FORGETTING the Digital to Analog conversion part, what is BEST Digital source?

[barrows]

54 minutes ago, beerandmusic said:

If a song is 10101010 is there a method or device that can transmit that 100% accurately up to the d to a converter?

All devices in common use by audiophiles actually do transmit the digital data accurately.  The only thing left to effect sound quality is the electrical noise which comes over the interface with the data.  In the case of USB audio, the noise appears to contribute to sonic problems, most likely by corrupting the internal DAC clock power supply (creating jitter) and perhaps even by the electrical noise mixing with analog signals creating intermodulation distortion.

In the very early days of USB audio there were actual data transmission problems, but these result in very audible tics, clicks, or even short dropouts of sound, thankfully these issues were solved long ago. 

[barrows]

@beerandmusic

 

I was referring to USB audio, which technically is superior to SPDIF.  I saw in another thread that you are not considering USB, I consider this a mistake and it may be that you just need a better USB source than a commercial computer.  Commercial computer products are not built to high standards for noise and audio, so they produce a lot of noise which can result in sound degradation (but this degradation is not due to data loss).

Additionally:  SPDIF is subject to jitter (not really data loss), that is why it has problems (jitter is timing error and not data loss), the SPDIF cable is important to reduce jitter, but using a commercial computer as an SPDIF source is not going to be ideal, just as it is not for USB.

This is why high quality audiophile digital sources exist: servers, and Ethernet renderers, to provide a much better source than any computer.

[barrows]

47 minutes ago, beerandmusic said:

So PC's are out in your opinion to accurately deliver the 1s and 0s...what is your suggestion for perfect digital transmission?

No, they will deliver bit perfect audio almost always, unless something is wrong with your settings.  The problem is they will also deliver a lot of noise which can make for problems in the DAC and sound degradation at the output.  As i mentioned before, this why there are audio source solutions like servers and Ethernet renderers with low noise.  I prefer using a very low noise Ethernet renderer with USB output.

[barrows]

Just now, beerandmusic said:

Isn't the usb output on the ethernet renderer also prone to noise issues?

Yes, but the point is that good Ethernet renderers are produced by companies with audiophile values, and address the noise issues specifically.  For example, a printer is not going to care about noise coming over on the USB cable, it is going to make sure the data is perfect, and then print the document, regardless of any noise.  Computer companies are building their products to the lowest possible price which will still work, not in order to produce audiophile quality sound from a connected DAC.

Audiophile companies (hopefully at least) are paying much more attention to small details which matter for audio, specifically making sure they have the best USB signal integrity possible with the lowest possible noise, this is why quality built digital source components are going to provide better sound quality (and why they cost more) than a consumer computer product. 

[barrows]

Just now, lucretius said:

Do you mean analog noise like hum?

No.  Although ground loop hum could be possibly caused by a poorly implemented USB interface.  I mean noise coming over USB and wreaking havoc in the DAC, all of the following are possible reasons:  noise on the ground plane of the USB interface, getting to the oscillators and causing additional jitter (short term clock error).  Noise coupling into analog circuitry and causing intermodulation distortion of analog signals.  Noise coupling into the DAC chip from the ground plane and disrupting internal operations (remember DAC chips are half analog components).

[barrows]

15 minutes ago, beerandmusic said:

but still, wouldn't it be better to go from ethernet in directly into the d->a circuitry without having to go out usb?

Ultimately I think that yes, Ethernet input DACs have the potential to be great.  The problem is the computing power necessary to receive Ethernet transmissions (especially hi res audio) is much higher than that required to receive USB audio, so with Ethernet you have to have a powerful computer processor in the DAC.  Having this processor produce a ton more noise, and as we have already explained, this is a problem.  You want the most "quiet" from a noise perspective, component internally for audio, so putting a powerful computer processor in there creates more problems.  The only DAC I have heard with direct Ethernet input that sounds really awesome is the very expensive Linn Klimax model, I could live with that one (except for its old school sample rate limitations), but it is way out of my league cost wise.  In the Klimax Linnhouses the Ethernet processor in its own shielded compartment carved out of a solid block of aluminum, this shields it  from interfering with the audio circuitry.  Unfortunately this approach is expensive to implement.

I use an Ethernet renderer in a separate chassis, at some distance from the DAC, to keep processing noise away from the DAC circuitry.

[barrows]

28 minutes ago, beerandmusic said:

what about the teac nt503 which has won Japanese awards?  I believe you said before you haven't tried it?  Can the "processor noise" be measured internal to the device, so a professional can tell whether or not the noise is an issue?

 

Or what about the many AVR's or spinners that have ethernet capabilities (e.g. marantz)?

 

Again, i want to keep the analog part out of it for now.

What about it?  You cannot keep the analog part out of a DAC, a DAC is an analog component and in the analog section is where most of the problems are going to be manifest.

As to whether or not the processor noise could be measured I do not know the answer to that, but it is probably so.  this is likely a question for an RF engineer, which I am not.  What I do know is that that airborne RF noise mostly scales with processor activity levels, and Ethernet requires much more processor power/activity than receiving USB audio does, hence it cause more noise problems, and a lot of this noise will just be broadcast into the air around the processor, to be picked up by any nearby circuitry.

No product is perfect from these standpoints.  High end audio exists precisely to address these lower level issues; things will still work if you do not address these issues, but they will not sound as good.  Ethenrt is added to most everything in the consumer electronics world now as consumers demand it, that does not mean it is a good idea to do so, nor that these products will produce awesome sound, most consumers do not care about that anyway.

[barrows]

3 minutes ago, beerandmusic said:

So are you here suggesting a hardware device with no ethernet circuitry is capable of producing better SQ?  E.g. a usb only dac will sound better than an ethernet only dac, if implemented properly by same best designer of both usb & enet...(obviously hypothetical)?

What I am saying is that with the very best implementation possible i like the idea of an Ethernet input DAC, but making it really good is very difficult and will be expensive.  What does the Linn Klimax which I reference go for, $20K or something like that.  But I cannot afford a Linn Klimax, even for DIY a chassis like Linn uses would be around $2K plus finishing costs. It is far more cost effective to have a good USB DAC, fed by an external Ethernet renderer, where by having the Ethernet stuff in a separate chassis, at some distance from the DAC, the noise problem is much reduced.

So, if you wan to get an Ethernet DAC, and have decided to spend much more money, and can live without DSD or sample rates above 192 PCM, get the Klimax and be happy.

[barrows]

1.  A NAS would be for storage.  The approach of using a renderer allows any commercial computer gear to be well away from the audio system.  I prefer to keep the computer gear in another room, far from the audio system, and then have just the high end renderer cost to the system.  The renderer is like what our company makes (the Rendu products), or competing products like the Auralic Aries, etc.

 

2.  I just mean on another shelf on a rack.  As with all audio gear, a little spacing between components is not a bad idea.

[barrows]

2 hours ago, beerandmusic said:

The dacs should be able to implement a way to receive the file perfectly regardless of the cable and discard the noise...so that all is left is the designer's method of guestimating the analog out.

Yes, that would be nice, but noise is everywhere.  Data integrity is not a problem, noise is.  Noise is not so easily "discarded".

A renderer is generally capable of DLNA, but that is just one way (the most common) of transferring audio files over Ethernet.  If you use DLNA, you can use an android tablet or phone as the control app, these are independent of the hardware in use (Renderer, like Sonore Rendu or Auralic Aries as examples).  A typical renderer set up looks like this:  There are multiple control apps to choose from.

 

NAS-Ethernet cable-Router- Ethernet cable-Renderer-USB cable-DAC, with a control app on a tablet or phone communicating via WiFi.

[barrows]

21 minutes ago, beerandmusic said:

yea, i don't want to have to use a phone...i suppose i could aways use jriver, upplay, foobar on windows....i am basically doing the same thing as the renderer in your example but using a sony bluray player as the renderer with bluray coax out to dac....sounds decent.

Yeah, if you prefer not to use  phone or tablet via Wifi you can control things via a computer on the network using JRiver, Roon, or LMS, etc.

[barrows]

8 minutes ago, beerandmusic said:

schiit's marketing says the eitr does it, and i thought galvanically isolated dacs do as well?

None of this is absolute.  Well done galvaniccally isolated USB inputs are certainly recommended, but the isolation is still not perfect and some of this noise gets through.  My DAC has a fully isolated, floating USB input, but still sounds better when a lot of attention is paid to reducing noise of the USB source.  Never think of "isolation" as being absolute, there is no perfect "isolation". 

[barrows]

I love the DS DAC, I have had a couple of them come through my hands and have listened to them quite a bit.  I would recommend them without the built in ethernet input and using a separate Ethernet renderer for all of the reasons previously stated.  But if you have the opportunity you should try it at home either way to decide for yourself.

[barrows]

3 minutes ago, beerandmusic said:

DS DAC as in Korg? 

PS Audio DirectStream

[barrows]

Hey Alex,

 

So you are suggesting that even the older XMOS is capable of this, and to what sample rates?  I know the newer XMOS chips have a lot more processing power and hence use a lot more power (equates to noise), but the original versions cannot do nearly as mush, and to what sample rates?  Even Convers can only do 192 PCM, which is really limited these days with people demanding DSD256, sometimes via DoP (so 752.8 PCM).  What I have seen from other solutions people have, without a quite powerful processor is limited in capabilities.  For customers to buy in I feel one needs wide ranging support for ROON RAAT, DLNA, HQPlayer NAA, and sample rates to at least 384 PCM and DSD256 native, if not DSD 512.  Or was you approach much more minimal?  Without universal support for all these approaches one risks leaving behind a lot of customers.

Every Ethernet interface I have worked with has always used at least twice the power of a USB interface, even ones limited to 24/192?  Although the newer XMOS (I think it is the X-208?) does seem to need more juice itself.

 

On the SDtrans, I am well aware of that approach, but it is too tweaky for most these days.  I have a Resonessence Mirus PRO DAC at home right now and it has a built in SD card player, but the user experience is too far from the convenience which most are used to gain much acceptance.  A library on SD cards is hard to imagine managing.  Sound quality is good on the Mirus Pro via SD card, but not better than the Signature Rendu SE via USB.

[barrows]

10 minutes ago, barrows said:

delete, its late here on the east coast, Maine.

 

[barrows]

@Norton, yeah, for my listening SD cards are just too fiddly.  While I mostly listen to entire albums, I am likely to go from Bach to Lorde to Tool in one listening session, so there is no way to assure that what i want to listen to next is going to be on the same SD card.  Plus I found navigating even a 32 Gb card a PIA compared to DLNA via Linn Kazoo.  I am not willing to take a step back for no apparent sonic gain either, but perhaps not everyone is having the SQ I already have via DLNA.

 

@Superdad, on Ravenna, my understanding is that Ravenna allows for any playback engine/software to work, if this is true it may be a really good solution.  It seems the only drawbacks may be price of the module and the fact that it appears to have somewhat more demanding network requirements than are normal for consumer home networks.  I guess we will see when the modules become fully available... 

[barrows]

2 hours ago, beerandmusic said:

don't know if it is noise or some other usb issue, but i swear every time i try s/pdif it sounds "cleaner".

Most of us have had the same experience if using a standard commercial computer as a source via USB, especially if the DAC in use has a really good SPDIF input.  This just indicates that the USB input in that DAC may not be so good, and a standard computer is a very noisy source.

If you build a SD card source, then how does that source connect to the DAC?  You still have the same problems, you have to get the data to the DAC somehow, and every way has its challenges, so you are trading problem for problem.  Every way can be done well, or poorly, but you do need to pay attention to whether it is done well or not.  I am confident in the performance of the Rendu products specifically because I have compared them directly to other sources, including the SD card reader on the Mirus Pro.  The SD card reader on the Mirus Pro does sounds great, but no better than the Signature Rendu SE.  this is an easy comparison to do if you have a Mirus Pro at hand with no cable swapping to get in the way of A/B switching.  If you really like the SD card approach, then a used Mirus might fit the bill.  One box solution which is nice as well.

I have not used Ravenna yet, but Chris C. has, please read his experiences with Ravenna on this site to learn more about its network requirements.  Ravenna sounds pretty promising to me, except for the expense (unknown at this time I believe)  For a DAC manufacturer to use it they must purchase the module for Merging, Merging is a Swiss company, and their DACs and ADCs come at premium prices, while being fairly ordinary (although well engineered for sure).  I suspect the Ravenna module will be fairly expensive to incorporate, enough so that it will not be seen in DACs under about $5K or so, but I cannot be sure yet.

@marce, on close coupling of components: I always remember John Atkinson's measurements of one of the MSB DACs, which stacked its DAC box directly on top of its separate power supply box: doing so induced AC frequency noise into the DACs analog output (the components in question were designed to be stacked this way), of course you may have been referring to HF noise re: chassis shielding.

[barrows]

2 hours ago, beerandmusic said:

I am sure the chinese or some other country will bring something similar to market for under $1K

Very unlikely.  The Chinese DACs are usually reversed engineered copies of existing products, unfortunately most of the interesting bits of the DirectStream are in the coding of the FPGA, maybe they could copy that, but I doubt they would bother.  They are better off copying DACs where they can just put a DAC chip in a box with some bits and bobs...

[barrows]

2 hours ago, beerandmusic said:

My current library i widdled down to 512GB...had over 1TB (many dsd files).  I have no problems managing that much data, it is organized nicely...and they make 1tb sd cards.

My library is not huge, currently about 2TB, but it is also always growing, not getting whittled down!  With the Mirus Pro I do not think you can use Foobar as shown here for management, it is kinda of slow to get through a bunch of albums to find what you are looking for.  I could not live with that after the experiencing the convenience of DLNA with Linn Kazoo app.  But it may be OK for you.

[barrows]

1 hour ago, beerandmusic said:

I am pretty sure that I have configured foobar in past to work with dlna server (e.g. minim server)...so i would think could be used with Mirus Pro.  I really don't think finding any album or song in a couple seconds...on any size library would be an issue.

I was referring to using Foobar as a way of browsing/library management for the D card player built into the Mirus Pro.  You can only find a song/album anyway if the correct card is in the Mirus Pro as well.  but, like i said, if these are not issues for you, you might very much the SD card approach.

[barrows]

12 minutes ago, marce said:

Its conducted noise into the DAC, should be catered for, we do it all the time on a variety of things that are going to have wires plugged into them, audio and instrumentation and of course the mil stuff where the requirements are tighter. Though I do agree with unbalanced it can be a PITA, balanced for analogue, LVDS for digital is best.

Agreed.  Unfortunately most using the LVDS/I2S approach have the masterclock in the source, it can be done the other way, with the masterclock in the DAC, but that is not what is available in commercial products (with the only exception I ma aware of being MSB's proprietary I2S approach).

[barrows]

3 hours ago, beerandmusic said:

I am just tired of dickin with usb.

For the third, (or fourth?) time:  You are not going to get great results out of USB by using a commercial computer as your source!

A well made audiophile source is in order here.  Like a good custom server, or good Ethernet renderer, etc...

 

The problem is not USB audio, it is using a noisy commercial grade computer as your source.  There is not anything wrong with USB audio in and of itself.

[barrows]

3 hours ago, marce said:

I would like to see some proof of the differences in noise outputs between standard equipment and some audiophile stuff, as this is the only way to determine how good or bad a device is regarding EMC compliance.

All this talk of noise but NO measurements, the only way to control noise, is know what noise there is, then you can design the correct filters, that's the only way, anything else is guessing.

@marce,  I respect your experience and always appreciate your posts, both here and at diyaudio.  I know you have a lot to offer the community on high speed circuit design, and thank you for participating here.  I am referring to the difference between the USB output of a regular commercial computer, and a high end audio source such as Sonore's Signature Rendu SE.  I do not know the best place, or method to measure noise on the USB output, but I do know that John Swenson looks at this.  If you have any tips I would appreciate that.

What I do know is that the sonic differences are not subtle at all, and everything is, of course, bit perfect in any case.  So the only plausible explanation for me is noise.  JS has a lot of information on USB noise, and some theories, although mostly unproven as to what is going on technically (and he is trying to verify those theories).

For proof of some kind of differences, even just between two commercial computer sources, I believe John Atkinson measured some significant differences in the output of a Chord Mojo between two sources: his laptop computer, and an iPad.  You can find those at stereophile.com in the digital processor reviews for the Chord Mojo.  Just one example.

I understand there is some controversy amongst some engineers about this, but listening to the difference on a good system will end the controversy immediately for anyone with any audiophile tendencies at all.  Certainly it would be nice to confirm exactly what is going on, and I welcome any help you might offer in that area.

 

I remember at the early days of high end computer audio playback, when Amarra first came out.  I could not believe that a playback software program could make a difference in SQ, as long as one confirmed things were set up to be bit perfect, but the sonic evidence was not questionable at all.  Then I learned about the engineers responsible for Sound Blade (pro audio software) and all their efforts to make better sounding software, and the opinions of recording engineers who all accept that different softwares sound different, even in bit perfect applications.

[barrows]

1 hour ago, beerandmusic said:

And others have not yet followed suit?

Many other DACs already have similar or better circuitry to what Schiit has.  The Schiit guys like to make up marketing terminology and then pretend they are doing something entirely new and different.  While their marketing approach and price points are a little different, none of their tech really appears to be, except perhaps their use of 20 bit ladder chips for an audio DAC.

Just because someone says their circuitry "eliminates USB noise problems" do not believe them.  Yes, a well done isolated USB input will REDUCE the susceptibility to USB noise, it will not eliminate it, as pointed out by @marce, even with full galvanic isolation noise can still couple through to some degree.

I am aware of no DAC which does not benefit from a better USB source, despite the claims of some manufacturers.  It is a belt and suspenders world out there, and noise issues need to be addressed at every stage of an audio system.  No free lunch as they say!