Je vous remets le texte de Sheffield Lab qui apparait de travers en pdf, et je me suis trompé sur le découpage de track, je voulais évoquer la perception différente du taux de détection entre un signal sur fréquence (simplifié) et un signal musique (complexe)
Sheffield Lab
Track 71. This track tests your system for radiated distortion. Sinusoidal wave of the so-called Pretsizionnoy Audiosistemy One (Audio Precision System One), and processed a special device which enables to regulate the required amount of distortion. The start signal was distortion 0.0004%. He has been processed and recorded on a track 71, as a score for the subsequent tracks. The track has the tone distortion 0,03%. In each subsequent track distortion is increased by about 10 dB. If you have a distortion analyzer, you can test the calibration, since the value of distortions in the following tracks very accurately. Naturally, if you have a "noisey" CD player or poor D-to-A converter, the coefficient of nonlinear distortion THD + N and influence, reducing the testimony of dough. If you do not have the distortion analyzer, you can use these tracks, as a test of "masking". The principle of "obscuring" the next. Beginning with low levels of distortion and increasing until it becomes audible, you get testimony threshold distorting your system or your ability to feel it. Start with a track 71. If you are unsure of what seems a distortion, immediately go to the track 76, because it has 10% distortion to be clearly heard any audience, even on a bad audio. The distortion is a weak sales, mixed in with clean tone. If you can not notice a difference between any of these tracks, the distortion of your system is very high (> 10%) or just poorly hear. Most people can easily hear the distortion of 1% to a good system to run in waves.
Track 72. Same as track 71, only with the distortion of 0,1%.
Track 73. Same as track 71, only with the distortion of 0,3%.
Track 74. Same as track 71, only with the distortion of 1%.
Track 75. Same as track 71, only with the distortion of 3%.
Track 76. Same as track 71, only with the distortion of 10%.
Track 77. This test, like the previous test, except that the signal is used as a piece of music. Source distortions affecting all frequencies equally and does not affect the standard. Therefore, the test is more informative than the tests distortions which depend on the frequency or level. We believe that the musical material with a distortion of 3%, some listeners will be able to hear the sound system, but hardly anyone will be able to detect a distortion of 1%. Of course, a distortion of 10% should be obvious to all systems.
Track 78. Same as track 77, only with the distortion of 0,1%.
Track 79. Same as track 77, only with the distortion of 0,3%.
Track 80. Same as track 77, only with the distortion of 1%.
Track 81. Same as track 77, only with the distortion of 3%.
Track 82. Same as track 77, only with the distortion of 10%.
Pas besoin de faire des ronds de jambe
comme je vous l'ai signalé (références générales) et Jean en a fait la distinction (mémoire),
la détection de distortion se fait, en situation réelle, la plupart du temps sans référence à "un original" et c'est ce que Audio Pavillon voulait simplement montrer... que cela est très difficile sans l'original... tout simple.
Sans référent, jusqu'à 10% c'est entendable mais incertain
Avec référent original, comme Sheffield Lab le précise, on devrait pouvoir détecter le 1% suivant le systéme+oreille
et on est plus discriminant sur signal pur
cela dépend aussi du type de distortion, zone de sensibilité...
...mais bien incapable de l'affirmer, et encore moins de quantifier.
![[Image: lnqm.png]](https://zupimages.net/viewer.php?id=23/30/lnqm.png)