP3AMC2

​

While the ODAC is quite sparse package with just a USB connector and 3.5mm output, the Fiio E10 has a headphone amplifier with both bass boost and gain settings, plus a S/PDIF coax output. They are both quite small and fit in your hand so are great portable solutions to take on the road if needed.

Construction quality on the outside is good and a small step above cheezy DIY packages from other companies.

Software
I simply plugged in both units and Windows 10 Creator's Edition on my laptop nicely recognized both and created the WASAPI end point which I used in Roon in exclusive mode. Roon Reported the correct sample rates for ODAC which was 44.1, 48, 88.2, and 96Khz.

Oddly, the Fiio E10K only advertises 44.1, 48 and 96 KHz, missing 88.2 KHz! Not sure why it doesn't show that sample rate as I am confident the built-in DAC chip supports it. Could be a firmware bug in which case, Fiio should address.

Both units worked reliably inside Roon which was nice. They did not require any of the hacks to make them operational such as not using event driven mode in WASAPI that some DACs require.

Measurements
Let's dive right in with our J-test signal to see levels of noise and spurious signals around our main tone of 12 Khz:

For reference, I am also showing the Topping D30 (in yellow) which is my currently recommended low-cost DAC. The Topping D30 retails for $120 including shipping (https://smile.amazon.com/dp/B01L1SD...t1_sce0_prc0_aapi0&ref=bit_scomp_smile_ddt_t1) and so is cheaper than the ODAC. It also has more features in the form of proper RCA jacks, S/PDIF input, external power supply, etc.

As we see here, the Topping has lower noise floor and what jitter it has, is hugging the main tone so is quite inaudible.

The Fiio E10 has higher noise floor and fair bit of spikes around the main tone but at higher levels than topping and more of them. But again, not an audible concern but objectively worse.

I was surprised by the poor showing of the ODAC here where it also had close-in noise but also spurious idle tones above and below as the green arrows show. One pair is jutter the other, some oscillator running at the frequency that is bleeding into the output of the DAC. For a DAC that is supposed to be objectively good, this DAC sadly is not.

To its credit, the JDSLabs ODAC has a bit higher output than the Fiio which is always good.

Next, let's check levels of harmonic distortion using a 1 Khz tone:

Again, our Topping D30 pulls way ahead with higher output (due to its use of external power supply) and lower harmonic distortions than both of the other DACs.

At the other extreme, the Fiio 10 Khz has harmonic distortion that keeps going and going! Now these are at very low levels but why???

The ODAC finishes in the middle of the pack. It has good 5 to 10 db higher noise floor than the Topping D30 but overall it is cleaner than E10K has harmonic order increases.

For low level linearity test, we use a -90 db sine wave at 24 bits/44.1 Khz:

What we are looking for is a clean sine wave. The Topping D30 and ODAC are very close to each other. The Fiio E10K however, has lower output which makes it more susceptible to noise pollution and more up and down jumps so definitely less cleaner than the other two.

Headphone Listening
I did not have time to do a listening test through the line out from these two DACs. Instead, I tested them with different headphones direct from their output. In the case of Fiio E10K, I used its front headphone output rather than the rear line out.

Here, the Fiio E10K pulled way ahead. It is able to get at least twice as loud as the ODAC, able to drive even my difficult load of Sennheiser HD650. Its output never gets distorted by the volume control levels out at 70-80% of its maximum setting. Turning it up further seemed to do nothing. At max volume, the output was loud enough but not uncomfortable. In that regard, more power could be had.

In comparison the ODAC was rather anemic. Yes, you could listen to it but just not at loud enough levels.

The winner here is clearly the Fiio E10K.

Summary
From pure objective point of view, the winner here is the Topping D30. It is cheaper than the ODAC, has far better feature set including support for such things as DSD. In that regard, I can't recommend the ODAC. Maybe at the time Nwavguy thought of it, it was good but time has past and newer entrants even from budget sources easily outperform it.

Likewise, the Fiio E10K easily outpaces the JDSLabs ODAC because of its inclusion of headphone amplifier. It is also half the price. So another recommended choice! If you need a little DAC and headphone for very little money, the Fiio E10K Olympus 2 is my current choice.

 

[Darwin]

Good timing. I just got some HD 650’s and have been looking at amps.

 

[amirm]

I also have a number of headphone amps which I will be testing soon. So if you already have a DAC and want an external amp, that review will be coming.

 

[Darwin]

I’m looking forward to it. I’m currently using a Meridian Explorer 2 which is fine and does MQA. But I’m looking at other options.

 

[danadam]

NwAvGuy's jitter measurements of ODAC, for comparison:

I’m pleased to report the production version has even lower jitter. In the last ODAC article I reported the jitter components adding up to –103.3 dB. In the production version, as you can see below, it’s over 3 dB better at –106.5 dB with far fewer components. In reality both of these are comfortably below the –100 dB threshold I’ve established for transparency, so I’m not suggesting this is an audible benefit, but it’s still nice to see:

 

[Eric Auer]

Thanks for this review, I already have a D30 that I have been very happy with for a few months now.

I was thinking of getting the odac for another setup in my house but now I will keep on looking.

Eric

 

[amirm]

NwAvGuy's jitter measurements of ODAC, for comparison:
View attachment 9575

Unfortunately he is using a narrow band analysis from 8 to 14 Khz versus my full bandwidth one. Both the jitter and spurious tone that I found is outside of his measurement range:

By using a narrower frequency span and applying far more FFT points in dSound analyzer he is using he is able to dig out those tiny spikes. That is also why his measured response is 10 db better than mine. That is an artifact of using higher density FFT in the analysis. It reduces the measured noise of the instrument/DAC allowing the peaks to show up.

Net, net the two data points are not against each other. He measured what I did not and vice versa.

 

[amirm]

Thanks for this review, I already have a D30 that I have been very happy with for a few months now.

I was thinking of getting the odac for another setup in my house but now I will keep on looking.

Eric

My pleasure. I will keep testing these DACs and see what the future brings us. For an ideal DAC, I really like to see more sample rates and formats supported than ODAC does. It is just an old design that is no longer current.

 

[Blumlein 88]

My pleasure. I will keep testing these DACs and see what the future brings us. For an ideal DAC, I really like to see more sample rates and formats supported than ODAC does. It is just an old design that is no longer current.

Just a thought, since it all depends upon you. How about a blind test with the Topping and the ODAC?

 

[amirm]

Just a thought, since it all depends upon you. How about a blind test with the Topping and the ODAC?

Sure. Stand by....

 

[amirm]

OK, I grouped the ODAC and Topping D30 together in Roon allowing them to play in sync. Audio Precision said the 1Khz 0 db level for Topping D30 was about 2.17 volts whereas the ODAC was 2.10. Connected the two DACs to two inputs of my Stax SRM-007t and used that to perform an AB test. The headphones are the "Pro" electrostatic Stax.

Immediately there was a small difference but positive one. Tracing that it was the Topping D30 that was the better of the two. I could hear a bit more detail and more impact.

I then attempted to match the levels using Roon's digital volume control. This was tough as it would do nothing between 97 and 100 but the moment it went to 96, the level dropped to 1.93 volts on Topping. Running with that anyway, the difference between the two vanished. I could no longer tell if there was any fidelity difference (there is some channel imbalance in each DAC that is different and that can be audible).

Another interesting comparison occurred by accident. Since Topping supports 192 Khz and ODAC does not, when I played one of my 192 Khz tracks, Roon automatically performed sample rate conversion for ODAC, bringing it to 96 Khz. Despite that conversion, I could not detect any difference between the two.

So in summary, subject quality is the same if levels are matched. Otherwise, the higher output of the Topping D30 will make it edge out in side by side comparison.

 

[Blumlein 88]

Thank you Amir.

So the aim of the ODAC was a subjectively faultless DAC. It seems that aim was achieved or mostly achieved. Or at least your test is some evidence to that effect.

 

[mindbomb]

nwavguy's measurements were for the original odac. The one in this review and the one being sold now is the rev b, which is basically a different dac. Jitter should be a lot better on the rev b because of updated usb receiver.

So the gap between the original odac and topping d30 would be even larger.

I think he was mainly focused on the headphone amp, and the dac was more a perfunctory effort to help people use the amp.

 

[amirm]

I have his amp so will be reviewing that soon....

 

[sssn]

Great review, many thanks.

I think the headphone listening comparison is a bit unfair as the ODAC was never advertised to be a headphone amp. It's a pure DAC.
As far as its performance is concerned, it performs decently but it seems outdated. Lot of better and cheaper DACs have come around since then.

By the way, comparisons with NwAvGuy's measurements cannot be made as he tested the ODAC rev. A (using an ES9023 chip) while the rev. B you have uses the PCM5102. They're two entirely different DACs.

 

[amirm]

Who did the changeover to PCM5102? The consulting company or JDS Labs?

 

[mindbomb]

Who did the changeover to PCM5102? The consulting company or JDS Labs?

The original was nwavguy and the consulting company. The rev b was jdslabs and the consulting company. jds labs has a newer entry level dac available now, the ol dac.

 

[Grave]

Don't DAC's have a really high output impedance so you should not drive headphones directly from them?

"I was surprised by the poor showing of the ODAC here where it also had close-in noise but also spurious idle tones above and below as the green arrows show. One pair is jutter the other, some oscillator running at the frequency that is bleeding into the output of the DAC. For a DAC that is supposed to be objectively good, this DAC sadly is not."

I find this comment misleading because jitter at -95 dB should not be audible, therefore it is objectively good.

 

[amirm]

Don't DAC's have a really high output impedance so you should not drive headphones directly from them?

They do. In this case, the unit has a headphone jack, not typical RCA and someone had asked if they could use it to drive a headphone so I tested it that way.

 

[amirm]

I find this comment misleading because jitter at -95 dB should not be audible, therefore it is objectively good.

You need to be saying "subjectively" good. Objectively means getting close to a perfect DAC and this one does not. There is no excuse for any high fidelity DAC to spit out a constant tone.