Reclocking USB signals - significant improvement !

[blueninjasix]

Do you think the Phoenix has a bigger or lesser impact on the Dave than the Mscaler? Or in other words, which of the 2 gives the biggest bang for the buck at £2,500 Vs £3,500?

 

[Cereal Killer]

I'm totally willing to beleive there's benefit to better cleaner 5v and psu. But that doesn't hold true for reclocking a signal prior to sending to an asynchronous DAC that reclocks anyway.

From my experiences this is a good call; The two differences I've experienced with a computer front end are a decent electrically silent and stable power supply, and recently to great results, the removal of C/M noise. (a DIY project aided by the PFM massive)

Thanks for taking the time.

Agreed. Thanks Nick

 

[AndyU]

This morning my friend and fellow PFM member @TheFlash invited himself to my house to drink coffee and to listen to the Innuos Phoenix.

The initial system set up was Innuos Zenith SE connected to Chord Dave by USB (Supra USB 2.0). The Dave is connected direct to Pass Labs XA60.8 mono power amps and they feed Spendor SP200 speakers. The system had been switched on for a couple of hours before Nigel arrive.

The purpose of the session was simply to establish whether inserting the Phoenix made a difference that we could hear and whether we preferred it. Nigel suggested a track called 'Salt Of The Earth' by Bill Fay on his 'Countless Branches' album so we used that. It has clear piano with some bass and a Bill Fay singing. I liked it and it was a good suggestion.

After listening to the track with the base system set up we inserted the Phoenix between the Innuos Zenith SE and the Dave. The Supra usb cable was taken from the Zenith to the Phoenix and the Innuos usb cable supplied with the Phoenix was used between the Phoenix and the Dave.

Within about 15 seconds Nigel and I looked at each other and asked if we could stop the track and go back to without the Phoenix because we both heard a clear and compelling difference with the Phoenix that we liked.

Going back to without the Phoenix clearly removed the difference that we had heard but we then added back the Phoenix just to make sure and back came the improvement. With the Phoenix we both felt that the bass was easier to hear and had more clarity and perhaps a richer sound. The mids and top end were somewhat sweeter and less 'digital'. We both liked the improvement a lot which is saying something because it already sounded excellent without the Phoenix.

The next stage was to add in the Chord Blu2 Mscaler between the Zenith and the Dave. Even without the Phoenix this moved the bass etc in the same direction as the Phoenix and also added the very special qualities that fans of the Mscaler will recognise. Today was not about the Mscaler so I will not dwell on that but what we wanted to see was whether adding the Phoenix to an Mscaler system would have as much of an effect as it did with our first session.

Adding the Phoenix to the Blu2 Mscaler / Dave did indeed give similar results to the first session but I think we agreed that perhaps it was not quite as obvious. However to our ears the improvements were definitely still there.

Finally we tried the Phoenix in the same system but this time with the Blu2 Mscaler connected to a Dave with a custom build full Sean Jacobs external power supply with separate LPS voltage rails to the digital and analogue circuits in the Dave. At this point we reached the stage where we were no longer sure whether the Phoenix was making any difference to the sound quality despite switching backwards and forwards several times.


There is no doubt in my mind that I could hear a difference by adding an Innuos Phoenix to my standard system and that I liked the sound.

Inside the Phoenix, as well as the Innuos USB board one can see one of the power supply boards clearly labeled "Innuos Statement Front End PSU" and this is in line with Innuos saying that the Phoenix takes other Innuos servers towards their top of the range Statement.

Innuos say this about the Phoenix,

"The Phoenix USB Reclocker takes the USB signal from ANY source (It's not limited to Innuos products) and completely regenerates it to an extremely high precision signal to feed into your DAC.

3 Components in 1
The PhoenixUSB offers in one unit the equivalent of 3 separate components:
• A USB regenerator
• Linear power supply
• External master clock with its own linear power supply.

Highly Regulated Linear Power Supplies
The USB chip regenerating the signal contains no switching regulators. All 3 independent voltages to the chip originate from an independent linear power supply with further regulation provided by 3 sets of LT305 regulators.

The use of a 3 ppb OCXO clock running directly at 24 mHz and connected via a board track just a couple inches away from the USB chip itself ensures the shortest possible connected path. No precision is lost within cables and connectors, as is the case when using an external master 10mHz clock with an additional 24 mHz clock generator.

Two independent Statement-level linear power supplies, one dedicated to the OCXO clsock and the other used for powering the USB chip 5V USB line."

We all know before we start that the defenders of the faith (aka the FM objectivists) will pounce and say that a reclocker cannot possibly make any difference to the sound quality but to be honest I am not even going to go there. They can start their own thread somewhere else if they wish and debate that in their own little safe place. This thread was started by @cpg who listened to the Phoenix at a show and posted that he liked what he heard. That grabbed my attention sufficiently for me to want to try one to see if I could also hear a difference.

Whether what the Phoenix does is mainly down to reclocking or is due to its power supplies I do not know and I do not really care but the clear result from today was that both I and another PFM could hear a compelling difference due to the Phoenix.

It is not cheap but I suggest that if anyone has an Innuos server that is not a Statement then it would be worthwhile asking your Innuos dealer if you can get a home demo of the Phoenix.

I totally accept that you heard the difference the Phoenix reclocker makes to the USB interface between your Innuos Server and your MScaler/DAVE.

It follows that either the USB output of the Innuos as is must be below par in some way, or the USB input of the M Scaler/DAVE must be below par in some way. There is more going on electrically than the transmission of data, to the detriment of sound quality.

It follows that a reasonable person would ask whether there was a way of totally avoiding the issue. I propose using the optical output of a Squeezebox Touch via optical. Whatever electrical intrusions there are from USB could not possibly exist with the Touch, since it is connected by toslink. The only issue might be jitter, but Rob Watts says the DAVE and M Scaler have stupendous jitter rejection, so that’s unlikely to be a problem.

I can’t understand why you say on the one hand that you don’t care whether what the Phoenix does is down to Reclocking or power supplies, yet at the same time you quote their sales spiel verbatim. Either you care or you don’t!

But anyway, if your listening test is valid, it seems to me a reasonable person wanting to improve the way they get data to an M Scaler/DAVE should look at a Squeezebox Touch, or any other streamer with an optical digital output, for this cannot ever suffer from the problems the Phoenix purports to solve. And save a couple of grand on the Phoenix, and potentially another £4k on the server.

 

[Cereal Killer]

Can you run roon through a touch?

 

[AndyU]

Can you run roon through a touch?

Yep. There’s a setting In Roon Core and that’s all you need. Select it, and Roon will see the Touch as an endpoint. Works smashing. Great way of using Roon all over your house if you want. You can’t control Roon through the touch, you still need a control point like an iPad or Phone or the thing you are running Roon Core on, the Touch itself is doing almost nothing just taking data from the network and putting it out over optical, s/pdif or USB as you fancy. Here’s how to do it

https://kb.roonlabs.com/Squeezebox_Setup

 

[booja30]

It would be interesting to do a comparison of the input stream to output stream, not wrt noise or anything, but to actual bits in/out. The difference described between with/without almost seems like something you'd get from a DSP. But since the product isn't described as a DSP, and has no measurement/configuration steps required like a DSP, it would have to be a fixed setting.

I'm not saying that's what is happening, but it would explain a lot.

 

[Cereal Killer]

Cool tool then.

 

[Fourlegs]

I totally accept that you heard the difference the Phoenix reclocker makes to the USB interface between your Innuos Server and your MScaler/DAVE.

It follows that either the USB output of the Innuos as is must be below par in some way, or the USB input of the M Scaler/DAVE must be below par in some way. There is more going on electrically than the transmission of data, to the detriment of sound quality.

It follows that a reasonable person would ask whether there was a way of totally avoiding the issue. I propose using the optical output of a Squeezebox Touch via optical. Whatever electrical intrusions there are from USB could not possibly exist with the Touch, since it is connected by toslink. The only issue might be jitter, but Rob Watts says the DAVE and M Scaler have stupendous jitter rejection, so that’s unlikely to be a problem.

I can’t understand why you say on the one hand that you don’t care whether what the Phoenix does is down to Reclocking or power supplies, yet at the same time you quote their sales spiel verbatim. Either you care or you don’t!

But anyway, if your listening test is valid, it seems to me a reasonable person wanting to improve the way they get data to an M Scaler/DAVE should look at a Squeezebox Touch, or any other streamer with an optical digital output, for this cannot ever suffer from the problems the Phoenix purports to solve. And save a couple of grand on the Phoenix, and potentially another £4k on the server.

Innuos say they have measured noise in optically linked systems and found more noise in the optical linked circuit than in a well implemented USB system. Optical is not a 'free lunch' and the downside of it is noise from the optical convertors getting into the receiving circuits which is well understood and known.

Hence why Innuos decided not to use optical in the Statement or Zenith servers and why they only have USB connections.

It would have been easy for Innuos to include optical if they thought it was superior or even equal to USB.

I am not saying they are right and you are wrong, just that there is a counter view.

 

[AndyU]

Innuos say they have measured noise in optically linked systems and found more noise in the optical linked circuit than in a well implemented USB system. Optical is not a 'free lunch' and the downside of it is noise from the optical convertors getting into the receiving circuits which is well understood and known.

Where do you get the idea that it is “well known” that noise from optical converters gets into receiving circuits? Can you give me an example or two? In whose DACs is this an issue? Or is it an issue at the transmitting end? Why equally wouldn’t noise from USB interfaces get into receiving circuits?

If you look at Stereophiles review of the Benchmark DAC2 you can see that it might also be the other way round - for their DAC, USB is measurably poorer than optical, at least for 24 bit data.

So, I don’t see how you can generalize. If it was me I’d check out optical into the M Scaler, and potentially save a fortune. Why spend money trying to fix problems you don’t need to have?

 

[booja30]

Also, even if the optical receiver generates noise at least it is a known quantity and you have a better chance to design around it than having to deal with the big unknown at the other end of a USB cable.

 

[Fourlegs]

Look guys. I said I was going to report on my listening to the Phoenix in response to the thread’s author saying he heard an improvement at a show.

I have also made it clear earlier I am not going to get in this sort of warrior dispute.

For those interested in the Phoenix I have said what I heard. For all others, chill out. It will be better for your health.

End of.

 

[sq225917]

All interfaces have their own issues, that's one of the reasons why theres so much plug and play variability in dacs, especially when a designer has a particular focus on their preferred interface to the detriment of the others. Personally, you cant beat firewire, dedicated hardware at both ends, totally controlled ecosystem, consigned to the bin of historic interfaces....pftt.


Once again nick thanks for taking the time. I dont doubt what you heard at all.

 

[Alex S]

That’s fine, FL. Out of interest though, have you auditioned any other USB purifiers, reclockers and whatnot in the past?

 

[Shadders]

Where do you get the idea that it is “well known” that noise from optical converters gets into receiving circuits?

Hi,
I checked the Innuos website. I was unable to locate the application note, study, or support article indicating this.

Does anyone have the links to the Innuos articles which discuss this ?. Thanks.

Regards,
Shadders.

 

[Fourlegs]

That’s fine, FL. Out of interest though, have you auditioned any other USB purifiers, reclockers and whatnot in the past?

No, this is the first time I have tried anything like this and I confess I only tried it this time because the OP raised the subject.

 

[Fourlegs]

Hi,
I checked the Innuos website. I was unable to locate the application note, study, or support article indicating this.
Does anyone have the links to the Innuos articles which discuss this ?. Thanks.
Regards,
Shadders.

You and AndyU clearly think this is a tag wrestling match from the telly the way you two behave. It makes me smile if nothing else.

 

[Shadders]

You and AndyU clearly think this is a tag wrestling match from the telly the way you two behave. It makes me smile if nothing else.

Hi,
I don't understand why you wrote what you did.

You clearly stated :

Look guys. I said I was going to report on my listening to the Phoenix in response to the thread’s author saying he heard an improvement at a show.

I have also made it clear earlier I am not going to get in this sort of warrior dispute.

I did not ask you specifically, so there is no need for you to respond.

Regards,
Shadders.

 

[Fourlegs]

Shadders, if you are potentially interested in the Phoenix then fine but if you just want one of your typical keyboard warrior point scoring arguments then kindly butt out. Has the thread started by @Cereal Killer not got through to you?

 

[Alex S]

No, this is the first time I have tried anything like this and I confess I only tried it this time because the OP raised the subject.

Probably just as well. You doubtless know now, if not a few days ago, that such products as the Ifi Purifier and Uptone Regen are some of the most contentious in the industry.

 

[AndyU]

You and AndyU clearly think this is a tag wrestling match from the telly the way you two behave. It makes me smile if nothing else.

Smile away. I was hoping it was a discussion and participants would be able to back up the things they say if asked. I have totally accepted your listening results. Can’t see what’s tag wrestling about that. Just don’t understand how you can claim that certain facts about optical are “well known” when me and Shadders don’t know them and you can’t corroborate your statement with a link or two. I gave a link showing it is possible for USB to measure worse than optical. There are plenty others. There are also plenty links that show them to measure as identical as possible. So I just don’t see how you can make sweeping statements about optical provoking noise when whether it does or not is down to design. Rob Watts, who designed the M Scaler and DAVE you use, regards optical as the gold standard. Here he says “optical is the best input - its really a question of getting the other inputs to match optical.”. Is he wrong?