Uptone Audio Regen - Clock Mod

[YashN]

Now that I have more hours on the clock mod I can report more detailed findings.

 

3. PC -> Adnaco fiber USB hub -> Regen -> Regen w/ Clock Mod -> Lampizator L7 DAC

 

 

- Cascading Regen and Regen with Clock Mod is the real surprise. The improvement, particularly in terms of musicality and resolution, improves quite significantly. This difference is bigger than going from no Regen to Amber Regen. In my system this brings about improvements in the order of changing to a different DAC.

 

I did two rounds of test just to ensure that I was not hallucinating and I am not. So now setup (3) is my preferred setup for the time being.

 

Awesome feedback, JL.

 

Henning, great of you to have tested this and shared your findings.

 

Thanks to both.

[Blizzard]

The reason I asked is because you were talking about a state of the art type of system (so presumably not an existing one) using software like HQPlayer. The software in such a system would eliminate the reason to have the chip or FPGA.

 

The renderer technology I'm talking about is existing. The Merging NADAC is a great example of a early generation DAC that uses this implementation.

 

Please share this DAC technology that uses no chip or FPGA in the signal chain. I'm unfamiliar with it.

[Jud]

 

Please share this DAC technology that uses no chip or FPGA in the signal chain. I'm unfamiliar with it.

 

Miska's DSC1.

[Blizzard]

Miska's DSC1.

 

What is Miska's recommended way to interface his DAC with a computer running HQplayer?

[tne]

 

Please share this DAC technology that uses no chip or FPGA in the signal chain. I'm unfamiliar with it.

 

The Lampizator DSD only DAC.

[Blizzard]

The Lampizator DSD only DAC.

 

Does it not use an Amanero USB interface?

[devs]

It's probably safe to assume the biggest contributing factor to the better sound experienced via the USB connection is due to the DSP/upsampling etc tweaks on the computer end that using USB allows you to take advantage of.

 

This is where most current internal Ethernet renderers fall short. Your limited to using UPNP/DLNA (or similar proprietary protocols) to interface to them. The next generation of these renderers (as explained in a previous post) they won't be limited by these factors.

 

No DSP or upsampling, just feeding the same bitperfect stream to either ethernet or USB.

 

AIR is indeed a proprietary protocol and not similar to or UPNP/DLNA/OH based. It uses a virtual sounddevice on the PC/mac to bridge to ethernet/WiFi and then the built in ethernet/WiFi receiver hardware.

 

The data transport between PC-devialet is asynchronous (PC is master), data is buffered inside the Devialet and 'clocked out' using the Devialet internal master clock.

[Blizzard]

No DSP or upsampling, just feeding the same bitperfect stream to ethernet or USB.

 

AIR is indeed a proprietary protocol and not similar to or UPNP/DLNA/OH based. It uses a virtual sounddevice on the PC/mac to bridge to ethernet/WiFi and then the built in ethernet/WiFi receiver hardware.

 

The data transport between PC-devialet is asynchronous, buffered inside the Devialet and 'clocked out' using the Devialet master clock.

 

You said "after tweaks to the USB and computer" did the tweaks that you applied to the computer not used while streaming via AIR?

 

What tweaks did you try with the USB?

[devs]

.

You said "after tweaks to the USB and computer" did the tweaks that you applied to the computer not used while streaming via AIR?

 

What tweaks did you try with the USB?

 

Yes, but what I meant to say was that a plain and standard PC outperformed USB using AIR but after optimizing and tweaking that same PC outperformed ethernet/AIR using USB. Of course the comparisons were made with all the same tweaks/optimizations and software present.

 

At a certain point I stopped comparing and defaulted to USB but at that point I had already installed W2012, AO, HDPlex linear PSU and 250W DC ATX converter. Turning point became the installation of a Kingrex dual headed USB cable and Kingrex battery PSU. Later came all the other stuff mentioned in my signature and also the Kingrex items were upgraded. The ethernet cable I tested with back then was an Audioquest Diamond.

 

The turning point also described here in more detail: Antoine's system - Page 4

[tne]

Does it not use an Amanero USB interface?

 

Snipped from Lampizator's 5th Generation DSD DAC

 

"The DAC uses the popular Amanero Technologies Combo 384 USB interface board to handle data transmission to the unit."

 

and also Lampizator says it takes a simplified approach that "works purely in the analog domain—with passive filters and active tube circuits—without a DAC converter chip. "

 

clearly i interpreted your query about chipless DACs too narrowly.

 

 

I do not own one, but have heard a few Lampi's. The DSD only version is the one that I most likely can afford one day. I convert to DSD128 in HQPlayer before sending to my DAC so the lack of PCM capability and the fact that my other DACs can handle PCM, makes this my current next DAC of choice.

[Blizzard]

.

 

Yes, but what I meant to say was that a plain and standard PC outperformed USB using AIR but after optimizing and tweaking that same PC outperformed ethernet/AIR using USB. Of course the comparisons were made with all the same tweaks/optimizations and software present.

 

At a certain point I stopped comparing and defaulted to USB but at that point I had already installed W2012, AO, HDPlex linear PSU and 250W DC ATX converter. Turning point became the installation of a Kingrex dual headed USB cable and Kingrex battery PSU. Later came all the other stuff mentioned in my signature and also the Kingrex items were upgraded. The ethernet cable I tested with back then was an Audioquest Diamond.

 

The turning point also described here in more detail: Antoine's system - Page 4

 

 

 

I think the issue with AIR is with how the virtual sound card is implemented. Based on what they claim, all PC\media player tweaks are completely bypassed when sent out of the PC via AIR. This means all the tweaks on the PC end that were made were implemented via USB output but not AIR. This could explain things. I don't think it's because there is less jitter introduced via the USB connection. As you said when doing an apples to apples comparison, Air was superior.

 

The great thing about the new generation of Ethernet renderers, is they are exactly like USB when it comes to what it sent out of the media player. If you apply DSP and tweaks in the software, you will hear them from the DAC. They can also stream up to DSD 1024 over 2 channels.

[devs]

No DSP or upsampling, just feeding the same bitperfect stream to either ethernet or USB.

 

AIR is indeed a proprietary protocol and not similar to or UPNP/DLNA/OH based. It uses a virtual sounddevice on the PC/mac to bridge to ethernet/WiFi and then the built in ethernet/WiFi receiver hardware.

 

The data transport between PC-devialet is asynchronous (PC is master), data is buffered inside the Devialet and 'clocked out' using the Devialet internal master clock.

 

Oops, above I wrote "PC is master" but I meant that the Devialet is master. Also I was describing ethernet AIR here but of course the USB implementation in the Devialet (XMOS based) is also asynchronous.

[Blizzard]

Snipped from Lampizator's 5th Generation DSD DAC

 

"The DAC uses the popular Amanero Technologies Combo 384 USB interface board to handle data transmission to the unit."

 

and also Lampizator says it takes a simplified approach that "works purely in the analog domain—with passive filters and active tube circuits—without a DAC converter chip. "

 

clearly i interpreted your query about chipless DACs too narrowly.

 

 

I do not own one, but have heard a few Lampi's. The DSD only version is the one that I most likely can afford one day. I convert to DSD128 in HQPlayer before sending to my DAC so the lack of PCM capability and the fact that my other DACs can handle PCM, makes this my current next DAC of choice.

 

It's a great DAC, but the Amanero does have a FPGA on it. I have 1 right here on my desk along with an excellent implementation of a Xmos based USB interface.

 

But I can assure you, once Lukasz gets his hands on the next generation of Ethernet renderers, the Amanero is gonna bite the dust

 

[devs]

I think the issue with AIR is with how the virtual sound card is implemented. Based on what they claim, all PC\media player tweaks are completely bypassed when sent out of the PC via AIR. This means all the tweaks on the PC end that were made were implemented via USB output but not AIR. This could explain things. I don't think it's because there is less jitter introduced via the USB connection. As you said when doing an apples to apples comparison, Air was superior.

 

The great thing about the new generation of Ethernet renderers, is they are exactly like USB when it comes to what it sent out of the media player. If you apply DSP and tweaks in the software, you will hear them from the DAC. They can also stream up to DSD 1024 over 2 channels.

 

With the current AIR 2 a virtual sound device is installed using a Devialet driver. It supports the regular WASAPI, ASIO standards and as it works as a sound card the bypassing you mention is not the case here. You can even use room correction software like Acourate etc. with it. HOWEVER this wasn't the case with AIR 1, here iTunes was used as a library browser and the music file itself was transferred (in chunks) to the Devialet. There the file was decoded inside the Devialet thus indeed bypassing any software player changes.

 

AIR (I've been discussing v2 today) was only superior on a non tweaked/optimized PC. USB became superior after tweaking but still then it was an apples vs apples comparison except of course for the hardware specifically used for the ethernet and USB connections themselves.

 

Of course there must be some bottleneck in Devialet's implementation of AIR preventing it to have the same (or superior) quality as USB. I'm not blaming ethernet here, my point was only that in reality ethernet isn't automatically and always better. Personally I'm just using the interface that gives me the best SQ/most enjoyable music reproduction. If that's USB today, I'll use that and if it's ethernet tomorrow I'll switch back to ethernet.

[Blizzard]

With the current AIR 2 a virtual sound device is installed using a Devialet driver. It supports the regular WASAPI, ASIO standards and as it works as a sound card the bypassing you mention is not the case here. You can even use room correction software like Acourate etc. with it. HOWEVER this wasn't the case with AIR 1, here iTunes was used as a library browser and the music file itself was transferred (in chunks) to the Devialet. There the file was decoded inside the Devialet thus indeed bypassing any software player changes.

 

AIR (I've been discussing v2 today) was only superior on a non tweaked/optimized PC. USB became superior after tweaking but still then it was an apples vs apples comparison except of course for the hardware specifically used for the ethernet and USB connections themselves.

 

Of course there must be some bottleneck in Devialet's implementation of AIR preventing it to have the same (or superior) quality as USB. I'm not blaming ethernet here, my point was only that in reality ethernet isn't automatically and always better. Personally I'm just using the interface that gives me the best SQ/most enjoyable music reproduction. If that's USB today, I'll use that and if it's ethernet tomorrow I'll switch back to ethernet.

 

I suppose the implementations I'm talking about are in a different league. Just like with USB interfaces, there's poor implementations and good ones. The fact that everything over 24/96 PCM is downsampled via AIR, definitely proves their implementation doesn't hold water to today's interfaces. Sounds like it has some great attributes but at the same time combines them with serious limitations.

 

But to be fair, since even Air 2 is 2013 technology, (hardware it can work on even older) it can't be expected to compare with 2016 technology in this fast moving sector.

[_JL_]

Just want to clarify a few things for those interested:

 

1. My PC is not particularly optimized, plain Windows 8.1 with JRiver running WSAPI event mode- that's it. No Win Server, no AO, etc. I prefer a working computer.

 

2. Not using any DSP or upsampling on the PC side so just simple bit-for-bit going to the DAC.

 

---

 

If there is anything I learned in this hobby, then it is that there is no rule. Logically one would expect single-box player to sound best as the data path is internal and only one master clock is used. I have my share of top-flight single-box players and still own a very good one but I can categorically say that the sound I am getting now is nothing I have experienced before.

 

Again I am not saying that the clock mod will do the same for others - no rule remember? I don't even recommend it unless one knows the risk involved. If the next-gen whatever is better then I am all for it but until then I am enjoying what I have now.

[Blizzard]

Just want to clarify a few things for those interested:

 

1. My PC is not particularly optimized, plain Windows 8.1 with JRiver running WSAPI event mode- that's it. No Win Server, no AO, etc. I prefer a working computer.

 

2. Not using any DSP or upsampling on the PC side so just simple bit-for-bit going to the DAC.

 

---

 

If there is anything I learned in this hobby, then it is that there is no rule. Logically one would expect single-box player to sound best as the data path is internal and only one master clock is used. I have my share of top-flight single-box players and still own a very good one but I can categorically say that the sound I am getting now is nothing I have experienced before.

 

Again I am not saying that the clock mod will do the same for others - no rule remember? I don't even recommend it unless one knows the risk involved. If the next-gen whatever is better then I am all for it but until then I am enjoying what I have now.

 

The beauty of the Lampi is since it uses an Amanero USB interface, it can easily be retrofitted with newer technology when available. That's a definite 1 up on DAC's with USB interfaces integrated into the main board.

[Superdad]

It's a great DAC, but the Amanero does have a FPGA on it. I have 1 right here on my desk along with an excellent implementation of a Xmos based USB interface.

 

But I can assure you, once Lukasz gets his hands on the next generation of Ethernet renderers, the Amanero is gonna bite the dust

 

Good grief Bliz, you really need to do some homework.

We all know that clean I2S/DSD with master clocking is ideal for feeding our DACs (and with a really good USB>I2S board--which neither of your pictured boards particularly qualify as--you can feed the master clock back in).

But as discussed elsewhere, AES67/Ravenna is not going to be the savior of computer audio. Yes, Ethernet is a good and convenient transport media, but it is not immune to a lot of the same problems we face with USB. And what sort of chips and processors do you think will be running those Ethernet renderers? Have you looked at Ethernet renderer boards? (Network Audio Renderer | ABC PCB)

And please don't give us this "oh, the next generation of Ethernet renderers are going to make all this obsolete." Have you looked at the Merging NDAC Ethernet input module?

 

Believe me, I am happy to embrace Ethernet solutions that eschew UPnP/DLNA. And a couple years back I spent a lot of time surveying all the proprietary protocols used by the broadcast/pro-sound/venue-sound companies (Ethersound, CobraNet, Dante, etc.) which lead up to AES67 because the shit didn't interoperate.

 

So if you know of some REAL advances happening with audio-over-IP solutions--other than AES67--please say so and come forth with proof of existence. Otherwise, let's all just wait and see.

 

And let's let these hardcore REGEN clock modders have their thread back!

[Blizzard]

Good grief Bliz, you really need to do some homework.

We all know that clean I2S/DSD with master clocking is ideal for feeding our DACs (and with a really good USB>I2S board--which neither of your pictured boards particularly qualify as--you can feed the master clock back in).

But as discussed elsewhere, AES67/Ravenna is not going to be the savior of computer audio. Yes, Ethernet is a good and convenient transport media, but it is not immune to a lot of the same problems we face with USB. And what sort of chips and processors do you think will be running those Ethernet renderers? Have you looked at Ethernet renderer boards? (Network Audio Renderer | ABC PCB)

And please don't give us this "oh, the next generation of Ethernet renderers are going to make all this obsolete." Have you looked at the Merging NDAC Ethernet input module?

 

Believe me, I am happy to embrace Ethernet solutions that eschew UPnP/DLNA. And a couple years back I spent a lot of time surveying all the proprietary protocols used by the broadcast/pro-sound/venue-sound companies (Ethersound, CobraNet, Dante, etc.) which lead up to AES67 because the shit didn't interoperate.

 

So if you know of some REAL advances happening with audio-over-IP solutions--other than AES67--please say so and come forth with proof of existence. Otherwise, let's all just wait and see.

 

And let's let these hardcore REGEN clock modders have their thread back!

 

The Amanero in my picture happens to be what's used in the Lampi. You may say it doesn't qualify as a really good USB interface, but it's what Lukasz chose for the Lampi anyways. Most others are based on the Xmos chip used in the board pictured beside it.

 

That out of date Ethernet renderer link is definitely not what the new DAC's are gonna be using.

 

AES 67 is just an interoperability standard. It doesn't define the maximum potential any of the audio over IP protocols are limited too. It just defines a minimum standard.

 

For high end audio DAC purposes it's pretty much irrelevant anyways. ALC Networx decided to make Ravenna adhere to the standard so it can interoperate with other audio over IP protocols, but it far exceeds the standards of any other audio over IP protocol.

 

Who else can do DSD 256 over 48 channels on 1 single gigabit network? Or DSD1024 over 2 channels on a single endpoint smaller than a credit card?

 

Since you are so knowledgeable in the area, which of the latest generation of OEM Ravenna renderer boards have you ran through the paces?

[zerung]

Good grief Bliz, you really need to do some homework.

We all know that clean I2S/DSD with master clocking is ideal for feeding our DACs (and with a really good USB>I2S board--which neither of your pictured boards particularly qualify as--you can feed the master clock back in).

But as discussed elsewhere, AES67/Ravenna is not going to be the savior of computer audio. Yes, Ethernet is a good and convenient transport media, but it is not immune to a lot of the same problems we face with USB. And what sort of chips and processors do you think will be running those Ethernet renderers? Have you looked at Ethernet renderer boards? (Network Audio Renderer | ABC PCB)

And please don't give us this "oh, the next generation of Ethernet renderers are going to make all this obsolete." Have you looked at the Merging NDAC Ethernet input module?

 

Believe me, I am happy to embrace Ethernet solutions that eschew UPnP/DLNA. And a couple years back I spent a lot of time surveying all the proprietary protocols used by the broadcast/pro-sound/venue-sound companies (Ethersound, CobraNet, Dante, etc.) which lead up to AES67 because the shit didn't interoperate.

 

So if you know of some REAL advances happening with audio-over-IP solutions--other than AES67--please say so and come forth with proof of existence. Otherwise, let's all just wait and see.

 

And let's let these hardcore REGEN clock modders have their thread back!

 

 

Alex, am I mistaken? Was John and you planning some interface for the Spdif?

[Blizzard]

Alex, am I mistaken? Was John and you planning some interface for the Spdif?

 

John's flagship product the Sonore Signature Rendu doesn't even use USB. It uses I2S over HDMI and SPDIF. What's that tell you there?

[Booster MPS]

Maybe we should keep this on topic of the Regen clock mods - I was enjoying reading about them....(hint)

[satfrat]

Maybe we should keep this on topic of the Regen clock mods - I was enjoying reading about them....(hint)

 

+1!!!!

[Superdad]

Who else can do DSD 256 over 48 channels on 1 single gigabit network? Or DSD1024 over 2 channels on a single endpoint smaller than a credit card?

 

Since you are so knowledgeable in the area, which of the latest generation of OEM Ravenna renderer boards have you ran through the paces?

 

You are going to have to show, not just tell. Here is a roundup of the latest cards from the AES67/Ravenna OEM developers, and not one of them can do more than 192Khz. And look at all those processors and sub-systems. And half of them run a Linux variant. Ah yes, these will all surely be the path to audio nirvana…

 

 

 

 

 

[Blizzard]

You are going to have to show, not just tell. Here is a roundup of the latest cards from the AES67/Ravenna OEM developers, and not one of them can do more than 192Khz. And look at all those processors and sub-systems. And half of them run a Linux variant. Ah yes, these will all surely be the path to audio nirvana…

 

[ATTACH=CONFIG]20332[/ATTACH]

 

[ATTACH=CONFIG]20333[/ATTACH]

 

[ATTACH=CONFIG]20334[/ATTACH]

 

[ATTACH=CONFIG]20335[/ATTACH]

 

[ATTACH=CONFIG]20336[/ATTACH]

 

Yes those are examples of a couple early generation Ravenna boards. But like you said they are very limited. The Archwave to 24/96 PCM, the ALCnetworX development board is limited to 24/192PCM and the Coveloz Bach limited to 24/192 PCM.

 

So it looks like you haven't hit the jackpot yet. Better keep looking. I'm bound by NDA so not about to share the info on a public forum. But soon enough they will be well known.

 

Not just Ravenna either, lots of proprietary Ethernet renderers are in the works as well by several high end DAC manufacturers.

 

I wouldn't put your foot in your mouth too deep. Because chances are you'll own a DAC with one of these in the near future.