Recent posts
https://software.intel.com/en-us/recent/544829
en"ippsFIR_32f_I" returns negative values, although neither taps nor the input are negative
https://software.intel.com/en-us/forums/intel-integrated-performance-primitives/topic/281875
<p>Hello,</p>
<p> I recognize negative values after filtering with "ippsFIR_32f_I", although neither my taps (feed forward coefficients) nor the input vector are negative.<br /> How is this possible?</p>
<p> Here is my code snippet, which reproduces this strange behaviour. <br /> Note: If changing "iNumIters" to a smaller value, this behaviour does not occur anymore.</p>
<pre>[cpp] // initializations and declarations
IppStatus IppStatus;
IppsFIRState_32f* pState;
int NFilt = 4096;
std::vector b(NFilt, 1.0f/NFilt); // Feed forward coefficients (taps)
std::vector bCheck(NFilt, 0.0f); // vector to check coefficients
std::vector delayLine(NFilt,0.0f); // initial condition
std::vector delayLineCheck(NFilt,0.0f); // vector to check final condition
std::vector x(NFilt+1024, 0.0f); // input signal
for (int k=2; k <br />IPP Version: 6.0.0.61<br />I would be very thankful for any help.<br />Sincerely yours</pre>Fri, 23 Sep 11 06:50:46 -0700poozir281875Behavior of ippsFIR_32f
https://software.intel.com/en-us/forums/intel-integrated-performance-primitives/topic/307405
<p>We have examined the performance of the ippFIR_32f implementation of the FIR filter and would like to confirm our understanding of its implementation. </p>
<p>First, we have noticed that when used as a single rate filter, the computational cost (number of clock cycles needed to process an input data sample)tends toriseas the number of filter coefficients isincreased, but drops significantly each time the number of coefficientsreaches a power of two. Second, when used as a multirate filter, we find that the computational cost is a roughly linear function of the number of coefficients, except for a real filter which shows a jump in computational cost when the number of coefficients is about 3x the decimation factor.</p>
<p>My inference is that the single rate filter is based on a frequency domain filter implementation using FFTs and that the multirate implementation is based on a polyphase filter architecure (i.e., a time domain approach).</p>
<p>Is this correct?</p>
<p>Thanks<br />
Robert</p>
Mon, 27 Feb 06 22:00:47 -0800robert.inkol@drdc-rddc.gc.ca307405