From: "Earl F. Glynn"
Subject: pf32bit enigma:  "alpha byte" indeterminate in new bitmap?
Date: 10 Apr 2000 00:00:00 GMT
Message-ID: <8ct9hb$38011@bornews.borland.com>
X-Priority: 3
X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2919.6600
Organization: efg's Computer Lab
X-MSMail-Priority: Normal
Reply-To: "Earl F. Glynn"
Newsgroups: borland.public.delphi.graphics

// Why does one Win 98 Pentium report
//   FFFFFFFF -1 NAN
// while two other Pentiums (one Win 95, one Win 98) report
//    1FFFFFF 33554431 0.00
// in the example below?

procedure TForm1.Button1Click(Sender: TObject);
  TYPE
    TRowSingle = ARRAY[WORD] OF Single;
    pRowSingle = ^TRowSingle;
    TRGBQuadArray = ARRAY[WORD] OF DWORD;  // could use TRGBQuad
    pRGBQuadArray = ^TRGBQuadArray;

  VAR
    Bitmap   :  TBitmap;
    RowSingle:  pRowSingle;
    RowDWORD :  pRGBQuadArray;

begin
  Bitmap := TBitmap.Create;
  TRY
    Bitmap.Width  := 100;
    Bitmap.Height := 100;
    Bitmap.PixelFormat := pf32bit;

    RowSingle := Bitmap.ScanLine[0];
    RowDWORD  := Bitmap.ScanLine[0];

    ShowMessage( Format('%8x %d %f',
                 [RowDWORD[0], RowDWORD[0], RowSingle[0]] ));

    Image1.Picture.Graphic := Bitmap;
  FINALLY
    Bitmap.Free
  END
end;


Both values are equivalent "white" colors, but why should the
values be different in the "alpha" byte?

I'm using a pf32bit bitmap to store a matrix of Single values,
and got bit by the NaN initialization on one machine.

--
efg

Earl F. Glynn
Overland Park, KS  USA

efg's Computer Lab:  http://www.efg2.com/Lab