moranlan

trully amazing project, my hats off to you!

I'm using an audigy2 zs PCI in a carpc environment doing all audio processing actively on the DSP (including crossover, eq, time delay etc..).
I'm not in need of any recording capabilities at the moment.

My speakers setup in the car is 4way using some amplifiers. I have a total of 7 channels. 1ch running the subwoofer, 2ch running the midbass woofers, 2ch running the midrange speakers and 2ch running the tweeters. no passive crossovers are used but those of the amplifiers which are set at safe marginal values not to kill my speakers in case of a kx issue.

My questions:

1. When playing an audio file with foobar into 10k2 0/1, even if trying to resample with SRC (foobar plugin) to 24/96kHz, the audio is still sent to the inferior 10k2 dac and resampled back to 16/48kHz, making the whole thing useless. Is that assumption correct?

2. If trying to play some 16/44.1kHz contect into 10k2 Wave HQ without resampling on the software, the audio will go to the p16v DAC and upsample automatically to 24/96 but will bypass the DSP and only play on the speakers connected to the rear channel (when front is swapped). Is that correct? Does it upsample everything or do I need to at least output at 24bit ?

3. This is more a theoretical question: While using high-end audio components, will upsampling FLAC files captured in 16/44.1kHz (956kbps) to say 24/96kHz improve the quality? I've been reading that it should due to better interpolation and some theorem.

4. Is there any way to be able and play 24/96kHz on more channels than the rear?

Really appreciate your help!!
-Mo

Maddogg6

Yes that is what is happening.

I think it depends on the kX driver compatibility setting 'P16V I2S Follows SRC' - it looks to me like, when not checked, audio will be upsampled - when checked the I2S DAC changes its sample rate to match playback sample rate - I think I read that in the help file as well.
You can have a look at the SPDIF/I2S status window while playing with various PB sample rates with Wave HQ and changing that option.

Not unless the up sampling employs some noise sharpening algo (can be CPU intensive) ... 'better' is subjective..
So you understand - converting audio to digital cause some high end freqs to be cut off, its the nature of digital audio and anti-aliasing filters all DACS employ.
Its argued by some the significance of those freqs that are cut out... 44Khz technically can cover the 20-20Khz freqs our ears can hear - while others say we perceive higher freqs - I would be one who argues the latter, and can verry well depend on the person, just as some people have better hearing in the 20-20Khz ranges, its seems logical the same can be true for higher freqs.

That said... think about it - how do you add something that was lost - noise sharpening is something a science called 'psycho-acoustics' came up with - noting the 'psycho' part. Seriously - the highs it may add that were lost is considered 'imaginary', in that it tricks your senses into thinking more high end is there, but its just noise.
So the easy/short answer is - no.
Wave HQ is multichannel capable - I have used a up-converting plugin to 5.1 with Winamp and Fubar2000 with success - but you can forget about all the DSP goodies (EQ, crossover, delays, routing etc... )
So the center/lfe works too... and also, in Sonar it gives me the headphone output of the Audigy Drive - so, theoretically you can do 8 channels with Wave HQ, but only sonar have I seen the headphone output ability, never in a media players upconversion.

The P16V is not a dsp and it is the DSP that allows us to do all those neat things with kX. So you would need to use hardware crossovers. Likely something active if your bi-amping.

moranlan

Thanks for the prompt reply.

Been playing with 'P16V I2S Follows SRC' and checking the SPDIF/I2S status and noticed the following (while playing to Wave HQ):

• Setting disabled - playing 44.1kHz/16bit I see this: p16v: PB: 96000Hz I2S Out:96000Hz. I guess this means that p16v really upsamples...
• Setting enabled - playing 44.1kHz/16bit or 96kHz/16bit using Best Sinc plugin I see this: p16v: PB: 96000Hz I2S Outbypass).
• Setting enabled - playing 48kHz/16bit using Best Sinc plugin I see this: p16v: PB: (bypass) I2S Outbypass).

1. What does PB stands for?
2. How do I know which resolution is being applied? I'm playing at 16 but can set foobar to play at 24, just wasn't sure if I have to do this on software or not...

I've been doing some reading, most recently, "The Art of Building Computer Transports v0.3". This guy claims that upsampling really helps even with CD-rips. Allegedly, this is achieved using the Secret Rabbit Code "Best Sinc" interpolation method. I found this to be odd, considering that these stuff were lost during the ADC.

“In digital audio, what matters is the audibility of interpolation error between samples. Since Shannon’s sampling theorem says it is possible to restore an audio signal exactly from its samples, it makes sense that the best digital audio interpolators would be based on that theory. Such “ideal” interpolation is called bandlimited interpolation. The problem is to correctly compute signal values at arbitrary continuous times from a set of discrete-time samples of the signal amplitude. In other words, we must be able to interpolate the signal between samples. Since the original signal is always assumed to be bandlimited to half the sampling rate, (otherwise aliasing distortion would occur upon sampling), Shannon’s sampling theorem tells us the signal can be exactly and uniquely reconstructed for all time from its samples by bandlimited interpolation.”

3. Any idea if the technique used on p16v any similar to what Secret Rabbit Code Best Sinc does?
4. Do you agree with those claims in the book?

I'm now in the midst of decision-making. If 24/96 or even 24/192 upsampling can actually help my SQ I would consider getting some different audio card, perhaps Asus's Xonar, and do all the DSP on software. If it doesn't provide any real-life improvements than I should probably just stick to the audigy+kx considering I get so much...

5. Any idea if it is now possible to do the following in kX:

• crossover slopes other than 12/24dB (i.e. 6dB, 18dB, 48dB)
• 30 band eq
• Speaker phasing

Thank you so much !!!

Maddogg6

I presume 'Playback' - the status screen shows the inputs in another section above. The P16V has a recording and playback device, so..

Yes you set the bit depth in software.
How can you tell??
1) Without external equipment, you cant be certain. Trust the software I suppose - I guess it comes down to this....
2) if you cant hear the difference - why is all this important to you? I dunno - test, listen, compare. Does it make any difference?

I mostly hear a difference between 48Khz and 96Khz by recordings made in the corresponding resolutions - I dont really hear a difference from upsampling -that's mostly the difference between a DSP config and Wave HQ (and no DSP processing at all).
Well there are 2 things that happen, lost fidelity when sampling - and the inherant sample value error because digital by nature has finite limits while analogs nature is infinite. This algo sounds like it addresses the latter. But its still based on math, not on actual original audio information - how could it be?, thus, there is still plenty room for error. But I am far from an expert - if it sounds, good use it.

I dont really know - so far the different interpolation algos I have heard made little to know difference with my equipment. Its not like I have reference amps/speakers/headphones or anything, normally you need such things to hear the difference.

Learning how to program ??.
Speaker phasing, erm, unless its rough phase, I presume you mean phase accross the band width, I have a feeling is not possible with the DSP resources, especially if using the cross overs and such. I think its the nature of these DSPs with filtering that undesired phase changes occur.

Chester01

as far as 'other' order eq's, a l-r 24db/oct filter is actually 2 butterworth filters cascaded together... so if you want to make a 48db/oct filter just cascade them (run them in series)

30 band eq... use parametrics; they are much more precise you can adjust the 'width', gain and peaking frequency. you can use win-isd pro to model the filters and 'see whats going on' (change the graph ranges from 10-22000 hz though and be sure to go to the transfer function filter page)

as for phase: yea its a bitch, I have been trying to figure out a system that has no phase artifacts, you will pretty much HAVE to use software processing to do this, look at Aixcoustic creations - THE HOME OF ELECTRI-Q and ELECTRI-Q (posihfopit) and get the Electri-Q plugin, then use "George Yohng's VST Wrapper for Foobar player Version 1.1," you can then take those parametric filters you make in win ISD and process them with no phase errors (use the 'linear phase' processing method). the only problem is that any linear phase or adjusted phase filter there will be a slight delay (generally not a big deal) but on switching songs in foobar there is a blip of the current song played (but advanced a sec or two from where you were listening) and then the new song starts

reguarding software resampling; I am not sure how the hardware resamplers in the creative cards compare but you can look here : SRC Comparisons to see most if not all software methods comapred. I use SoX 14.2 for foobar2000 (search in the foobar forums); this has a linear phase method and with the best modes selected (read the help file) i believe it achieves 99% accuracy.

__________________
COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium)
STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands...

dj_stick

don't you mean in series?

__________________

Chester01

yea, my bad; fixed it tho

__________________
COMP: Dual Intel PIII 733 Mhz; GeForce 4 Ti 4600; 1.128 Gb RAM; SB0350 (Audigy 2 ZS Platnium)
STEREO(I UPGRADED):Crown Audio K1 and K2 amplifiers (4000 watts at .1% THD ), JL 13w7 Subwoofer (6.5 CF) (2) 18" PR's, Klipsch SB-1's, some cement blocks for speaker stands...