• 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). There are daily reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Lynx Aurora 8 Review (8-channel DAC/ADC)

Rate this DAC:

  • 1. Poor (headless panther)

    Votes: 6 5.8%
  • 2. Not terrible (postman panther)

    Votes: 36 35.0%
  • 3. Fine (happy panther

    Votes: 55 53.4%
  • 4. Great (golfing panther)

    Votes: 6 5.8%

  • Total voters
    103

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
39,401
Likes
179,047
Location
Seattle Area
This is a review and measurements of the discontinued Lynx Aurora 8 audio interface (DAC & ADC). I am told you can find them used for $700 although the price I see on ebay is $1000.
Lynx Aurora 8 Review External Interface AES DAC ADC.jpg

The owner wants to know if this can be competitive with modern 8-channel DACs so that is the functionality I focused on. Back panel has DB25 which means you have to also make or buy pigtails to XLR connectors:
Lynx Aurora 8 Review Back Panel DB25 External Interface AES DAC ADC.jpg


These cables are quite thick although the ones the member send me are surprisingly supple. While you can add a USB input card to it, as you see this one doesn't have one so I tested using AES output of my analyzer. That output is stereo only so you are only going to see channels 1 & 2 in the measurements.

Lynx Aurora 8 Measurements
Full output level is a healthy 8 volts so I knocked the digital input by 6 dBFS to get our nominal 4 volt out:
Lynx Aurora 8 Measurements External Interface AES DAC ADC.png


While I am not happy with the shape of the noise floor, which varied up and down, overall sum of distortion+noise lands the Aurora 8 in competent category.

Letting the output go up to 8 volts doesn't get you much more:
Lynx Aurora 8 Measurements THD vs Level External Interface AES DAC ADC.png


Dynamic range does improve though:
Lynx Aurora 8 Measurements DNR External Interface AES DAC ADC.png


IMD distortion is good:
Lynx Aurora 8 Measurements IMD External Interface AES DAC ADC.png


Jitter is not:
Lynx Aurora 8 Measurements Jitter External Interface AES DAC ADC.png


Then again distortion is kept in check:
Lynx Aurora 8 Measurements Multitone External Interface AES DAC ADC.png


Dac filter is what we usually see as the default in DAC chips:

Lynx Aurora 8 Measurements Filter External Interface AES DAC ADC.png


That produces competent THD+N vs frequency:

Lynx Aurora 8 Measurements Distortion vs Frequency External Interface AES DAC ADC.png


Linearity is near perfect:
Lynx Aurora 8 Measurements Linearity External Interface AES DAC ADC.png


Conclusions
Lynx had quite a brand in 1990s. It was revered by both professionals and audiophiles building PCs as the time. Their PCI cards were very much sought after. Good to see performance is competent sans whatever is going on with jitter.

For audio playback use, you won't be using the inputs yet paying for them. Add the cost of the USB interface and you don't really have a bargain here compared to likes to Okto 8 DAC.

-----------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/
 

Veri

Master Contributor
Joined
Feb 6, 2018
Messages
8,937
Likes
11,066
Overall not bad at all. If you can buy one for a good price, could be worth considering. But if you're building a new 8-ch set-up might as well get something (even) better performing. IMO.

Thanks for the review @amirm!
 

tktran303

Addicted to Fun and Learning
Joined
Mar 27, 2019
Messages
514
Likes
837
^^^

It would be very similar.

This is a 15 year old converter for studio use.
Pretty good in it’s day, I would say.

The timeline might be off a bit, Amir. Lynx is relative newcomer to the studio market; only been established in 1998.

They offered an affordable solution to the established players, the likes of Digidesign, Prism or Apogee converters.

You might be thinking of E-MU or ESI or Terratec; who have been around in the 90s.
 
Last edited:

martijn86

Active Member
Joined
Aug 30, 2019
Messages
217
Likes
730
Location
The Netherlands
That's a very interesting jitter graph. I call it Mount Jitter. The most interesting thing about it might be how overall performance still seems to be perfectly reasonable. So yeah, something is wonk, but you can likely use it in a variety of professional applications and never notice it.

I don't mind that it lacks USB because AES is such a great interface. I mean, it likens USB and equals consumer S/PDIF, without the need for drivers, with a low-cost, large, locking connector. I'd love it if more <€500 consumer devices used it.
 

Lambda

Major Contributor
Joined
Mar 22, 2020
Messages
1,626
Likes
1,366

martijn86

Active Member
Joined
Aug 30, 2019
Messages
217
Likes
730
Location
The Netherlands
You might be thinking of E-MU or ESI or Terratec
Jesus, you just took me back to 2004! My first DAC was a Terratec Aureon 7.1 FireWire. I had totally forgotten that my first DAC was an 8-channel DAC! 17 year old me loved the rocket engines.

Yes. It had rocket engines.
terratec-aureon-7-1_2.jpg


You think a Chord DAC looks strange? You should see this thing in a dimly lit bedroom. It drove a Sony MDR-7560 like a champ.
 

Lambda

Major Contributor
Joined
Mar 22, 2020
Messages
1,626
Likes
1,366
Terratec Aureon 7.1 FireWire.
Nice. FireWire and optical digital IO.
This thing is doing probably a good (perfect) job recording bit perfect digital of SPDIF with FireWire DMA and stay in sync.
(If you have the right firewire chipset/hardware , OS and drivers installed)
A thing many modern USB interfaces struggle with.

FireWire was Awesome for AV stuff but bad in many other ways.

Jitter is not:
Lynx Aurora 8 Measurements Jitter External Interface AES DAC ADC.png
@amirm was this measured using the BNC Word Clock in?
How audible is or better inaudible would this be?
We now we can’t read of a noise level from a FFT because of FFT gain.
This jitter looks verry "random distributed" and not periodic therefore true level jitter can’t be read from this graph?

it loos like peaks are at -110dB but they might actually be way higher but fft gain is "averaging" them down
 

mdsimon2

Major Contributor
Forum Donor
Joined
Oct 20, 2020
Messages
1,653
Likes
2,030
Location
Detroit, MI
I assume synchrolock was enabled and locked for the measurements? Would be interesting to see measurements with it enabled and without to see if Lynx's jitter suppression claims hold any water.

For measurement purposes is it possible to slave your AP to Lynx to see what this looks like using the Lynx internal clock?

Am also very interested in seeing dashboard and dynamic range measurements with the output trim set to -10 dBV, I think you should be able to get ~2V out at that setting and am curious if noise performance improves as expected.

Michael
 

martijn86

Active Member
Joined
Aug 30, 2019
Messages
217
Likes
730
Location
The Netherlands
FireWire was Awesome for AV stuff but bad in many other ways.
It was at times a driver hell in the Windows XP/Vista era but I was a starting AV student and it was so much better than USB 2.0. Recording DV tapes over FireWire 400, storage arrays with FireWire 800, 4 drives to get a terabyte. You know, because I needed it for school That's when I started getting spoiled for high quality audio and video equipment.
 

DualTriode

Addicted to Fun and Learning
Forum Donor
Joined
Oct 24, 2019
Messages
746
Likes
498
Hello All,

There is a Lynx sound card in my computer with XLR in and outs that I use with AISO and APx500 software that I use to measure speakers sans the APx555 fan noise.

Thanks DT

Lynx stuff is okay.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
39,401
Likes
179,047
Location
Seattle Area
I assume synchrolock was enabled and locked for the measurements?
I cycled through all the modes for clock. On anything but AES0, i.e. input signal, it would lose lock after a few seconds. Then the cycle would repeat. I want to say for the time it was stable, the output was a bit cleaner. I am thinking that their lock mode uses an internal clock that it only resyncs once in a while with the input causing that glitch (when the buffer over or underflows).
 

GTAXL

Member
Joined
May 15, 2020
Messages
33
Likes
36
Location
Mount Vernon, Ohio
Why wasn't the ADC tested in this? That would be the primary usage of this in the professional studio. Do note this is their discontinued Aurora from 2005. The newer (n) version has Hilo-like performance and has a Audio Precision SYS-2722 report.

Back panel has DB25 which means you have to also make or buy pigtails to XLR connectors:
Which is pretty common on multi-channel professional converters. Do you expect them to fit 16 or 32 XLR connectors for both input, output, and the AES/EBU channels?
Lynx had quite a brand in 1990s. It was revered by both professionals and audiophiles building PCs as the time. Their PCI cards were very much sought after. Good to see performance is competent sans whatever is going on with jitter.
They were founded in 1998 and are still very much relevant today, their Hilo and Aurora (n) are used in many commercial studios and used for top-records to this day. Their PCI Express AES16e is used by many studios that aren't wrapped around Avid's ProTools ecosystem. The studios that are end up almost always using Avid's 192 or HD I/O.

I cycled through all the modes for clock. On anything but AES0, i.e. input signal, it would lose lock after a few seconds. Then the cycle would repeat. I want to say for the time it was stable, the output was a bit cleaner. I am thinking that their lock mode uses an internal clock that it only resyncs once in a while with the input causing that glitch (when the buffer over or underflows).
Was the SynchroLock LED illuminated? I wonder what would of happened if you used the wordclock input.
 

GTAXL

Member
Joined
May 15, 2020
Messages
33
Likes
36
Location
Mount Vernon, Ohio
I don't mind that it lacks USB because AES is such a great interface. I mean, it likens USB and equals consumer S/PDIF, without the need for drivers, with a low-cost, large, locking connector. I'd love it if more <€500 consumer devices used it.
Back then USB audio wasn't as mature as it is today, plus the target market is the commercial studio, which if they aren't tied into the protools ecosystem, will use AES/EBU or ADAT. It makes sense this converter has AES/EBU only. I'm personally using AES/EBU on my home rig because my workstation has had issues with USB regardless of DAC or ADC I throw at it. I have a Lynx AES16e that I use to interface with my Topping D90 and Lynx Hilo and it's flawless, regardless if people say jitter will be slightly worse, it still sounds excellent and I wouldn't hear jitter at such levels.
 

mdsimon2

Major Contributor
Forum Donor
Joined
Oct 20, 2020
Messages
1,653
Likes
2,030
Location
Detroit, MI
I cycled through all the modes for clock. On anything but AES0, i.e. input signal, it would lose lock after a few seconds. Then the cycle would repeat. I want to say for the time it was stable, the output was a bit cleaner. I am thinking that their lock mode uses an internal clock that it only resyncs once in a while with the input causing that glitch (when the buffer over or underflows).
This is what they say about synchrolock in the manual. It has a wide range PLL that will lock quickly but synchrolock will take one to two minutes to achieve a final lock at which point the synchrolock LED will be solid.

The Aurora incorporates SynchroLock clock synchronization technology to provide extreme tolerance to noisy external AES/EBU and word clock signals while generating an ultra-low jitter clock. This technology is especially useful for combating noise induced on cables in complex studio installations. SynchroLock provides clock synchronization while insuring bit-perfect digital transmission. When the Aurora is connected in an AES/EBU daisy chain, SynchroLock acts like a jitter firewall to prevent the propagation of jitter to downstream devices.
By coupling statistical analysis with low-noise clock generation techniques, SynchroLock is capable of attenuating jitter on incoming AES/EBU signals by a factor of 3000:1. Compare this to attenuation of 100:1 or less for professional quality analog phase-lock loops (PLL). SynchroLock can easily handle AES/EBU signals with jitter levels in excess of 800 nanoseconds.
The SynchroLock sample clock is a two-stage system that is comprised of a fast-locking, wide-range analog PLL and digitally controlled crystal-based secondary stage. Due to extensive number crunching of the secondary stage, SynchroLock typically requires one to two minutes to achieve final lock. While the secondary stage is working, the analog PLL loop maintains lock, but with much less jitter attenuation than the secondary stage.

When the final lock state is achieved, the secondary stage is switched on line and becomes the system clock source. In some cases this switching process may cause a momentary disruption in digital I/O. Because of this, it is recommended that recording or playback not be started until the SynchroLock LED is on solid (see Section 2.1 Front Panel Controls and Indicators, item e).
SynchroLock works on any external word clock signal. By default, SynchroLock is active when the Sample Clock source is set to a clock source other than Internal. SynchroLock may only be disabled by external control of the Aurora. When disabled, the front panel SynchroLock LED will NOT be illuminated.
SynchroLock is capable of locking to word clock frequencies within +/- 100ppm of 44.1 kHz, 48 kHz, 88.2kHz, 96 kHz, 176.4 kHz, or 192 kHz. Signals that fall outside of the lock range will cause the SynchroLock LED to flash. In this case, the analog PLL is active and will provide the system sample clock. To re-engage SynchroLock, simply tap through the SYNC SOURCE options to return to the original source.

Michael
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
39,401
Likes
179,047
Location
Seattle Area
Why wasn't the ADC tested in this?
Because owner's interest was as an 8-channel DAC and reason I accepted it. With this product being discontinued, I don't see the purpose in doing a bunch more work when I have such a huge backlog.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
39,401
Likes
179,047
Location
Seattle Area
Which is pretty common on multi-channel professional converters. Do you expect them to fit 16 or 32 XLR connectors for both input, output, and the AES/EBU channels?
The interest from this owner and general membership is for hi-fi use. There, a pig tail requirement is competing with this:

index.php


I want to make sure people don't think the cost of the unit is all they incur.

That aside, if you are going to have a pigtail, give me at least a stereo set of XLRs. This is what RME does:

ADI-2-ProFSr-C.png


DB-25 was never designed for this use. It was a parallel digital I/O port. Extracting 8 channels out of this connector makes for one hell of a breakout cable. The individual wires have to be very thin to fit in the cable assembly making susceptible to breakage. Go ahead and make the box 2 or even 4RU and give me nice XLR connections. I don't want to mess with specialized cables like this. It is a hack, plain and simple.
 
Top Bottom