
---------------------------------------------------
IEEE-1180 test, for [fdctmm32.c]  'AP922 MMX FDCT'
---------------------------------------------------
IEEE test conditions: -L = -256, +H = 255, sign = 1, #iters = 10000
Peak absolute values of errors:
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   2   1   1   1   1   1
   1   1   1   1   1   1   1   1
Worst peak error = 2  (FAILS spec limit 1)

Mean square errors:
   0.0720   0.0743   0.0717   0.0773   0.0724   0.0720   0.0694   0.0737
   0.0710   0.0698   0.0687   0.0739   0.0677   0.0709   0.0694   0.0704
   0.0737   0.0695   0.0697   0.0662   0.0644   0.0709   0.0748   0.0707
   0.0666   0.0603   0.0623   0.0705   0.0683   0.0652   0.0700   0.0648
   0.0657   0.0685   0.0663   0.0665   0.0683   0.0627   0.0639   0.0694
   0.0636   0.0674   0.0690   0.0666   0.0688   0.0664   0.0691   0.0678
   0.0663   0.0706   0.0720   0.0733   0.0703   0.0638   0.0735   0.0661
   0.0814   0.0729   0.0712   0.0773   0.0774   0.0722   0.0679   0.0774
Worst pmse = 0.081400  (FAILS spec limit 0.06)
Overall mse = 0.069627  (FAILS spec limit 0.02)

Mean errors:
   0.0214  -0.0009   0.0037   0.0217   0.0192  -0.0002   0.0012   0.0205
  -0.0056  -0.0022  -0.0003   0.0003   0.0017  -0.0011  -0.0050  -0.0008
  -0.0055  -0.0031  -0.0079  -0.0002  -0.0062   0.0007  -0.0030  -0.0095
   0.0160  -0.0057   0.0009   0.0161   0.0223  -0.0024  -0.0084   0.0162
   0.0233   0.0007   0.0037   0.0203   0.0263  -0.0007  -0.0001   0.0256
   0.0046   0.0068   0.0018   0.0030   0.0054   0.0000   0.0053  -0.0002
  -0.0017   0.0026   0.0038   0.0021  -0.0001   0.0024   0.0005   0.0027
   0.0242   0.0033  -0.0002   0.0209   0.0226   0.0024   0.0037   0.0210
Worst mean error = 0.026300  (FAILS spec limit 0.015)
Overall mean error = 0.005155  (FAILS spec limit 0.0015)

0 elements of IDCT(0) were not zero



---------------------------------------------------
IEEE-1180 test, for [fdctmm32.c]  'AP922 MMX FDCT'
---------------------------------------------------
IEEE test conditions: -L = -5, +H = 5, sign = 1, #iters = 10000
Peak absolute values of errors:
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
Worst peak error = 1  (meets spec limit 1)

Mean square errors:
   0.0714   0.0730   0.0705   0.0759   0.0747   0.0672   0.0674   0.0780
   0.0671   0.0715   0.0723   0.0721   0.0715   0.0717   0.0659   0.0683
   0.0650   0.0744   0.0672   0.0678   0.0669   0.0718   0.0694   0.0679
   0.0655   0.0630   0.0642   0.0698   0.0713   0.0660   0.0619   0.0698
   0.0693   0.0643   0.0639   0.0674   0.0666   0.0599   0.0618   0.0659
   0.0741   0.0724   0.0659   0.0669   0.0655   0.0671   0.0669   0.0679
   0.0706   0.0684   0.0691   0.0687   0.0679   0.0675   0.0682   0.0666
   0.0780   0.0754   0.0711   0.0760   0.0750   0.0740   0.0725   0.0777
Worst pmse = 0.078000  (FAILS spec limit 0.06)
Overall mse = 0.069264  (FAILS spec limit 0.02)

Mean errors:
   0.0210   0.0016  -0.0001   0.0245   0.0231  -0.0006   0.0034   0.0224
   0.0019  -0.0009  -0.0027   0.0021  -0.0009   0.0007   0.0009   0.0019
  -0.0038  -0.0034  -0.0060   0.0014  -0.0021  -0.0046  -0.0026  -0.0019
   0.0139   0.0020  -0.0026   0.0156   0.0177  -0.0062  -0.0037   0.0156
   0.0207   0.0019  -0.0027   0.0218   0.0192  -0.0003   0.0024   0.0169
   0.0027   0.0038   0.0007   0.0017   0.0059   0.0037   0.0045   0.0027
  -0.0012   0.0012  -0.0045  -0.0017  -0.0005  -0.0017  -0.0034   0.0020
   0.0222   0.0036   0.0033   0.0220   0.0198  -0.0004  -0.0019   0.0241
Worst mean error = 0.024500  (FAILS spec limit 0.015)
Overall mean error = 0.004939  (FAILS spec limit 0.0015)

0 elements of IDCT(0) were not zero



---------------------------------------------------
IEEE-1180 test, for [fdctmm32.c]  'AP922 MMX FDCT'
---------------------------------------------------
IEEE test conditions: -L = -300, +H = 300, sign = 1, #iters = 10000
Peak absolute values of errors:
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
Worst peak error = 1  (meets spec limit 1)

Mean square errors:
   0.0643   0.0595   0.0618   0.0676   0.0656   0.0603   0.0650   0.0610
   0.0592   0.0600   0.0607   0.0555   0.0612   0.0601   0.0639   0.0579
   0.0561   0.0559   0.0605   0.0549   0.0606   0.0553   0.0598   0.0605
   0.0543   0.0528   0.0550   0.0581   0.0618   0.0545   0.0521   0.0557
   0.0574   0.0542   0.0531   0.0574   0.0588   0.0550   0.0574   0.0576
   0.0562   0.0618   0.0559   0.0606   0.0558   0.0597   0.0591   0.0603
   0.0553   0.0596   0.0599   0.0621   0.0611   0.0545   0.0629   0.0590
   0.0661   0.0637   0.0614   0.0648   0.0629   0.0628   0.0643   0.0574
Worst pmse = 0.067600  (FAILS spec limit 0.06)
Overall mse = 0.059213  (FAILS spec limit 0.02)

Mean errors:
   0.0167   0.0029   0.0030   0.0226   0.0170   0.0025   0.0010   0.0192
  -0.0040   0.0012   0.0027   0.0023  -0.0016  -0.0049   0.0011  -0.0041
  -0.0035  -0.0017  -0.0017  -0.0035  -0.0006  -0.0005  -0.0030  -0.0023
   0.0133  -0.0012  -0.0038   0.0177   0.0196  -0.0053  -0.0027   0.0183
   0.0156   0.0000   0.0043   0.0184   0.0174   0.0030   0.0042   0.0170
   0.0036   0.0070   0.0055   0.0020   0.0052   0.0017   0.0063   0.0015
   0.0011   0.0006  -0.0005   0.0021  -0.0009   0.0029   0.0035   0.0026
   0.0131  -0.0023  -0.0026   0.0156   0.0125   0.0004   0.0017   0.0148
Worst mean error = 0.022600  (FAILS spec limit 0.015)
Overall mean error = 0.004594  (FAILS spec limit 0.0015)

0 elements of IDCT(0) were not zero



---------------------------------------------------
IEEE-1180 test, for [fdctmm32.c]  'AP922 MMX FDCT'
---------------------------------------------------
IEEE test conditions: -L = -256, +H = 255, sign = -1, #iters = 10000
Peak absolute values of errors:
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
Worst peak error = 1  (meets spec limit 1)

Mean square errors:
   0.0703   0.0736   0.0720   0.0739   0.0751   0.0690   0.0716   0.0763
   0.0701   0.0683   0.0683   0.0708   0.0669   0.0709   0.0715   0.0692
   0.0731   0.0688   0.0704   0.0686   0.0667   0.0689   0.0728   0.0684
   0.0664   0.0648   0.0651   0.0666   0.0667   0.0656   0.0692   0.0667
   0.0695   0.0686   0.0650   0.0640   0.0690   0.0620   0.0633   0.0647
   0.0655   0.0701   0.0713   0.0692   0.0690   0.0687   0.0701   0.0679
   0.0656   0.0709   0.0735   0.0721   0.0670   0.0635   0.0715   0.0639
   0.0745   0.0741   0.0703   0.0767   0.0725   0.0696   0.0694   0.0733
Worst pmse = 0.076700  (FAILS spec limit 0.06)
Overall mse = 0.069264  (FAILS spec limit 0.02)

Mean errors:
   0.0159   0.0032   0.0000   0.0185   0.0211  -0.0008  -0.0008   0.0217
  -0.0011   0.0001  -0.0031  -0.0060  -0.0035   0.0025   0.0021  -0.0024
  -0.0009  -0.0034   0.0002  -0.0044  -0.0003  -0.0051  -0.0060  -0.0038
   0.0210  -0.0014   0.0007   0.0216   0.0151   0.0004   0.0026   0.0193
   0.0277   0.0028   0.0020   0.0246   0.0240   0.0024   0.0041   0.0173
   0.0041   0.0025   0.0035   0.0028   0.0038   0.0031   0.0025   0.0077
   0.0076   0.0003   0.0017   0.0045   0.0022   0.0017   0.0059   0.0001
   0.0179  -0.0045  -0.0031   0.0209   0.0147  -0.0014  -0.0028   0.0173
Worst mean error = 0.027700  (FAILS spec limit 0.015)
Overall mean error = 0.005327  (FAILS spec limit 0.0015)

0 elements of IDCT(0) were not zero



---------------------------------------------------
IEEE-1180 test, for [fdctmm32.c]  'AP922 MMX FDCT'
---------------------------------------------------
IEEE test conditions: -L = -5, +H = 5, sign = -1, #iters = 10000
Peak absolute values of errors:
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
Worst peak error = 1  (meets spec limit 1)

Mean square errors:
   0.0734   0.0736   0.0758   0.0802   0.0784   0.0668   0.0672   0.0749
   0.0681   0.0710   0.0696   0.0721   0.0681   0.0704   0.0673   0.0684
   0.0672   0.0744   0.0691   0.0696   0.0683   0.0687   0.0699   0.0673
   0.0707   0.0628   0.0623   0.0654   0.0677   0.0632   0.0624   0.0648
   0.0690   0.0644   0.0686   0.0681   0.0695   0.0641   0.0635   0.0693
   0.0727   0.0666   0.0688   0.0701   0.0667   0.0653   0.0688   0.0688
   0.0694   0.0722   0.0714   0.0660   0.0629   0.0719   0.0711   0.0674
   0.0768   0.0715   0.0664   0.0763   0.0761   0.0739   0.0723   0.0722
Worst pmse = 0.080200  (FAILS spec limit 0.06)
Overall mse = 0.069394  (FAILS spec limit 0.02)

Mean errors:
   0.0176   0.0024   0.0054   0.0176   0.0224   0.0012  -0.0012   0.0213
  -0.0001  -0.0028   0.0016  -0.0009  -0.0009  -0.0016  -0.0021  -0.0006
   0.0012   0.0010  -0.0013  -0.0020  -0.0011   0.0001   0.0011  -0.0019
   0.0195  -0.0038  -0.0001   0.0168   0.0173  -0.0038  -0.0004   0.0160
   0.0198  -0.0030   0.0044   0.0191   0.0181   0.0047  -0.0021   0.0207
   0.0019   0.0038   0.0048   0.0059  -0.0009   0.0013  -0.0006   0.0034
   0.0016  -0.0038   0.0028  -0.0012   0.0005  -0.0005   0.0001  -0.0020
   0.0202  -0.0045   0.0000   0.0195   0.0177   0.0035   0.0019   0.0150
Worst mean error = 0.022400  (FAILS spec limit 0.015)
Overall mean error = 0.004844  (FAILS spec limit 0.0015)

0 elements of IDCT(0) were not zero



---------------------------------------------------
IEEE-1180 test, for [fdctmm32.c]  'AP922 MMX FDCT'
---------------------------------------------------
IEEE test conditions: -L = -300, +H = 300, sign = -1, #iters = 10000
Peak absolute values of errors:
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
   1   1   1   1   1   1   1   1
Worst peak error = 1  (meets spec limit 1)

Mean square errors:
   0.0623   0.0587   0.0623   0.0640   0.0645   0.0604   0.0636   0.0615
   0.0607   0.0622   0.0608   0.0558   0.0590   0.0617   0.0625   0.0591
   0.0551   0.0536   0.0644   0.0536   0.0569   0.0594   0.0604   0.0647
   0.0560   0.0519   0.0557   0.0546   0.0604   0.0554   0.0556   0.0540
   0.0564   0.0532   0.0539   0.0580   0.0571   0.0542   0.0572   0.0589
   0.0578   0.0602   0.0582   0.0590   0.0566   0.0599   0.0590   0.0596
   0.0582   0.0572   0.0582   0.0636   0.0613   0.0569   0.0628   0.0585
   0.0627   0.0626   0.0607   0.0661   0.0598   0.0618   0.0610   0.0601
Worst pmse = 0.066100  (FAILS spec limit 0.06)
Overall mse = 0.059086  (FAILS spec limit 0.02)

Mean errors:
   0.0161  -0.0011  -0.0013   0.0146   0.0169  -0.0032  -0.0012   0.0157
   0.0005  -0.0040  -0.0056  -0.0040  -0.0040  -0.0021  -0.0039   0.0009
  -0.0021  -0.0026  -0.0040  -0.0020  -0.0079  -0.0042  -0.0010  -0.0083
   0.0188   0.0001  -0.0003   0.0102   0.0134   0.0006  -0.0022   0.0130
   0.0176   0.0032   0.0019   0.0182   0.0197   0.0006  -0.0006   0.0169
   0.0008   0.0000   0.0044   0.0048   0.0016   0.0055   0.0012   0.0070
   0.0012   0.0020   0.0040   0.0016   0.0045   0.0003   0.0004   0.0041
   0.0193  -0.0006  -0.0003   0.0155   0.0168  -0.0016  -0.0004   0.0167
Worst mean error = 0.019700  (FAILS spec limit 0.015)
Overall mean error = 0.003783  (FAILS spec limit 0.0015)

0 elements of IDCT(0) were not zero


Profile: Function timing, sorted by time
Date:    Sat Jul 22 21:22:31 2000


Program Statistics
------------------
    Command line at 2000 Jul 22 21:22: "C:\work\proj\fdct1180\Debug\ieee1180" -256 255 1 10000
    Total time: 9061.329 millisecond
    Time outside of functions: 5.598 millisecond
    Call depth: 2
    Total functions: 26
    Total hits: 680010
    Function coverage: 38.5%
    Overhead Calculated 6
    Overhead Average 6

Module Statistics for ieee1180.exe
----------------------------------
    Time in module: 9055.731 millisecond
    Percent of time in module: 100.0%
    Functions in module: 26
    Hits in module: 680010
    Module function coverage: 38.5%

        Func          Func+Child           Hit
        Time   %         Time      %      Count  Function
---------------------------------------------------------
    5437.130  60.0     5437.130  60.0    20000 _ref_idct (ieeetest.obj)
    2703.203  29.9     2703.203  29.9    10000 _ref_fdct (ieeetest.obj)
     519.943   5.7      519.943   5.7   640000 _ieeerand (ieeetest.obj)
     374.948   4.1     9055.731 100.0        1 _main (ieeetest.obj)
      20.274   0.2       20.274   0.2    10000 _fdct_mm32 (fdctmm32.obj)
       0.179   0.0        0.179   0.0        1 _init_fdct_3dnow (fdct3dn.obj)
       0.026   0.0        0.026   0.0        1 _ref_dct_init (ieeetest.obj)
       0.011   0.0        0.011   0.0        1 _init_fdctdaan (fdctref.obj)
       0.010   0.0        0.010   0.0        1 _intfdct (fdctint.obj)
       0.008   0.0        0.008   0.0        5 _meets (ieeetest.obj)