• 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!

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

Rate this DAC:

  • 1. Poor (headless panther)

    Votes: 6 5.7%
  • 2. Not terrible (postman panther)

    Votes: 36 34.0%
  • 3. Fine (happy panther

    Votes: 56 52.8%
  • 4. Great (golfing panther)

    Votes: 8 7.5%

  • Total voters
    106

Tom C

Major Contributor
Joined
Jun 16, 2019
Messages
1,501
Likes
1,370
Location
Wisconsin, USA
This is equipment for studio use though, that's why to evaluate it and review it for a different application and to ding it for not holding up to said application doesn't make sense.

Also, why does this site strictly have to be geared towards just home hi-fi usage? I thought this was "Audio Science Review" a site where audio equipment was objectively measured. I would love to see more professional units measured, and to be honest I think that's where such measurements have more weight, in the studio where the music is being created and quality must be top-notch. Not to mention home studios that have racks of professional gear. Not everyone is a normal home hifi consumer and into playback only. I am looking into creating a commercial studio and would love to see professional units objectively measured.
He who pays the piper gets to call the tune.
 

H-713

Senior Member
Joined
Jan 24, 2021
Messages
379
Likes
666
Why is it a premium? I will happily accommodate a taller unit so I don't have to mess with these specialized and pain in the neck cables.

95% of the places buying this type of equipment are not a "ground up" build. It's an upgrade to an existing install in a studio / concert hall / whatever, and finding 4U of free rack space is never easy in any established system. If there is 4U of free space, who's to say it's all in the same spot? I would much rather deal with a custom cable than remove everything from a rack, re-install it and re-wire / cable manage the entire rack in order to consolidate available space. When this happens I'm going to be really pissed if the hardware I'm installing is 80% air.

No matter what, using dedicated connectors would have almost certainly necessitated a second PCB in the chassis, which adds a lot of cost and doesn't really benefit the end user of this device.
 

hollis

Active Member
Joined
Sep 3, 2018
Messages
117
Likes
84
The garden hose cable would be a bear to dress in a rack as well. Good luck getting them in the length you want or need as well.
This is my unit and, this is the rear of my rack:
 

Attachments

  • 66397906332__565B03FD-4628-41A0-9421-13E7EA4D135B.jpg
    66397906332__565B03FD-4628-41A0-9421-13E7EA4D135B.jpg
    324.8 KB · Views: 124

hollis

Active Member
Joined
Sep 3, 2018
Messages
117
Likes
84
My primary purpose was my rejection of the Minidsp 88A analog inputs and ADC stage. That "forced" me into the 100% digital 88d, which meant I had 2x problems to solve.
1) how to get signal into the 88d: when I ran 4-way mains I used a Gustard U12 then U16 (both sounded identical).
2) how to get signal out of the 88d: 4x AES outputs so, I didnt care at all about usb, the Lynx Aurora 8 very nicely took all my outputs and converted them to analog. It also later gave me a very clean turntable D/A, which I then processed with Dirac and then out to the mains.

The front of the rack is much more visually appealing:
64913381317__73F217FF-AD12-482C-B192-BCE9F073117D.jpg
 

hollis

Active Member
Joined
Sep 3, 2018
Messages
117
Likes
84
Full gear list:
- PCIE lynx aes card
- 88d (top left)
- Topping A90 (top right)
- wattbox swiches for led lighting amps, and powered monitors
- allways on Furman for PC, DAC, networking
- dbx compressor/gate for Zoom mic
- switch
- Aurora8
- metal mesh office desk organizer (the Aurora is hot)
- VTV purify amp
- DIY Murphy Corner Line Array mains
- Behringer 2031a powered monitor surrounds

Right now I am testing out an Okto Dac8 instead of the Lynx, it would provide me an upgrade path beyond Dirac (into Audiolens/Accurate) and simply my chain considerably (at the cost of losing standalone non-pc playback). So, a great time to send this in for review.

The Dac8 (as a 4x AES DAC for the 88d) testing not going smooth. I have been running into issues where PC playback is fine but the Dirac test impulse generator wont play. Once I got that solved (by switching the Dac8 to USB mode and using the PC to change its sample rate and bit depth) I encountered very strong distortion after the Dirac file was loaded. After much trouble shooting I found the solution was unplugging all AES except 1-2 from the Dac8, playing some signal, then plugging in the remaining channels one at a time.

The Lynx Aurora8, when first purchased, had quite a few problems as well. Mainly stemming around DB-25 standards. There are 2x different standards for the pinout, Yamaha vs Tascam, and the unit itself is not labeled for what pinout it requires. I purchased multiple DB25 connectors over a series of weeks in search of the right one, when I found a working one, I stopped my search and celebrated. That is why Amir got stuck with a very tidy 2' output DB25, and a stupidly long 16' input DB25. Also my ebay-used Aurora8 has no screw in terminals on one DB25 (I forget which), but I had to reconfigure my rack to add a physical support below the unit so that the cable had strain relief.
 
Last edited:

hollis

Active Member
Joined
Sep 3, 2018
Messages
117
Likes
84
One last warning if you go this route, buying a rack is a must. Here is variant 2 of the rear (back when I had 4-way mains). Note, in this case the Aurora8 was clocked to the Gustard U16 with the purple BNC cable.

IMG_2970.jpg
 

McFly

Addicted to Fun and Learning
Joined
Mar 12, 2019
Messages
903
Likes
1,861
Location
NZ
image-asset.jpeg

Nothing says "unprofessional" more than a 1u rack unit that needs 2 power cables.

its a bit like buying a Unimog 405 as my grocery getter SUV and complaining about comport features
Wait, really? Wouldn't the two inputs be from separate PDU's fed from separate circuits in the rack, one being general mains and the other being an essential UPS/(+Generator) supply, so in a power outage, you don't lose all your processing, much like i dunno, every data center in existence?

To me, nothing would scream professional louder than processor with a redundant power supply setup.
 

mdsimon2

Major Contributor
Forum Donor
Joined
Oct 20, 2020
Messages
2,477
Likes
3,315
Location
Detroit, MI
I wonder how its ADC/DAC loopback measurement fares against other converters from that period?

Here's the measurement from my Pacific Microsonics Model Two:

View attachment 211516

Mani.

Any chance you know the FFT length of the measurement? THD looks good but difficult to judge noise without FFT length.

Michael
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,386
Location
Seattle Area
No matter what, using dedicated connectors would have almost certainly necessitated a second PCB in the chassis, which adds a lot of cost and doesn't really benefit the end user of this device.
There is a hell of a benefit to the end user in being able to find a cable in an instant to wire things up as opposed to the pig tail. As to second PCB, the new version of this interface costs over $3,000. Excuse me if I am not shedding a tear over the cost of a PCB.

I used to manage hardware development at broadcast audio companies (one was sister to SSL). We always knew it was a hack to use a DB connector for analog audio. We did it to save money for us, not because it was good for consumer. That the end user has gotten used to it is sad and not a reason to continue when even the computer industry has abandoned these connectors.
 

Lambda

Major Contributor
Joined
Mar 22, 2020
Messages
1,785
Likes
1,519
Wait, really? Wouldn't the two inputs be from separate PDU's fed from separate circuits in the rack, one being general mains and the other being an essential UPS/(+Generator) supply, so in a power outage, you don't lose all your processing, much like i dunno, every data center in existence?

To me, nothing would scream professional louder than processor with a redundant power supply setup.
Yes. i was sarcastic.
 

Geert

Major Contributor
Joined
Mar 20, 2020
Messages
1,937
Likes
3,526
Let's take a look at a few top interfaces to see what the status is 20 years later.

Apogee Symphony 8x8 (5.800€)
Symphony-8x8MPConnectSeries-module.jpeg

RME M-32-Pro (3.600€)
M-32-Pro-C.png


Metric Halo ULN8 (4.399€)
ULN-8_3d_top_Back.png

You don't like it? I understand, but it is a (the) standard in pro studio audio. And looking at the price of some of this equipment there's more going on than cost cutting.
 
Last edited:

Tom C

Major Contributor
Joined
Jun 16, 2019
Messages
1,501
Likes
1,370
Location
Wisconsin, USA
32 channels sounds like a good start.
8 channels in plus 8 channels out for 4.400 or 5.800 euro sounds like a lot, doesn’t it? Do you really have to pay that much?
 

xema

Member
Joined
Dec 1, 2019
Messages
93
Likes
80
32 channels sounds like a good start.
8 channels in plus 8 channels out for 4.400 or 5.800 euro sounds like a lot, doesn’t it? Do you really have to pay that much?
do remember this is a 20 years ago equipment.
channels were not as cheap as nowaday.
 

Lambda

Major Contributor
Joined
Mar 22, 2020
Messages
1,785
Likes
1,519
Do you really have to pay that much?
"Nobody Gets Fired For Buying IBM"
Every enterprise IT is mostly HP,DELL,Lenovo... and specifically there way more expansive Enterprise/ business line.
of cause they could have it cheaper if they get Acer PCs from bestbuy and they would maybe perform even better.
go figure

If you are a pro studio you buy a whole system from a trusted vendor with support. maintenance and warranty.
the cost of having someone doing maintenance is high, Down time is even more expansive and the worst thing is trouble shooting douing a live event/ recording.
 

Waxx

Major Contributor
Joined
Dec 12, 2021
Messages
1,931
Likes
7,688
Location
Wodecq, Hainaut, Belgium
It's also, Lynx is a known brand, and a proven design that is high quality, reliable and standardised (hence the use of DB25 connectors, but not only that). That is more important for bussinesses than the last 1% in performance or the cheapest price. Some newer brands may give better specs, but it's not proven in design and reliability so a bussiness risk like said above.
 

manisandher

Addicted to Fun and Learning
Joined
Nov 6, 2016
Messages
652
Likes
607
Location
Royal Leamington Spa, UK
If you still have a PM2 you should send it in or do an RMAA test.

Sold to a mastering studio in Paris 7 odd years ago.

Any chance you know the FFT length of the measurement?

I don't have the FFT size for the loopback measurement, but I do for the separate DAC and ADC measurements:

DAC (FFT 512):
DAC THD Test -10dB.JPG

ADC (FFT 1024):
ADC THD Test -10dB.JPG

Bear in mind that the unit was already a few years old when these measurements were taken back in 2003.

Mani.
 

mdsimon2

Major Contributor
Forum Donor
Joined
Oct 20, 2020
Messages
2,477
Likes
3,315
Location
Detroit, MI
Sold to a mastering studio in Paris 7 odd years ago.



I don't have the FFT size for the loopback measurement, but I do for the separate DAC and ADC measurements:

DAC (FFT 512):
View attachment 211608

ADC (FFT 1024):
View attachment 211609

Bear in mind that the unit was already a few years old when these measurements were taken back in 2003.

Mani.

Thanks for sharing!

For the ADC and FFT length of 1024, FFT gain is 10 x log(1024/2) = 27 dB. Looks like signal level at the ADC input is about -10 dBFS and noise floor on the FFT is around -130 dBFS. So noise is something like -130 + 10 + 27 = -93 dB, imagine this might be better if a higher level input signal was used and of course may also be limited by the analyzer noise floor.

For the DAC I think they are saying 0 dBr on the analyzer is 18 dBu (6.16 V) but it also looks like there may be a 30 dB notch. The input signal level on the FFT doesn't quite make sense to me at -49 dBr. My best guess at interpretation would be input signal level is -49 dBr but this is with a 30 dB notch so actual signal level is -19 dBr. FFT gain here is 10 x log(512/2) =24 dB and noise floor is around -133 dBr. So noise is something like -133 + 19 + 24 = -90 dB. Again signal level seems kind of low as -19 dBr to 18 dBu is 0.7 V.

If we assume your loopback (which must be a DAC to ADC loopback based on the analyzer configuration) is using a similar FFT length on the ADC I would say loopback noise is -129 + 7 + 27 = -95. So potentially better than the individual ADC / DAC measurements but also done at a higher level based on -7 dBFS at the ADC input.

I have always heard very good things about the PM2 so maybe my interpretation is a bit off or it could be that the measurements were limited by the analyzer. If anyone is more familiar with the R/S analyzers I would be happy to get another opinion.

Michael
 
Last edited:

msmucr

Member
Joined
Aug 16, 2019
Messages
34
Likes
53
Location
Prague, Czech Republic
Thanks for the review, Amir. It's nice to see studio gear among your measurements. It's a rare and I always appreciate it be it at your site or at some magazines (like UK's Sound on Sound). It's a pity you didn't have a time to measure also inputs like for all previous audio interfaces or A/Ds, even just for comparison at your reference measurement setup. Anyway, thanks for D/A :).

We had similar Aurora 16 many years ago in studio and I can confirm what L-Train posted before. This weird looking dither jitter test (increased skirting with clear sideband peaks) are most likely due to Synchrolock not completely settled. I used our unit with WC external sync from studio master generator and it behaved similarly, until it was completely locked. After that it was clean as a whistle and stable. Quite a few other vendors has similar mechanism to ensure quick device locking to external sync source with higher initial corner frequency of PLL, which is then progressively lowered to provide higher external jitter rejection. Essentially the mechanism is very similar to analog TV tuners when switching channels :) Of course it could be pretty elaborate implementation with two stage PLL, custom stuff going on FPGA etc.
If it was somewhat wonky on particular tested piece, then Aurora firmware version might be also worth of checking, it might well alter behavior of this. If I recall that correctly, all updates for standalone units (eg. without some USB or Firewire L-Slot card) were done via MIDI ports. So some decent (with good handling of SYSEX messages) MIDI interface can be necessary for the procedure.
Also Lynx has fantastic support engineers IMHO and they take their gear very seriously, which could help even with legacy models. That is one of reason I always like the company. Similarly like RME - no BS products with very solid performance you can rely on, fixing bugs and solving issues over product lifetime (even after that if there is no technical obstacle), forward thinking with regards to modularity.

Michal
 
Last edited:
Top Bottom