• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Lynx Hilo DAC/ADC Review and Measurements (part 1: DAC)

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
13,493
Likes
5,662
Location
Seattle Area
#1
This is a review and detailed measurements and comparison of Lynx Hilo professional DAC and ADC. The unit is on kind loan from member DallasJustice who uses it as an 8-channel DAC to drive his active JBL M2 speakers (together with Benchmark DAC3). In this review, I will be evaluating its DAC performance and comparing it to Benchmark DAC3. In Part 2, I will be measuring its ADC performance and comparing it to RME ADI-2 Pro.

The Lynx Hilo is a few years and today it retails for $2,300 from US major supplier of pro products, sweetwater. The Benchmark DAC3 despite being the same price, is DAC only. The Lynx Hilo has a large (though a bit grainy and low contrast) color touchscreen. That makes navigation far easier than competing devices. Its companion host control panel allows easy control just the same. So if you want a pro soundcard that is not a nightmare to navigate, this is it.

The unit is quite chunky but still desktop sized. Here you see it compared to Benchmark DAC3:

Lynx Hilo DAC and ADC review and measurement.png.jpg


The unit only has balanced connectors so that is how I tested it. Levels of course can be adjusted and you can convert the balanced lines to levels compatible with consumer gear. For this testing, I focused on +24 dBu to match the Benchmark DAC3 I had just reviewed.

While the Lynx Hilo goes through some kind of USB plug-n-play configuration it does not expose itself as a sound card that way. As such, installation of drivers is mandatory for any operation which I did.

Let's get into the measurements and see how she does.

Measurements
As usual, let's look at our dashboard:
Lynx Hilo DAC and ADC dashboard measurement.png


Looking at the frequency response it is ruler flat (earlier post was incorrect):

Lynx Hilo DAC and ADC frequency response measurement.png


Comparing the THD+N distortion products versus Benchmark DAC3 we see higher levels:
Lynx Hilo DAC and ADC distortion versus frequency measurement.png


It is unfortunately that its differential rises above DAC3 in mid-frequencies where our hearing is most sensitive.

The DAC3 out distances itself just the same in intermodulation distortion versus level:

Lynx Hilo DAC and ADC IMD distortion vs level measurement.png


That is a 9 dB difference. The DAC3 also doesn't go into clipping until -5 dBFS whereas the Lynx Hilo does so at -8 dBFS.

Our favorite measurement, the linearity, shows essentially perfect output from both DACs:

Lynx Hilo Linearity Measurement.png


For pedantic reasons I have marked the 0.1 dB deviation points if you want to split hairs (in which case the Benchmark DAC3 goes to -114 dB vs -110 for Lynx Hilo). But again, both are at the limits of the measurement and variations can occur in each run at these levels. Error even at -120 dBFS is well under 0.5 dB.

Finally, here is the waveform for -90 dB sine wave:

Lynx Hilo DAC and ADC -90 db sine wave measurement.png


Sans noise that is expected, it is an ideal looking sine wave indicating that reproduction of CD rate will be near ideal (the test however is run at 24 bits, not 16).

Summary
The Lynx Hilo is essentially without faults in its DAC performance. Yes, the Benchmark DAC3 is a bit cleaner with lower distortion but lacks ADC functionality, touchscreen UI, etc. So value wise, the Lynx Hilo is much better and is a recommended product.

Edit: fixed the incorrect frequency response graph.
-------------

As always, questions, comments, recommendations, 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:
Joined
May 15, 2018
Messages
15
Likes
4
#2
Wow, I didn't expect Hilo to be measured so soon. :eek: Thank you for your work, Amir!
As for the results I guess this DAC is competing with the best. But still I think this video is appropriate :D

 

svart-hvitt

Major Contributor
Joined
Aug 31, 2017
Messages
1,320
Likes
467
#3
@amirm , can you measure Hilo’s pure digital operation, like USB to AES-EBU?

How flawless is its pure digital conversion?

How well/flawlessly works Hilo’s volume attenuation in pure digital operation?
 

dallasjustice

Major Contributor
Patreon Donor
Joined
Feb 28, 2016
Messages
1,190
Likes
577
Location
Dallas, Texas
#4
Thank you @amirm

Lynx released the Hilo in 2011. My unit you tested was purchased 5 or 6 years ago. I’m glad it’s still an excellent workhorse in a multi channel setup.
 

SIY

Addicted to Fun and Learning
Technical Expert
Joined
Apr 6, 2018
Messages
625
Likes
508
Location
Chicago-Milwaukee axis
#5
How are you injecting the test signals? I.e., what's the AP output and DAC input?

One day, you'll set the graphs to turn off the distracting time stamp and show the title better. ;)
 

dallasjustice

Major Contributor
Patreon Donor
Joined
Feb 28, 2016
Messages
1,190
Likes
577
Location
Dallas, Texas
#7
Is there a jitter test you can post too?
 
Joined
May 8, 2018
Messages
18
Likes
14
#8
Thank you @amirm

Lynx released the Hilo in 2011. My unit you tested was purchased 5 or 6 years ago. I’m glad it’s still an excellent workhorse in a multi channel setup.
@dallasjustice, while I realize this site is measurement orientated, since you own both the Hilo and the DAC3 do you have any brief comment on the subjective sound quality of the two?
 

dallasjustice

Major Contributor
Patreon Donor
Joined
Feb 28, 2016
Messages
1,190
Likes
577
Location
Dallas, Texas
#9
@dallasjustice, while I realize this site is measurement orientated, since you own both the Hilo and the DAC3 do you have any brief comment on the subjective sound quality of the two?
I honestly can't tell the difference. My setup is very unusual. I have a 4 way digitally active setup for the JBL M2 and 4 subs. The Hilo only has 6 analog outs, so I need another DAC for the horn tweeter on the JBL M2. The M2 horn tweeter gets the DAC3 and the M2 midwoofer gets the Hilo balanced out. The integration is perfect. I like the DAC3's high output and it locks perfectly over AES; there's no drift or other sync issues with it over AES.

Having said all that, if I could find a DAC/ADC that performed as well as the Lynx Hilo with 8 analog outs, at least 24dbu output and could do ASIO multi-client, I'd consider buying it. For now, that doesn't exist.
 
Joined
Apr 7, 2018
Messages
8
Likes
11
Location
DE
#10
Hi, did you already notice, that RME released a new firmware of interest ? ;)

https://www.forum.rme-audio.de/viewtopic.php?id=27295

"Firmware DSP 21 brings an interesting new feature - well, interesting for geeks and nerds (like us). It addresses the ripple seen from the AD converter chip with frequency response measurements at higher resolution, at sample rates 44.1 up to 192 kHz. While this is nowhere near audible, the ADI-2 Pro is also used as measurement front-end. With today's high quality devices and measurements using a very high resolution, that ripple becomes visible, and is both distracting as well as a problem for documentation.
Other ADC chips typically have only a fourth of that ripple, which is why it mostly stays invisible. Firmware version DSP21 now adds a unique, phase-linear compensation filter, that reduces the ripple significantly. Not completely, but enough that in normal measurements the ripple is no longer visually obvious. The filter has zero effect on the audio performance. No phase shift, no change in SNR, THD or bandwidth, not in impulse response and not in latency. It just reduces the ripple by up to 86% (depends on mode and sample rate). That's why it is not an option but always on, and only switched off automatically with DSD recording.
The compensation filter also works on the already quite linear 176/192 kHz sample rates, improving linearity to a fabulous level. Note that 352.8 kHz and up do not have any (!) ripple, which is one of the reasons the manual recommends 384 kHz for audio measurements.
So you can shrug your shoulders and continue using the ADI-2 Pro as it is, or add a small pristine improvement on top of the jewel that the ADI-2 Pro already is. Your choice!
"


Download of the firmware update files (original update thread): https://www.forum.rme-audio.de/viewtopic.php?id=27280

Pro and DAC:
- Class Compliant compatibility for Windows 10 and Squeezebox Touch adjusted
- Dual EQ: save state to preset slot Temp after activation in the I/O menu

Only ADI-2 Pro:
- Level Meter of analog inputs simultaneously track pre and post EQ. This way overloads stay visible and Auto Ref Level works correctly even when the EQ is used to attenuate the signal.
- Added compensation filter for the AD-converter to reduce ripple in the frequency response for PCM, 44.1 up to 192 kHz, filter modes SD Sharp, Sharp, SD Slow, Slow.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
13,493
Likes
5,662
Location
Seattle Area
#12
5 dB increments are way too much for FR measurements :(
Sorry. I am short of time and so used the stock AP frequency response measurement. These are the components from Michael's system and without it, he can't play anything so I am trying to return them to him as fast I can.
 
Joined
Feb 11, 2018
Messages
48
Likes
41
#13
How are you injecting the test signals? I.e., what's the AP output and DAC input?
Agreed: always state signal source, as this makes a difference in most dacs.

A few other thoughts, Amir:
Dashboard should have FFT Y axis in dB.
In-band frequency response should use 1dB/div Y axis.
Stop-band response should be displayed with a linear scale 20Hz-Nyquist at 10 or 20dB/div Y axis. Stop-band response is very indicative of how well the designers understand the purpose of filtering (or not).
THD is ok but 2-tone 19&20K with FFT is more revealing. A 5 or 6 tone cluster would be even better.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
13,493
Likes
5,662
Location
Seattle Area
#14
Hmmm. Seems like both AES and S/PDIF are unable to lock on input signal. The frequency counter jumps between a few hertz up and down, the analog waveform has glitches in it and moves left and right, and the FFT on the right is a disaster. Everything other than output level is jumping up and down which tells me it is not really locking correctly to the input:

1530119162993.png


I tested it with RME using the same config and all is perfect. So it is not anything messed up in the configuration of AP.

Turning on "SRC" completely eliminates the issue which should if it thinks the input rate is varying.

This is what the control panel says:

1530119278986.png


It says it is "locked" so there is no sync issue with input. Not seen above but in real-time the digital input meter is fluttering on its last few values so clearly it is not able to extract the audio samples correctly.

Here is the output with SRC Mode on:

1530119472338.png


Everything is clean now. Frequency is locked on 1 kHz as it should.

Quite surprising. As I said, both S/PDIF and AES are doing this.

Weirdly, you can set the sample rate way on top to a different value. If I set it to 48 kHz, it still captures audio the same way but it goes even more wild with far more values changing. Yet it still says it is "locked" to input.

Any ideas anyone? Hate to declare it completely broken but it seems so.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
13,493
Likes
5,662
Location
Seattle Area
#15
How are you injecting the test signals? I.e., what's the AP output and DAC input?
In USB testing which is what I post, AP is generating digital signals in software in the PC and sending it to the DAC using ASIO bit-perfect transfer mode. And of course the input is from the DAC, in this case, balanced as I mentioned in the review.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
13,493
Likes
5,662
Location
Seattle Area
#16
One day, you'll set the graphs to turn off the distracting time stamp and show the title better. ;)
Believe it or not, people demanded timestamps on measurements and hence the reason it is there!
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
13,493
Likes
5,662
Location
Seattle Area
#17
OK, looks like I had posted the ADC measurements instead of DAC for frequency response. Ran it with higher resolution and it is ruler flat and updated in the first post. Sorry about that. :)
 

bennetng

Active Member
Joined
Nov 15, 2017
Messages
212
Likes
58
#18
The sine wave still looks jagged when compared with the first post, but then the FFT looks normal. What is going on?
 

dallasjustice

Major Contributor
Patreon Donor
Joined
Feb 28, 2016
Messages
1,190
Likes
577
Location
Dallas, Texas
#19
Hmmm. Seems like both AES and S/PDIF are unable to lock on input signal. The frequency counter jumps between a few hertz up and down, the analog waveform has glitches in it and moves left and right, and the FFT on the right is a disaster. Everything other than output level is jumping up and down which tells me it is not really locking correctly to the input:

View attachment 13429

I tested it with RME using the same config and all is perfect. So it is not anything messed up in the configuration of AP.

Turning on "SRC" completely eliminates the issue which should if it thinks the input rate is varying.

This is what the control panel says:

View attachment 13430

It says it is "locked" so there is no sync issue with input. Not seen above but in real-time the digital input meter is fluttering on its last few values so clearly it is not able to extract the audio samples correctly.

Here is the output with SRC Mode on:

View attachment 13432

Everything is clean now. Frequency is locked on 1 kHz as it should.

Quite surprising. As I said, both S/PDIF and AES are doing this.

Weirdly, you can set the sample rate way on top to a different value. If I set it to 48 kHz, it still captures audio the same way but it goes even more wild with far more values changing. Yet it still says it is "locked" to input.

Any ideas anyone? Hate to declare it completely broken but it seems so.
Interesting. I’ve never used any digital input other than USB. I would say the sample rates don’t match but it sounds like you already tried that.
 
Top Bottom