What is wrong with internal soundcards?

[ROBSCIX]

There seems to be a great amount of "debate" in this fourm regarding internal over external gear.

I have been asking quite a few people, if you could "fix" what you think is wrong with a internal PC sound card what would you add/remove or modify?

There has been a few ideas that are common among many I have spoke with such as inputs for external power. Higher quality clean-up section for ripple suppression and leveling of power input.

Other ideas that have floated around among those already involved in the discussion through PM:

1. Separate high quality DAC's for L & R channels
2. Improved jitter correction
3. Higher quality Clock generators
4. EMI/RFI shielding
5. Possible Balanced output such as XLR
6. Different output section TRS, TS
7. Improved output sections using higher quality components
8. Different daughter cards allows specific features.

This is a bit of a project of mine so research is in order.
So, what do you guys think, what are you suggestions to improve what you think is currently wrong with internal sound cards?

 

[tintin47]

All of those changes would clear most of the issues up. It seems to me that most don't need the DAC actually in the computer/arent willing to go through heavy mods to make your vision happen. That or they believe that non-soundcard DACs can be much higher quality for cheaper. I don't know enough about DACs to comment on that, though.

 

[ROBSCIX]

Lets here what you have to suggest...

 

[leeperry]

Lynx2/RME is all you need I think.

they have unsurpassed clock accuracy and shielding.

 

[ROBSCIX]

I don't need suggestion for which card to buy.
I am looking for ideas as to what people think is wrong with a internal sound cards and what could be done to improve them.

I have listed a few ideas that I think would improve them and others have suggested ideas also so do you have any ideas?

 

[leeperry]

well clearly, the RME soundcard are where the money is for me

they allow you to change the masterclock, so you can use Reclock in "bitperfect" :
Bit exact audio in Reclock without any resampling - SlySoft Forum

apparently the Lynx can also do that.

RME even throws in bit-perfect MME drivers(luckily my M-Audio soundcard does too).

when you buy a top end RME/Lynx, I believe you get the quintessence of PC soundcards...it doesn't get any better than that, they got it all figured out nicely for you already

 

[smrtby123]

It is usually a combo of the reasons you listed. When you pay about $250 for a motherboard, a small portion of that is for the audio section. The RealTek chips are pretty good while the Soundmax stuff sucks. My Asus m2n-32 SLI has a soundmax chip and surround sound doesn't work, and it clips at low levels. When you have a proper transport you get high quality chips and components, engineering dedicated to audio and not adding audio to a larger system, better software support, and more power available. You can't just "fix" the problem with good engineering. The same thing is true in the situation of on-board graphics and a proper PCI-E GPU card.

 

[Acix]

The best thing that you can do is to buy a nice sound card for your PC...

The best ones in the lower prices are,
1. Kontrol 1
2. Lexicon omega
3. Echo Dj indigo only for laptop).

For better converters and more pro options,
1. RME FireFace 400/800
2. Apogee mini dac.

As well you can check PCI sound card versions.

 

[ROBSCIX]

I must not be making myself clear with the original post.

Consider this a theoretical discussion.

If you could build a internal soundcard designed for the best sound possible what would you addto make it have the highest audio quality?

Woudl you want a:
A great clock chip with low jitter tied to L&R sperate DAC units al powered form a external PSU? ....etc... seperate add ones for 7.1 etc?


Nobody has any ideas on how to improve the sound quality of internal soundcards?

 

[null_pointer_us]

I've heard of mods for internal sound cards.

But I don't see the point in discussing what would need to change for internal sound cards to improve. Unless I'm missing something, none of the input you might see here would actually make it into the hands of sound card developers.

I'd definitely read the thread, though.

 

[Nebby]

Unless he's working for a sound card company....

 

[royalcrown]

In all honesty, the whole "internal soundcards are inferior" strain of thought is totally false - An external DAC is just another box that sits on top of your computer instead of within it. Bus noise, RFI, EMI have all been claimed to negatively impact the SQ of internal cards, yet the noise levels (as well as every other aspect) on a properly designed soundcard are well beneath the threshold of audibility. As a comparison:

RightMark Audio Analyzer test : E-MU 0404 USB loopback bal 16b 44k
EMU 0404 USB
vs.
RightMark Audio Analyzer test : [MME] E-DSP Wave [EC00]
EMU 0404 PCI (with balanced TRS cabling)

Despite the belief that the USB is superior in terms of sound quality, the two are indistinguishable and any differences are well within audible limits. In fact even in terms of the small differences in measurements, the PCI card actually beats the USB DAC in all but IMD.

Really, I think the whole belief that PCI cards are inferior stemmed from people looking at them and saying "oh it's inside a PC, so it MUST be inferior" without actually looking at the cards' specs and listening to them. I've blind tested a level-matched 0404 PCI to a DAC1 USB and couldn't tell the difference whatsoever - that's just me, but I doubt that I'm the exception to the rule given that my inability to distinguish the two lines up perfectly with what the measurements predicted.

Then again, PCI soundcards aren't "audiophile" and I assume that has a lot to do with it.

 

[leeperry]

these RMAA tests are flawed coz they use the onboard ADC anyway...

look at these, they used a high end Roland rack and measured THD/SNR :
DOS/V POWER REPORT | Impress Japan
DOS/V POWER REPORT | Impress Japan
DOS/V POWER REPORT | Impress Japan
DOS/V POWER REPORT | Impress Japan
DOS/V POWER REPORT | Impress Japan
DOS/V POWER REPORT | Impress Japan
DOS/V POWER REPORT | Impress Japan
DOS/V POWER REPORT | Impress Japan

most internal cards are lousy, but some of them actually have properly shielded PCB's

 

[ROBSCIX]

Quote:

Originally Posted by leeperry /img/forum/go_quote.gif these RMAA tests are flawed coz they use the onboard ADC anyway... look at these, they used a high end Roland rack and measured THD/SNR : DOS/V POWER REPORT | Impress Japan DOS/V POWER REPORT | Impress Japan DOS/V POWER REPORT | Impress Japan DOS/V POWER REPORT | Impress Japan DOS/V POWER REPORT | Impress Japan DOS/V POWER REPORT | Impress Japan DOS/V POWER REPORT | Impress Japan DOS/V POWER REPORT | Impress Japan most internal cards are lousy, but some of them actually have properly shielded PCB's

Many times, people don't understand what they are seeing on RMAA or how to properly interpret the results.
In loopback mode, the distortion ratings are results of both sections.
Where you run into issue in loopback, is when the ADC is of lower quality then the DAC in the playback section. This causes lower measurments in the results and gives people a incorrect impression of the cards ouptput measurments.

 

[leeperry]

RMMA results use loopback, which can hide any groundloop/EMI/noise problem you could/would encounter w/ a third party input.
these tests are worthless IMHO, this Japanese site did it really well, I just dunno if they used a bit-perfect player...but even if they didn't their results are still plenty more meaningful than some random RMAA results