Computer Audio Asylum

Music servers and other computer based digital audio technologies.

Return to Computer Audio Asylum

best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona

98.210.153.55
Posted on October 31, 2009 at 20:23:40
andrej
Audiophile

Posts: 23
Location: San Francisco
Joined: October 31, 2009
 I have finally connected my Elgar Plus DAC to the Lynx AES16e-SRC sound card in a Mac Pro running iTunes and Snow Leopard. My system also includes a Verona clock and a Paganini transport (connected in the standard way, using DSD, with Verona as the master clock). In the future, I will upgrade to Scarlatti DAC and Upsampler (and maybe the clock), but for now I'd like to squeeze the most out of what I have. So I would be most grateful if you reaffirmed or corrected the following.

The Verona can only do 44.1 / 48 kHz. Elgar clock out is only 44.1 kHz. Lynx can only output at the clock's rate. So, for Lynx as a source, I cannot use Verona for anything but red book CD audio (presumably, it does not make sense to use Elgar's clock instead of Verona). Based on this, it seems there are two types of situations for listening to Lynx:

1. When listening to 16/44.1 from Lynx, I should use Verona as the master clock for Lynx and Elgar, with audio arriving to Elgar via AES1.

2. When listening to audio with sampling rates higher than 48 kHz, I need to clock Elgar based on the AES digital input from Lynx (via Elgar's MS:Slav menu option). And I should use the dual AES link for audio.

In both cases, I need to set the Core Audio rate on the Mac to the source file sampling rate, using Audio MIDI Setup. It seems I don't have to re-run iTunes after changing the Core Audio rate anymore?

Should I do something else to optimize the sound quality? The Lynx cable seems pretty flimsy, for example. I should probably use a RAID array of solid state disks, not hard disks? Do you know of a good document / web page / forum where somebody else has sorted this all out specifically for my hardware?

thanks,

Why compare Lynx AES to a Paginini??, posted on November 2, 2009 at 11:04:30
Presto
Audiophile

Posts: 2946
Location: Canada
Joined: November 10, 2004
 There are a few levels of PC based D/A conversion that would probably be a more fair comparison. The AES is a $1000 solution. I think the $2K Apogee Rosetta 200 (at a minumum) or perhaps something from Lavry Engineering would be a starting point.

Two things to consider here:

1) If you're going to compare a twenty thousand dollar standalone upsampler to a $1000 upsampling PCI card, you SHOULD expect something twenty times as expensive to win and win by more than just a neck!

2) If you're that concerned with upsampling that spending $20K is even an option, why not spend the $20,000 to hire a acoustical engineering summer student and PAY him by the hour to resample your music collection up to whatever you feel is the latest euphoric number (24/176, 24/192, 32/384...) using a NO compromise offline resampler program? There will always be compromises when you do resampling of high-res material in real time. Depending on quality level and dithering method selected, an off-line resampler can take 4x to 10x the playback time to do a proper job of resampling. Then, you have ALL of your files in ONE "super meticulously upsampled" format and you can then build a system around this sample rate and have no NEED for realtime resampling at all.

And then read this:

http://www.theaudioworks.co.uk/article.php?newsitem=44

"The Paganini system still lacks the absolute levels of resolution, scale and authority of the much more expensive Scarlatti system, and for those that require the very best available, only the Scarlatti will do."

Let's be serious here for a minute. The $20,000 Paginini is 'not so bad' but if you're a REALLY serious audiophile you need the Scarlatti? Puhlease. This is price-tag marketing.

The $20K upsampler is "ok".
The $40K upsampler is better.
the $100K upsampler is for REAL audiophiles.

...and the $30 offline resampling program which negates the need for a realtime upsampler altogether may be the ultimate solution - depending on how you value time versus how you value money. The only difference between time and money, in some instances, in the willingness to hire someone!

A single sample rate. Optimum EXHAUSTIVE upsampling. No realtime upsampling.

For what it's worth, I've never heard the Lynx AES upsampler. Then again, I've heard a lot of different realtime software upsamplers. The "free" ones available for music players as plugins (even the rare and exotic secret rabbit code) all have one thing in common:

I prefer them when they're not in use.

As far as upsampling using Resample? It's so good that in all the cases where I have a/b compared files before and after I simply can't tell them apart. I do NOT believe that higher numbers = better sound though, so I might be biased when it comes to upsampling. Then again, I don't have a $100K rig here so you might argue that my system can't resolve differences offered by the various types and offerings of upsamplers. I am not inclined to argue with "dCS guys" because they may be living in a realm of greater magnification of micro-detail and are adjusting a much finer knob on the 'audio microscope', per se.

Drop $30 on Resample. Resample some files. Compare an online resampled track using the Paganini to an offline resampled file using Resample (with no realtime Upsampling enabled).

I for one would be EXTREMELY interested in your findings.

If Resample matches your Paganini or even comes close, I would be inclined to invest the time in doing offline resampling.

Cheers,
Presto

It's not working!!, posted on November 3, 2009 at 10:29:46
Presto
Audiophile

Posts: 2946
Location: Canada
Joined: November 10, 2004
 I went to do some resampling last night to bit-compare different methods.

My Resample program is not working at all any more (I have not used it in years).

I will re-install and try again. I know it used to work flawlessly...

Cheers,
Presto

RE: Why compare Lynx AES to a Paginini??, posted on November 2, 2009 at 19:35:21
andrej
Audiophile

Posts: 23
Location: San Francisco
Joined: October 31, 2009
 | Why compare Lynx AES to a Paginini??

A very subjective reason only: Because I can use either Paganini or a computer as my source. So I wanted to know what audio quality I am sacrificing by taking the latter, more convenient option.

| There are a few levels of PC based D/A conversion that would probably be a
| more fair comparison. The AES is a $1000 solution. I think the $2K Apogee
| Rosetta 200 (at a minumum) or perhaps something from Lavry Engineering
| would be a starting point.

You are correct, the comparison is "unfair" given the prices of Lynx AES16 and Paganini, but it can technically still be made :)

I shall look into these other options (I do want a 24/196 solution, not just 24/96).


| Two things to consider here:
|
| 1) If you're going to compare a twenty thousand dollar standalone upsampler
| to a $1000 upsampling PCI card, you SHOULD expect something twenty times as
| expensive to win and win by more than just a neck!

Could not agree more in principle. The $20k Paganini actually cannot be used as an upsampler for anything but the disks playing in it because it does not have any inputs (unlike most of the other dCS transports). The Scarlatti Upsampler is actually about $10k. But it can only do 24/96 via USB from a computer, not 24/192 (which the dCS DACs can take via DSD or AES). And I want at least 24/192 if I am going to pay big bucks for the upgrade. Otherwise, I will just play with Lynx, maybe RME, to learn a thing or two and have some fun :)


| 2) If you're that concerned with upsampling that spending $20K is even an
| option, why not spend the $20,000 to hire a acoustical engineering summer
| student and PAY him by the hour to resample your music collection up to
| whatever you feel is the latest euphoric number (24/176, 24/192, 32/384...)
| using a NO compromise offline resampler program? There will always be
| compromises when you do resampling of high-res material in real time.
| Depending on quality level and dithering method selected, an off-line
| resampler can take 4x to 10x the playback time to do a proper job of
| resampling. Then, you have ALL of your files in ONE "super meticulously
| upsampled" format and you can then build a system around this sample rate
| and have no NEED for realtime resampling at all.

Again, I agree; and software upsampling is certainly something I will explore carefully (without hiring somebody; why would I pay somebody to have all the fun :) ). However, the question of how to do upsampling is currently not a primary limitation for getting my desired system. The primary limitation is getting 24/192 (or even DXD at 32/384) accurately out of the computer into a great DAC such as dCS Scarlatti (which I will probably purchase soon to replace my Elgar Plus). And unfortunately, even the dCS Upsampler does not do that with computer audio, it is limited by the 24/96 limit of USB. Therefore, one is reduced to using dual AES into the DAC, by cards like AES16 or RME equivalents. Optimizing this connection was my primary focus. And I wanted to see how far I can get compared to using Paganini.

| And then read this:
|
| http://www.theaudioworks.co.uk/article.php?newsitem=44
|
| "The Paganini system still lacks the absolute levels of resolution,
| scale and authority of the much more expensive Scarlatti system, and for
| those that require the very best available, only the Scarlatti will| do."|
| Let's be serious here for a minute. The $20,000 Paginini is 'not so bad'
| but if you're a REALLY serious audiophile you need the Scarlatti? Puhlease.
| This is price-tag marketing.
|
| The $20K upsampler is "ok".
| The $40K upsampler is better.
| the $100K upsampler is for REAL audiophiles.
|
| ..and the $30 offline resampling program which negates the need for a
| realtime upsampler altogether may be the ultimate solution - depending on
| how you value time versus how you value money. The only difference between
| time and money, in some instances, in the willingness to hire someone!

Perhaps. Still, I have to admit I am so impressed with dCS that I am not convinced that it will be easy
to find a software solution that does as good a job as they do in their hardware, even if one has in pri
nciple more time to do it offline. I doubt it that a person with knowledge sufficient to compete with dC
S would sell their software for $30 :) In addition, it is not clear to me what hardware dCS use for thei
r upsampling, so I cannot exclude a possibility that it would be difficult to reproduce their algorithms
on a general purpose computer in a reasonable time (sometimes hardware acceleration is worth a factor o
f 1000+).


| A single sample rate. Optimum EXHAUSTIVE upsampling. No realtime
| upsampling.
|
| For what it's worth, I've never heard the Lynx AES upsampler.
|
Lynx AES-SRC doesn't really have any upsampling for practical playback. The short descriptions of the AES16-SRC card are somewhat misleading in this regard. The card can only do sample rate conversion on the
inputs, not on the outputs.


| Then again,| I've heard a lot of different realtime software upsamplers. The "free" ones
| available for music players as plugins (even the rare and exotic secret| rabbit code) all
| have one thing in common:|
| I prefer them when they're not in use.
|
| As far as upsampling using Resample? It's so good that in all the cases
| where I have a/b compared files before and after I simply can't tell them
| apart. I do NOT believe that higher numbers = better sound though, so I
| might be biased when it comes to upsampling. Then again, I don't have a
| $100K rig here so you might argue that my system can't resolve differences
| offered by the various types and offerings of upsamplers. I am not
| inclined to argue with "dCS guys" because they may be living in a realm of
| greater magnification of micro-detail and are adjusting a much finer knob
| on the 'audio microscope', per se.
|
| Drop $30 on Resample. Resample some files. Compare an online resampled
| track using the Paganini to an offline resampled file using Resample (with
| no realtime Upsampling enabled).
|
| I for one would be EXTREMELY interested in your findings.
|
| If Resample matches your Paganini or even comes close, I would be inclined
| to invest the time in doing offline resampling.

Yes, this is exactly the sort of thing I will be trying out a bit more, though it will take some time. I heard that the Mac OS resampling is pretty good, so I will try it too. And Saracon (if it is not too expensive).

But even with resampling, I will still be fighting the lack of a single external clock for all components, because Lynx AES16 can only output at the rate of the clock, which in the dCS Verona clock case is fixed to 44.1 kHz (or 48 kHz), so I am forced to put the dCS DAC into the slave mode with respect to the AES input data from Lynx, a highly unsatisfactory solution. So, making full circle, I don't expect Lynx to be competitive with Paganini no matter what I do, which, as you said, is not surprising given the price differential.

Thanks, Andrej
There's fun to be had comparing various offline upsamplers..., posted on November 2, 2009 at 14:43:37
carcass93
Audiophile

Posts: 2634
Location: NJ
Joined: September 20, 2006
 I only compared 3 (r8brain, SoX and Adobe Audition) - and they all were clearly different, with Audition being best by wide margin, and r8brain just plain bad.

And if you take into account different settings in each app - that's even more fun! For instance, I preferred Audition with Pre/Post Filter enabled, while fmak - without (and graphs are on his side, BTW).

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 1, 2009 at 13:34:21
andrej
Audiophile

Posts: 23
Location: San Francisco
Joined: October 31, 2009
 Listening test:

Equipment:

- dCS Paganini SACD/CD transport, dCS Verona clock, dCS Elgar Plus DAC; all dCS equipment connected in the standard manner via firewire;

- Lynx AES16e-SRC in a Mac Pro (headless, without graphics card, Snow Leopard 10.6.1, 1 TB Seagate 7200 rpm in the JBOD mode on the Apple RAID card, 16 GB RAM, iTunes 9.02). AES16 is a clock slave to Verona, connected to Elgar via AES with standard Lynx cables.

Listening setup:

The A/B listening setup was quite easy to establish: I played exactly the same 16/44.1 material from Paganini and AES16. The comparison is not quite fair to AES16, as I have used upsampling to DSD in Paganini (but what I can do, this is what I would do in real life without a software upsampler on the Mac). It was then possible to start both sources at the same time, set the tracks to the repeat mode, and simply switch between them at any time with the Elgar remote; the switch takes a few seconds, with Elgar achieving the task very gracefully, with a nice fade in / out, without fault.

Source material: I used Chesky's The Ultimate Demonstration Disc red book CD, mostly Track 3, Spanish Harlem by Rebecca Pidgeon - it is a good track because it has high-resolution, airiness, and three-dimensional imaging. It also contains a good variety of instruments, including female voice, bass, violin, shakers, acoustic guitar, and piano. I also used other tracks on the Ultimate Demo disk, but I learnt all I did from Spanish Harlem alone. The SACD of the same track provided the ultimate reference.

Results:

The bottom line is that AES16 is not bad (without upsampling), but it is not as good as Paganini (with upsampling to DSD). The sound of AES16 is "mushier" and not as "clean" as that of Paganini. This is especially evident in the starting bass notes as well as in the singer's voice. The resolution is not as high, as evident from the shakers in the right channel. And the locations of instruments in space are not as well defined. It also seemed that Paganini was a little louder (perhaps for about 1 unit on the Elgar's volume control, ranging from 0 to -20), but this could be only a subjective impression caused by the slightly different sound quality.

For reference, the SACD from Paganini of course excelled over the red book CD in the Paganini. It simply made the attributes I listed above even better. The sound as a whole was higher resolution, clean, luscious, pinpointed in space. Bass was articulate, well defined.

Trying to assign subjective grades, if the Paganini SACD and Lynx CD define the range in sound quality from 1 to 5, the Paganini CD was much closer to Lynx CD (say at 4). I suppose this is not that bad, for a $800 card. But the difference is large enough that I will not actively listen to red book CDs from Lynx; it is good enough for listening in the background though.

In addition to the Paganini / Lynx CD comparison, I was again reminded of how great the SACDs are. There is no way that I would abandon Paganini SACD as the source, unless something better in sound (as well as convenience) comes along. I wish for Sony / Phillips coming to their senses and allow us to rip our SACDs., and for dCS to come up with an interface solution that will allow us to feed these ripped SACD (as well as other hi res) files into their DAC.

Finally, I believe various other people who said that 16/44.1 via dCS Scarlatti Upsampler USB connection sounds as good as a Paganini CD, but I can't believe that a company as great as dCS will not soon go beyond the current USB limitation of 24/96, so I am stuck with waiting on that to happen before spending big bucks to make my stereo computer friendly :)

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 1, 2009 at 22:13:18
cics
Audiophile

Posts: 1095
Joined: November 9, 2006
 Verona clock would work nicely with traditional transports like the Paganini as it's a high jitter source. Conventional wisdom suggests computers are too noisy - this is only true for default setups. Computer "Transports" (i.e. optimised solely for audio) changes jitter and realtime upsampling performance significantly for the better.

I run a cMP² + Lynx AES16 (192k dual wire output) > Scarlatti DAC. My Scarlatti Clock is not used as sound quality deteriorates notably (see here). I would suggest changing your AES16SRC to AES16. Comparing to other soundcards: ESI's $150 Juli@ runs it close (when comparing 96k input) and RME's HDSP9652 lags both by a small margin. EMU offers good hardware buy unfortunately has bloated drivers & software and doesn't offer true low latency.

Sometimes the Lynx AES16 is blamed for poor sound quality. My experience suggests it's the "messenger" and is more reflective of a poor underlying computer setup.

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 2, 2009 at 08:50:38
andrej
Audiophile

Posts: 23
Location: San Francisco
Joined: October 31, 2009
 i'll get informed about cMP, thanks for the link (btw, thanks also for the extensive jitter tests you published on another thread here). the problem with cMP for me is that I prefer to use a Mac (I have a Mac). Maybe I could run cMP inside a Parallels Desktop Windows emulator, but I am quite sure this would be horrible given how optimized cMP is.

Digging on the RME web site, I wonder if anybody has thoughts on the (relatively new) HDSPe AES card, which can do dual AES (the PCIe version of 9632, HDSPe AIO, cannot actually do dual AES).

Also, I appreciate your suggestion to replace AES16e-SRC with AES16e, presumably driven by the desire to minimize the unnecessary stuff on the card. But do you know the difference between AES16e-SRC and AES16e is significant? I just hate to go back to the dealer and bother him with replacement ...

FInally, I would not mind making my Mac Pro as good an environment for audio as possible. Is there a great thread or a web site that would get me going. I already did the common sense things, such as running the Mac Pro in a headless mode without even a graphics card, removing all expansion cards but disk RAID (and I should possibly remove even that one). I have 16 GB of RAM, and a 7200 1TB SASA disk (shall convert into SDD at the first opportunity). I turned off most services, including Spotlight indexing. But it has a fan which I don't know I will ever be able to get rid off without melting the machine. Is there a relatively simple objective way to measure how good a source a given configuration is (not by listening, but by measuring something coming out of AES16 or Elgar)? I do have some other equipment (eg, Mobile IO ULN2) and software (eg, SpectraFoo) that may have some functionality needed here.

Thanks! Andrej

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 2, 2009 at 22:46:45
cics
Audiophile

Posts: 1095
Joined: November 9, 2006
But do you know the difference between AES16e-SRC and AES16e is significant? I just hate to go back to the dealer and bother him with replacement ...

Less is more. Lynx uses FPGAs in its design. I would assume additional processing power would be needed for SRC processing - don't have specific details though. If you have the option to swap it out for a less complex AES16e, do it. Also, in your settings, disable Lynx's Synchrolock feature (its a digital PLL that's not used in playback).


Is there a great thread or a web site that would get me going. I already did the common sense things, such as running the Mac Pro in a headless mode without even a graphics card, removing all expansion cards but disk RAID (and I should possibly remove even that one). I have 16 GB of RAM, and a 7200 1TB SASA disk (shall convert into SDD at the first opportunity). I turned off most services, including Spotlight indexing.

The last time I worked on Macs was far too long ago (Motorolla based, Lightspeed C, etc..). You're on the right track but I don't know of any material for Mac audio optimisation. Removing the graphics card is great. Some suggestions:

  1. Lower RAM installed - in cMP, lowest possible RAM is recommended (256MB). Mac users get better results when adding more RAM which is a surprise however 16GB is overkill.

  2. Power your HDD externally - you can use similar components used in cMP. Another common option is to use eSATA and store noisy drives in a separate unit.

  3. BIOS setup is very important - don't know if Mac offers options here.

  4. Power supply improvements have large impact to sound quality - if standard ATX PSUs are used, you have options here.


At the software level, more things could be done to OSX and the player but this needs Mac specific knowledge. XP with Minlogon is a superb platform. Linux offers similar tuning capabilities (done by Soundchek) just not sure about the Mac.


Is there a relatively simple objective way to measure how good a source a given configuration is (not by listening, but by measuring something coming out of AES16 or Elgar)? I do have some other equipment (eg, Mobile IO ULN2) and software (eg, SpectraFoo) that may have some functionality needed here.

Improvement will be seen in less Jitter Distortion at the analogue outputs. Use a HF input tone (at lease 14kHz) and resulting spectrum will tell a great deal. For example, in my measurements, adding a USB memory stick caused a 3db rise in the noise floor. Likewise, adding the Scarlatti Clock added 4ps to Jpp.
RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 2, 2009 at 00:08:03
fmak
Audiophile

Posts: 4099
Joined: June 1, 2002
 The 9652 is a 96k card. You cannot compare its sound quality with higher sampling rate cards like you seem to have done. The Julie@ doen't even have a proper spdif output connector (miniplug) and looks difficult yo alter. The AES 16 has a HD26 plug which is inherently poor. The 9632 can have a 75R BNC fitted in a few minutes with 2 cm of wire. There is no point in comparing sound card digital outputs without clcok cleaning or regeneration of the clocking stream.

Software cannot compensate for hardware defects as endemic in PC/MAC ware.

You are also not using the dCS equipment as intended.

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 2, 2009 at 06:24:04
Dawnrazor
Audiophile

Posts: 7499
Location: N. California
Joined: April 9, 2004
 Fred,

The Julie@ doen't even have a proper spdif output connector (miniplug) and looks difficult yo alter.

Assuming you meant to say "...difficult TO alter" I beg to differ.

There are many many posts on the Juli@ about adding linear psus, taking I2s out of it, and here is a post that adds a bnc output. You may have missed all these as they show up in the cMP2 and cPlay threads, which you have dismissed out of hand


RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 6, 2009 at 05:02:00
fmak
Audiophile

Posts: 4099
Joined: June 1, 2002
 The Juli@ also causes the kiss of blue screen death! Look at it in the esi forum.

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 2, 2009 at 09:08:12
fmak
Audiophile

Posts: 4099
Joined: June 1, 2002
 I only respond to posts as they are written. If you wish to speak for cics having made these changes, please first consult him and then make this clear. cMP makes no mention whatever of hardware issues in relation to the digital output of sound cards. It also changes from edition to edition.

In the link you have given, connectors with the wrong impedance (50R) have been fitted. This affects the sound significantly in a good system.

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 4, 2009 at 08:06:05
Dawnrazor
Audiophile

Posts: 7499
Location: N. California
Joined: April 9, 2004
 Fred,

In the advanced section there are many things. One of which is a section that has links to other peoples updates and additions to cMP2.

Some are powersupplies and some are mods to the Juli@ card. These are not from cics but he HAS included them for people who want to address these points.

There are no editions of cmp2. It continues to evolve though.
I prefer this, posted on November 2, 2009 at 06:42:59
ForgotPassword
Audiophile

Posts: 58
Joined: September 21, 2009
 Shorter connection, and sturdier pins.



"We should no more let numbers define audio quality than we would let chemical analysis be the arbiter of fine wines." N.P.
RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 1, 2009 at 20:34:32
Dawnrazor
Audiophile

Posts: 7499
Location: N. California
Joined: April 9, 2004
 You might want to try the Amarra demo instead of iTunes to see the difference. That probably would help your lynx alot and I think you'll be surprised at the effect software has on the sound.


RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 1, 2009 at 09:26:48
drrd
Audiophile

Posts: 80
Joined: January 5, 2007
 I'd agree with fmak about the Lynx card. I've owned a Lynx Two and AES16, I liked the sound of the Two just as a sound card but as a digital source I just don't think either are up to much.

dCS make great dacs but god they let you know if the source isn't up to it. My previous 954 and Elgar dacs were not very tolerant though the Scarlatti sounds very good whatever you put in. It can sound great though but not with AES/SPDIF input in my experience. I tried an empirical audio offramp and pacecar, better than the Lynx but you can do significantly better.

I'd probably have to disagree with fmak on the 974 having owned one too. It's good, you can use 3-wire sdif/dsd to the Elgar which beats AES/SPDIF but I don't think it offers any great step up in SQ, just a 'different' sound to my ears.

The new USB upsamplers sound really excellent. With the async USB connection to a PC or MAC and firewire to the DAC I find even a 2 quid firewire cable sounds better than high-end AES or SDIF BNC cables.

I made an interesting discovery recently. You're supposed to need to sync dCS units up properly or you get glitching, either word clock or have the DAC sync to AES/SPDIF signal. Well with my Scarlatti DAC and Upsampler I set them both to internal clock (master) and firewire connection and USB to the PC. No glitching whatsoever and it sounds better than either dac -> upampler sync or with the Scarlatti clock sync. I'd be interested to know what happens if you try the same thing with your Paganini transport and Elgar.

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 1, 2009 at 10:35:13
andrej
Audiophile

Posts: 23
Location: San Francisco
Joined: October 31, 2009
 Thanks a lot for your comments!

Both you and fmak recommend re-clocking between a sound card (either AES16 or 9632) and a DAC (if that is what fmak meant by relocking). What would be a good device to do so?

I would prefer a solution that works at > 24/96 and with Lynx AES16 (via AES, I suppose), although a 24/196 USB-based device for a Mac that results in better sound than Lynx or 9632 would be ideal (on its own or in combination with something else).

Without the need for more than 24/96 (which I have), a Scarlatti Upsampler for a USB connection to a computer would be my choice today (paired with a Scarlatti DAC, as opposed to Elgar Plus), knowing that I'd come pretty close to the best anybody can do, at any price.

I'll try to answer your question at the bottom a little later today.

thanks, andrej

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 3, 2009 at 04:35:06
drrd
Audiophile

Posts: 80
Joined: January 5, 2007
 What happened in your system with the unclocked firewire connection I described?

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 3, 2009 at 07:26:26
andrej
Audiophile

Posts: 23
Location: San Francisco
Joined: October 31, 2009
 sorry, so many things to try, have not done it yet - but i will, probably this afternoon.

i also learnt a few other things in the meantime that i can easily do to improve the situation, so i will have to repeat my listening tests anyway (for example, take the hard disk out of the computer, disable Synchro Lock in the Lynx, take out the RAID card, try software upsampling).

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 1, 2009 at 11:05:32
drrd
Audiophile

Posts: 80
Joined: January 5, 2007
 I wouldn't recommend 'reclocking' between a sound card and the DAC to be honest. As I said I think you get a different sound but not necessarily better and nothing like what you would get with the Scarlatti USB connection.
I think fmak has a dCS 972 but I'm not certain on that, I've owned and used a 974. You can't buy either new as dCS don't make them now but the Scarlatti upsampler is better even with standard connections, with the async USB connection it's a lot better.
RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 1, 2009 at 08:35:13
andrej
Audiophile

Posts: 23
Location: San Francisco
Joined: October 31, 2009
 many thanks for your response!

for now, my overall attitude is to start playing with a computer source for my stereo system, but still go back to the Paganini transport if the computer is not as good (until dCS or somebody else figure out a way to get a competitive 24/192 audio from a computer conveniently, presumably via USB, maybe USB3). i just installed the Lynx card yesterday and i haven't done real listening tests yet, but I am reading this forum to get ideas and to limit things to try. in this light, i am less motivated to get a dCS 974, since this means serious money, and I am just playing with computer audio (for now).

thanks for the RME 9632 suggestion - I may try it since it is not horribly expensive and would not change the overall "logic" of my setup. i also would not mind at all processing my files with a software upsampler; disk space is cheap. What is the best software upsampler (for a reasonable price)? Saracon?

What is the best computer audio card for feeding the Elgar DAC up to 24/192 kHz, presumably via dual AES?

i used the apple RAID card for a single 1 TB Seagate 7200 rpm disk to put it into enhanced JBOD mode (more buffering, battery backup during crash), because i just happened to have the card, but also motivated by my vague notion that it is good to make the output from the computer as "smooth" as possible (I glanced through a technical discussion on this forum about jitter being minimized primarily by focusing on the computer side). But I have no idea if RAID makes any impact on audio quality, haven't done listening tests yet. That said, I would love to remove as much stuff as possible from the Mac Pro (I am already running it in the headless fashion, without the hot graphics card, logging in via Screen Share from my MacBook), so I will probably remove the hot RAID card too. This should reduce the fan action at least. A lot of people are recommending SSDs, but that would be a little too expensive for now, although I may try playing with it with a subset of my music.

thanks a lot again, andrej

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 1, 2009 at 09:06:01
fmak
Audiophile

Posts: 4099
Joined: June 1, 2002
 It takes a lot to get PC audio to sound as good as your dCS system.

I use a PC and I have progessively gone 'backwards' in terms of computers. I am run the 9632 on an Atom 330 fanless although it was a pain to get an adequately low cpu temperature. I am not even using XP SP3; just SP2 which is some 2 GB less than SP2. I paid a lot of attention to the power supplies within the system as well. This way I have good sound with the dCS954, a Twisted Pear Buffalo dac with 7 power supplies, and a UA 2192 dac. All PC outputs are relocked and non PC electronics go through mains regenerators. PC electronics go thru an isolating transformer.

If you use a 9632, do relock the output. It is the same with the Lynx or any other soundcard.

Just for the hell of it, I have just tried a VIA Epia ME6000 fanless PC with the Musiland 192k usb box. the whole thing cost less than $200 and sounds quite good. This is with the same attention to details. The ME6000 actually runs on 60% cpu load playing 192k files but is fanless and 'cool'! The Atom 330 runs at about 3-4% and gets 'hot'. No matter what they say about quiet fans, there is a difference in sound due to interference with the LS output. Basically there are no quiet fans.

Enjoy yourself but pay little attention to the hypes here. Just trust what you have experienced yourself.

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 1, 2009 at 09:49:49
andrej
Audiophile

Posts: 23
Location: San Francisco
Joined: October 31, 2009
 how would you relock the output from AES16 or 9632?

i appreciate the idea of a fanless PC, but for now, at least, i will stay with a Mac. i will try to minimize the "noise" in it, by removing the RAID controller as well as the graphics card, and eventually using SSDs. it might be beneficial to use a single low speed CPU, to further reduce the heat, assuming one cpu is plenty of compute power for audio. i have a dual CPU quad core Xeon Power Mac running at 3.2 GHz, with 16 GB of RAM (late 2008 model).

thanks again, andrej

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on November 1, 2009 at 10:58:07
fmak
Audiophile

Posts: 4099
Joined: June 1, 2002
 The dCS974 relocks, but the ppls are quite wide intentionally to enable the unit to lock onto poor sources. Unlike drdd, I find the 972 much better than the Purcell both sonically, controlwise and in terms of waveshape fidelity. You should explore the sdif 2 75 Ohm 3-1ire connection if you get one. This way you avoid the Firewire cable and can do dsd.

There are not many good relockers around. I use a Big Ben, with improved internal power supply. I also have the Digital Lens and the Meridian 518. Inmates here will have you believe that usb and I2S are the only good 'modern' ones but these are just as, or even more sensitive to cabling and so bang goes this hype about low or no jitter being the reason.

This clocking thing a a bit of a lottery and no 2 clocking schemes sound the same. I have no issues whatever with dCS sync at all, except for the AES16 (many have this problem). Trial and error and cabling are essential.

RE: best way to use Lynx AES16e (in a Mac) and dCS Elgar Plus / Verona, posted on October 31, 2009 at 23:20:52
fmak
Audiophile

Posts: 4099
Joined: June 1, 2002
 You may benefit from using a dCS972/974 as the processor for the AES16 output. This will relock the Lynx and allow you to do sample rate and word length changes with all manners of control. The 'Audio rather than Pro' dCS upsamplers do not allow you the control or provide the facilites.

I am not impressed by my AES16's sound quality, which I have tried quite hard to improve, including making my own dual wire cable, disabling all other outputs and leaving the WC input/outputs only. The power supply regulation onboard the Lynx is also not good, using 78Mxx chips to power the clock, for example. Comapred to the RME 9632, the Lynx lacks 'transparency' and bass resolution. It is pleasant and can sound nice, but not 'great'.

Another strategy is to sell the Lynx; move to another card and do SRC offboard or use a good software package such as Audition 3.01, Saracon, or whatever. I just have a no of 2.5 in 500GB external Firewire/Sata drives to hold the upsampled music which I plug in at will.

There is no need for RAID in a music system and this also adds to possuble backup and restoration issues.

Page processed in 0.228 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2009, Audio Asylum, all rights reserved ] [ Go Green ] [ Take5 ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]