External audio interface - scam, or necessity? (intel mac)

Not digging through 120 posts to see if it was mentioned... :)
...but digital bits and interface functionality features aside, all-internal audio cards are more prone to noise and interference from the electronics within the computer.
By bringing the analog portion and audio conversion portion of an interface outside of the computer, that is avoided.
Some interfaces may still have a card that goes into the computer, but they keep the analog audio and conversion portion of their electronics outside in a separate unit, or you have it all outside of the computer in your interface, and it's a just some type of cable link back to the computer.

Computers are full of electronic "hash". Which is why it's best to move them as far away from your audio gear as possible in a best-case scenario.

That "hash" may not be something that's very obvious...so it's not as simple as "I can't hear any noise".
It may be cumulative and it may be intermittent. Basically...computers were not built as audio devices, with sonic purity in mind.
Once you've converted the audio, then it's just another file in the computer and not a problem.
 
miroslav said:
Not digging through 120 posts to see if it was mentioned... :)
...but digital bits and interface functionality features aside, all-internal audio cards are more prone to noise and interference from the electronics within the computer.
By bringing the analog portion and audio conversion portion of an interface outside of the computer, that is avoided.
Some interfaces may still have a card that goes into the computer, but they keep the analog audio and conversion portion of their electronics outside in a separate unit, or you have it all outside of the computer in your interface, and it's a just some type of cable link back to the computer.

Computers are full of electronic "hash". Which is why it's best to move them as far away from your audio gear as possible in a best-case scenario.

That "hash" may not be something that's very obvious...so it's not as simple as "I can't hear any noise".
It may be cumulative and it may be intermittent. Basically...computers were not built as audio devices, with sonic purity in mind.
Once you've converted the audio, then it's just another file in the computer and not a problem.
Click to expand...

No need for digging to my mind, thanks for the thoughts.

If you know of handy links to real-world test results demonstrating internal noise or 'hash' of computers with internal soundcards surfacing in audio recordings, it would be interesting to review. My particular concern would be hash or pc electronic noise in mac computers surfacing to the point where it was disrupting recordings, but any evidence would be helpful.

Certainly any pot or physical connection can and eventually is likely to contribute to noise, which is why connections should be exercised regularly imho, and electrical interference from cabling is a constant fight. Noise in a 24 bit internal sound circuit, that I could not get away from without its affecting my signal, would be a huge concern for sure.

Intermittent issues can surface in about any system without redundant processing, and even with, and external USB AI's rely on computer processing to write files so they do not provide immunity from that potential issue.

A guy here in the thread, Dave, did me a huge favor in analyzing a few files for noise. My working understanding is that in my case I do not have a glaring internal noise issue, but I have to say, I did not completely understand his result and did not note a focused conclusion regarding noise - so I might have misinterpreted his result. Sure appreciated the favor, sometimes it just takes a lot to understand all of the implications of a technical test result.
 
Google is your friend...:)...and this is just a small sampling of web page links talking about computer EMI noise:

Solving Computer Audio Problems

Forum - Main Forums : PC Music

Sci.Electronics FAQ: Repair: RFI/EMI Info

Computer noise in audio system | element14

The Well-Tempered Computer

This video is not specific to internal sound cards...but it gives you some idea the kind of EMI noise that is in and around computers. So check it out and decide if you still want your analog audio anywhere near your computer.




Though I honestly get the feeling that Stephen Hawking could come here an provide equations and examples and you would still be skeptical...mainly because you seem dead-set on using your internal sound card. ;)
 
miroslav said:
Google is your friend...:)...and this is just a small sampling of web page links talking about computer EMI noise:

Solving Computer Audio Problems

Forum - Main Forums : PC Music

Sci.Electronics FAQ: Repair: RFI/EMI Info

Computer noise in audio system | element14

The Well-Tempered Computer

This video is not specific to internal sound cards...but it gives you some idea the kind of EMI noise that is in and around computers. So check it out and decide if you still want your analog audio anywhere near your computer.




Though I honestly...get the feeling that Stephen Hawking could come here an provide equations and examples and you would still be skeptical...mainly because you seem dead-set on using your internal sound card. ;)
Click to expand...


Thank you for taking the time to provide this information, I will dutifully review all of these links. I see that the Mac g5 is referenced in the first link.

I am familiar with electromagnetic noise, as a network technician for Dartmouth I managed their video over IP solution for many years, among many other roles that dealt with this, obviously not the least of which was cat 3 to cat 5 & cat 6 cabling often run in cable chases along with power cabling. Please do not presume that I am a stranger to the general issue of electronic devices generating interference.

I look forward to digging through the rest of the information that you have so kindly provided.

I am not invested in any particular belief, only in growing my understanding of the issue and the evidence-based science behind it. It is true, that I take a skeptical perspective toward most things - I would not expect that, to be offensive and do not mind your poking fun at it. I do not have to use my internal sound card; I have a DR-05 as well, and have been using that, since it is connected to a solid state mixing board whereas the mac mini is connected to a pmp4000, which has several fans. I am simply in no rush to buy an external AI given I have alternatives and also larger issues to resolve. Thanks again for having taken the time to add the links, much appreciated.
 
Maybe it was some your initial posts that made it seem like you had your mind made up about cards vs external options.

Since you have a background in computing...the the video alone should drive home the point about computers and EMI.
Sure, you could do things on the audio side to fight the EMI...but IMO, it easier to try and avoid it, and then Stephen Hawking won't have to get involved. ;)

YMMV...
 
miroslav said:
Maybe it was some your initial posts that made it seem like you had your mind made up about cards vs external options.

Since you have a background in computing...the the video alone should drive home the point about computers and EMI.
Sure, you could do things on the audio side to fight the EMI...but IMO, it easier to try and avoid it, and then Stephen Hawking won't have to get involved. ;)

YMMV...
Click to expand...

Thanks for your advice Miroslav, I appreciate it. Good luck with your music.
 
It's worth saying that the noise inside a computer can take many forms.

To check for a couple of common ones...

First record some silence (connect your mixer as normal but turn everything down to zero). After you have 30 seconds or so, zoom in on the noise floor in both the vertical and horizontal axis. If you see anything that looks a bit like a varying square wave superimposed on the analogue "hash" that's digital noise. As I mentioned many posts ago, when I used to receive audio files from a guy with a very high end Mac (can't recall the model) this was a near constant issue. Luckily, Audition's NR system could eliminate it.

Second, generate a clean sine wave. Again zoom in until you can see individual waves and, again, zoom on the vertical scale. If you see any form of extra frequency superimposed on the main sine wave this is probably one of the clock generators "leaking" into the audio card.

Note (before ecc83 tells me off) that this isn't a perfect test--it depends on how your DAW draws the waveforms how much you can see so you may get a "false negative".

However, as has been said many times, if you don't notice the problems in your recordings, then why worry? Working with an internal card isn't ideal but this is home recording after all!
 
Last edited by a moderator:
That video was probably the worst bit of pseudo-scientific claptrap I have seen in years.

First of all the loon was using a device designed to detect and demodulate RADIO frequencies. Surprise, surprise! Computers with a Xtal controller RF oscillator in them radiate some RF! Switch Mode Power Supplies* are also fekking great power oscillators running at 30kHz or more and chucking out harmonics into the MHz range.

Secondly the radio was not tuned to a station. This meant the AGC was opening the gain to max and the sensitivity was probably better than one MICROVOLT! Effectively the "noise to signal" ratio was close to infinite!

Thirdly, radios have a ferrite aerial inside them DESIGNED to couple to RF fields, last time I looked, none of my soundcards did?

Sure, audio systems CAN detect and demodulate RF but we take great pains these days to avoid that. (there are stringent tests imposed here in the EU. I guess the US has similar?) Had the guy waved an SM57 about that laptop feeding a 1/2 decent pre amp he would have picked up Jack Shit.

But the results for internal sound cards speak for themselves. As I have said, my 2496 card achieve a noise floor of around -93dBFS unweighted. A weighted figure would be better than -100dB, mainly because it would reduce the influence of 50Hz hum and some mains harmonics..NOT you see RF artifacts! More expensive cards, e.g. RME, have even better noise floors.

In any case, what IS an external AI but a box of mic pres with a fekking COMPUTER in it!

*Waggle a mic, better a guitar within a foot of a 50/60 Hz linear wall rat/line lump. No RF there but BOY there is a big hum field!

Dave.
 
Not going to tell you off Bobbs, you covered the caveats pretty well!

There are few sources that could stand such intense scrutiny. The output of a stereo FM tuner will have easily detectable 19 and 38kHz residues. Tape signals are "sampled" at the bias frequency. And you could not even DO the test to black disc because vinyl is so frigging noisy!

Many monitors, especially subs now use "class D" switching amplifiers. Anyone care to open theirs up and 'scope the speaker terminals?

Old AM telly sound had 10,125 Hz imposed. 625 colour screamed 15625 at you (people, especially the ladies DID notice that as cabinets degenerated to 3mm plastic!) NICAM was wonderfully rugged but low bit digital.....

Dave.
 
I'm pretty sure RFI and EMI are different. Something like RFI needs something to act as a detector circuit while EMI can be induced directly into a cable from the source. But then I'm not really the expert here.
 
Dave,

I think we have to differentiate between well designed specialist products like the 2496 (and there were other good PCI sound cards in the days when PCI ruled) and the sort of single chip 39 cent device that comes installed on most mother boards these days. There's a chalk and cheese difference, both in terms of audio performance and screening against interference, between specialist cards and the cheapies. After all, why would a manufacturer worry too much about a sound chip that's mainly for watching Youtube and the occasional Skype call? Performance and screening isn't a priority...but price is.

The 2496, on the other hand, cost a bit of money and was aimed at a totally different market.
 
miroslav said:
I guess then...computer EMI is a myth....
Click to expand...

Not at all my friend! That computers and virtually all electronic devices chuck out RF noise is a fact of life, the first problem in studios was probably fluorescent lights.

But like a lot of things "in life" it is how we deal with what we cannot change that matters. Audio circuits CAN be made RF proof to almost any degree you want and like most things in life again, the more you pay, the better they tend to get! (you CAN make computer systems nuclear EMP proofed if you have the odd $M.)

Valves, well triodes, are by nature a bit RF proofed. Transistors are not and I was in on the ground floor of the start of silicon transistor gear where even AUDIO devices would amplify very well into 100mHz and above. Most makers of "consumer" audio were tightarsed bastards and would not fit even one cap and one R to at least try to mitigate the intrusion and detection of radio signals. I spent a goodly proportion of my time retrofitting EMI protection in radiograms, music centres and yes, even some hi fi amps!

Then! Taxi drivers got PMR (and transistors got better at RF) but the biggest problem perhaps was CB burners.

We live immersed in an EM spectrum, from low grade heat to cosmic rays. It is how engineers DEAL with that spectrum that matters. Only in the audio field do people design for silly bandwidths! You don't make a ship's steering servos able to respond at 1kHz so why make a mic pre with a 3dB point at 250kHz? Speaker systems that reach 50k when there is only more clap trap to support such a response!

Dave.
 
"I'm pretty sure RFI and EMI are different" Well BSG, one is a subset of the other!
ElectroMagnetic Interference embraces all radiation whereas Radio Frequency Interference starts at the RF end. But, where's that? 20kHz and below will penetrate the sea and was (is?) used for submarine communication.

You are quite correct to say RF needs a detector but this can be the first PN junction of the input transistor in a preamp (or on the chip of same) but very high RF fields, from a close mobile phone say, can cause local overload and anytime a circuit gets non-linear it becomes a rectifier and that is what a radio detector is!

Bob, the 2496 is not shielded! I guess they keep the ***t out by careful, low Z path design and appropriate filtering and RF stoppers (Ooo! The card is old enough for there to be a schematic in the ether? Never looked!).

Dave.
 
miroslav said:
But that's the point in this thread...that you CAN change by simply using and external audio interface to avoid EMI, and improve your audio quality.
Click to expand...

No, you can't. EMI will still be there but a better external AI (such as the KA6) will be a few dB better than a 2496 and THAT better than a £0.5 soundchip.

No matter, it is the subjective effect that matters and even with the "inferior" 2496 noboby is going to hear noise at -90dB in a properly setup system and CERTAINLY not in the presence of wanted sound at even neg40!

Of course we all strive for the best we can do but some folks still bear tape!

Dave.
 
Last edited:
Shhhtop! You're both right!

Yes, there's EMI around us everywhere, from mains wiring in our walls to the cell phones we carry to the video monitors we use, etc. etc.

However, the AMOUNT varies greatly. In a normal room, the levels are low enough to rarely be a big problem. On the other hand, if you have high voltage power lines over head and a TV transmitter next door, the levels are pretty high. At the TV facility I work at in London, we once had to run a bunch of mic cables from one of the studios out to our reception. The only available route was via the air conditioning plant room for two of the studios...an area full of several hundred amps of 3 phase power, motors switching on an off, control boxes, etc. We had to go to huge lengths to avoid EMI on that cabling (star quad cables in a sealed steel conduit if you're interested).

The inside of a computer is an example of another place that is electrically noisy. This doesn't mean it's impossible to have audio in that sort of environment but it does mean that care has to be taken with circuit design an screening to minimise the problems. That's why Dave's 2496 works okay...but also why the one-chip-wonders can suffer badly.

Moving to an external interface is a very cost effective way to reduce the amount of EMI you have to deal with but not spend a lot of money on it. It's not essential for a good interface to be external...it's just easier to build one that way.
 
"Shhhtop! You're both right!"

Ok, I will....In a minute! The external AI HAS to be a good one! You are still putting high gain audio circuits and CPUs (and ram I guess?) in close proximity, closer maybe than a soundcard chip is to the MOBO's CPU in some cases?

Later I shall run a noise test of the iO2 as an example of perhaps the very cheapest but still acceptable end of the AI market?

Bottom line, "we" can only make it better, we cannot eliminate EMI effects entirely but once we get it to -130dBu (if we ever can) that is as far as we can go.

There is no digital audio system on the planet that I am aware of that gets within 10dB of that?

Dave.
 
ecc83 said:
No, you can't.
Click to expand...


No one's talking about changing the computer in some way so it doesn't have EMI.

It really is a very straightforward solution.
Use a long cord....and put your computer and interface and other audio gear as far apart as you can.

If you're going to place your interface on top of or next to your PC....then you're not "changing" anything.
If you move it away...there's the "change" that matters.
That may not appeal to some of the ITB crowd that has everything piled up within arm's distance...but it does make a difference if you keep certain gear apart.
It's what single-coil guitar players learned a long time ago...it matters where you stand and what direction you turn in if you want to minimize noise.
 
miroslav said:
No one's talking about changing the computer in some way so it doesn't have EMI.

It really is a very straightforward solution.
Use a long cord....and put your computer and interface and other audio gear as far apart as you can.

If you're going to place your interface on top of or next to your PC....then you're not "changing" anything.
If you move it away...there's the "change" that matters.
That may not appeal to some of the ITB crowd that has everything piled up within arm's distance...but it does make a difference if you keep certain gear apart.
It's what single-coil guitar players learned a long time ago...it matters where you stand and what direction you turn in if you want to minimize noise.
Click to expand...

Actually when I did the noise floor tests on the KA6 it was sitting ON TOP of the computer that hoses the 2496 and was working!

My point is, external AIs CAN be quieter because they are better designed. I have pointed out at least twice, AIs ARE computers with pre amps! The guitar analogy fails because there is no CPU inside an electric guitar!

Dave.
 
You must log in or register to reply here.
Back
Top