Pr,
Makes sense to me.
Cheers! |
MP,
Yes, you keep saying it was to do with sound processing and I'm agreeing with you. Networking, and streaming of audio files, has nothing to do with sound processing: apples and oranges.
Any computer can have problems playing audio files if it's trying to run too many processes at once, especially Vista because of the high overhead of the Aero interface and the increased number of security threads. However, I agree with you about recommending Vista for audio playback because MS replaced kmixer with core audio which sounds a lot better. To summarize, If I had to choose an operating system to listen to music on the computer, I'd choose Vista. If I had to choose an operating system to stream music, I'd choose XP or Linux.
I have an 1.6Ghz Atom based nettop computer, running XP, that I use as a music server. I've had it streaming lossless files to six different devices and it doesn't even break a sweat. However, I know it couldn't decode and play five audio files at a time.
You see, were saying the same thing. You say potato and I say potato....... |
PR,
I recall the issue was with certain core programs/components bundled into Windows that dealt specifically with sound, not windows IO functionality in general. It had to do with maintaining throughput with older, slower computers not necessarily designed for high performance A/V.
Don't recall the details nor can I confirm that the problem truly exists, but it makes sense to me that newer OS versions have more refined software components for A/V since that is now a core application area that people buy the computers for. |
Think about it. If older versions of Windows were re-sampling data packets traveling over a network, then all data would be adulterated or corrupted whether it be a picture, music, excel spreadsheet etc. Because Windows, or any operating system, does not have the ability to distinguish the contents of a data packet, how would it be able to single out audio packets and re-sample them? The only thing that windows looks at in a data packet is the header information to determine where a packet came from, where it's going and the checksum to determine if it was corrupted during it's journey.
Now, you can get third party software that will do deep packet inspection, but even the best versions have very low levels of accuracy. The RIAA, and ISP's, have been doing trials with deep packet inspection software with poor results. In order to get high accuracy rates, they really need to capture the whole file being transferred, determine the type of file and if it's legitimate or illegal, and then either allow or stop the transfer. |
Prpixel,
I do know the difference between the two but not the details of which MS software does what when, so that is helpful.
I only recall reading that it was an issue with the older versions of Windows in general but not XP.
Since it is very difficult often to determine what OS bundled software is actually doing what when something does not sound right, I recommend Vista in general as a more robust and less problematic platform for digital audio over older versions of Windows, that's all. |
Mapman,
Windows win95, win98 and WinXP do not down sample network packets on a wireless network.
Windows XP kmixer does re-sample audio to 44.1 or 48k.
Your confusing digital to analog conversion in a soundboard with streaming packets of data over a network. |
2chnlben,
It is possible that software on your PC source could reduce the sample rate even though the router is not an issue.
For example, I have read (unconfirmed) that some older (pre Vista) versions of Windows may reduce sample rates implicitly in some cases. I believe this is possible though I do not recall specifics.
To avoid these kinds of problems, stick with newer computer platforms that are well suited for multimedia type apps, like Windows Vista or newer similar server versions of the OS (yes Vista gets a bum rap in corporate corners but it is essential to problem free playback off of PCs, I believe) or Apple. |
Thanks Blindjim. I'm going with the Mapman's take on this matter - If I experience no dropouts and the music being played sounds fine, I will assume that I am hearing the entire sample rate of the file being played (with exception to the limitations of the other electronics - i.e.: SB3 and Duet are limited to 24/48). |
with respect to routers, it is all about thru put. Packets. The content of these packets isn't the issue.
it's like loading a truck, on in WY Fi, maybe loading a plane with materials. What the amterials are isn't important. It's their size, and or overall weight, and speed of the plane that matters... With Vista you may need to make some other tweaks to speed up the passage of these 'packets'. Another thread here about that is more well suited to describe the steps needed to do this.
Signal strength as was said is important for securing good speed and avoiding drop outs or intermittent playback over a wireless connection. |
This is excellent - thank you all, for clearing that up. After reading that, "...most WiFi routers are limited to 16/44.1..." I was concerned that all of my 24/48, 24/96 downloads, not to mention the additional cost, had been in vain.
My music server is just not at the level I want it to be, so before I do any more tweaking/modding, I wanted to cover the basics to make sure I dont gratuitously stimulate the economy! |
Also routers only route and do not implement algorithms for down sampling data streams so if you are hearing the music play without dropouts, you are most likely getting all the bits delivered that should be. |
WiFi speeds will vary dramatically with both signal strength and the wireless standard used by the equipment (802.11a,b,g,n). 802.11g, which is the standard most commonly used by recent equipment, should handle pretty much any audio data comfortably, provided that the signal strength is good. (A wifi link will operate at reduced speed if signal strength is weak). 802.11g will invariably, though, provide much lower speeds than its theoretical 54mbps maximum, even under ideal signal conditions.
To test your speed, transfer a large computer file across your wireless link, and measure how long it takes. Multiply the number of megaBytes of the file by 8 to get megabits, and divide by the number of seconds to get mbps.
Regards,
-- Al |
16 x 44,1000 x 2 (stereo) = approx 1.4Mbs.
May be a problem with the very first version of 802.11 circa 1997, but no problem with later WLAN 802.11a or 802.11g, the latter having a theoretical bandwidth of 54Mbs.
Regards, |
