Carl,
I wasn't indicating that Phils comments were incorrect, I was more so just trying to clarify my point, or better yet, viewpoint..
I must be completely lost as to what we're talking about.. When I say that that all desktops have sound cards, I'm talking about the "sound cards" that generally slip into a pci slot in the back of the machine and are easily interchangeable, and often times are removed and replaced with higher quality sound cards - usually by gaming addicts and high-fi nuts.
The sound circuit your referring to is a different creature, which I assume must work directly with the sound card.
But oh man, it was such a simple question which beg to ask; can the originating audio source be improved? It's the very foundation of the audio chain.
Hello Rich Powers:
I think I can say that we all feast on the questions that come up, because we can not only share what we know, but we can also check whether we know a good answer.
As to your starting question, I always believe that the "originating audio source" can be improved. But that's only the beginning step, because everything that follows in the sound chain must also be carefully designed so that a change to the source is correctly carried all the way to the output.
I have never had a plug-in soundcard, except prior to the Windows PC, back during the Amiga computer days. There we have two Sunrise Sound Cards, a 16-bit and a 12-bit, both of them excellent. I'll have to restore that system because they are so great. But they don't work in Windows PC, so it would mean putting the Amiga on life support.
Hi Again Rich Powers:
I just carefully re-read your opening message and see a question no one else asked.
Are you sure the output you called a "headphone output" is really that, or could it actually be a "line output." The expected technical difference between the two is in PhilB's note.
Then later on, you mentioned you weren't sure about how to interpret the impedances that have been tossed around, and this dislodged some deep buried information I've not used in awhile, which may give you a frame of reference:
There are two types of impedance inter-connections: matching and bridging.
An example of matching is when the output impedance must exactly match the input impedance, as with an audio amplifier (8-ohms) or a transmitter's RF output (50-ohms).
But today most audio connections use the simpler bridging method, and the rule of thumb is simple.: A low impedance can drive a higher impedance so long as the higher impedance is at least 5-times greater than the lower source impedance.
So if the source output impedance is 1,000-ohms, it can drive anything 5,000-ohms or higher. That's called "bridging."
One of the earlier notes did touch on this, but simply used slightly different words to describe it.
Which brings us back to where it started... "Is the earphone jack sufficient?
The answer is yes.
We are talking about AM and the nuances brought up and discussed, except for Phil's and my comments about software "enhancements", won't be a factor.
I think an external USB sound I/O device is a good choice due to the ability of easily switching from one computer to another without disassembly. Your concern about "hiccups" with a USB device is worth considering and this is one of those "try it to find out" situations. If it helps, I use data capture hardware through the USB port and have had no problems with dropout or hanging. Granted, this is different from sound devices but I would expect that if the driver is provided by the sound device manufacturer there would not be a problem. What happens later in the software processing may raise issues.
Neil
My best advice when using computers as audio source is always going to be to keep it in the digital domain until its safely out of the computer. At the very least, a Griffin iMic. And use only high quality shielded cable with an RF filter from its line out.
I haven't been able to get to it, but if I could, I would use audio optical cables because they are virtually impervious to interference.
Rich Powers,
I fully understand the confusion about impedance matching. There is a huge amount of lore on the internet about audio impedance matching. It mainly goes back to the days when transformers were used between audio stages in a typical circuit. That is now relegated to history.
A transformer coupled stage requires careful impedance matching between stages. Transformers are not the preferred option in modern design.
So, what really matters is the voltage level produced by the computer audio output, whether it be headphone or line out. When the destination load it 10k ohms, the voltage at the destination will be essentially the same regardless of the driver source resistance of 10 or 200 Ohms.
There is no information contained in the AC'97 specification that would demonstrate that the headphone out is any different from the line out when driving a 10k ohm or higher load.
PhilB