aboutsummaryrefslogtreecommitdiffstats
path: root/utils/bsincgen.c
Commit message (Collapse)AuthorAgeFilesLines
* Ensure some macros have the correct sizeChris Robinson2017-08-281-1/+4
|
* Add a higher quality bsinc resampler using 24 sample pointsChris Robinson2017-08-271-4/+7
| | | | | | | This improves the transition width, allowing more of the higher frequencies remain audible. It would be preferrable to have an upper limit of 32 points instead of 48, to reduce the overall table size and the CPU cost for down- sampling.
* Properly open the output file for writingChris Robinson2017-08-271-1/+1
|
* Use a common header for Unicode-awareness on WindowsChris Robinson2017-08-271-79/+1
|
* Allow specifying the output filename with bsincgenChris Robinson2017-08-261-31/+140
|
* Rename the bsinc resampler to bsinc12Chris Robinson2017-08-251-46/+48
|
* Parameterize the filter order and rejection in bsincgenChris Robinson2017-08-231-28/+29
|
* Parameterize the bsinc table nameChris Robinson2017-08-221-8/+9
|
* Properly postfix the filter order numberChris Robinson2017-08-211-2/+6
|
* Simplify bsincgen's Kaiser functionChris Robinson2017-08-181-8/+3
|
* Keep bsinc filter quality more consistent between scalesChris Robinson2017-08-171-7/+9
| | | | | | | This generates the filters using the proper size and scale. The 'a' divisor should represent the +/- sample range (and thus be a whole number), with the number of sample points being double that. Increasing the filter size to a multiple of 4 (for SIMD) can be done by padding in 0s afterward.
* Workaround Android not having log2Chris Robinson2017-08-171-0/+4
|
* Correct the bsinc filter orderChris Robinson2017-08-171-6/+7
| | | | | | Despite the claim that it was an 11th order filter, the transition width was generated by specifying 12th order. A 12th order filter would need 14 sample points rather than the 12 it had.
* Make the sinc4 table staticChris Robinson2017-08-161-1/+1
|
* Make the bsinc table layout more efficientChris Robinson2017-08-161-99/+20
| | | | | | | | | | | | | | | | The old layout separated filters, scale deltas, phase deltas, and scale phase deltas into separate segments that each contained a numbers of scale and phase entries, Since processing a sample needed a filter and one of each delta entry relating to a particular scale and phase, the memory needed would be spread across the whole table. And since subsequent samples would use a different phase, it would jump around the table a whole lot as well. The new layout packs the data in a way more consistent with its use. The filters, scale deltas, phase deltas, and scale phase deltas are interleaved, such that for a particular scale and phase, the filter and delta entries used are contiguous. And the phase entries for a particular scale are kept together, so the ~500 to ~1000 samples processed per source update stay within the same 3KB to 6KB area of the 70+KB table, which is much more cache friendly.
* Keep bsinc info together in a structChris Robinson2017-08-151-37/+61
|
* Pre-compute the sinc4 resampler coefficient tableChris Robinson2017-04-081-1/+39
|
* Add a tool to generate the bsinc tablesChris Robinson2015-11-101-0/+374