manisandher wrote:

Yes, understood. But does it self-terminate? If not, you will need to use a T-connector and a 75 Ohm BNC terminator at the wordclock input.

Mani, I do not think that it is necessary to self-terminate it. Lynx own mixer allows “shutting down” the BNC cable and making the card do not “see” it, well theoretically. Take a look how Lynx mixer manages it.

manisandher wrote:

But I also wanted to know what the mastering engineer you were refering to thought about the sound - in what way did he think it was better after his digital manipulations. I just thought this would be interesting...

You know my attitude on this subjects.

http://www.goodsoundclub.com/Forums/ShowPost.aspx?postID=10866

It is imposable to deal with those people. They demand, in fact insist, you to kiss them in ass and they convert any specific sonic criticism about their sound into some sort of insult directed against them personally. This is unavoidable pattern for those people and it is absolutely impossible talk with them. They are too much fixed un on the fight for own place in hierarchy of “demonstrated respect” and too much abscessed about the “damage control”. To talk with them about the nuances of sound is like to pee against wind. I do not play these games for years and when I meet another “celebrated industry Moron” who again begin to express his/her patronizing attitude then I customary send then to fuck themselves. It might be no socially friendly but it is geographically very accurate destination for them.

Romy, I'm not sure we're talking about the same thing.

In your current setup, you are using the internal Lynx wordclock. In this case, the Lynx may well 'shut down' the external BNC wordclock cable. But that's not what I'm concerned with.

When you are slaving the Lynx to the Model Two, you will select 'external' for the 'Preferred Clock Source', attach a BNC cable between the two units and set the Model Two to 'MASTER'. Now, what I'm saying is that you need to make sure that the BNC input to the Lynx is properly terminated @ 75 Ohms. It may well be that the Lynx already does this automatically (the Weiss AFI1 does, but the Fireface800 has a switch on the back which has to be pressed manually).

I was reminded of the importance of this in an email from Dave Peck. Dave had asked Keith Johnson a question on my behalf and as part of his reply, Keith said, " I did observe that line termination must be present to prevent severe overshoot that could mis-clock systems."

And it's this 'line termination' that I'm refering to.

Apologies if I misunderstood your previous post... I really don't mean to 'teach a grandmother to suck eggs' (strange expression - hopefully it's used in the US too!).

 I see, I did not understood what you said as I was under impression that ANY I/O clocks must be conventionally 75Ohm terminated. Sure, Lynx id 75R terminated.  The page 24 …

http://www.lynxstudio.com/manuals/AES16eUserManual.pdf

I thought you say that UNUSED input clock cable needed to shunt with 75R.  BTW, if I ma not mistaken then the importance of the 75R load for clock in Master mode was stressed in Pacific manual.

There was one of those PC Audio demo show in CA. The AA guy, John Swenson, when there, posted his comments

http://www.audioasylum.com/forums/pcaudio/messages/5/55409.html

I do not extend a lot of credit to those demos; I also do not extend a lot of credit to Mr. Swenson as he is declaratively associated with Chris Connaker, who is a stinky high resolution hi-fi whore. Still, there were 2 interesting moments that picked my attention in John’s comments.

First I was wondering why the Berkeley Boys who were in there did not run their new converter but used the 10 years old Pacific DAC? That is against the rule!

Second, John Swenson wrote:

“This consisted of 4 systems, a Matan server (custum built server) a silent PC running XP, a G5 Mac and a modern MacPro. All the systems had Lynx AES 16 cards (the MacPro was PCIe, the others were PCI) each connected to its own Pacific Microsonics Model 2 machine via AES. The model 2s sent a clock back to the Lynx cards so the Model 2 was the master.”

This was the exact subject of my interest as in this case the following component – DAC driver the source component (Lynx card). With all assurance that Pacific’s clock is better I still do not feel comfortable to put the carriages ahead of horns. If Pacific and Lynx support asynchronous type of interface then the location of the clock I presume would not be a mater. In the AES case however, I do not know…

Sure, Keith Johnson knows what he does…but why in fact he does? There are many alternative voices of his level of expertise that would find this view controversial. Still, it looks like Mr. Johnson is another person who runs Pacific in master clock configuration is D/A mode…

Romy the Cat wrote:

First I was wondering why the Berkeley Boys who were in there did not run their new converter but used the 10 years old Pacific DAC?

Romy the Cat wrote:

Still, it looks like Mr. Johnson is another person who runs Pacific in master clock configuration is D/A mode…

manisandher wrote:

If I were feeling benevolent, I'd say the organisers didn't want to turn the meeting into a marketing event for any one particular manufacturer (although it seems Amarra did quite well it this respect). It's more likely that as KJ was using Model Twos as ADCs, they decided to use them as DACs. Also, I don't think KJ is involved with Berkeley Audio...

I think the reason why they use Pacific is because they record something in there. Pacific, has a very good recorder has a very “proper” feature letting the DAC clock to run as a salve from the A/D clock. That is THE VERY right operation mode and if I use Pacific as DAC havier then I would unquestionably use it. I have to note that the demonstration of recording and then instantly to play the same music in the same space is VERY moronic by nature and it is shame then they go so low to do it…..

manisandher wrote:

In what way is this putting the carriage before the horse? If the DAC clock is the 'horse', then surely by being the Master, the DAC clock is 'pulling', and not 'pushing'. In any event, have you tried this yet?

What is the source of the signal, DAC of PC? I always would like the sores to me master and the load to be slave. It might be not right association but when I see a reversed scenario I always visualize salmon that swim upstream to lay eggs… Still, the validity of my sentiments might be greatly overridden but the specifics of a given DAC design… Somebody with knowledge how Pacific DAC works need to overview this… I probably need to call to Lynx people and to Pacific and to ask them. I more know Lynx/Pacific tandem as A/D. The Lynx/LavryGold tandem needs no clock slaving…

Romy the Cat wrote:

... but for a person who works in the marketing you sound too naïve...

Romy the Cat wrote:

Somebody with knowledge how Pacific DAC works need to overview this…

Romy the Cat wrote:

I always would like the sores to me master and the load to be slave.

manisandher wrote:

Surely, you're not going to find anyone more knowledgeable than Keith Johnson in this respect... and he uses the Model Two in Master mode (with the Lynx card)!

Which does not necessarily mean everything in my view.

manisandher wrote:

You know, I'm really surprised at your attitude. I understand that if experimenting were to be costly that you'd want to ask 'experts' first to make certain that your investments might be worthwhile. Now, it seems you've spent an inordinate amount of time experimenting in other areas of audio and yet in this case you're reluctant to. But trying Master mode would cost you nothing. NOTHING! Why not just try it and see what happens? Then by all means see if you can explain the result by asking the 'experts'. This seems a complete 'no-brainer' to me.

Sure, it is valid argument and I will definitely try, all that I need to do it going to storage and find my custom Lynx cable with clock lead that I ordered for a while back. There are however some “minor things” that I question.

Lynx card in a way is freaky. Whoever the cloaking mode the card is it switches to internal clock if the external clock is not there. They did it very nicely and very transparent and it is in way a main bitch. When Lynx locks the external clock with it’s SynchroLock (it takes 30-40 seconds) then it is fine. Then when the external clock is switched the Lynx turned to run from own internal clock and the SynchroLock is still stay engaged. My point is that the configuration and indication of Lynx suggests that it use external clock but in really it is not. In the mode what the outputs the signals everything is even more confusing as I have no clear reference where the Lynx’s clock sourced from. The fact that I have the cable connected from Pacific is NOT an assurance that Lynx USES Pacific’s clock. What I would like to see in Lynx is that when the externals clock is no good then the card shall not work and throw some kind of error. Lynx does not do it. Furthermore when I switch Lynx in high resolution clock tracing then you might see different rate of the clocks variation between Lynx and Pacific. So, when I see the Pacific’s AES line clock variation then I presume that Lynx clock in slave mode shall inherent the same variations. But it does not and the variation more like the Lynx’s internal clock Lynx. That all makes me doubt with what is doing on.

Anyhow, here is Lynx’s clocking diagrams, as you can see they do not treat external clock any different from any other clock. BTW, the recording mode (and I tested it many times in 1X mode) the LINE 1 clocking has no disadvantage over external clock from Pacific.

Anyhow, I agree that I need to try it…

First of all I had a good conversation with Lynx people. Pacific is very well known to them unit and many of their customers use it. They suggested as the most typical application for their customers to slave Lynx from Pacific in A/D mode and to slave Pacific from Lynx in D/A mode. That does makes sense to me and it was how I felt. What however I have learned was that I was in way a fool as I thought that Lynx SynchroLock is the main slaving mechanism, it is in fact not. The SynchroLock is the secondary optional locking mechanism that is much narrower then Lynx’s PLL. The SynchroLock removes jitter from clock but Lynx folk advised that if I use a very high quality clock then it would make sense to defeat the SynchroLock and let the Lynx’s PLL to see the “naked” external clock. Also, the Lynx people expanded to me how to figure out what clock Lynx is running if the SynchroLock is not engaged

So, I brought my custom Lynx cables with clock feed and set up the DAW and Pacific in a new configuration with Lynx slaved. Today the WGBH broadcasted live the Choir of Christ's College with David Rowland conducting and I recorded it in a new configuration. The result was good, if you wish here is the fragment (88/24, 55meg)

http://www.mediafire.com/download.php?iidnnnz2nde

Then I have a session when I was trying to found out how the Pacific sound all together in this new configuration. Here are some conclusions:

1)      In A/D mode the Lynx must be slaved. The AES-line-clock or External-clock did not make a lot of different but my A/D mode is FM, which is not so critical. There is a minor difference in favor of External clock but only with no SynchroLock used. When the AES-line-clock used then SynchroLock might be used.

2)      In A/D mode with External clock when Pacific runs as Master the SynchroLock must not be used. The sound with and with SynchroLock is different. Without the SynchroLock sound becomes more mucro-contrasty and more playful.

3)      In D/A mode I did not try to slave Pacific from Lynx as it would require to re-plug the clock cables – I will not do it anyhow.  So, I run the Pacific and Lynx with independent clocks and then Lynx slaved from Pacific.

4)      Lynx slaved from Pacific is way more preferable configuration then Pacific and Lynx with independent clocks. I did not detect that when Pacific serves clock then only the amount of HF was changed as Mani reported. What I heard was that the whole Sound was changed, the sound become less mechanical and more liquid - unquestionably become better.

That is all for now. I made a number of listening tests between the Pacific in new Master mode and Lavry Gold. I will post the update later. I did not figure out if I will do it he in this thread of in the “Recording options: Pacific Microsonics vs. Lavry Gold”

Romy the Cat wrote:

Lynx slaved from Pacific is way more preferable configuration then Pacific and Lynx with independent clocks. I did not detect that when Pacific serves clock then only the amount of HF was changed as Mani reported. What I heard was that the whole Sound was changed, the sound become less mechanical and more liquid - unquestionably become better.

This mystery that I would like to get answer….

Both Pacific Microsonics Model One and Model One have adjustable DAC output. Although Pacific claims the ultra-low distortion and super-high quality of their digital attenuation my negative position about DSP attenuation is well know.

So, Pacific has 24dB output gain. My question is where they have 0dB gain in this thing, the location where the DSP will be presumably not engage.

It is possible that it would be at +24dB, where the digital attenuator is wide open and is in “unity gain”. The Pacific recommend that Output Level Monitor  be set at +18dB to provide the best digital resolution and best analog stage performance. Would it be that at 18dB Pacific “eats” 6dB by DSP means?

It is in way very peculiar as the guys who designed Berkeley Alpha DAC (the same as did Pacific) made Berkeley to have the same behavior. Berkeley had DSP attenuator at output and Berkeley folks recommend using -5dB of full scale as the “best preferable setting”.

I do not care what they “recommend” and they might have other agendas, but I would like to know where Pacific has 0dB and where DSP is not engaged. I think the answer for Pacific would be the same as for Berkeley. I do not think anybody would push an answer in context of Pacific – not a lot of people use it “for Sound” and mostly professionals use Pacifics to pay their mortgages. In context of Berkeley the answer might be “pushed” but so far I heard public voices about Berkeley DACs only from idiots. I spoke with Berkeley guys about the subject but their answer was not concussive, ot put this way – they made it not conclusive.

So, where the Pacific has 0dB in Output Level Monitor?

Romy the Cat wrote:

I find it very covenant to run analog source and then be able to switch the source signal passing across Pacific A/D-D/A where both D/A and DAW run from the same Pacific’s A/D clock. In fact it is so convenient that I asked myself if it worth for me to run Pacific from my Preamp outputs and just save to my DAW some of my most precious LPs and Tapes.

manisandher wrote:

Have you tried this yet? If so, in percentage terms, how close do you think the digitized signal is to the original analog signal? (This is something that I'm hoping to try myself at some point, but I don't have a good analog source right now.)

Nope, I did not try AD any better source then FM with Pacific. I did try it with Lavry Gold, the result was so-so but I knew that I did it very badly and I was not going for “best” results. I need to rewire everything to do it “properly” and it is not very inconvenient for me. Also, I do not have a lot of motivation to convert LP or Tape to high-resolution files. If I have an AD permanently plugged into one of the outputs of my preamp, as I proposed above then it might be a game changer…

Also, there was recently an interesting observation that I am not sure how to interpret. When I had Lavry Gold AD and Lavry Gold DA plays into each other and compared the sound of my FM from SU-1X direct with the sound of AD-DAW-DA chain then I heard absolutely no difference. With replacing the Lavry Gold AD-DAW-DA to Pacific AD-DAW-DA I did observe a minor difference, not improvement of courses…

manisandher wrote:

Related to posts on other threads, I recall noting that the Model Two's clock frequency is slightly off 44.1 KHz. With the Model Two in Master mode, the Weiss AFI1 reads the clock frequency as something like 44.08 KHz (I'm not in the UK right now, but will check the exact freq. when I get back). However, when the Model Two is switched to 48KHz, the AFI1 reads the clock frequency as exactly 48.00 KHz. Romy, perhaps your Model Two exibits the same behavioiur? If so, could this 0.05% difference account for the anomalies that you're hearing between the Model Two and Lavry? (Of course, this should only be the case when the Model Two is in Master mode.) To test this, you may want to record and compare a piece at both 44.1 (x1/x2) and 48 (X1/X2). I know the latter will cause you a lot of pain, but keep reminding yourself that it's in the pursuit of knowledge, and you should pull through OK... [do you allow smilies on this site?]

Romy the Cat wrote:

The problem is that I detect that if I record in 44/24 or 88/24 in single-wire mode but the second wire is connected from Lynx to Pacific (and the line is muted by Lynx mixer) then I do have worsening of sound quality. The sound auditable improves when I plug out the second, not used line.

manisandher wrote:

This is very interesting and something that I have not noticed. Because I cannot double the clock frequency going into the Weiss AFI1, I have tended to ignore playing/recording 4fs files. But, I've kept the second AES/EBU wire attached nevertheless... because I didn't think it could make any difference. I will try unplugging it and determine whether or not I can hear a difference too.

Well, frankly I do not truly understand what that Weiss AFI1 does. I think this devise is bogus by a concept. Here is how I interested it: If you take Firewire of USB out of your PC then the other side of Firewire of USB shell be right in DAC, right next to the master clock on the DAC side. If your Firewire of USB devise have some kind of proxy box, with own PS and own clock, then I think it defeats all purpose.  I do not insists that I am right – I just think it is architecturally does not look kosher to me. You insist that Pacific shall run as master clock referent, OK, but the referent to what? To Weiss AFI1, right? Then you have a Firewire link to PC but Firewire does not care clock data – it is asynchronous.  So, my question would be – what manages the clock of your playing/recording software? Then you play/records then can you do it right to your Weiss AFI1 device? Will, the change of the Pacific clock trigger the automatic reset the sampling rate of your playing/recording software. The most important: when you sent from Pacific reference clock of 88K but play a file of 44K, outputting the stream to a third audio devise, then will you have the half-speed effect? If you do not have the same effect as reel tape of 15ips played at 9ips then your software is not truly slaved by Pacific/ AFI1 chain.

Anyhow, to mimic the problem that I describe do the following:

1)      Play a good analog source to line A of your preamp
2)      One of the outputs send to Pacific A/D
3)      Set Pacific  in 176/24  mode, dual-wire
4)      Send signal out of Pacific D/A to line B of your preamp
5)      Set the Pacific output (of whatever menace you have) to be identical volume to line A of your preamp
6)      Flipping  between the line A  and B of your preamp confirm that you have identical sound (or note whatever delta it would be)
7)      Now switch the Pacific into 44/24 mode, single-wire mode, without touching anything else.
8)      Note that this time that the output from line B of your preamp (across Pacific) will be slightly louder (let say 2dB, but I did not measured)
9)      Noted that the output from line B of your preamp (across Pacific) will be a bit brighter and flatter then the line A of your preamp. It will be no death of imaging and sound would be almost broken up – almost like from a computer voice.
10)   Now, unplug the second-wire and note the difference.  In my case the louder 2dB will be gone and the sound will be identical to what was in paragraph # 6 above. 

During this experiment the minor differences between 44 and 176, if you have any,  shall be discarded as the amplitude of the problem I descried above is much higher then the dealt between  44 and 176

Romy the Cat wrote:

If you do not have the same effect as reel tape of 15ips played at 9ips then your software is not truly slaved by Pacific/ AFI1 chain.

Romy the Cat wrote:

Anyhow, to mimic the problem that I describe do the following: