Onboard vs Cheap USB SoundCard? Options

[Garf]

Well, it's a 3.5 jack in both sides. Is that a Stereo cable? :-)

Yes, the >80dB stereo separation you measured with the internal card clearly means that's fine.

http://i48.tinypic.com/24pb3n6.jpg

This is total failure. Basically signal in != signal out.

http://i46.tinypic.com/34zzwv8.jpg

This is closer, but the distortion values are extreme. Something is either clipping, or the soundcard is applying some effects to the signal or so.

[Tall-Guy]

OK. I found the problem. It seems like the problems only happen when I select Direct Sound. MME stop the issues I had. Any idea why? Do I need to configure the device differently due to that in windows?
Here is the results. Soooo, which one is better?

http://i46.tinypic.com/29wpee8.jpg

This post has been edited by Tall-Guy: May 25 2012, 20:22

[Tall-Guy]

OK. I did some homework and spent some time reading about the values and what they mean. I also bundled all the Results in one picture:



Few things to remember:
* The USB Sound Card is connected to the Onboard Line in
* The Onboard Card has a Speaker jack in the back, and a Headphones jack in the front (I tested both as you can see above)
* For some reason, after a test or two - the Usb Dongle gets really hot. I tried to run the test on a 'cool' dongle vs 'heated' and results are the same (still wonder why its getting so hot though)

It seems that surprisingly the Onboard scored better on all field (correct me if I'm wrong). Guess I'll just connect the Headphones to the Onboard card.
It also seems like the 'Speaker' jack is better than the 'Headphones' jacks (except for THD?). And questions time again :-)

1. Why is Headphones jack preforming worse? Isnt it the same card at the end of the day?
2. Is the difference in the Headphones and Speakers jacks are big enough that I should only use the Speakers Jack?
3. In case the above answer will be using the Speakers Jack, I will probably need to buy (or make) a small box that let me switch between the Speakers and the Headphones easily without pluging/unpluging. Will that extension hurt the quality much? is it suggested at all?
4. I think someone mentioned previously that the USB Soundcard is probably also acting as an amp. Does the onboard card also amping the sound? should I worry about that?
5. The rule of a thumb will be always make sure the volume is highest possible on Windows 7 Mixer and adjust the volume to my need using the Speakers/Headphones built-in knobs?

Thank you!

This post has been edited by Tall-Guy: May 25 2012, 23:19

[saratoga]

Note that unless you're using an external amp you need to test with your headphones connected to get a reasonable idea how good the devices are driving a real load.

[Tall-Guy]

Note that unless you're using an external amp you need to test with your headphones connected to get a reasonable idea how good the devices are driving a real load.

How is it possible to test it with headphones connected? I have no option to plug those as I am force to connect the card out into the line in.

[saratoga]

How is it possible to test it with headphones connected?

Wire cutters.

Well that or a splitter.

Take your pick.

[Tall-Guy]

So, here's the results when I add a Splitter to the out jack - one leg connected to the Line in and the other to the headphones.

No splitter:



Splitter + Headphones



Seems like the results over the Headphones output has improved with the splitter...

This post has been edited by Tall-Guy: May 26 2012, 00:23

[Arnold B. Kruege...]

So, here's the results when I add a Splitter to the out jack - one leg connected to the Line in and the other to the headphones.

No splitter:



Splitter + Headphones



Seems like the results over the Headphones output has improved with the splitter...

I see differences, but unless I'm missing something, nothing significant. Am I missing some measurement that changed enough to be audible?

[Tall-Guy]

Your speaking about the changes between the USB Card and the Onboard card? or the Test with and without the headphones?
I'm not sure to be honest. I knew the Onboard results are better, but I'm not sure if the gaps are indeed audible or not.

[Garf]

Seems like the results over the Headphones output has improved with the splitter...

It's essentially the same within measurement error. I see the "you need a headphone amp" meme a lot, but what I haven't seen much is data supporting it. People who disagree are welcome (and encouraged) to prove me wrong.

From your data, the onboard is quite capable of powering the headphones without an amp.

[Tall-Guy]

So lets some it up:

Are the results supporting the use of a specific card or slot (Headphones vs Earphones jack)?
If not, I should probably stick with the USB Card, as it's easier to 'switch' between.

[Garf]

So lets some it up:

Are the results supporting the use of a specific card or slot (Headphones vs Earphones jack)?
If not, I should probably stick with the USB Card, as it's easier to 'switch' between.

Your onboard card with speaker out gives better quality than the headphones out or the USB card.

The difference is probably audible, but certainly not easily.

[Tall-Guy]

Your onboard card with speaker out gives better quality than the headphones out or the USB card.

The difference is probably audible, but certainly not easily.

Fair enough. I'll use this just to be sure. Is a cheaper switcher (so I can switch between the headphones and my speakers) will have a meaning on Quality?
EDIT: or even a splitter instead of switch...

This post has been edited by Tall-Guy: May 26 2012, 11:21

[Brand]

The results between no splitter and splitter+headphones are very close.. I wonder if it was tested properly.

Did you turn on the volume on the headphones, did you hear the test signal?
Which headphones did you test with?


When I tested two soundcards, or actually an onboard and a powered mixer, the differences with headphones were significant. Stereo crosstalk was affected the most.


EDIT: Also, a tip: you can export the HTML report and upload it to something like Dropbox, which allows you to host HTML files (and the associated images).

This post has been edited by Brand: May 26 2012, 11:29

[Garf]

By the way, your result on the headphone out being clearly worse than the speaker out surprised me, but I can confirm that my onboard soundcard (ALC898) gives similar results. This means that either:

a) The electronics powering the headphone port are generally of a lesser quality than the ones powering the speaker out, or they're designed to cope with higher output impedances at a cost of quality.
b) The headphone out tends to be connected by quite feeble wiring running quite some length in the computer case and passing all the other electronics in there. That could hurt. (The speaker out is probably better shielded as part of the mainboard PCB).

I'm less surprised about the onboard sound outperforming the USB card. The chipsets used in these have evolved over a number of years. They were initially quite horrible (old mainboard), but gradually improved to things like this (modern mainboard audio). The cost of development of these can be spread over many models and quite some quantity of boards as they're mass produced. Aside to that, better audio quality is nowadays one of the few remaining parts where a mainboard manufacturer can make a difference, and it's sometimes used in advertising. I'd also suspect the people designing the mainboards are quite capable of avoiding signal interference. Compare this to the USB card which is a free gadget you got with your headphones.

My speaker out also has no problems powering even high-end Sennheisers, though I don't have a splitter so I can't be sure that the audio quality drops if you do this - but given your result, I have no particular reason to suspect that.

(Edit: BTW, I can confirm that the "hotfix" for RMAA does solve the issue I described!)

This post has been edited by Garf: May 26 2012, 11:35

[Tall-Guy]

Ermm, I'm not sure about the headphones Volume actually, It had a knob on the cable itself. Let me recheck it.

[Tall-Guy]

You were right Brand, the Volume on the Headphones itself as not at the max. When I tested it using the Speakers port, I wasn't able to reach all the way to 1dB (only around 2.5-3dB). It was possible with the Headphones port and most easily with the USB Card. Results below:

Without Splitter (same test, just a reference to compare):



With Spliiter and Headphones:



Does this chance the concept of which port I should be using with my headphones? (Because it seems like that for speakers, the normal our should still be the best).

This post has been edited by Tall-Guy: May 26 2012, 11:55

[Brand]

Based on what I see here I'd probably choose the onboard headphones out. Although stereo crosstalk is significantly worse compared to the USB, which in turn has other problems, like high THD and IMD.

I actually don't know how bad THD and IMD are, in the audible sense. I'm guessing they, well, distort the waveforms, but I'm not sure if those artifacts are easy to spot or how to hear them, especially if they're not extremely high.. I'll let someone else comment on that.

As you probably noticed yourself, the volume on the headphones affects the performance. If you usually listen at, for example, half volume, you could also test them at that level and you could get more "practical" results.

[Garf]

Here are some results from an ALC898 (thanks to Case for running these):

http://www.saunalahti.fi/~cse/RMAA/Intel%20DZ77GA-70K.html
http://www.saunalahti.fi/~cse/RMAA/Intel%2...Headphones.html
http://www.saunalahti.fi/~cse/RMAA/Intel%2...0+%20HD650.html
http://www.saunalahti.fi/~cse/RMAA/Intel%2...20+%20K601.html
http://www.saunalahti.fi/~cse/RMAA/Intel%2...20+%20K601.html

The conclusion is the same: the output suffers from having to drive the headphones (300 and 120 Ohms respectively), but even so the speaker port does outperform the headphone out by quite a bit. I think this is a quite interesting conclusion!

Tall-Guys result is the opposite by a small margin, with the speaker out suffering more from having to drive the headphones.

I'm slightly surprised the results for the higher impedance headphones (HD650 at 300Ohm) are better than those for the 120 Ohm model (AKG K601).

[Tall-Guy]

Brand,
Yes - I hardly believe I'll be listen to Music with max volume (I could hear the RMAA tests clearly when the headphones were on the table).

Garf,
Just as a reference, It was tested using a 32 Ohm headphones.

Below you can find the results with 'Optimal' volume. On the Speakers output the Max was optimal. In the Headphones out and the USB Card I tweaked it down a bit.
It seems like the USB Card is performing much better with volume at 2/3 than max:



Still, I'm assuming the Headphones Onboard port is still performing best?

[saratoga]

I'm slightly surprised the results for the higher impedance headphones (HD650 at 300Ohm) are better than those for the 120 Ohm model (AKG K601).

High impedance == less load == easier to drive.

If you want to test a device, use 16 ohm headphones. That is where you separate good from bad.

[saratoga]

Brand,
Yes - I hardly believe I'll be listen to Music with max volume (I could hear the RMAA tests clearly when the headphones were on the table).

Unfortunately, to do that measurement you'll need a preamp, or else an ADC with very large dynamic range. Otherwise, your measurements will probably be limited by the dynamic range of your line in.

Below you can find the results with 'Optimal' volume. On the Speakers output the Max was optimal. In the Headphones out and the USB Card I tweaked it down a bit.
It seems like the USB Card is performing much better with volume at 2/3 than max:

Yes, the results you posted above show massive distortion (look at the THD figures, anything near 1 is probably being overdriven, you were 10x that).

Still, I'm assuming the Headphones Onboard port is still performing best?

The headphone and USB output perform very similarly in that test, and both are probably good enough that your measurements are being limited by the sensitivity of the line in to at least some extent (which is a good thing). Unless one of them has some audible background hiss at regular listening levels (which RMAA will usually miss), I don't think there is a reason to prefer one over the other. I guess you could also try 16 ohm headphones and see if they load the devices enough to bring out some differences.

[Tall-Guy]

saratoga,

Would the Headphones port be more 'Safe'?
When the volume was around 2/3, the difference between the USB and the Headphones are indeed similar like shown below:



But between 2/3 volume and Volume at max, the sound get very unstable on the USB Sound card compare to the Onboard Headphones port:



What I meant by being 'Safe', is that some Audio will require me to play around with the Volume, this 'limitation' will make it unpossible to reach the unstable state. Right?
That in case that indeed there is no real benefit in using either the usb or headphones audio and quality wise.

[saratoga]

What I meant by being 'Safe', is that some Audio will require me to play around with the Volume, this 'limitation' will make it unpossible to reach the unstable state. Right?
That in case that indeed there is no real benefit in using either the usb or headphones audio and quality wise.

Its not "unstable", you're just overdriving the output. Above some maximum volume it'll start to distort, and the performance gets worse. Does that distortion occur at a level you'd be comfortable listening to ? If so, then I would stick to the onboard. If its ear-splitting loud, I wouldn't worry about it.

[Tall-Guy]

It's not ear splitting, there is a good chance I'll reach there with a beer or two ;-)
There is some 'Noise' when no sound is being played on the Headphone Jack, but I don't think it's notice-able why Audio is being played. I'll test this again just to me sure.

Can I plug both my speakers to the out port and my headphones to the headphones port at the same time? or will it overload the onboard card and I should plug one in and disconnect the other?