Chris@72: 
Chris@72: Thinkpad T540p i5-4330M @2.80GHz with 16GB RAM, plugged in
Chris@72: Arch Linux, gcc 4.8.2
Chris@72: Using sonic-annotator v1.0 (commit:41c4de1e05d8), release build
Chris@72: 
Chris@72: Debug flags: -g -fPIC
Chris@72: Release flags: -O3 -ffast-math -msse -mfpmath=sse -ftree-vectorize -fPIC
Chris@72: 
Chris@73: Release flags for qm-dsp also include -fomit-frame-pointer
Chris@73: 
Chris@73: The input file is 1-channel 16-bit PCM at 44100Hz, duration 0m43.5s.
Chris@72: 
Chris@72: 
Chris@72: DEBUG/RELEASE:
Chris@72: 
Chris@72: commit:ce64d11ef336, release build of Silvet, release build of qm-dsp
Chris@72: 
Chris@73: real	1m44.456s
Chris@73: user	1m44.343s
Chris@73: sys	0m0.210s
Chris@72: 
Chris@72: commit:ce64d11ef336, debug build of Silvet, release build of qm-dsp
Chris@72: 
Chris@73: real	14m16.124s
Chris@73: user	14m16.907s
Chris@73: sys	0m0.217s
Chris@72: 
Chris@72: commit:ce64d11ef336, release build of Silvet, debug build of qm-dsp
Chris@72: 
Chris@73: real	1m55.204s
Chris@73: user	1m55.053s
Chris@73: sys	0m0.253s
Chris@72: 
Chris@72: Subsequent tests use release builds of both.
Chris@72: 
Chris@72: 
Chris@73: VAMP FEATURE SUPPRESSION:
Chris@73: 
Chris@75: commit:7133f78ccbf6, as commit:ce64d11ef336 but with CQ output feature
Chris@75: return commented out
Chris@75: 
Chris@75: real	1m46.162s
Chris@75: user	1m46.093s
Chris@75: sys	0m0.157s
Chris@75: 
Chris@77: commit:78a7bf247016, as commit:ce64d11ef336 but with CQ output and FCQ
Chris@75: output feature return commented out
Chris@77: 
Chris@77: real	1m45.206s
Chris@77: user	1m45.153s
Chris@77: sys	0m0.147s
Chris@77: 
Chris@77: conclusion: no advantage in removing these
Chris@77: 
Chris@77: 
Chris@77: DEBUG PRINTOUTS:
Chris@77: 
Chris@77: commit:f3bf6503e6c6, as commit:ce64d11ef336 but with debug printouts
Chris@77: removed
Chris@77: 
Chris@77: real	1m43.744s
Chris@77: user	1m43.657s
Chris@77: sys	0m0.203s
Chris@77: 
Chris@81: conclusion: obviously we want to remove these eventually, but might as
Chris@77: well keep in during testing
Chris@77: 
Chris@80: 
Chris@80: EM ITERATIONS:
Chris@80: 
Chris@80: commit:5314d3361dfb, as commit:ce64d11ef336 but with only 6 EM
Chris@80: iterations instead of 12
Chris@80: 
Chris@80: real	0m59.055s
Chris@80: user	0m58.897s
Chris@80: sys	0m0.193s
Chris@80: 
Chris@80: conclusion: EM dominates the time taken, not CQ or note forming
Chris@80: 
Chris@81: 
Chris@81: CQ DECIMATOR CONFIGURATION:
Chris@81: 
Chris@81: Uncommitted revision (because changes are in CQ subrepo) that is as
Chris@81: commit:ce64d11ef336 but with resampler SNR=30 and BW=0.04 instead of
Chris@81: SNR=60 and BW=0.02
Chris@81: 
Chris@81: real	1m43.176s
Chris@81: user	1m43.067s
Chris@81: sys	0m0.190s
Chris@81: 
Chris@81: conclusion: supports the previous test
Chris@81: 
Chris@84: 
Chris@84: EM TWEAKS:
Chris@84: 
Chris@84: commit:a0dedcbfa628, as commit:ce64d11ef336 but with variables hoisted
Chris@84: out of loops and consts added wherever applicable
Chris@84: 
Chris@84: real	1m44.548s
Chris@84: user	1m44.460s
Chris@84: sys	0m0.183s
Chris@84: 
Chris@84: conclusion: compiler already knows this stuff
Chris@84: 
Chris@86: commit:64b08cc12da0, as commit:ce64d11ef336 but with loops merged so
Chris@86: as theoretically to reduce intermediate calculations
Chris@86: 
Chris@86: real	3m46.969s
Chris@86: user	3m46.850s
Chris@86: sys	0m0.220s
Chris@86: 
Chris@87: commit:6075e92d63ab, as commit:64b08cc12da0 but with innermost loop
Chris@87: reverted to three loops with simple bodies instead of one with a more
Chris@87: complex body
Chris@87: 
Chris@87: real	1m44.767s
Chris@87: user	1m44.490s
Chris@87: sys	0m0.190s
Chris@88: 
Chris@89: commit:97b77e7cb94c, as commit:6075e92d63ab but with templates stored
Chris@89: as doubles instead of floats (doubling the size of the plugin binary)
Chris@89: 
Chris@89: real	1m40.135s
Chris@89: user	1m39.820s
Chris@89: sys	0m0.230s
Chris@89: