ULN-8 shootouts: Next!

Barry,

Thanks for sharing your comparison results.

I still haven't listened to the '8' yet. And since I didn't partake of the loyalty discount, I may not want to since $6k for a single component is out of my league.

Good to see you posting here.
Clay

Out of interest, how did you connect the Berkeley? And what other components did you use in the test?

Eloise

Hi Clay,

Thank you for your kindness.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Hi Eloise,

We listened to the Berkeley two different ways:

One way was with a recent vintage MacBook Pro running Amarra, feeding a Weiss DAC2 via Firewire and taking the AES out from the Weiss to the AES input of the Berkeley.

The other way was with the AES out from the ULN-8 feeding the AES input of the Berkeley.

The AES feed to the Berkeley was a Kubala Sosna AES cable.
All other cabling was by Nordost.

Monitoring was via my Magnepan 3.6s, augmented below 30 Hz by a pair of Outlaw LFM1-EX subs.
A photo of part of the system can be seen on the home page at
www.barrydiamentaudio.com and on the "Studio" page of the same site.

Source material was from various CDs and high res sources, including some HRx recordings by Keith Johnson and some of my own recordings. Having created the latter myself, I find these extremely informative in evaluating components, rooms and systems.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Thankyou for that Barry ... I was just curious.

Eloise

Did you try that setup, too?

I am curious how you felt the ULN8 went against the Dac2 and also the Dac2 vs the Berkeley. I guess the answer would be not so well - the fellow is using a Berkeley!
Thanks for posting.
Cheers
A

I'm also interested to see how the DAC of the ULN-8 compares to the Weiss DAC2 (sonically).

When reading the ULN-8 specs I was disappointed to see that the drivers are Mac only. the Weiss has a big plus there with Mac and Windows support.

Respectively wondering if you had the time and resources to interface a Zalman PC using Sampitude 10 Master and an Alpha DAC in your comparisons. BTW, one of my Zalman computers uses 4 SLC SSD using hardware raid 0. Have you come to the conclusion the best sound from a Mac is via firewire and a MacBook, Weiss, Metric Halo, Sonic, or TC Electronics?

Which version of Amarra did you use? Perhaps a software I/O version may influence the sound as greatly as a digital filter. Just thoughts......

Regards,

Tim Marutani
Emeryville, CA

Cool post! Great pics of the studio and the gear. I love the Magneplanar speakers....they are my favorite speakers. Thanks for your help with the vibration issue. I was going to order the tubes, but desided to try a square air bag and got busy and drop the ball on the project. Its coming though.

I'll send you an e-mail with some ideas regarding "I don't fully trust the comparison yet and am trying to see if something I did put the Berkeley at an unfair disadvantage."

Regards
Jesus R
www.sonore.us

Hi Ted,

We didn't listen to the Weiss in this test.
Our prime interest was in comparing the BADA and the '8 as I'd never heard the BADA and my guest had never heard the '8.

He did mention he thought the Weiss sounded bandlimited and lacking in definition compared to the BADA.

I heard the Weiss in one of the blind comparisons done with the ULN-8 a few years ago. I chose to live (and work) with the '8.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Hi Tim,

I use Macs and PCs in my life but for audio, I have chosen to use a Mac only. For my ears, a Mac with a Firewire connection to a Metric Halo interface has provided the best sound I've yet experience. When listening to recordings I made myself, this combination easily get me the closest to what I heard at the sessions.

To answer your other question, we used the latest, full version of Amarra.

While I hear different sounds from each software application I have used, I would not agree the influence is of the same magnitude as switching the digital filter. At least in the instance of last Friday's filter change on the BADA, the difference was much larger than any change in software would provide.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Hi Barry,
Did you have a chance to compare ULN-8 with Prismsound? Seems that they are of the same league but the comparison between the 2 is lacking. Your input is appreciated

Yours

Athur

"...some ideas regarding "I don't fully trust the comparison yet and am trying to see if something I did put the Berkeley at an unfair disadvantage."

JR,

This is something that the whole group would be interested in hearing. Why not post it here?

Oh wait, maybe you think that AES through a Lynx card is better than Firewire? :)

clay

Hi Athur,

The Prism was one of more than a dozen "contenders" involved in the blind comparison with the ULN-8 I heard a few years ago.

I found it sweet sounding, perhaps even "silky" is not a bad descriptor. Definitely a fine unit.
I purchased the ULN-8.

I find many of the best units to sound quite fine, especially when they are heard without switching directly to a ULN-8. ;-}
It is that switch to the ULN-8 (and even more so, back to the "other") that I find to be the key.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

"Did you have a chance to compare ULN-8 with Prismsound? "

Barry, maybe you wanna recap the devices that were tested in the large blind shootout? I believe it was referred to on gearslutz, but I never saw the thread.

Clay

Hi Clay,

I would seriously doubt the Berkeley absolutely requires the Lynx card to perform the Firewire to AES conversion. If it does, I would see that as a big red flag. But I give the designers much more credit than that.

I think they *recommend* the Lynx card. I seriously doubt (and really, really hope) they don't "require" it.

Frankly, I personally see the absence of Firewire on the Berkeley as a limitation.

Just my perspective.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

I should have an opportunity to compare both Weiss Dac II and Sonics Model 4 this week, both will be firewire from a mac pro running Amarra, If any one's interested I could post something ? I have been trying to get hold of a Berkeley , I am hopeful that they will hae an ROHS compliant version soon,for us Euros.
Keith.

Post away!

"I would seriously doubt the Berkeley absolutely requires the Lynx card"

Understood, Barry, I was chiding/joking with JR. I was guessing that the 'disadvantage' he might see is in not using the Lynx card from the Mac (G5 required) directly to the BADA, without the 'conversion' from Firewire.

Personally, I view S/PDIF & AES as an inferior interface (due to non use of local clock) to Firewire from a theoretical perspective, world class implementations notwithstanding.

Clay

I do hope people will use the Lynx AES16 into the BAD or what ever other AES input they may have. This may seem self promoting to you. However, I have reached out to Berkely and THEY seem to have a preference in the Lynx AES16 audio card. I asked them about another card and they said we could do better with the Lynx AES16 audio card in to the DAD. They expained one had 35 times more jitter.... I don't presume to tell you or anyone else that firewire is better or worse than AES. I think they are both the best options and I also like Gordon's usb implementation.

FYI I don't use the AES or the firewire......I use the L22.

JR
www.sonore.us

an that cd ended up being the source. That is unfortunate (driver issues I guess). I believe both the BADA and Weiss (which I own) shine in HiRez, although I'm sure the "8" does to. Any plans to redo? Thx
Ted

"I do hope people will use the Lynx AES16 into the BAD or what ever other AES input they may have. This may seem self promoting to you."

Hype would probably be considered as self-promoting, but not hope.

I probably shouldn't even have posted that last bit. I was hoping (there's that word again) that the potential self-promotion concern would not prevent you from posting the 'ideas' you had, and then I guessed - light-heartedly, I thought - what those ideas might be. :) Apparently that didn't come across.
Apologies
Clay

hi everyone, I thought I'd add my 2c to the feedback from friday evening's listening with the BADA and ULN-8. First a big thanks to Barry for being a great host. A wonderful guy and was fun a time hearing his studio setup. A very clean, revealing and transparent sound - I think I could have stayed there a long time listening to music (a very different sound with the Maggies than I'm used to)!

The two DAC's were set up with a preamplifier such that they could be level-matched and instantly switched (no changing cables, setup or level-matching between switches, which was a big advantage in hearing the differences).

We were having issues getting the Mac+ULN8 play high res, but apparently this had to do with the latest beta drivers for the 8 and so were unable to do a comparison using hi res and ended up using CD's via a Sony CD player. I would personally really like to hear the setup again configured for hi rez (esp HRx and some 192 music).

As Barry mentioned, initially the 8 easily outperformed the BADA (transparency and detail, instrument separation, better attack and decay, deeper soundstage). This was using filter 1.16 which is the default filter for the BADA using redbook material (1.24 is the default for all other sample rates). When we switched the filter to 1.24, the differences were far smaller, and to my ears much more subtle and required careful listening. Interestingly according to the BADA manual, these are the same filters with different spacing of the HDCD flags (unsure of implications). With choral music (Rutter's Reqiuem; Stephen Layton's Lux Aeterna), the separation of the singers was greater and the imaging more accurate (felt more in the recording space). In the album Melos (Vassillis Tsabropoulos), the leading edge of the violin in tracks 1 and 4 we listened to had greater detail, and could hear the subtle cues like the breathing of what I'm assuming was the lead violinist in the left of the soundstage. We also listened to a CD-R of Lift, which Barry mastered, and could hear the subtle differences, particular track 4 which we spent more time on (guitar string attack, soundstage imaging, etc.).

In addition to checking the drivers/firmware of the ULN-8 to make sure they play nicely with the BADA, there are 2 other points of focus for me to make sure the comparison is fair: (1) the break-out cable from the 8 which was connected to the Kubala Sosna Emotion AES/EBU cable into the BADA, and (2) vibrational isolation of the BADA. On point one, Barry wanted to take another look to make sure it was configured properly. On point 2, given the limited rack space we had to put the BADA on top of the CD player. The CD player was the only source we used for comparisons and so we had the drive vibration coupled to the BADA (I put Stillpoints between the two, but the ULN8 was both floated and had rollerball-style isolation beneath; the CD player also was floated and had the rollerball-style isolation beneath it). The differences were close enough that they could be explained by issues in any of the 3 points, but since I haven't listened to the sonic and resolving differences of a dozen+ DAC's in a comparison, I'm not in a position to judge this. But I can say when I better isolated the BADA, I did get a big change in sound (stage, detail, etc.). Changes in power cords also made very significant differences in listening to the BADA and Weiss during my period of ownership. Another upside for the 8 or Model 4 in this regard is it's use of a DC input - one could add a custom power supply from Paul Hynes or Bolder Cable and probably achieve even better results.

A "mini-rave" in NYC is being planned by Ciamarra in Sept, and will have the same DAC's (as well as the Lavry and Weiss) setup for listening, so will be another chance to listen for differences between them. In this case they'll be compared based only on AES output from a Lynx card and so may be a better comparison of just the DAC function of these units.

Congrats to both Metric Halo and Berkeley Audio for such fantastic products! Look forward to hearing more from both of them :) We didn't have time to connect the Weiss DAC2 as we were running late into the evening but I'd love to do that as well in the same system (maybe we'll get hirez and the weiss up and running for another session). The Weiss in my system has a wonderful, rich sound that I can see many loving and potentially picking in blind test as well.

I find isolation on a CD player/Transport has it biggest effect on the separation between instruments, a black background, and attack/decay. Subtle but significant enough... more so as you move up to the level of gear represented by the MH ULN-8 and the BADA.

Thanks Barry and silverlight for the much needed comparison.

They make cellos sound like violins. That's one bright top end! LOL...just kidding.......but Melos's Anja Lechner plays the cello, not violin, Geoff :) It's one of my favorite late night discs to spin. Srajan (6moons) turned me onto it.

I'm very interested to hear about more comparisons like this, using hi-end DACs and hi-end ears. :) I hate the word "shootout" but in this case know that Barry and Geoff meant no harm with it; these DACs...like Weiss/Berkeley/Metric are all upper class achievers from strong professional pedigrees, and are pushing the art.

I think Ted mentioned this and I was going to ask, "Wasn't that a cello?" ;-}

I'd just like to add a few things of my own if I may.

While the change in filter did bring the two much closer together, I still heard the same differences between them, still in favor of the ULN-8. To my ears, the filter change just made the changes relatively small compared to the rather large differences we heard with the initial filter.

Having listened extensively to various vibration control measures (and finding good vibration control to be important enough that I use it throughout the system), I personally do not attribute the differences I heard to vibration induced issues. The filter change alone made a larger difference than any vibration control methodology (or software change) in my experience.

All that said, I would like to do the comparison again some time, perhaps with completely independent systems playing the same files. One path through the ULN-8 and the other through the BADA, taking the analog outputs of both through the common line stage and on to the amplifiers, as we did last Friday. That would certainly eliminate any question regarding the DB25 to AES adapter I use. (Incidentally, the following day, I used the same adapter to take AES out of the ULN-8 and feed it back in, using the HRx files. There was no problem.)

I would certainly not reach any grand conclusions based on the one test - other than to suggest that anyone interested in the BADA might want to seriously audition the ULN-8. And maybe for myself, smile a bit (once again) at having a ULN-8 as a core component in my system.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Barry-

For playback with the ULN-8, wouldn't you normally use the Metric Halo software rather than Amarra/iTunes?

Aloha,

Dan

The other cool fact about the ULN-8 / Model 4 of course is that it's a studio workhorse and provides a lot of functionality (recording, multi channel, etc.), and for audiophiles with great analog rigs, a great means to convert to high quality digital. As for my Melos mention, that's what I get for writing a set of notes after a string of late nights and long days! thanks for the catch - will have to enlist a proof reader :)

Aloha Dan!

The Metric Halo software is not playback software per se (i.e. one cannot assemble "playlists" etc.).

To audition recordings or gear, I use any of the applications I would normally use in my work (soundBlade, Wave Editor, Peak, DSPQ, etc.).
Any of these will be routed via the MIO Console software, which I use with the ULN-8. (Know however, the ULN-8 can be configured to run standalone if desired.)

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Barry-

Mahalo for the wikiwiki reply.

How does the playback SQ of soundBlade, Wave Editor, Peak, DSPQ, etc. compare to Amarra/iTunes?

Aloha,

Dan

Barry says:
"Know however, the ULN-8 can be configured to run standalone if desired."

Ditto for the ULN-2, this means that after setup, the DAC will re-boot to the last 'state' WITHOUT need for any user intervention. Great for audiophiles. I never touch my ULN-2, not even for changing inputs from Firewire to Toslink (from Apple TV).

clay

I own the Amarra Model 4, which is a re-branded and slightly modded ULN-8, and I have done two separate comparisons of that unit to the Alpha DAC. The methodology was different, in that I used a Mac Mini with Amarra software connected by Firewire into the Model 4/ULN-8, and then took the digital output in AES to the Alpha DAC. I am not aware of any downside to this method, but I would be curious to hear if that handicapped the Alpha DAC (or helped it) by feeding it something different than what the Model 4 received.

In any event, I preferred the sound of the Model 4/ULN-8 in both instances over the Alpha DAC. Most of the source material was hi-rez, including HRX, Kent Poon's disc, HD Tracks, and Blue Coast recordings. I was using headphones and high quality headphone amps for the comparisons, and I heard the same kind of differences in extension, detail/definition, and soundstaging that Barry D mentions above. They both sound fantastic, but I preferred the Model 4 and thought it had the edge in sound quality.

I really was amazed, to be honest, because I expected to prefer the Model 4 plus Alpha DAC setup after the hype and an earlier demo of the Alpha in a great system, and also from not knowing the Metric Halo gear personally. I do not expect to add the Alpha or any other DAC, although I am still hoping to borrow the Alpha from Tim Marutani for a demo in my home system with speakers. Plus, there is the anticipated flagship Spectral DAC that may materialize someday. Always curious, of course. ;)

Hi Voltron,

Since the Alpha cannot except Firewire (a lack from my perspective), AES is the optimal way in.

Your hookup methodology sounds the same as for part of the evening, I AES out from the ULN-8 feeding the Alpha.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Hello Barry. I agree that the Alpha DAC should strongly consider Firewire or at least USB2 given its most obvious use. My main concern with using the Model 4 as a transport was whether it did anything to the signal before sending it to the Alpha. I did not think it did anything to change the digital signal (and certainly nothing to degrade it) but I was unsure.

I also use the Model 4 for needle drops, and it is fantastic. I am still trying to figure out the nuances of SoundBlade and how best to cut up tracks in the Sonic Console, but the Amarra software is supposed to unify all of that in future releases which will be good for audiophools like me.

Hi Dan,

I wish I kept some test files I put up last year, where I started with one of my own recordings (at the time 24/96) and applied a simple gain change in each of the four programs you mentioned, then saved the results.

All four sounded different from each other, with soundBlade being the most "open" sounding (and Wave Editor, surprisingly, not that far behind). That said, none were "night and day" differences but they were clearly audible. Still, I find differences between DACs to be considerably more significant. (In fact, if you compared the files with some of the apps, differences were very difficult to hear. Once again, the open quality of soundBlade made the comparisons easier than any of the other apps.)

While I have not directly compared soundBlade and Amarra, since the latter is based on code from the former, I have every reason to believe they are at their core, the same application (as far as playback - Amarra is not a mastering/editing application).

To date, I have not heard the sonic equal of the "engine" used in Sonic's applications. Perhaps the closest *might* be Reaper but this too, is not a playback app per se and is designed more for music production, where I see soundBlade (at least in my uses) as more of a post-production tool. (Reaper might be a steal among great sounding apps.)

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

for the terminally curious, Reaper is only $60, Wave Editor is $79.

Wave Editor also includes the top notch iZotope upsampling (and dithering) software.

Both apps are 64-bit fluent.

clay

Hi Clay,

I believe this is true for Reaper but is "64-bit fluent" applicable to Wave Editor?
I thought it is a 32-bit app.

By the way, I have heard (subject to verification) Audiofile Engineering, makers of Wave Editor, is working on music server software too. That should be quite interesting if it turns out to be the case.

Yes, Wave Editor's inclusion of iZotope's 64-bit SRC and MBIT+ dither/noise shaping only makes it an even better value at $79. (At the end of last year, it was "on sale" for $59.)

Audiofile Engineering's wonderful batch processor, Sample Manager ($79), also has the iZotope SRC and MBIT+.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Barry says:

"I believe this is true for Reaper but is "64-bit fluent" applicable to Wave Editor?
I thought it is a 32-bit app."

Fluent is maybe the wrong word, although I chose it as opposed to say 64-bit capable, or other more "strongly worded" phrases. What I meant was that both can handle 64 bit information, as in Wave Editor's use of the 64-bit iZotope modules. Wave Editor probably is only 32-bit.

apologies for any confusion.
Clay

Hi Clay,

The 64-bit information is internal to the SRC algorithm.
To my knowledge, the program itself is 32-bits.

This is similar to the Console software with your MIO. It utilizes an 80-bit wide data bus but the 80-bit part is a reference to the precision of its internal math.

Hmm. How much hard disk would we need to store 60 minutes of 80/192 audio? ;-} (I'd guess around 12-13 gigs.)

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Has anybody compared the MH ULN-8 against Amarra Model Four Converter, I assume the Model Four has been released yet.

Hi Ear,

The Model 4, along with the 303, 304 and 305 are all based on the same platform as the ULN-8.

They are essentially the same device. Some have mic preamps, some don't, etc. I believe Sonic adds some of its own software, for example their own EQ but am not aware of any hardware differences or any other substantive differences that would make for differences in sound.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Hi Barry, thank you for your reply. Regards Werner

"I wish I kept some test files I put up last year, where I started with one of my own recordings (at the time 24/96) and applied a simple gain change in each of the four programs you mentioned, then saved the results."

Barry-

I still have the four test files(z1-4), but you never posted their ids on the Crooked Path forum.

I looked at Wave Editor, it's amazing how much you get for $79. It makes me wonder why Amarra should cost $400.- $1000.

Let's hope that Audiofile-Engineering develops some audiophile playback software.

Aloha,

Dan

"Let's hope that Audiofile-Engineering develops some audiophile playback software."

That's certainly the rumor making the rounds.

Officially, Matthew Foust has said, when I asked:

"Typically, we don't comment on unannounced products. I'm not sure where that rumor started. Suffice it to say, if we were to be developing such an application, it would be very high quality, work with any DAC, and be priced *very* competitively."

you do the math
clay

Hi Dan,

If you see this post, I'd love to get those files again. (It would take a while for me to recreate them.)

The programs used to create them were:
DSP Quattro, Peak Pro XT, soundBlade and Wave Editor.

I didn't identify the files at the time because I wanted folks to hear them first without knowing which application created them.

If I can hear them again, I can identify them.

For folks who don't know what Dan and I are discussing, a year or so ago, I was comparing four different editing/mastering applications for sonic differences (or more precisely, to see if there are any).

I use an original 24/96 recording I made of a small jazz ensemble, using my standard array of a single matched pair of Earthworks QTC-1 microphones separated by a disk/baffle of my own design. The recording was made with a Metric Halo ULN-2 feeding its own Record Panel software on my Mac laptop. Both the Metric Halo and the laptop were powered via a power conditioner. The microphone cables were Nordost Valkyrja.

The test was simple and did not really "stress" any of the applications. Since I record with a lot of headroom, I took the two original mono AIFF files and imported them into each of the four applications. Then, within each app, I made a gain change of approximately 12 dB. I saved the results and those were what we compared.

What I found was that each app indeed has its own "sound". I still use all of them but for different functions as each has strengths the others lack. Interestingly, some apps did not reveal the sonic differences as much as others did - perhaps due to their own sonic fingerprint outweighing the differences between apps.

To be sure, what I heard was NOT of the night-and-day order but still, with the apps that revealed the differences (not surprisingly, the ones that created the most transparent of the files), the differences were not hard to hear. Primarily, they were in the area of open-ness, with some apps sounding "darker", less revealing of the space around the players and the finer details of instrumental harmonics.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Barry-

I will try emailing them to you one at a time.

If that doesn't work, I burn a copy on a DVD-R and mail it to you.

Aloha,

Dan

Thank you Dan.

Much appreciated.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

Hi Dan
can you comment on Sonic Studio's App vz the other Apps you have tried. I am curious whether Amarra is the only App that will give you the playback SQ I have experienced. It is outstanding. Will any other App do the same for playback or are they just for recording?
Thank you for your posts. I learn alot.
Cheers
Andrew

Andrew -

Amarra is the only playback software I've tried other than iTunes and SqueezeCenter.

Best PC playback I've heard was with an ART Legato Asynchronous USB to SPDIF Converter connected to my wife's Macbook. I didn't have the Legato when I tested Amarra. I returned the iLOK for a refund.

My only experience with DSP Quattro, Peak Pro XT, soundBlade and Wave Editor, are the sample files Barry posted earlier this year.

I don't know which is which, but they all sounded different.

Dan