Onboard vs Offboard Audio

N

NeoMatriX724

Limp Gawd
Joined
Jul 26, 2007
Messages
269
Any thoughts on onboard vs offboard audio?

I'm currently driving HD650's off of onboard Realtek sound. I'm running Optical Out into a Schiit Modi/Magni Uber DAC/Amp stack.

I'm wondering if there's any quality gain to be had from going to an standalone sound card.

There's a fair amount of Movies, Music and Gaming involved.
 
Your Modi basically is a standalone card. Just external and with no DSP. Get the drivers from Schiit, connect the Modi to your PC through a USB cable and see if you hear a difference. Though, running optical out, you're bypassing the realtek pretty heavily as is.
 
Yea, I figured it'd be marginal just wanted to make sure I wasn't missing anything. RealTek for just straight audio is pretty solid, more the connection points and on-board noice become an issue.

Optical pretty much bypassed all those from what i hear.

Wasn't sure if discrete cards did anything else.
 
A discrete card would be an alternative to your DAC and headphone amp. No sense in using both, unless you need surround sound or something like that separate from your stereo headphones. In most cases, a setup like yours would be preferable to an internal sound card because keeping that circuitry outside the computer reduces the likelihood of electrical noise and interference from the computer components.
 
Call it an upgrade itch lol, but I'm happy with my stack of Schiit and my cans. Not much else in that upgrade path then.
 
Neo,

I have almost the exact same setup as you. What I did to A-B test the optical from my Realtek ALC1150 vs USB output to my Modi is this:

Connect both the optical and USB cables to your Modi at the same time. Get the drivers setup for the USB device. Now you can do a quick A-B test by using the button on the front of the Modi to switch input sources. When I did that sound test, the USB sounded drastically better. Much less thin and "tinny" sounding. As much as the optical out shouldn't be affecting the sound, it's obvious there's a negative impact. There was no going back for me.
 
For sure you should test what works best with your setup.
I have an Oppo 105D with the Sabre 9018 Reference DAC.
Optical out from PC (from Auzentech Prelude) at 96K sounds better than USB out at 96K, slightly warmer sound and a little more detail.
All processing is turned off on the Prelude, its as clean as can be.
I cant do higher than 96K on optical due to Auzentech limitation but USB 192K sounds a bit better than 96K optical as you would hope.

Try your own kit and see what you prefer.
 
For gaming, the sound quality to be added is some type of 3d surround processing. For example creative SBX.
 
Astral Abyss said:
When I did that sound test, the USB sounded drastically better. Much less thin and "tinny" sounding. As much as the optical out shouldn't be affecting the sound, it's obvious there's a negative impact. There was no going back for me.
Click to expand...

Interesting that two digital signals would sound markedly different. I wonder if that actually has to do with the design of the DAC.

Nenu said:
I cant do higher than 96K on optical due to Auzentech limitation but USB 192K sounds a bit better than 96K optical as you would hope.
Click to expand...

Again, interesting that the 192K sounds better. Is it a different recording? My understanding is that 192/24 is more of a marketing play than an actual improvement.

http://people.xiph.org/~xiphmont/demo/neil-young.html
 
Same recording downsampled and different recordings.
Although its clear that nothing can be directly established from different recordings, it does often follow the same pattern.
Good 192K recordings are more detailed.
I'm sure I have got some 96K and 192K of the same recording as well but I cant remember which albums.

Not all equipment can expose the difference and not all people can tell or care for it.
People who hear my equipment can tell, but most gloss over it because its not something they would pay extra for, which is fair enough.
A few friends have upgraded their hifis because of it.
 
I guess I'm skeptical. To quote the link I posted:

192kHz digital music files offer no benefits. They're not quite neutral either; practical fidelity is slightly worse. The ultrasonics are a liability during playback.

Neither audio transducers nor power amplifiers are free of distortion, and distortion tends to increase rapidly at the lowest and highest frequencies. If the same transducer reproduces ultrasonics along with audible content, any nonlinearity will shift some of the ultrasonic content down into the audible range as an uncontrolled spray of intermodulation distortion products covering the entire audible spectrum. Nonlinearity in a power amplifier will produce the same effect. The effect is very slight, but listening tests have confirmed that both effects can be audible.
Click to expand...

I would be curious if you can hear any intermodulation on the test files in the link I posted. That could explain some of the differences you're hearing with 192kHz files.
 
Fair enough, determine for yourself.
I dont need test files with an unknown agenda.
 
Sorry, I dont mean to be dismissive but this forums HATES people who can tell a difference, it is sure to start another war.

A test file doesnt prove whether I can or cannot hear what I do.
Also consider that some test files are placebos to try and discredit people.
But even then there are other reasons why 96 vs 192K can show an improvement.
ie if the clocks timing isnt linear enough at 96KHz or has a slower rise time, it may sound a bit better at 192KHz when playing the exact same audio data should those 2 improve with the faster clock.

I used to be sceptical to some extent until I upgraded to the kit I have now.
This is because there was so much bullshit in the industry as I grew up that didnt materialise into anything you could hear, and man did it cost.

I had a few lower quality DACs and amps prior to my current kit and the difference from CD to 96KHz was apparent but not by much and I wasnt bothered. I didnt go out of my way to get 96KHz material.
96 to 192K was not worth the effort, there was a small difference but when listening to music, it didnt matter.
(previous testing in my home was with an Audiolab 8000a amp, Onkyo 875 AV amp and its DAC, Emotiva UMC-1 DAC, Auzentech Prelude, X-Fi and onboard PC audio, Mission 754 and Acoustic Energy Aegis Evo 3 speakers)

With my current kit, CDs sound really fantastic, 96KHz has a fair bit more detail and 192KHz a bit more.
The extra detail from all music is incredible.
I have 2 different Sabre 9018 Reference DACS and both sound so similar and do the same to expose more detail.
The amp upgrade made a big difference in headroom, detail and bass extension.
The speakers are something else!
Not all music benefits from 192K thats for sure and some recordings appear to just be upsampled because I cant tell the difference.
But I'm more than happy with the sound I get.

I am using:
Transmission Line 10" bass. 5" mid and ribbon tweeter fronts.
Emotiva XPA-2 amp
Oppo 105D BD Player, stereo and 7.1 DAC / volume control.
Minimax Tube DAC plus, using Foobar 2000 for music volume control.
Silver, cotton dielectric RCA cables from DAC to amp, soldered directly into the amp and with silver bullet connectors to the DAC. (this cost about £100 and I only made it out of curiosity/boredom but its bloody good!)
Van Damme Hifi O2 free 6mm speaker wire. (very cheap for high quality and very good wire)
 
Nenu said:
This is because there was so much bullshit in the industry as I grew up that didnt materialise into anything you could hear, and man did it cost.
Click to expand...

On this we can agree :)

The purpose of the test files in the link I posted it to determine if a given listening setup is vulnerable to "nonlinearity causing audible intermodulation of the ultrasonics," something that is much more likely to crop up on a 96 or 192kHz file than a 44.1kHz file. Thus, if someone claims to hear a difference with 192kHz files, they may be right, but it could be distortion rather than improvement.

Do you have a link to an article explaining the clock differences? I haven't been able to dig up any information on that subject.

Regarding the agenda of the authors, Xiph.org is: "a non-profit corporation dedicated to protecting the foundations of Internet multimedia from control by private interests. Our purpose is to support and develop free, open protocols and software to serve the public, developer and business markets."

If recordings at 96 or 192kHz are simply mastered better, that's a perfectly good reason to go down that road. In fact, the reason I have a record player (besides the fact I think records are cool) is that some recordings are simply mastered better for the vinyl release (less dynamic range compression) than the digital release. I just don't think 96 or 192kHz offers any benefit on its own. Well conducted blind testing seems to verify this.

http://archimago.blogspot.co.uk/2014/06/24-bit-vs-16-bit-audio-test-part-ii.html

Your setup looks great. Enjoy the music :cool:
 
It was a comment off the top of my head as I am an EE engineer.

One of the things that makes one DAC or CD player sound different is the timing of the clock pulse, whether it be due to the clock itself, what the DAC does with it, what other circuits do with it or interference.
A common term for this type of interference is jitter.
I was pointing out that not all DACs are of the same quality and have varying abilities.
One that is a bit lax at 96KHz in either the timing of the start of the clock pulse, or if its rise time is non uniform such that it doesnt trigger the clock at uniform time intervals, it could provide a sonic improvement if the faster clock improves on either of those.
The faster clock is more likely to have some tolerances tightened, that is all.


Yep, better masters are highly sought after. But also matching the master to how it is played back can matter.
For example, if the master is bitstream and it is not played back as bitstream, there can be a slight drop in detail.
I have verified this using a bitstream file and a 176K / 24bit rip.
My Oppo can play the files directly or can be fed bitstream via USB.
I generally play files/disks directly on the Oppo so its not something I worry about.
I have even tested FLAC vs WAV and in a few occasions, there is a tiny bit more detail from WAV but you have to stretch your hearing to care. It is repeatable though and doing blind tests with my brother on his kit we got it right every time. I havent done that test here.
(he also has an Oppo 105)

Interestingly, I'm about to be given a Thorens TD-150 record deck and am just shaping up for it.
The downside of records is the reduced dynamic range sadly.

Prior to my current DACs, I preferred vinyl.
With the Oppo and Minimax, I prefer well mastered digital, it is more lively.
(This was compared to a Linn Sondek LP12 with Valhalla PSU, RB350 arm I think and a MC cartridge, cant remember which. It was 2 years ago, I no longer have access to it)
The 9018 reference DAC really is that good, its worth checking out.

Cheers :)
 
Nenu I hope you or someone else will call out any bullcrap in my reasoning, as I have stopped at 16 bit, flac, old school power sections

The power supply would introduce distortion with a 192kHz quantized sound stream? :) being played back?

No.

Audio chips on cards that do the upsampling/downsampling have their own means of adapting to an ad-hoc power supply.
Power is one thing, signal sampling rate is completely different and could even prove invisible to the power section.

Upsampling, say, 96 to 192 is not a fruitless endevaour because the DSP on the card may have some dithering algorythms running that detect rapid transients and pad them with corrective signals. I have personally been satisfied with 48kHz and flac.

Also 192 is required for some multichannel operations, like surround sound, home recording, better working equalizers (imagine an instant of a song divided into 192 thousand beeps during a second. Plenty of opportunity to add psychoacoustic equalizers and such.

Furthermore, this ultrasonics thing. A 192kHz sampling rate is not the same as outputting a whopping 192kHz audio signal into the pipeline.
One is the 'firing rate', the other is the 'caliber'.

You should probably be concerned with low frequency hum more, like from ground-loops. Because that has a tendency to induce itself all over the place, 'microphone' stray signals into wires and 'steal' power from the mosfets because a low level signal requires more power to bring to audible levels.

If you want to fool around you might want to simply listen to other gear.

After like 10 years of listening to a FM801, SB Live!, Audigy 2LS, Audigy 2ZS, X-fi USB, a PCM 2704 based module, a prodigy 7.1 hifi, and dozens of source-amp analog signal cables I will tell you this: learn to listen first.
Wind outside can make you adore a song on a particular evening while mains power disturbances will leave you dumbfounded and confused another day.

I picked apart my prodigy 7.1 hifi card like a hawk, testing easily around 7 opamps. Even after all those years I caught myself in an expectation bias, exchanging a component that had no business with the signal path.
Get just.. whatever. Try a low volume on source and high volume on amp combination and vice versa. Try some equalizers. Make sure all your gear shares a common receptacle.
I know you're using digital, but keep in mind - everything plays music. From a capacitor in your power supply to a loose door on a cabinet somewhere. Learn to listen.

In your case I wouldn't plug a pcie card but maybe try a USB solution. More fun to tinker with, and it's possible to power it from batteries instead of relying on dirty USB 5v.

Sorry I couldn't have been of more help. Currently I am more into learning electronics from ground up. I know what I want to build (chip-amp+self-powered DAC complete with digital power, dedicated virtual ground ICs and such.
 
michalrz said:
Nenu I hope you or someone else will call out any bullcrap in my reasoning, as I have stopped at 16 bit, flac, old school power sections
Click to expand...
I'll answer off the top of my head, hope it helps :)

The power supply would introduce distortion with a 192kHz quantized sound stream? :) being played back?

No.
Click to expand...
It can cause noise to be transferred from one part of the circuit to another along the power or ground rails.
If the clock doesnt have an independent power supply, it can cause excess jitter from the clock.
The PSU can cause problems if its voltage does not remain extremely stable with load.
Shielding around the PSU and other components can reduce noise transferring electromagnetically.

Audio chips on cards that do the upsampling/downsampling have their own means of adapting to an ad-hoc power supply.
Power is one thing, signal sampling rate is completely different and could even prove invisible to the power section.
Click to expand...
Bad effects are cumulative, one doesnt wipe out the other.

Upsampling, say, 96 to 192 is not a fruitless endevaour because the DSP on the card may have some dithering algorythms running that detect rapid transients and pad them with corrective signals. I have personally been satisfied with 48kHz and flac.
Click to expand...
Yes, it is possible to provide better interpolation than some DACs do internally.

Also 192 is required for some multichannel operations, like surround sound, home recording, better working equalizers (imagine an instant of a song divided into 192 thousand beeps during a second. Plenty of opportunity to add psychoacoustic equalizers and such.
Click to expand...
Nope :)
If you are dividing up the 192K samples between all your surround speakers, you might have a point.
But each speaker can be driven at 192K with current hardware.

Furthermore, this ultrasonics thing. A 192kHz sampling rate is not the same as outputting a whopping 192kHz audio signal into the pipeline.
One is the 'firing rate', the other is the 'caliber'.
Click to expand...
Sampling rate is twice the maximum frequency sampled as per Nyquist Theorem.

You should probably be concerned with low frequency hum more, like from ground-loops. Because that has a tendency to induce itself all over the place, 'microphone' stray signals into wires and 'steal' power from the mosfets because a low level signal requires more power to bring to audible levels.
Click to expand...
If you have those problems yes.

If you want to fool around you might want to simply listen to other gear.
Click to expand...
Absolutely!

After like 10 years of listening to a FM801, SB Live!, Audigy 2LS, Audigy 2ZS, X-fi USB, a PCM 2704 based module, a prodigy 7.1 hifi, and dozens of source-amp analog signal cables I will tell you this: learn to listen first.
Wind outside can make you adore a song on a particular evening while mains power disturbances will leave you dumbfounded and confused another day.
Click to expand...
Not experiences I have had myself but I wont knock what works for you.

I picked apart my prodigy 7.1 hifi card like a hawk, testing easily around 7 opamps. Even after all those years I caught myself in an expectation bias, exchanging a component that had no business with the signal path.
Get just.. whatever. Try a low volume on source and high volume on amp combination and vice versa. Try some equalizers. Make sure all your gear shares a common receptacle.
I know you're using digital, but keep in mind - everything plays music. From a capacitor in your power supply to a loose door on a cabinet somewhere. Learn to listen.
Click to expand...
If your gear has a fair bit of variance in its operation, some playing around of this nature may benefit.
I changed the Opamps in my Minimax DAC but ended up going back to the original. I could probably find a more suitable opamp but its not that cheap a project.
However, there are certainly better opamps than those on sound cards.
Its usually better to use as high a driving signal (voltage) as you can without exceeding the specs of the receiving device / causing distortion.
If your amp can take 2V pk-pk on RCA, you may not be getting the best from it. Many sources max output is less.

In your case I wouldn't plug a pcie card but maybe try a USB solution. More fun to tinker with, and it's possible to power it from batteries instead of relying on dirty USB 5v.
Click to expand...
I follow a similar thought pattern.
 
Last edited:
You are already running "offboard sound", your external dac / amp is connected through optical so the onboard sound is doing basically nothing. You could use USB if your dac has async USB, which is better, or you could use coaxial from a sound card, which might be better than optical as it does not have to convert the signal / is a dedicated audio connection.

Chances are that they will all work fine and you will not notice much / any difference. Optical is immune to noise from PC which is the only good thing about it for audio. I use async USB and cannot say it is that much difference from the coaxial connection I had before, but the optical connection seems to sound worse than coaxial or USB on my Xonar STX > external dac, the coaxial on my xonar STX was picking up noise from my graphics card so I switched to async USB.
 
Nenu said:
One of the things that makes one DAC or CD player sound different is the timing of the clock pulse, whether it be due to the clock itself, what the DAC does with it, what other circuits do with it or interference.
A common term for this type of interference is jitter.
I was pointing out that not all DACs are of the same quality and have varying abilities.
One that is a bit lax at 96KHz in either the timing of the start of the clock pulse, or if its rise time is non uniform such that it doesnt trigger the clock at uniform time intervals, it could provide a sonic improvement if the faster clock improves on either of those.
The faster clock is more likely to have some tolerances tightened, that is all.
Click to expand...

Thanks for clarifying. In that case, it would seem the benefit of 96 or 192kHz is not inherent in the file itself, but rather in the clock used by the DAC. Thus, a better clock would benefit 44.1kHz files just as it would 192kHz files.

The downside of records is the reduced dynamic range sadly.
Click to expand...

You're correct that vinyl allows reduced dynamic range when compared to digital. On the other hand, most music doesn't take advantage of the increased dynamic range digital offers. Quite the opposite - for a long time recordings have been trending toward reduced dynamic range - dynamic range compression - to give the perception of a louder recording. This technique serves a purpose, but when overdone makes music fatiguing to listen to. Vinyl's upper limit for dynamic range is lower than digital, but its lower limit is higher - it doesn't allow for the same degree of dynamic range compression as digital. Therefore some recordings that are heavily compressed on the digital release sound much better on vinyl. Take Stadium Arcadium from RHCP. The digital release is horribly compressed. The vinyl release was mastered by a different engineer - Steve Hoffman - and has better dynamic range. It sounds vastly superior.

Vinyl: http://dr.loudness-war.info/album/view/93198
Digital: http://dr.loudness-war.info/album/view/86171
 
Epos7 said:
Thanks for clarifying. In that case, it would seem the benefit of 96 or 192kHz is not inherent in the file itself, but rather in the clock used by the DAC. Thus, a better clock would benefit 44.1kHz files just as it would 192kHz files.
Click to expand...
Yes, tighter clock timing will benefit all digital audio.
But upsampling a 96K file to 192K will not increase detail unless the clock has improved as a consequence. ie there was already a noticeable problem at 96K.
A true 192K file will contain more information than a 96K file.
The right equipment will expose it.

You're correct that vinyl allows reduced dynamic range when compared to digital. On the other hand, most music doesn't take advantage of the increased dynamic range digital offers. Quite the opposite - for a long time recordings have been trending toward reduced dynamic range - dynamic range compression - to give the perception of a louder recording. This technique serves a purpose, but when overdone makes music fatiguing to listen to. Vinyl's upper limit for dynamic range is lower than digital, but its lower limit is higher - it doesn't allow for the same degree of dynamic range compression as digital. Therefore some recordings that are heavily compressed on the digital release sound much better on vinyl. Take Stadium Arcadium from RHCP. The digital release is horribly compressed. The vinyl release was mastered by a different engineer - Steve Hoffman - and has better dynamic range. It sounds vastly superior.

Vinyl: http://dr.loudness-war.info/album/view/93198
Digital: http://dr.loudness-war.info/album/view/86171
Click to expand...

I agree, they made a right bloody mess.
I'm not getting into the loudness wars, I find what I like and listen to it.
I will do some comparisons when I get my deck, I didnt get a lot of time with the Linn.
 
Nenu said:
I'll answer off the top of my head, hope it helps :)
Click to expand...
Oh it does, thanks.


It can cause noise to be transferred from one part of the circuit to another along the power or ground rails.
Click to expand...
That I suspected, but what's so special about higher sampling rates? more susceptible?

If the clock doesnt have an independent power supply, it can cause excess jitter from the clock.
Click to expand...
Looking at the PCM2704 datasheet, plenty of room around the oscillator but it's unpowered.

The PSU can cause problems if its voltage does not remain extremely stable with load.
Shielding around the PSU and other components can reduce noise transferring electromagnetically.
Click to expand...
Call me excessive but I really wanted a full magnetic USB separator at one point with its own power input options but it was like 100$? I want battery power :(
As for the shrouds, I use them and love them. But I don't gloat because people go bananas when they see it.

Bad effects are cumulative, one doesnt wipe out the other.
Click to expand...
I did not realize one would affect the other. I mean I would probably think about it and then shrug it off as difficult to implement. And I'm surprised this scenario is not countered by design at IC level.

Sampling rate is twice the maximum frequency sampled as per Nyquist Theorem.
Click to expand...
Flashback from 'college' !

If your gear has a fair bit of variance in its operation, some playing around of this nature may benefit.
I changed the Opamps in my Minimax DAC but ended up going back to the original. I could probably find a more suitable opamp but its not that cheap a project.
However, there are certainly better opamps than those on sound cards.
Its usually better to use as high a driving signal (voltage) as you can without exceeding the specs of the receiving device / causing distortion.
If your amp can take 2V pk-pk on RCA, you may not be getting the best from it. Many sources max output is less.
Click to expand...

Yes I was quite unlucky in that department. The Prodigy 7.1 hifi was bearly capable of 1V. The current PCM2704 toy is only 0.5 or so. I understand the receiver has a high input impedance and that's one benefit for higher driving voltaage and the other would be noise rejection, right?

Thanks for your time sir!
 
michalrz said:
That I suspected, but what's so special about higher sampling rates? more susceptible?
Click to expand...
The ability to encode more information.
A downside is a faster clock and more detail so more noise to be dealt with.
ie there is more noise and the noise floor needs to be lower to allow the higher detail to be exposed.

Looking at the PCM2704 datasheet, plenty of room around the oscillator but it's unpowered.
Click to expand...
I'm not intending building my own DAC :p
I'm not sure what you are asking here.
Its not necessarily a bad thing to use the same power supply for the clock and DAC etc, everything has to be made to a budget.
The budget may be better spent isolating noise rather than on another high quality PSU.

Call me excessive but I really wanted a full magnetic USB separator at one point with its own power input options but it was like 100$? I want battery power :(
As for the shrouds, I use them and love them. But I don't gloat because people go bananas when they see it.
Click to expand...
I would avoid magnetic devices, they are prone to interference and causing interference.
Look for optical USB but take care because many of them still pick up noise and transmit it with the signal optically.
If you come across the JPlay forums, some hardcore enthusiasts built some amazing setups with Optical USB. They went as far as using 2 PCs, one to play/decode transmit the audio data and the other to only convert raw data into analogue.
JPlay itself is designed to reduce processing noise within the PC while playing audio so induced jitter is reduced.
Last time I tried it, there were 2 general ways of using it, with an audio player plugin or as a standalone but rudimentary player (higher quality). There was a difference.
It damnwell does have a positive effect on detail but is a bit of a pita to use and isnt cheap. There is a demo to try.
I found the difference it made wasnt quite as high once I built my own USB cable from decent copper wire and this made me decide to not buy it.
ie the new USB cable seemed to give some of the advantage that JPlay was fixing, so JPlay had less problem to fix.
http://jplay.eu/
It may be even better now, its over a year since I last tried it.


The reason I made my own USB cable was because I tried a few lying around and found they did sound different.
I didnt expect this as my DACs isolate the USB input and the signal is clocked into an internal memory buffer where a new clock signal is generated.
So the USB clock isnt used going forward but it seems that corruption could still occur somewhere with lower quality USB leads.
Bits are not necessarily bits hey. Or perhaps it is still entirely a jitter related issue?

Its a mute issue for me now because I play music directly from the Oppo most of the time.
It can take USB pens/drives or will read from network shares.
I only use USB from PC when I am being lazy or doing stereo gaming (and cant be asked to power everything else up/configure PC) Its quick and easy to use USB out.
I'm not bothered about improving the quality for this.

edit:
if you are using an external DAC, you can still use a battery to power it and is a good idea if it takes power from USB or if its power supply is crap.
Make a 2 wire USB lead with just the signal wires attached, no power connections (this is what I did).
If this allows it to function, it should give some improvement.
If it doesnt function, you need to provide 5V on its USB in as well (from battery)
You will likely need to twist the data cables otherwise they will pick up too much noise and fail to work.
I did a fairly loose twist which worked ok on the Oppo but would occasionally crash the Minimax DAC.
I tightened them a bit and all was well.

I did not realize one would affect the other. I mean I would probably think about it and then shrug it off as difficult to implement. And I'm surprised this scenario is not countered by design at IC level.
Click to expand...
Many issues are fixed at IC level, it depends what you buy.
The 9018 Sabre32 ref DACs jitter elimination circuits do a very good job.
But it can still be fed a crap clock signal so great care has to be taken to get the best from it.
(There are many ways of getting jitter)

Yes I was quite unlucky in that department. The Prodigy 7.1 hifi was bearly capable of 1V. The current PCM2704 toy is only 0.5 or so. I understand the receiver has a high input impedance and that's one benefit for higher driving voltaage and the other would be noise rejection, right?
Click to expand...
Its good you are aware of the low voltage out. Verify what your amps input can be driven with.
Yep, inputs have to be high impedance to prevent loading the very low power output of the source.
The signal voltage it can take is down to the components used and/or design, not its impedance.

Yes higher signal will improve the s/n ratio and also allow higher gain for quieter music.
 
Last edited:
The only practical changes that are possible are Optical, USB or a separate sound card. I can tell the difference to some extent, but at the end of the day I'm looking for a good balance of price vs performance.

I'm plenty satisfied with the Magni 2/Modi 2 Uber combo and the only thing I can really adjust as mentioned above, is the Source.

I mainly listen to music off Spotify but do have a sizable collection of FLAC as well. So while sound quality matters to me, I know there are limitations due to my source material. I will do an A-B test of USB vs Optical setup though since now I'm curious.

Gaming wise, I wondered how much of a benefit an offboard card would give me if i was just feeding it into my Magni/Modi.
 
AFAIK TosLink can pass through DSP from audio hardware. I remember being able to send gaming-related DSP from some RoG drivers over optical.

Given this nuance it wouldn't surprise me if optical wasn't as transparent as believed.
 
You can enable dsp and effects that are sent over spdif, these can be disabled but you are still at the mercy of the Windows mixer.
To prevent any changes to the original audio, use a direct streaming protocol in your music player like Kernel Streaming or WASAPI exclusive mode.
This bypasses the Windows mixer and volume control (for most DACs) so the audio player needs to be the volume control if your DAC doesnt have one.
 
I.e. in order for a direct comparison between SPDIF and USB to be valid, you should use an exclusive protocol rather than WMixer/DirectSound.
 
Yes.
You should be playing audio in this way anyway otherwise it wastes the potential of your sound system.
When I play from Foobar2000 I use a WASAPI plugin to get WASAPI exclusive mode.
There is a distinct loss of detail and imaging when I use Windows DirectSound.
 
I'm wondering if the experiences cited earlier in this thread were really due to superior USB handling or due to poisoning of the SPDIF output by way of non-exclusive output modes.

I have a reasonably upscale system (HD 800s not pictured) but CBA to care about these nuances outside of rhetorical discussions. For example I usually stick to DirectSound via MusicBee because I don't like exclusive access to my outboard audio when I need to play something in a different program (e.g. videos in Chrome, or just playing BGM while in-game).
 
my FLAC Library is configured via Foobar2000/ASIO already and that's what I use for my A-B testing so I'm clean on that front. Spotify is pretty much at the mercy of DirectSound from my understanding.

The TOSLink cable not being as transparent as USB does come as a surprise but I should have expected as much. I'll dig a bit more and check it out.

On top of that...I multi-task...A LOT. So the mixer is relatively important, but worst case I can use an external mixer and run my PC through another output into that if I want to bypass the Windows Mixer for Music.
 
If you know of anyone with other cables, its worth trying those as well.
I have an optical cable that sounds horrible and one that sounds good.
I mentioned earlier that USB cables can sound different, the one I made sounds the best on my system.
 
Nenu said:
A downside is a faster clock and more detail so more noise to be dealt with.
ie there is more noise and the noise floor needs to be lower to allow the higher detail to be exposed.
Click to expand...

The Prodigy card I had and keep using as an example had a 192kHz rate. With such a wide spectrum, have you yourself noticed sound degradation solely due to a badly implemented attempt at a high sampling rate?

Nenu said:
I'm not intending building my own DAC
I'm not sure what you are asking here.
Its not necessarily a bad thing to use the same power supply for the clock and DAC etc, everything has to be made to a budget.
Click to expand...

Well, I am currently using this:
And you've intrigued me as to USB cabling and clocking issues.
The device I have I got really cheap and now I have the itch to tweak it.
I was asking about the oscillator - in this case. You can see it to the left of the DAC chip. The caps are all squares - are they MKP? Are they worth being replaced by quality electrolytes such as OSCONs and such?
But the real kicker is now for me - could I remove that oscillator, grounded on both sides through the SMD caps between it and the IC.
I see a lot of the high quality stuff has a simple 5V supply and use oscillators that are temperature independent and have other advantages. Does that relate in any way to your worries about clock jitter? Is it as important as the cable you figure?
If it's nonsense just let me know :)
I only use the analogue paths.


nenu said:
(cables) tried a few lying around and found they did sound different.
I didnt expect this as my DACs isolate the USB input and the signal is clocked into an internal memory buffer where a new clock signal is generated.
So the USB clock isnt used going forward but it seems that corruption could still occur somewhere with lower quality USB leads.
Bits are not necessarily bits hey. Or perhaps it is still entirely a jitter related issue?
Click to expand...

Looking at the PCM2704's datasheet again I only see reference to that onboard oscillator. The only thing it needs from USB is DATA+ and DATA-. It has a pin to select powering mode from USB BUS to self-powered. So there's room to play.

I don't know if timing issues in the USB RX/TX would manifest at all. I remember some experiment I've read about where people were able to A/B test sources with slightly modified timings (one channel lagging behind the other by like microseconds). And random people did perceive the 'right' pace as the nicer sounding.

Edit: Where are my manners - there's no opamp between the card and receiver. (Yamaha AX-550 RS) and there are coupling caps. Thinking of frankensteining a simple low gain pre-amp after the DAC and decoupling DC at the very end.

Crap you made relapse :D
 
michalrz said:
The Prodigy card I had and keep using as an example had a 192kHz rate. With such a wide spectrum, have you yourself noticed sound degradation solely due to a badly implemented attempt at a high sampling rate?
Click to expand...
Yes, most PC soundcards fall into this category.
Its not that 192K sound worse than 96K, rather that it doesnt bring enough benefit.
To say badly implemented is probably a bit of a stretch because the DAC and components used have to fall within budget. Its pretty good the option is even there as it is sure to improve over time.

Well, I am currently using this:
And you've intrigued me as to USB cabling and clocking issues.
The device I have I got really cheap and now I have the itch to tweak it.
I was asking about the oscillator - in this case. You can see it to the left of the DAC chip. The caps are all squares - are they MKP? Are they worth being replaced by quality electrolytes such as OSCONs and such?
But the real kicker is now for me - could I remove that oscillator, grounded on both sides through the SMD caps between it and the IC.
I see a lot of the high quality stuff has a simple 5V supply and use oscillators that are temperature independent and have other advantages. Does that relate in any way to your worries about clock jitter? Is it as important as the cable you figure?
If it's nonsense just let me know :)
I only use the analogue paths.
Click to expand...
I am not the best person for advice on upgrade part selection, I have been out of the industry for quite a few years and havent kept up that closely.
I recommend start from a good base if spending good money on components. The benefit will be greater and you wont be re-using high performance caps in other designs as much, reducing heat damage.
Get a DAC with a DAC chip that has been demonstrated to give exceptional results when partnered with quality components in a well designed circuit.
Look out for Sabre products like this
https://hifiduino.wordpress.com/2013/01/03/99-es9018-diy-dac-board/

Looking at the PCM2704's datasheet again I only see reference to that onboard oscillator. The only thing it needs from USB is DATA+ and DATA-. It has a pin to select powering mode from USB BUS to self-powered. So there's room to play.
Click to expand...
You can get away with a 2 wire USB cable, worth a try.

I don't know if timing issues in the USB RX/TX would manifest at all. I remember some experiment I've read about where people were able to A/B test sources with slightly modified timings (one channel lagging behind the other by like microseconds). And random people did perceive the 'right' pace as the nicer sounding.
Click to expand...
For sure they occur.
What you dont know is where the major sources of jitter are occurring and whether there are any other over-riding issues that would mask improvements.
The DAC chip itself will limit how good the sound can be so its not worth spending much unless you have a decent chip to start with.

Edit: Where are my manners - there's no opamp between the card and receiver. (Yamaha AX-550 RS) and there are coupling caps. Thinking of frankensteining a simple low gain pre-amp after the DAC and decoupling DC at the very end.

Crap you made relapse :D
Click to expand...
Good luck :p
 
Epos7 said:
Interesting that two digital signals would sound markedly different. I wonder if that actually has to do with the design of the DAC.



Again, interesting that the 192K sounds better. Is it a different recording? My understanding is that 192/24 is more of a marketing play than an actual improvement.

http://people.xiph.org/~xiphmont/demo/neil-young.html
Click to expand...

Yup my HDMI to amp was way much better than the Optical toslink
 
Nenu said:
If you know of anyone with other cables, its worth trying those as well.
I have an optical cable that sounds horrible and one that sounds good.
I mentioned earlier that USB cables can sound different, the one I made sounds the best on my system.
Click to expand...

Ah I see
 
You must log in or register to reply here.
Back
Top