IPB

Welcome Guest ( Log In | Register )

2 Pages V   1 2 >  
Reply to this topicStart new topic
Is anyone here using onboard audio?, Computer audio (Analog out)
uart
post Jun 5 2013, 19:34
Post #1





Group: Members
Posts: 810
Joined: 23-November 04
Member No.: 18295



Onboard audio (built into PC motherboard) seems to be tons better these days than it was a decade or so ago. Analog output seem pretty decent on recent systems I've owned, with the one one exception of driving headphones (though I've got a headphones amplifier so that doesn't bother me).

Currently I've got a gigabyte motherboard that uses an ALC889 codec that's supposed to have an output SNR of 108dB. Honestly I don't know how close it actually comes to that with this particular board's implementation, but it does sound very good to me and loopback tests that I've done look fine as well.

Anyway, it's time to upgrade my media computer and the motherboard I'm looking at using has an ALC887 codec which only claims to have an output SNR of 97dB. I kind of hate to "downgrade" here, but I'm wondering if I'd actually hear any difference or not. Somehow I suspect that the real make or break factors are going to be in the actual implementation (layout etc) rather than the theoretical codec specs.

I was looking at an alternative motherboard with a VIA2021 codec that claims an impressive 110dB SNR. Has anyone got systems with either of these two codecs or got any idea of their relative sound quality?

This post has been edited by uart: Jun 5 2013, 19:37
Go to the top of the page
+Quote Post
saratoga
post Jun 5 2013, 19:45
Post #2





Group: Members
Posts: 5039
Joined: 2-September 02
Member No.: 3264



The claimed SNR numbers are meaningless, so I would not even look them up, let alone let them weigh into your decision making.
Go to the top of the page
+Quote Post
uart
post Jun 5 2013, 19:50
Post #3





Group: Members
Posts: 810
Joined: 23-November 04
Member No.: 18295



QUOTE (saratoga @ Jun 5 2013, 11:45) *
The claimed SNR numbers are meaningless, so I would not even look them up, let alone let them weigh into your decision making.

What then do you think are the most important considerations? Is the layout? Or the quality of capacitors etc?
Go to the top of the page
+Quote Post
skamp
post Jun 5 2013, 19:50
Post #4





Group: Developer
Posts: 1450
Joined: 4-May 04
From: France
Member No.: 13875



My ALC663 chip is very decent, in terms of dynamic range, noise levels and distortion. Pretty much perfect 16 bit performance, it wouldn't make sense for me to ask for more. My laptop's headphone out has a big downside though, which is a 75Ω output impedance, which either distorts some headphones / IEMs, or produces output with not enough volume in cases where I use attenuation, for one reason or another (Replaygain, EQ). But with an external headphone amp with a low output impedance, it would be just fine.

So yeah, if you have a good headphone amp, you can probably use your onboard audio and get fine audio. Run an RMAA test if you're really worried.


--------------------
See my profile for measurements, tools and recommendations.
Go to the top of the page
+Quote Post
saratoga
post Jun 5 2013, 19:52
Post #5





Group: Members
Posts: 5039
Joined: 2-September 02
Member No.: 3264



QUOTE (uart @ Jun 5 2013, 14:50) *
QUOTE (saratoga @ Jun 5 2013, 11:45) *
The claimed SNR numbers are meaningless, so I would not even look them up, let alone let them weigh into your decision making.

What then do you think are the most important considerations? Is the layout? Or the quality of capacitors etc?


Measured performance is the most important, and most likely, only consideration.
Go to the top of the page
+Quote Post
uart
post Jun 5 2013, 20:00
Post #6





Group: Members
Posts: 810
Joined: 23-November 04
Member No.: 18295



QUOTE (saratoga @ Jun 5 2013, 11:52) *
QUOTE (uart @ Jun 5 2013, 14:50) *
QUOTE (saratoga @ Jun 5 2013, 11:45) *
The claimed SNR numbers are meaningless, so I would not even look them up, let alone let them weigh into your decision making.

What then do you think are the most important considerations? Is the layout? Or the quality of capacitors etc?


Measured performance is the most important, and most likely, only consideration.

Haha, I know that. cool.gif I'm interested in what factors effect the measured performance. Anyway I know what you mean, I'll probably just have to try it and see. If I don't like it then I can always drop in a soundcard.

This post has been edited by uart: Jun 5 2013, 20:01
Go to the top of the page
+Quote Post
uart
post Jun 5 2013, 20:08
Post #7





Group: Members
Posts: 810
Joined: 23-November 04
Member No.: 18295



QUOTE (skamp @ Jun 5 2013, 11:50) *
My ALC663 chip is very decent, in terms of dynamic range, noise levels and distortion. Pretty much perfect 16 bit performance, it wouldn't make sense for me to ask for more.


Thanks for the info skamp. I believe that the ALC663 has a listed SNR of 95dB, so the ALC887 should be as good or better. Yeah it's pretty impressive how much onboard audio has improved these days, those RMAA results are very good. The RMAA results I measured with my current motherboard are quite similar to that as well. smile.gif

I can remember some horrendous onboard implementations back a decade or so ago. I had one that was so bad that you could literally hear the mouse cursor move and windows scrolling (somehow the changing CPU load actually converted these events in quite audible noise in the onboard audio). That was truly an awful implementation sad.gif

This post has been edited by uart: Jun 5 2013, 20:18
Go to the top of the page
+Quote Post
uart
post Jun 5 2013, 20:29
Post #8





Group: Members
Posts: 810
Joined: 23-November 04
Member No.: 18295



QUOTE (skamp @ Jun 5 2013, 11:50) *
My laptop's headphone out has a big downside though, which is a 75Ω output impedance, which either distorts some headphones / IEMs, or produces output with not enough volume in cases where I use attenuation, for one reason or another (Replaygain, EQ). But with an external headphone amp with a low output impedance, it would be just fine.


Yes headphones seem to be the main bug bear for onboard audio. The ALC889 codec on my current system claims to have an inbuilt headphones amp with a 2 ohm output impedance! Specifically they claim a 200 ohm output impedance for "lineout" mode and a 2 ohm impedance for "headphones" mode. Unfortunately, in the actual implementation "they" (the motherboard manufacturer) have used way undersized coupling capacitors to actually support such a low impedance.

The irony is that the headphone amp part of the codec actually works well, the driver/ui allows you to select "lineout" or "headphone" and it really works and gives a huge boost in volume if using low-z phones. BUT, at the same time it completely screws up the low frequency performance because the coupling caps can't handle the low impedance.

This post has been edited by uart: Jun 5 2013, 20:31
Go to the top of the page
+Quote Post
saratoga
post Jun 5 2013, 20:48
Post #9





Group: Members
Posts: 5039
Joined: 2-September 02
Member No.: 3264



QUOTE (uart @ Jun 5 2013, 15:00) *
Haha, I know that. cool.gif I'm interested in what factors effect the measured performance. Anyway I know what you mean, I'll probably just have to try it and see. If I don't like it then I can always drop in a soundcard.


Performance basically comes down to the analog design of the system.

QUOTE (uart @ Jun 5 2013, 15:08) *
QUOTE (skamp @ Jun 5 2013, 11:50) *
My ALC663 chip is very decent, in terms of dynamic range, noise levels and distortion. Pretty much perfect 16 bit performance, it wouldn't make sense for me to ask for more.


Thanks for the info skamp. I believe that the ALC663 has a listed SNR of 95dB, so the ALC887 should be as good or better.


I don't agree with that. Comparing numbers that don't mean anything does not somehow make them meaningful. Garbage in, garbage out.
Go to the top of the page
+Quote Post
carpman
post Jun 5 2013, 22:16
Post #10





Group: Developer
Posts: 1319
Joined: 27-June 07
Member No.: 44789



QUOTE (uart @ Jun 5 2013, 19:08) *
I can remember some horrendous onboard implementations back a decade or so ago. I had one that was so bad that you could literally hear the mouse cursor move and windows scrolling (somehow the changing CPU load actually converted these events in quite audible noise in the onboard audio). That was truly an awful implementation sad.gif

I still can hear these things:
Win 7 x64 > Motherboard (Gigabyte GA-970A-D3) > On-board Realtek HD Audio > Denon Amp > Headphones (with volume pretty high and nothing playing ---- if there's any audio playing I can't make out any of the system noises you mentioned).

Versus, via my M-Audio 2496 (no system noise at all).

When you talked about hearing such "audible noise", did you mean even when playing music you could still hear it?

I'm wondering now, if there's something wrong with my on-board audio. I was a bit surprised when I heard the difference between on-board Realtek and my m-audio card (only via headphones).

Don't wish to hijack thread so perhaps this is a separate topic?

C.

This post has been edited by carpman: Jun 5 2013, 22:17


--------------------
TAK -p4m :: LossyWAV -q 6 | TAK :: Lame 3.98 -V 2
Go to the top of the page
+Quote Post
saratoga
post Jun 5 2013, 22:30
Post #11





Group: Members
Posts: 5039
Joined: 2-September 02
Member No.: 3264



Headphones are a different story. Many on board and even dedicated sound cards are not great with all headphones.
Go to the top of the page
+Quote Post
ktf
post Jun 5 2013, 23:38
Post #12





Group: Members
Posts: 391
Joined: 22-March 09
From: The Netherlands
Member No.: 68263



To answer the question, yes, there are. I do. My laptop has according to ALSA a AD1984A, the specsheet says that one has an output impedance of less than 1 ohm. I use it with a Sennheiser HD595, works great.


--------------------
Music: sounds arranged such that they construct feelings.
Go to the top of the page
+Quote Post
carpman
post Jun 5 2013, 23:39
Post #13





Group: Developer
Posts: 1319
Joined: 27-June 07
Member No.: 44789



QUOTE (saratoga @ Jun 5 2013, 21:30) *
Headphones are a different story. Many on board and even dedicated sound cards are not great with all headphones.

I'm a bit dumb when it comes to this, but:
Motherboard has: 3 x audio jacks (Line In/Line Out/Microphone)

I'm going from Line Out (mini jack) to RCA > Denon Int. Amp > Headphones
Why should that cause an issue (the amp is driving the headphones)?

Forgive my ignorance.

C.

EDIT: To make it clear I'm replying to saratoga

This post has been edited by carpman: Jun 5 2013, 23:41


--------------------
TAK -p4m :: LossyWAV -q 6 | TAK :: Lame 3.98 -V 2
Go to the top of the page
+Quote Post
saratoga
post Jun 6 2013, 00:22
Post #14





Group: Members
Posts: 5039
Joined: 2-September 02
Member No.: 3264



QUOTE (carpman @ Jun 5 2013, 18:39) *
QUOTE (saratoga @ Jun 5 2013, 21:30) *
Headphones are a different story. Many on board and even dedicated sound cards are not great with all headphones.

I'm a bit dumb when it comes to this, but:
Motherboard has: 3 x audio jacks (Line In/Line Out/Microphone)

I'm going from Line Out (mini jack) to RCA > Denon Int. Amp > Headphones
Why should that cause an issue (the amp is driving the headphones)?


I think I misunderstood. If you're using the amp, than really the only thing you care all that much about is the SNR. In your case it sounds like its not good enough if you hear some hiss.
Go to the top of the page
+Quote Post
ChronoSphere
post Jun 6 2013, 12:55
Post #15





Group: Members
Posts: 518
Joined: 11-March 07
Member No.: 41384



I have the ALC889 (also a gigabyte board) and it was pretty fine noise-wise, before I upgraded my CPU - the new one uses a higher voltage and this slight bump introduced interference noise. Which is pretty annoying when not listening to anything. So I'd say it depends on your board and how well the onboard card is shielded.
Go to the top of the page
+Quote Post
yourlord
post Jun 6 2013, 23:08
Post #16





Group: Members
Posts: 225
Joined: 1-March 11
Member No.: 88621



Yep, my onboard audio feeds into my receiver via S/PDIF..
Go to the top of the page
+Quote Post
Zarggg
post Jun 8 2013, 02:12
Post #17





Group: Members
Posts: 560
Joined: 18-January 04
From: bethlehem.pa.us
Member No.: 11318



Ever since my Audigy 2 crapped out, I've started using the onboard output on my motherboard, which uses VT1708S. I haven't noticed a difference, positive or negative.

This post has been edited by Zarggg: Jun 8 2013, 02:12
Go to the top of the page
+Quote Post
stephan_g
post Jun 9 2013, 17:27
Post #18





Group: Members
Posts: 214
Joined: 6-June 10
From: Bavaria
Member No.: 81240



ALC262 on a Fujitsu-Siemens D2587 (Celsius W360) here. The chip does have its quirks (like a digital filter that scales with fs, highpass unfortunately included, and about 0.1-ish dB of passband ripple), but the implementation appears to be very good. The HD580s that I'm using on the computer aren't overly fussed by a 75 ohm output impedance either. Sounds just fine to me, with no detectable noise even with my most sensitive 'phones, and I can use the front panel connectors (which unfortunately not many aftermarket sound cards support) rather than having to crawl behind the computer when switching between headphones and speakers. The old Audiotrak Prodigy 7.1 drivers could output the same signal on all output pairs, that was neat. I think the card needs a recapping job though.

When going back from BTX to ATX, FSC unfortunately switched from Realtek to Conexant chips on their boards (they still have Realteks in their notebooks, my work Lifebook S761 has an ALC269). I don't know how good those are hardware wise (no datasheet for the general public, boo), but the drivers have even less features than the standard Win7 x64 HD Audio drivers (in case of my PC @work I can actually make use of the bass extension, which makes the crappy old powered speakers there somewhat more acceptable). What a total joke.

This post has been edited by stephan_g: Jun 9 2013, 17:33
Go to the top of the page
+Quote Post
Nessuno
post Jun 10 2013, 08:11
Post #19





Group: Members
Posts: 423
Joined: 16-December 10
From: Palermo
Member No.: 86562



QUOTE (stephan_g @ Jun 9 2013, 18:27) *
I can use the front panel connectors (which unfortunately not many aftermarket sound cards support) rather than having to crawl behind the computer when switching between headphones and speakers.

Caveat: I've seen (and heard) front panel connectors served by a pair of unshielded and not twisted cables, a perfect antenna to pick up noise and static from every EMI emitter it encounters along its way inside the case so that a good sound card could become unlistenable in a few cm! wink.gif


--------------------
... I live by long distance.
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Jun 10 2013, 13:13
Post #20





Group: Members
Posts: 3929
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Nessuno @ Jun 10 2013, 03:11) *
QUOTE (stephan_g @ Jun 9 2013, 18:27) *
I can use the front panel connectors (which unfortunately not many aftermarket sound cards support) rather than having to crawl behind the computer when switching between headphones and speakers.

Caveat: I've seen (and heard) front panel connectors served by a pair of unshielded and not twisted cables, a perfect antenna to pick up noise and static from every EMI emitter it encounters along its way inside the case so that a good sound card could become unlistenable in a few cm! wink.gif



The above analysis ignores at least one or two of the favorable attributes of this transmission line when it comes to resistance to noise pick up, which are low impedance and high signal voltage.

Low impedance high signal voltage lines have considerable resistance to noise pick up. The wiring to the front panel output, and perhaps to a lesser degree the wiring to the front panel line input share this favorable characteristic.

Even the mic input is generally resistant to noise pickup.

Examples of lines that are far more susceptible to noise pick up would be professional mic cables and LP magnetic cartridge wiring.
Go to the top of the page
+Quote Post
stephan_g
post Jun 10 2013, 14:37
Post #21





Group: Members
Posts: 214
Joined: 6-June 10
From: Bavaria
Member No.: 81240



When people have trouble with the front outputs, it generally means the case designer has messed up. Some cases tie all the grounds together at the front panel (USB, audio, everything), which gives the best ground loop imaginable. Audio ground must be separate from e.g. USB ground or case at this point. I find it interesting that HD Audio apparently gets along with a shared ground for the output and mic input. Would be interesting to measure the xtalk.

IMO pro mic cables should be an example for something that is not particularly susceptible to noise pickup - when connected to an amplifier with high CMRR, their rejection of noise can be excellent. They can effectively be much longer than any kind of unbalanced mic connection would ever dare to be. However, great care must be taken when mixing balanced and unbalanced inputs/outputs and cables - otherwise performance can be truly atrocious. If you need to go from balanced to unbalanced or vice versa, the cable must always be of the unbalanced variety.
For example, a typical dynamic mic comes with balanced wiring. It would be expected to pick up all kinds of crap when connected to a typical unbalanced consumer soundcard input (BTDT), while being quiet as a mouse on a balanced input. Rewire it with some coax (unbalanced), and things would be reversed.

BTW, the problematic part of a phono system is in the tonearm, where cabling is pretty much balanced in topology. That's why the tonearm itself is typically grounded to act as a shield, much like the extra grounding on balanced mic cable. Some cartridges are notorious for creating hum, btw - they apparently tie the returns of the two channels together, thereby creating a nice ground loop. Which brings us right back to where this post started...
Go to the top of the page
+Quote Post
Nessuno
post Jun 10 2013, 14:52
Post #22





Group: Members
Posts: 423
Joined: 16-December 10
From: Palermo
Member No.: 86562



QUOTE (Arnold B. Krueger @ Jun 10 2013, 14:13) *
QUOTE (Nessuno @ Jun 10 2013, 03:11) *

Caveat: I've seen (and heard) front panel connectors served by a pair of unshielded and not twisted cables, a perfect antenna to pick up noise and static from every EMI emitter it encounters along its way inside the case so that a good sound card could become unlistenable in a few cm! wink.gif

The above analysis ignores at least one or two of the favorable attributes of this transmission line when it comes to resistance to noise pick up, which are low impedance and high signal voltage.

It was an ex post analysis, after having heard a lot of noise from a front panel jack which disappeared just connecting the headphone to the rear panel jack. Opening the case, I noticed the above situation, that maybe improved a little disconnecting the cable from the motherboard and twisting it. The soundcard was integrated in the motherboard. Anyway the rear jack was practically dead silent, so I took an extension cord and kept using it.
Do you think the noise was already there at the board side of the cable? Or maybe, as stephan_g says, a ground loop? I remember the noise had a kind of variable but periodical pattern, not just static.

Edit: add reference to stephan_g post.

This post has been edited by Nessuno: Jun 10 2013, 15:10


--------------------
... I live by long distance.
Go to the top of the page
+Quote Post
Nessuno
post Jun 10 2013, 17:54
Post #23





Group: Members
Posts: 423
Joined: 16-December 10
From: Palermo
Member No.: 86562



ERRATA CORRIGE:

QUOTE (Nessuno @ Jun 10 2013, 09:11) *
a pair of unshielded and not twisted cables

It was a couple of channels, so three cables, unshielded and not twisted. Sorry for the very gross and misleading mistake! My fault, I wrote that wrong sentence without reading it again, till now.


--------------------
... I live by long distance.
Go to the top of the page
+Quote Post
stephan_g
post Jun 11 2013, 23:19
Post #24





Group: Members
Posts: 214
Joined: 6-June 10
From: Bavaria
Member No.: 81240



Impedances on a headphone out are so low that you should not generally be picking up any audible interference from inside the computer, even with modest unshielded ribbon cable. Within the audible range, electromagnetic field impedance is very high (electric field), so coupling remains low. You get a bit of capacitive coupling, certainly most of it still on the PCB. But what's a few pF if you've got a 75 ohm output impedance? Attenuation ought to be >=60 dB even at 20 kHz, and 14 dB more wouldn't surprise me.

(What might still happen, of course, is strong RF being caught by the wiring and AM demodulated by nonlinearities in the amplifier. Hence why smartphones are hard on dedicated headphone amps. Though usually it is the input that's the problem.)

Once a ground loop enters the picture, things change rather drastically. Our ground conductor now has to carry part of potentially quite substantial, rapidly fluctuating return currents. USB power, graphics card... stuff with return currents in the amps. Nothing you want to share a ground return with if you're audio circuitry. Even a fraction of an ohm still easily translates into the 10s to 100s of ÁV across the audible range. (It has to because people can obviously hear the result...)
A ground loop also makes a good antenna for the low-impedance (magnetic) fields from other ground loops with substantial circulating AC currents. You can apparently find those on graphics cards at times.
Go to the top of the page
+Quote Post
Arnold B. Kruege...
post Jun 12 2013, 12:02
Post #25





Group: Members
Posts: 3929
Joined: 29-October 08
From: USA, 48236
Member No.: 61311



QUOTE (Nessuno @ Jun 10 2013, 09:52) *
was already there at the board side of the cable? Or maybe, as stephan_g says, a ground loop? I remember the noise had a kind of variable but periodical pattern, not just static.


Yes. In fact I'm experiencing this kind of a thing with the computer that I am presently typing on.

I recently upgraded the system board on this PC. I've built several machine with the same system board, and the other 3 examples have had good noise performance. This one has variable noise, and even the back panel jack is noisy.

I breadboarded the PC outside of the present case, and it was quiet. I conclude that the case itself has a grounding problem.

This apparent grounding problem has persisted over a period of maybe 7 years, 3 different system boards and 2 different power supplies. I can often find the source of grounding problems by means of visual inspection, but so far this one has eluded me.
Go to the top of the page
+Quote Post

2 Pages V   1 2 >
Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 21st October 2014 - 00:01