What is the best method to get digital audio from your PC to your DAC
Aug 2, 2020 at 8:23 AM Post #31 of 104
I'd like to chime in here but please forgive me if you find something disagreeable. I'm no expert by any means and am only sharing my experiences when it comes to digital. Just a bit of background: I was previously into analogue only (casettes and CDs; typical 90s kid). Over time I found it be difficult to sustain my collection and the final nail on the coffin was when I moved to our new place and the move destroyed 2/3rds of my collection. I have since moved to a digital only setup and have converted whatever analogue I had to digital. I'll admit it doesnt sound as "magical" but it does the job.

So initially i was struggling quite a bit. The wavs and flacs that I had created were being pushed into a DAC from foobar on windows. The quality of audio was pitiful to say the least.

The next step was ASIO and it upped the game but I wasn't satisfied. At this time I upgraded my DAC to a maverick audio tubemagic d2 (my first serious DAC). The change was a revelation and I think that's what motivated me to keep it digital and not giving up on it so fast.

I used the d2 for a few years and in the mean time I experimented with the software side of things. I tried Audirvana, Jriver and a plethora of other windows based options. I think the best was Bughead emperor. But it was just too clunky functionally. Not to mention it required a fairly decent machine. This was the time I got into experimenting with linux distros focused on audio. Tried a lot of them and that's when I stumbled on Daphile. I won't go into the nitty gritty details but it's been my goto audio playback and library management tool but I'll come to it later.

I think this is where I lost track too and gave into the Raspberry pi based solutions. First got a pi and then the allo based solutions. I upgraded to better and better dacs and also auditioned a few dedicated streamers from bryston, sotm, etc. I think the cambridge cxn v2 was the last one I tried and even had it in my system for a few weeks. They are all good solutions but they were either too expensive or didn't sound significantly better than the allo based units.

During this I was also in the process of setting up my reference DAC, I chose to go for a NOS DAC (PCM1794 based). The dac which I wanted to build only had a spdif input onboard and required a seperate usb to i2s interface to convert the signal so that it could be fed to the dac main board. I got a luckit usb-i2s board and powered it using a seperate linear power source. My dac build was completed last year. I used this dac for testing with a lot of equipment.

I was using an allo usbridge at that time and the sound was pretty acceptable. One day the lps powering the usbridge went bust and I had to use an old laptop. I don't know what hit me and i just wiped it clean and installed a live version of daphile on it; remebering it to sound best amongst the pc based solutions. Set up upnp/dlna and played music. I kid you not, the laptop ran circles around the allo. Details and dynamics went into another level altogether.

I compared that laptop to a number of commercial players i mentioned with help from a lot of friends. The laptop was almost always superior (but only with my DAC). I was finally getting the audio quality I wanted.

I then played around with hardware: got myself a dedicated fanless audio pc, linear power supply for the pc, mutec usb reclocker (MC3+ usb; although i just recently sold it too coz i found it to be not affecting the sound much), did some fiddle-foo with daphile and here I am. My source sounds better than most four figure solutions. I'm a happy camper.

TLDR;
1. Software is as important as hardware. Ditch windows if you can. No matter what any software claims, or no matter what bypassing you do; it still sounds inferior to most linux distros let alone daphile.
2. Daphile is good but limited in functionality. There's no free lunch.
3. Avoid pricey streamers, allo stuff sounds pretty good and costs less.
4. The main game is played by your usb to i2s interface. As long as it has a seperate power supply and is isolated from the usb power, you don't need any fancy reclockers or other voodoo-doodoo that companies seem to be selling these days. Check this before you buy your dac. A similar solution will be a dac that takes i2s from an outboard usb to i2s interface. But i found that it only prolongs the signal path.
5. Good grounding and a linear power supply for your PC is almost always ideal. But differences won't be very huge.

Thanks for reading and good luck!😊

100% agreed that linux audio is a crazy upgrade over windows. I use wtfplay, similar to daphile but this time commandline instead of fancy ui.

Now regarding upgrade with dedicated streamers, I haven't explored but I'm inclined to believe they may sound better than PC direct when done well for many DACs. So the thing about these is that we are mostly playing against stress along the phy or ground plane noise, either due to periodicity issues of the data or due to emi along the data. I had as much improvement moving from stock cable to uptone uspcb cable as much as I did moving from windows asio players to wtfplay. Now I understand some DACs might have super well isolated systems from usb input to i2s conversion making them impervious to these, and great internal clocks, but do remember that you'd have been paying quite a bit for that design. Isolating power supply, optocoupling etc cost quite a bit. Schiit sells just interface upgrades for 300$, and this is excluding the power supply design you already paid for. If I bought a quite expensive dac I guess I could take it as a given, but for the DACs I could afford, I doubt.

Regarding NOS, unfortunately I share a different sentiment. I'm in for extreme oversampling for as close to accurate reconstruction as possible (closest to perfect sinc, minor windowing changes aside). I agree audio is ymmv, but don't let a poor OS reconstruction filter get you to thinking all OS digital filters will be bad.

And last but not the least don't waste your time with the guy with preconceived notions that windows audio is better supported. Alsa is UAC compatible from the get go, windows didn't support so they had to write driver. On top of that linux has special audio priority kernels developed just for the craziest of us. For interfaces, manufacturers care for Mac the most and we are talking about things like round trip latency etc which are not directly related to just sq perspective. Next thing you'll hear from those guys, they'll try to tell you windows directsound is an audio enthusiast "feature".
 
Last edited:
Aug 2, 2020 at 8:57 AM Post #32 of 104
100% agreed that linux audio is a crazy upgrade over windows. I use wtfplay, similar to daphile but this time commandline instead of fancy ui.

Now regarding upgrade with dedicated streamers, I haven't explored but I'm inclined to believe they may sound better than PC direct when done well for many DACs. So the thing about these is that we are mostly playing against stress along the phy or ground plane noise, either due to periodicity issues of the data or due to emi along the data. I had as much improvement moving from stock cable to uptone uspcb cable as much as I did moving from windows asio players to wtfplay. Now I understand some DACs might have super well isolated systems from usb input to i2s conversion making them impervious to these, and great internal clocks, but do remember that you'd have been paying quite a bit for that design. Isolating power supply, optocoupling etc cost quite a bit. Schiit sells just interface upgrades for 300$, and this is excluding the power supply design you already paid for. If I bought a quite expensive dac I guess I could take it as a given, but for the DACs I could afford, I doubt.

Regarding NOS, unfortunately I share a different sentiment. I'm in for extreme oversampling for as close to accurate reconstruction as possible (closest to perfect sinc, minor windowing changes aside). I agree audio is ymmv, but don't let a poor OS reconstruction filter get you to thinking all OS digital filters will be bad.

And last but not the least don't waste your time with the guy with preconceived notions that windows audio is better supported. Alsa is UAC compatible from the get go, windows didn't support so they had to write driver. On top of that linux has special audio priority kernels developed just for the craziest of us. For interfaces, manufacturers care for Mac the most and we are talking about things like round trip latency etc which are not directly related to just sq perspective. Next thing you'll hear from those guys, they'll try to tell you windows directsound is an audio enthusiast "feature".
You hit the hammer on the head my friend😅👍
I haven't used wtfplay but I believe you're definitely enjoying the good sound as long as you're getting a bitperfect stream.
Yes the usb to i2s interface and its power supply did cost me but it was every dollar well spent. I have noise free and clean usb conversion and the clocks on board are exceptionally good.
To oversample or not-to-oversample is the question; and I guess it comes down to taste, since both have pros and cons of their own. I find the NOS dacs to be quite relaxed and to my tastes. My dac in particular takes me closest to my memories of enjoying music on my analogue gear.
In conclusion I'd like to reiterate that linux is far better and easy to configure for good quality usb audio. You're also spoilt for choices on implementation in the form of soo many distros. Better yet linux provides unmatched control over the technicalities of playback and most of these distros are very configurable in that sense.
 
Aug 2, 2020 at 1:32 PM Post #33 of 104
The guy has been spewing his anti linux agenda in too many places. Especially against people who have actually tried all and found linux was better 😔.

Having owned surface book and an avid fan of windows phones before they were shown the door, I use windows for almost everything. Except audio, where I use wtfplay.
 
Last edited:
Aug 2, 2020 at 2:00 PM Post #34 of 104
The guy has been spewing his anti linux agenda in too many places. Especially against people who have actually tried all and found linux was better 😔.

Having owned surface book and an avid fan of windows phones before they were shown the door, I use windows for almost everything. Except audio, where I use wtfplay.
Oh I see. Wonder why people do that? What's to gain by misguiding someone?
I have nothing against windows personally and have 3 windows devices at home. I think its a capable OS with tons of apps and a solid repute amongst the business and tier 1 service sectors.
It's just not built for serious audio, that's it.
 
Aug 8, 2020 at 10:48 AM Post #36 of 104
So guys, I use a dac too.
And hqplayer
But I use windows and it's all I can, if I get a curious cable with thr 5v power external am I ok?
Or do I need an uptone thing?
Before you jump straight into uptone, which is a costly affair, please try an allo sparky as hqplayer endpoint to begin with. Same with cables, curious cables is like 300$. You can start with something inexpensive, I enjoy my uptone uspcb which is a 35$ pcb/cable.

Tldr: don't jump in head first. Dip and move in slowly correlating what works and what doesn't in your setup.
 
Last edited:
Aug 9, 2020 at 4:02 AM Post #37 of 104
Hi

Currently, I am using a PC Foobar (WASAPI) -> USB -> Fostex HP-A8 (internal clock).
I am thinking of buying a CD Transport or would it be the same as using the PC CD-ROM, since there is an option on the Fostex HP-A8 to use the unit's internal clock?

Also, should I output from the PC using USB or TOSLINK? Or no difference since the output is in WASAPI mode?
 
Last edited:
Aug 21, 2020 at 2:06 AM Post #39 of 104
Don't most studios use Windows for recording and mastering? Surely it's sufficient for home audio.
But they also use crazy expensive dacs, dedicated clocks and whatnot.

It's about the end result. Maybe they have circuits either in the middle or inside the dac like these regenerator equivalents or opto couplers.

Windows has well supported daw tools that they need to use.

For home use, if you can go as expensive as studios probably windows would do great since the issues will be mitigated somewhere down the line by expensive components.

But if you don't have serious budget, it's better to cut off noise at the source, where Linux is king. You can take full control there.
 
Aug 21, 2020 at 9:14 AM Post #40 of 104
But they also use crazy expensive dacs, dedicated clocks and whatnot.

It's about the end result. Maybe they have circuits either in the middle or inside the dac like these regenerator equivalents or opto couplers.

Windows has well supported daw tools that they need to use.

For home use, if you can go as expensive as studios probably windows would do great since the issues will be mitigated somewhere down the line by expensive components.

But if you don't have serious budget, it's better to cut off noise at the source, where Linux is king. You can take full control there.


Based on that answer, I seriously doubt you've ever been in a professional recording studio. Good luck finding an expensive "audiophile" DAC in a one - what you will find is moderately priced professional equipment.
 
Last edited by a moderator:
Aug 21, 2020 at 4:14 PM Post #41 of 104
Based on that answer, I seriously doubt you've ever been in a professional recording studio. Good luck finding an expensive "audiophile" DAC in a one - what you will find is moderately priced professional equipment.
I think the words "studio" or "recording studio" is being used rather loosely here.
In the professional music business, a studio might either be plainly a recording room and a console room (where the singers/musicians and recording artists/music directors; record music) OR it could be a recording studio with a dedicated area/room for mastering/mixing (most often the console room will have mixing/mastering equipment as well). In the case of the first scenario the recorded music is mixed and mastered in a seperate place (also called as studio or mastering studio).

A studio doesn't usually require a front end Windows based machine as most of the work is done in the analogue domain. Or atleast this is how it was in the (g)olden days. However these studios required huge spaces and it wasn't easy to maintain the equipment either.
With software based DAWs taking off, most analogue consoles have been replaced by digital consoles where the workflow involves digital equipment which interfaces with a software front end. These softwares can be windows or mac based (macs being more popular).

Having said this, I must emphasize that any serious or professional recording studio has pretty expensive digital consoles which have adc/dac/preamp/mixing capabilities. These consoles also come with their own proprietary softwares that have drivers/service layers that help reduce latency and distortion.

TLDR: A windows machine with a usb based interface and FLstudio as DAW is not the ideal recording studio. The audible product of such a setup will always be inferior to a dedicated console based rig and faaaar inferior to an analogue recording workflow. Windows is not optimised for music recording or playback. Period. If you're happy with what you're hearing, great. Congrats. If you still want something better ditch widows.
 
Aug 21, 2020 at 7:54 PM Post #42 of 104
My own experience with this led me pretty fast to using a separate streamer, and pushing data using Roon or LMS, but in the meantime, I found the following....

USB covers every eventuality in terms of data transport capability - other connections may deliver better sounding results, but usually because the timing and noise of the async/isosync stream is being cleaned up in some form of filter or conditioner in the box. Nearly all coax/AES convertors also do some conditioning - this is entirely possible with USB as well. Where people have done tests (plenty on Headfi) using an AES connection and USB connection between source and DAC, the results are usually imperceptibly different (where people took care to use identical manufacturers cables)

In some tests where I used a USB conditioner, provided it was independently powered and delivered a clean power and ground into the USB connection, it provided a noticeable step up in SQ from direct connection to the PC. Less convincing were tests with conditioners that 'reclocked' the stream - marginal gains at best, fractions of noticeable SQ improvement.

Now, this held good for some computers, but not all - With a custom-built. high wattage high power PC, the SQ improvement was marked. With an Apple MacBoook pro, it was minimal, and with a lenovo desktop, it made no difference at all.

In summary:

USB can deliver the same SQ as CO-Ax or AES, there's no 'magic' in the connection format - perceived SQ wins in AES or CoAx are really wins from conditioning or reclocking within the convertor. Unless your DAC doesn't support USB, it's the most flexible choice.
It's worth testing a device to see if it makes any perceptible difference - it varies vastly between PC's and DACs - test it in your system.
It's entirely possible that the USB out from your computer is fine, and won;' necessarily be improved by any conversion, clocking or isolation.

Have fun!!
 
Aug 22, 2020 at 6:02 PM Post #43 of 104
Missing knowledge on how our perception and interpretation fully work, is a strong reason to double check with controlled methodology before claiming to know anything for a fact. Ignorance about some human mechanisms is certainly not an excuse to pretend that sighted impressions prove objective facts about fidelity.
Your point actually makes a case for measurements as a way to avoid the possible issues and unknown unknowns in the human subject. Not that it was needed in the first place as fidelity is not a matter of opinion, subjective experience, or even hearing. Fidelity is an objective concept, a degree of change against what we're trying to reproduce. So from the start, measurements were the answer for matters of fidelity.

Is Linux better than Windowz for audio playback? Can anybody really claim to have one definitive answer to that? Just considering the settings available on each OS, the great many computer builds and potential devices plugged to them, the various distros of Linux, the possible differences between motherboard USB and a USB card, the many different DACs with at least a few different protocols and drivers... It makes any form of generalized claim look like someone didn't really take the time to consider the complexity of situation.

OP's questions and some of his self answers, can only be considered with cautious rule of thumbs IMO. like what @Roseval started here https://www.head-fi.org/threads/wha...rom-your-pc-to-your-dac.932199/#post-15724274
The actual gear involved will always play an active role in determining what's the best solution(if there is a specific issue to mitigate in the first place). So again, measuring a bunch of options on a given system would be the right way to approach those questions, or other possible problems we encounter on a system. The answers would most likely not translate onto all other playback systems, but at least on that one system under that particular set of conditions, we'd have an actual answer instead of whatever that thread has turned into.
 
Aug 23, 2020 at 5:11 AM Post #44 of 104
Missing knowledge on how our perception and interpretation fully work, is a strong reason to double check with controlled methodology before claiming to know anything for a fact. Ignorance about some human mechanisms is certainly not an excuse to pretend that sighted impressions prove objective facts about fidelity.
Your point actually makes a case for measurements as a way to avoid the possible issues and unknown unknowns in the human subject. Not that it was needed in the first place as fidelity is not a matter of opinion, subjective experience, or even hearing. Fidelity is an objective concept, a degree of change against what we're trying to reproduce. So from the start, measurements were the answer for matters of fidelity.

Is Linux better than Windowz for audio playback? Can anybody really claim to have one definitive answer to that? Just considering the settings available on each OS, the great many computer builds and potential devices plugged to them, the various distros of Linux, the possible differences between motherboard USB and a USB card, the many different DACs with at least a few different protocols and drivers... It makes any form of generalized claim look like someone didn't really take the time to consider the complexity of situation.

OP's questions and some of his self answers, can only be considered with cautious rule of thumbs IMO. like what @Roseval started here https://www.head-fi.org/threads/wha...rom-your-pc-to-your-dac.932199/#post-15724274
The actual gear involved will always play an active role in determining what's the best solution(if there is a specific issue to mitigate in the first place). So again, measuring a bunch of options on a given system would be the right way to approach those questions, or other possible problems we encounter on a system. The answers would most likely not translate onto all other playback systems, but at least on that one system under that particular set of conditions, we'd have an actual answer instead of whatever that thread has turned into.
With all due respect sir, it seems you're trying to answer a question and then raising doubts over your own answer.
On one hand you want objective evidence of something being better than the other by conducting a test on a specific system and then you're saying that results of such a test might be difficult to extrapolate to other systems/settings.
This doesn't solve any problems either and definitely doesn't answer the OPs question under any circumstances. So this comment on the redundancy of the previous posts is as redundant as the posts themselves.

I do agree that a gross generalisation about one thing being superior to the other is a hollow claim, but given that we're talking about music playback here, I think it wouldn't be wrong to present/share our subjective opinions/experiences and let the OP and other respected forum members draw their own conclusions by trying the solutions mentioned here themselves. I mean I'm not using my gear at home to listen to frequency sweeps, test tones and noise samples, right? Music and art is as subjevtive a topic as food and drink.

Anyway, I am not claiming anything here. I just tried to share my experience, that's all. Over the years I have spent a large amount of money to get my system to sound "good" (read into this as you please). I have made my fair share of mistakes and losses. To that end, I've found that I was much more satisfied with the sound coming from linux (in my case; Daphile) than windows. I cannot present any objective data to support my observation and as you rightly said, even if I do; the same results might not be replicable in every setup.
Let me reiterate: windows is an operating system meant for a very wide variety of usage/applications. At its core and in its vanilla form, it is not optimised for "high fidelity" audio playback. Linux distros which are meant for the sole purpose of handling music are optimised for this task. They offer much more flexibility and control over handling audio data than windows. Most importanly the latter is free (most distros anyway). Given the above, I think it would be unjustified to not give linux a chance.
Thanks.
 
Aug 24, 2020 at 1:24 AM Post #45 of 104
After playing around for a while, here's my findings:

1) WASAPI is a must, without it is just not listenable
2) Playing CDs through my PC Blu ray ROM is superior to loseless in WAV or FLAC format
3) Optical out is better than USB

The optical cable i used is 2USD (yes two, delivered) vs my Wireworld starlight USB cable (can't remember how much i paid, 100USD?).
Next I feel like trying the glass toslink cable to see if there's any difference.
 

Users who are viewing this thread

Back
Top