Announce: Enhanced Digital Output app - USB Dac and 192k Digital Ouput

[PAOGORMAN2001]

Since I know basically nothing about digital music but know what sounds good to me I would like to ask a question.

I have a squeezebox touch connected via wireless router and then output by coax to a DACmagic which is then connected to a NAD master series integrated amp by balanced xlr cables.

My question is whether this upgrade that Triode has developed will work in my setup

Thanks in advance

[Triode]

Installation was very smooth, and the integration with the UI on the SBT is well done. Just a thought - is it possible to write an LMS add-on to allow selection of digital interface options, and display DAC interface via the Web Gui?. I am going to start a thread in audiophiles re: SQ impressions to keep this thread clean, but sounding pretty good to me at the moment!.

Not easy without adding to the protocol used between the server and the player which really means this would need to be part of core firmware and LMS rather than addons. At present this app is based on what I can do without changing the core code as Logitech do not yet see the marketing requirement for this feature...

[edwardian]

Just curious, forgive me if this has been answered already, but after a quick scan of 25 pages I didn't notice - has anyone had success using the Touch USB out to a Wavelength DAC?

[jksbt]

Hi,

I'm new to these forums but have really learned a lot from them. Triode your applet works for peachtree idac although the dac uses an adaptive usb 1.0 and so sounds better using the coax from the sbt. Very easy to use and I appreciate all the effort you put into it.

I have used some of soundcheck's suggestions (quite helpful) and your edo applet and could definitely notice a significant improvement in SQ. Thank you both.

I had an idea to greatly reduce the jitter of an sbt that entails modifying your applet ( or maybe it doesn't?). This might not make any sense so please feel free to throw darts or whatever at this idea. In my situation and possibly others here, the sbt is not close to the pc/nas setup. Perhaps 20 or 30 ft. I just run a long cat 6A eithernet cable (thanks soundcheck--it does make a difference although I can't believe I am saying that) from my pc to the sbt. Then I could go to the dac via either usb or coax which is a short 1m cable. Given the jitter in the sbt is good but perhaps not as good as we all might like I was wondering if the following idea is feasible: Could you use the usb output of the sbt (using some version of the edo applet) to go to a low cost usb to spdif box that would clean it up significantly and avoid making any hardware mods to the sbt (thereby avoiding violating the logitech warranty). The output of the spdif box goes into the coax input of the dac but now the jitter of the sbt has been greatly reduced (hopefully) and no one has to mess with hacking into the hardware innards of the sbt. Although if that is your thing then more power to you.

I read from triode and John Swenson's excellent and well thought out explanations that the edo applet may not be able to do this currently because it needs to query the dac to help set up the usb output channel in the sbt. I suppose a usb to spdif box in between is going to make that impossible. However, maybe a simpler usb 1.0 pass through version (or applet option) of the edo code would work to accomplish this? An added bonus is that this would make the edo code even more versatile since their would not need to be a dac query. Thus, the concern about making it work with various dac brands would be avoided thereby creating an even larger user base for this mod. Although it does increase the cost to buy a usb to spdif box. Or maybe I have no idea what I am talking about (quite possible).

A question one might ask here is why not go directly from the pc to the dac via usb? I have tried it and I can tell you that long usb cables are not OK for transmitting high quality audio. Many dac mfr's warned me against it and they were completely correct. The sound degradation was obvious and not pretty. On the other hand ethernet is adept at traveling long distances and does possess transformer isolation at its receiver. So what I am attempting to do is transmit to the sbt as cleanly as possible over longer distances using ethernet and then reducing the sbt jitter with a usb to spdif box (with short 1m usb cable) and then go to the dac via spdif coax (again short cable).

Would appreciate any comments, etc. on this idea.

[JohnSwenson]

Hi,

I'm new to these forums but have really learned a lot from them. Triode your applet works for peachtree idac although the dac uses an adaptive usb 1.0 and so sounds better using the coax from the sbt. Very easy to use and I appreciate all the effort you put into it.

I have used some of soundcheck's suggestions (quite helpful) and your edo applet and could definitely notice a significant improvement in SQ. Thank you both.

I had an idea to greatly reduce the jitter of an sbt that entails modifying your applet ( or maybe it doesn't?). This might not make any sense so please feel free to throw darts or whatever at this idea. In my situation and possibly others here, the sbt is not close to the pc/nas setup. Perhaps 20 or 30 ft. I just run a long cat 6A eithernet cable (thanks soundcheck--it does make a difference although I can't believe I am saying that) from my pc to the sbt. Then I could go to the dac via either usb or coax which is a short 1m cable. Given the jitter in the sbt is good but perhaps not as good as we all might like I was wondering if the following idea is feasible: Could you use the usb output of the sbt (using some version of the edo applet) to go to a low cost usb to spdif box that would clean it up significantly and avoid making any hardware mods to the sbt (thereby avoiding violating the logitech warranty). The output of the spdif box goes into the coax input of the dac but now the jitter of the sbt has been greatly reduced (hopefully) and no one has to mess with hacking into the hardware innards of the sbt. Although if that is your thing then more power to you.

I read from triode and John Swenson's excellent and well thought out explanations that the edo applet may not be able to do this currently because it needs to query the dac to help set up the usb output channel in the sbt. I suppose a usb to spdif box in between is going to make that impossible. However, maybe a simpler usb 1.0 pass through version (or applet option) of the edo code would work to accomplish this? An added bonus is that this would make the edo code even more versatile since their would not need to be a dac query. Thus, the concern about making it work with various dac brands would be avoided thereby creating an even larger user base for this mod. Although it does increase the cost to buy a usb to spdif box. Or maybe I have no idea what I am talking about (quite possible).

A question one might ask here is why not go directly from the pc to the dac via usb? I have tried it and I can tell you that long usb cables are not OK for transmitting high quality audio. Many dac mfr's warned me against it and they were completely correct. The sound degradation was obvious and not pretty. On the other hand ethernet is adept at traveling long distances and does possess transformer isolation at its receiver. So what I am attempting to do is transmit to the sbt as cleanly as possible over longer distances using ethernet and then reducing the sbt jitter with a usb to spdif box (with short 1m usb cable) and then go to the dac via spdif coax (again short cable).

Would appreciate any comments, etc. on this idea.

In the case of a USB to S/PDIF converter the specs of the converter itself are what the Touch will send to the server, S/PDIF is unidirectional, there is no way the Touch can query a DAC over an S/PDIF link.

On your main question, unless you spend a lot of money on the USB to S/PDIF converter (ie get a really well implemented asynch USB interface) it is most likely not going to have as low jitter as the Touch does already. Now there is a whole different aspect of S/PDIF interfaces called impedance matching.

The spec calls for 75 ohms for the source, receiver and cable. The cable is the only part that usually comes out close to 75 ohms. Most transmitters and most receivers come nowhere close to 75 ohms. Even the cables very rarely have 75 ohm connectors. The differences between source and receiver are what causes the problems (I don't have time to get into why that is). So even if both are not 75 but are close to each other you can have a decent link. So say if both the source and receiver are near 50 ohms and you have a pretty close to 50 ohm cable, this part of the equation will be good.

Unfortunately there is no way to know what a particular Source or DAC is, you just have to try things and find out what works best. In a lot of cases someone will say "this is a really good source" when in reality it's nothing special, it just happens to match the DAC they are using, for someone else with a different DAC it might be terrible.

And also to confuse the issue DACs vary greatly in how much they are sensitive to these effects, some are quite sensitive and some are fairly insensitive. So for some all these sorts of things do make a significant difference and for others all this stuff is almost irrelevant.

So IF you have DAC which is sensitive to this stuff both the inherent jitter of the output AND the impedance match between source and DAC come into play. The Touch has pretty good jitter level as is, certainly not the best on the planet, but quite good. Its a total crap shoot as to whether its going to be a good impedance match to any particular DAC.

Compare that to a USB to S/PDIF converter, very few inexpensive ones even come close to the jitter level of the Touch. You usually have to spend quite a bit to get one that has significantly lower jitter than the Touch (and I mean quite a bit more money than the cost of the touch itself). They DO exist, but not inexpensive which seems to be part of your inquiry. And then you have the crap shoot of impedance matching all over again. So you could have a situation where the Touch is not a good impedance match with your DAC and you find a USB->S/PDIF converter that is about the same jitter as the Touch but just happens to have a good match with your DAC and it sounds better. You could also find a really good converter with very low jitter, which has a poor impedance match with your DAC and it sounds no better than the Touch. The Touch may have a good match with your DAC and sounds better than any of the converters, even the really good ones.

You really don't know what the situation is going to be until you start trying things.

In my opinion the best way to go is get rid of the S/PDIF part of the equation all together and get a GOOD DAC with a really good asynch USB receiver, connect it to the Touch with a fairly short cable and let the ethernet cable to the touch be your long connection. If you do that you get rid of a LOT of the "you take your chances" aspect of dealing with S/PDIF. BUT you have to make sure you don't get a DAC with a proprietary USB implementation such as the M2Tech ones. A numbers of companies use the M2Tech OEM board in their DACs, you can't use those with the Touch.

I know it's not as cut and dried as you would like it to be, but unfortunately that's the way it is.

John S.

[jksbt]

John, Thanks for an outstanding and well thought out reply. I agree with everything you said. And silly me forgot about the impedance mismatch issue between source and destination. One of the reasons I wanted the short spdif cable between the output of the usb to spdif box and the spdif input of the dac was to greatly reduce the possibility of bad impedance mismatch from excessive cable length (due to reflections, etc.). And short cables do solve that issue.

However, I totally thought (incorrectly) that the mfr.'s of all of these boxes would adhere to some sort of standardized spec input and output impedance for the spdif I/O. I guess I was naive. Or perhaps the I/O impedance spec on spdif is poorly specified or not at all? Quite sobering. Maybe disappointing is more like it.

Well, if true, it sucks and explains a few more of the gotchas in this hobby (I notice there are many). But I also infer that very good design in the receiving dac and the touch or usb to spdif box can mitigate some of this potential source to destination impedance mismatch.

So I gather I had part of the solution ---let the ethernet handle the long jog and keep the usb or spdif short. But to avoid the source/destination impedance mismatch issue it would be far better to go from the touch using a short usb cable via asynch usb. And, the receiving dac still has to be pretty well designed to perform the asynch correctly.

OK. So now I am back to wondering which dacs the EDO applet will work on. And I think you are saying to just try it and see and report back.

Does the above summarize correctly the gist of your reply (to make sure I understand correctly)?

And since the dacs I am interested in pursuing (the peachtree idac I have now is a loaner for audition purposes ) are the NAD M51 and Anedio D2 and possibly the NAD C390DD -- does anyone have info if the USB input of any of these will work with the EDO applet?

[jksbt]

John, just one more question...

If I decide to "roll the dice" with this crazy idea of using a usb to spdif converter at the output of the touch (because the sbt doesn't talk to the particular dac usb input that I have via the EDO applet) ...then from your reply is it possible that the touch could be ok talking to the usb to spdif converter even though it doesn't talk to the usb input of the dac? For example, might it be ok with talking to a MF Vlink? I am not saying this is necessarily the best idea -- just trying to verify that it could possibly work in the same sense as the touch via the EDO applet could possibly talk to the usb input of a given dac. In other words, the fact that its a usb to spdif converter (vs. a usb dac) doesn't necessarily increase or reduce the probability that the touch could communicate properly with it via the applet.

Sorry if I am being repetitive or verbose...just trying to make sure I understand.

Thanks.

[Triode]

If I decide to "roll the dice" with this crazy idea of using a usb to spdif converter at the output of the touch (because the sbt doesn't talk to the particular dac usb input that I have via the EDO applet) ...then from your reply is it possible that the touch could be ok talking to the usb to spdif converter even though it doesn't talk to the usb input of the dac? For example, might it be ok with talking to a MF Vlink? I am not saying this is necessarily the best idea -- just trying to verify that it could possibly work in the same sense as the touch via the EDO applet could possibly talk to the usb input of a given dac. In other words, the fact that its a usb to spdif converter (vs. a usb dac) doesn't necessarily increase or reduce the probability that the touch could communicate properly with it via the applet.

Yes there are usb -> spdif converters which conform to the usb audio spec and will work with the app. As John says whether this results in a better spdif connection to the dac than the direct one from the touch will be very dependant on the converter and also the sensitivity of your dac to spdif source variences. If you do want to go down this line, make sure you look at usb -> spdif converters which work with linux without any additional drivers. I would search this thread and the async usb experiements thread to confirm if anyone has tried specific ones (though this will validate that they work, not that they are better than your current direct spdif)

[Baronvonkarp]

Hi Triode,

I de-installed de EDO app, to hear the difference between 192 and 96Khz. After that I wanted to install de EDO app again, but after i installed the EDO app and de SBT reboots, I got the message that the kernel could not be installed? When I look to the installed apps, the SBT says the EDO app is installed, but at rebooting I do not get the message (in the upper right of the start window) that the EDO app is installed. Is the EDO app installed correctly or do I need to do some onther things to let it work correctly.

In between of all of this I got my SBT set back to factory and installed the last software updates. But I still get the message that the kernel can not be installed, when i am installing the EDO app.

[SteveCresswell]

Hi Triode,

I de-installed de EDO app, to hear the difference between 192 and 96Khz. After that I wanted to install de EDO app again, but after i installed the EDO app and de SBT reboots, I got the message that the kernel could not be installed? When I look to the installed apps, the SBT says the EDO app is installed, but at rebooting I do not get the message (in the upper right of the start window) that the EDO app is installed. Is the EDO app installed correctly or do I need to do some onther things to let it work correctly.

In between of all of this I got my SBT set back to factory and installed the last software updates. But I still get the message that the kernel can not be installed, when i am installing the EDO app.

Exactly the same with me.

Since this morning, internet radio and the BBC iPlayer app started failing to connect to the BBC with the message could not resolve the ip address for bbc.co.uk
Then I started having all sorts of issues with menu items failing to select, cursors failing to move up or down to select options.
I had no alternative but to do a factory reset after which the enhanced digital output app failed to update the Kernel. There is now no message in the top left hand corner on the reboot screen and the EDO won't work without the kernel, which won't install so the EDO won't work anymore.
I think they call it catch 22