<div>Hi,</div>
<div> </div>
<div>I have some questions about exact semantics of anti-alias.</div>
<div> </div>
<div>General anti-alias steps:</div>
<div>1. calculate coverage of a pixel</div>
<div>2. apply coverage to source pixel</div>
<div>3. composite with destination</div>
<div> </div>
<div>So my question is</div>
<div> </div>
<div>In step 2, how are the coverages applied to source pixels?</div>
<div>If the source pixel has no alpha (something like rgb_24), does cairo treat it same with rgba_32 with 1.0 alpha?</div>
<div>Coverage will be multiplied to source alpha channel as well as source rgb color channels to keep them in premultiplied form, right?</div>
<div> </div>
<div>In step 3, if the operator is SRC and the destination format is rgb_24, what about source pixels with coverage less than 1.0?</div>
<div>They are blended or not?</div>
<div>I think this is the same problem with compositing rgba_32(premultiplied) to rgb_24 with operator SRC.</div>
<div>Are they converted into unpremultiplied form or just copyed rgb to rgb?</div>
<div> </div>
<div>I think just turning on anti-alias should not cause huge visual difference. (that's just a hint as written in cairo API reference manual)</div>
<div> </div>
<div>Thanks in advance.</div>
<div><br clear="all">-- <br>Best Regards,</div>
<div>Taekyun Kim</div><br>