[EnOYiN]

Quote:

Originally Posted by tbritton /img/forum/go_quote.gif (BTW, DTS stands for "Digital Theater Systems" sound.) http://www.google.com/search?client=...=Google+Search Terry

Yeah I knew that.

Quote:

SB Live! 5.1 (SB0100), latest drivers from Creative (5.12.2.252), Foobar2000 v0942, ASIO4All v2 (ASIO Virtual Devices), SB Live! Audio [9000] 1 - 32 bit - Left SB Live! Audio [9000] 2 - 32 bit - Right SB Live! Audio [9000] 3 - 32 bit - <None> SB Live! Audio [9000] 4 - 32 bit - <None> SB Live! Audio [9000] 5 - 32 bit - <None> SB Live! Audio [9000] 6 - 32 bit - <None> + and - in Foobar works volume fine, Windoze WAV/MP3 Volume does not work ... does that mean I got it working correctly? And the Offline settings are what you said to do.

I guess it does. You can always try to set the latency and buffers really low to check it. If you will get distortion then it most likely works.

The best way is still to play a DTS file from a CD.

 

[hugz]

One thing I never understood, that hopefully you can answer

If ASIO replacements (eg asio4all) aren't bit-perfect, then whats the point of them at all? why did someone go to the effort of making a program that doesn't acheive its purpose?

Thanks

 

[EnOYiN]

Here is something I posted in another thread:

Quote:

Asio4all was not made to fullfill the audiophile needs. It was made for recording in studios. Using asio4all will enable you to sychronise output and input and mix this the right way without having to worry about windows which will give different latencies to different devices. So if you've got multiple devices you can give them all the same latency and latency compensation. I'll give you an example if you have lost it here somewhere. Example: (something that might happen in studios) Suppose you've got 3 devices. Two soundcards for recording and another one for another one for playback. If you are using windows this is what will happen: Device 1: Latency 128 (samples)/ Latency compensation 0 Device 2: Latency 1024 (samples)/ Latency compensation 0 Device 3: Latency 512 (samples)/ Latency compensation 0 You can see that all of them have a different latency and will have to be synchronised afterwards. Windows will give the appropriate latency for every single device. ( it will also change the signal from and to the device - but lets not whine about that right now) When using asio4all this is what you can do: Device 1: Latency 1024 (samples)/ Latency compensation 1024 Device 2: Latency 1024 (samples)/ Latency compensation 1024 Device 3: Latency 1024 (samples)/ Latency compensation 1024 Now you can see that all the devices are using the same latency and because of this every signal will "play" at the same time. Using latency compensation you can compensate the latency. (pure logic there ) So the latency in the end will be 0. You will not have to synchronise anything anymore. Note that you will have to take the highest latency to get this to work. If you take a lower latency the device which needs 1024 will start distorting. This will not alter the playback quality. (it will still be bit-perfect) The only thing which is really changing is the latency. Note that it is only interresting to do this when you are recording from more than 1 device. ( so that would be in a studio most likely) Audiophiles use asio4all just because asio can bypass windows. (to make the output signal bit-perfect) It does not matter what latency you are using and because of this you want to use the highest latency possible. (2048) It does not matter since you are not trying to synchronise anything. You just don't want windows to touch your signal. You can compensate the latency but again: It doesn't matter since you are not trying to synchronise anything.

 

[hugz]

makes sense, thankyou. time to return to linux on my music box

 

[EnOYiN]

Quote:

Originally Posted by hugz /img/forum/go_quote.gif makes sense, thankyou. time to return to linux on my music box

Vista will handle audio completely different. (better) And there is also a driver from usb audio. This driver will provide true ASIO and is therefore bit-perfect. Sadly it is far from free.

 

[hithere]

Quote:

Originally Posted by tbritton /img/forum/go_quote.gif I'm lost in one regard. Why does resampling to a higher bit rate negatively affect sound quality? Rounding errors? Phase errors? Nyquist errors and jitter? How? I hope we are not confusing sampling rate (number of samples per second) with bit depth (16-bit vs 24-bit vs 32-bit) - using a larger "word" size should not resample anything, should it, but would merely fit the top 16 bits up front into the 24-bit word, right? The extra bit depth is very useful for performing DSP calculations with greater accuracy

...I thought we were talking about sample rate conversion from 44.1 khz to 48, where the important thing is the math used when the number of samples at the new, higher, rate are not integer multiples of the number of samples in the rate being converted from, therefore some kind of interpolation must be used lest the new output show patterned errors and not fairly represent the original input to the best of it's ability. Separate issue from bit depth (16 vs. 24 vs. 32, etc.) Sure, increasing bit depth is as easy as adding zeros, then using whatever math to cut the sample down for volume control and the like.

I don't know if the k-mixer makes an audible difference in most cases, I've read reports of audible differences, but then I've read reports of people claiming to hear audible differences between digital cables, which I've never seen demonstrated in a blind test (and suspect I never will). I have read things that lead me to believe that it doesn't work properly for someone who wants nothing more than to pass exactly what's on a CD over a digital interface to an outside decoding mechanism. According to some documentation, k-mixer doesn't perform SRC if there is only one stream present and/or hardware mixing is enabled on the card.

 

[EnOYiN]

Quote:

Originally Posted by hithere /img/forum/go_quote.gif ...I thought we were talking about sample rate conversion from 44.1 khz to 48, where the important thing is the math used when the number of samples at the new, higher, rate are not integer multiples of the number of samples in the rate being converted from, therefore some kind of interpolation must be used lest the new output show patterned errors and not fairly represent the original input to the best of it's ability. Separate issue from bit depth (16 vs. 24 vs. 32, etc.) Sure, increasing bit depth is as easy as adding zeros, then using whatever math to cut the sample down for volume control and the like. I don't know if the k-mixer makes an audible difference in most cases, I've read reports of audible differences, but then I've read reports of people claiming to hear audible differences between digital cables, which I've never seen demonstrated in a blind test (and suspect I never will). I have read things that lead me to believe that it doesn't work properly for someone who wants nothing more than to pass exactly what's on a CD over a digital interface to an outside decoding mechanism. According to some documentation, k-mixer doesn't perform SRC if there is only one stream present and/or hardware mixing is enabled on the card.

I do not mean to offend you or anything, but I simply have no idea what you are trying to say here. If you mean to say that the post I wrote contains errors or wrong statements please tell me.

However I doubt there is any faulty information in there since I checked everything 3 times from different credible sources.

 

[seefeel]

Quote:

Originally Posted by xenithon /img/forum/go_quote.gif There is the commerical USB driver from USB-Audio, which replaces the generic USB driver and allows the application to talk directly to the hardware, bypassing the kernel, EVEN if the hardware does not have its own ASIO driver.

Is this really true? Does the USB-Audio driver really bypass the kernel even if the USB DAC you're using doesn't support ASIO?

 

[EnOYiN]

Quote:

Originally Posted by seefeel /img/forum/go_quote.gif Is this really true? Does the USB-Audio driver really bypass the kernel even if the USB DAC you're using doesn't support ASIO?

It's true. This application was invented by steinberg as an addition to Cubase. I think it was version 3.5 (of Cubase)

Here is the site of steinberg. There are some parts in german though.
http://www.steinberg.de/23_1.html

 

[seefeel]

I'm referring to the usb-audio.com driver. If I use this driver with a Scott Nixon USB DAC that doesn't support ASIO I'll be bypassing the kmixer and getting bit-perfect output?

 

[EnOYiN]

Quote:

Originally Posted by seefeel /img/forum/go_quote.gif I'm referring to the usb-audio.com driver. If I use this driver with a Scott Nixon USB DAC that doesn't support ASIO I'll be bypassing the kmixer and getting bit-perfect output?

I am reffering to the same driver.

This driver was invented by steinberg.

Yes, you will get bit-perfect output. You can download a trail, if you like, to see if it will work with your DAC. I would like to advice you to try it first before you buy it. It isn't cheap. The beep is kinda annoying but at least you can try it.

 

[xenithon]

seefeel - yes it is true. Unlike ASIO2KS and ASIO4ALL, which are wrappers for KS, the usb-audio driver talks directly to the hardware completely bypassing the OS audio stack.

 

[EnOYiN]

So.

I am pretty much out of ideas about any other stuff I can add.

If you've got any suggestions please tell me. There should be something else I can add here. If there will be no other suggestions I will quit editing the original post since this has been somewhat timeconsuming.

 

[Joshatdot]

I was getting distortion and what not when I set buffers and stuff to 0

btw, what are some examples of a DTS CD track? .. does that mean any CD's that are DDD? you mean .dts files extracted from a DVD?

 

[tbritton]

Quote:

Originally Posted by EnOYiN /img/forum/go_quote.gif So. I am pretty much out of ideas about any other stuff I can add. If you've got any suggestions please tell me. There should be something else I can add here. If there will be no other suggestions I will quit editing the original post since this has been somewhat timeconsuming.

Thanks, you've done a truly great job with this!

Terry