what happened when CImage load a 16-bit BMP file

[fitatc]

I am try to deal with 16-bit bmp file with CImage class. but when I load a 16-bit bmp file with CImage::load(LPCTSTR pszFileName) and get bits-per-pixel value by CImage::GetBPP, the return is 32:confused:.

CImage imgOriginal;
hResult = imgOriginal.Load(m_FileName);
if(FAILED(hResult)){
// hanle error
}

int iBpp = imgOriginal.GetBPP();

I am using VS2008 in Windows Xp. I can not find any comment about this from msdn, Could anyone tell me what happened in CImage::Load? for a 16-bit bmp sample file, see : http://wvnvaxa.wvnet.edu/vmswww/bmp.html[^]

[Mark Salsbery]

iBpp is 16 for me. Mark

Mark Salsbery Microsoft MVP - Visual C++ :java:

[fitatc]

Thanks very much for answer my qustion. 1.Did you test in windows xp? I tryed it again, and is still 32. 2.And I finded another stranged thing: If I right click that sample 16-bit bmp file to see it's property, It's 32 too. :omg: Maybe this two things is connected, And if your OS not show 32-bit, That could be my OS problem.

[Mark Salsbery]

fitatc wrote:

1.Did you test in windows xp? I tryed it again, and is still 32.

I first tested only on Vista Ultimate SP1. This morning I tested on XP Pro SP3...same result... 16bpp

fitatc wrote:

2.And I finded another stranged thing: If I right click that sample 16-bit bmp file to see it's property, It's 32 too.

Are you sure you downloaded the right file? I right-clicked the 16bit sample from the link and saved it to my harddrive to test. Mark

Mark Salsbery Microsoft MVP - Visual C++ :java:

[fitatc]

// --- file's property --- //

Mark Salsbery wrote:

Are you sure you downloaded the right file? I right-clicked the 16bit sample from the link and saved it to my harddrive to test.

Yes, I check it by Read it in Hex, it's 16.

42 4D 42 40 00 00 00 00 00 00 42 00 00 00 28
00 00 00 7F 00 00 00 40 00 00 00 01 00 10 00
03 00 00 00 00 40 00 00 00 00 00 00 00 00 00
...

but, In my OS, the file property is still 32. I think it's may be my OS's problem, XP-Professional Version 2002 Service Pack 2. And since I works in japan, my OS is janpanese version of WindowsXP. // --- CImage --- // Others in MSDN forum advised me not to use CImage, because it's for display. And I find that I can't get "the type of compression" by CImage class, :confused:So, is it possible that Cimage class change the "BI_BITFIELDS compressed 16-bit bmp" to "uncompressed 32-bit bmp" in CImage::load?

[Mark Salsbery]

fitatc wrote:

Others in MSDN forum advised me not to use CImage

CImage just wraps a DIBSection and uses GDI+ for loading and saving images. If that's what you need it can be used for anything DIBsection-related, not just for display.

fitatc wrote:

is it possible that Cimage class change the "BI_BITFIELDS compressed 16-bit bmp" to "uncompressed 32-bit bmp" in CImage::load?

It's possible, but I don't know why it would be changed. Run it in the debugger and step into the CImage::Load() code. I did, and it looks like it is possible (in CImage::CreateFromGdiplusBitmap()) to be converted to a 32bpp bitmap but that code is bypassed. This shouldn't be OS related - GDI+ didn't change for years before Vista. It could be the version of CImage you're using however. What Visual Studio version are you using? Step into the code - it doesn't change magically :) Mark

Mark Salsbery Microsoft MVP - Visual C++ :java:

[fitatc]

Mark Salsbery wrote:

It could be the version of CImage you're using however. What Visual Studio version are you using?

My visual Studio version is: 2008, Version 9.0.21022.0 RTM. I steped into the code, as your said, It seems like the point is in

Bitmap::Bitmap(
IN const WCHAR *filename,
IN BOOL useEmbeddedColorManagement
)
{
...
lastResult = DllExports::GdipCreateBitmapFromFile(filename, &bitmap);
}

;P but I am not very familiar with GDI+, so I determined to study it for my question. Thanks very much for your help, and I know how I can deal with the question like this now. :)

[Mark Salsbery]

fitatc wrote:

Bitmap::Bitmap

What is this? I thought you were using CImage? :)

Mark Salsbery Microsoft MVP - Visual C++ :java:

[fitatc]

Mark Salsbery wrote:

What is this? I thought you were using CImage?

Yes, I just steped into the CImage::Load() It's like below:

atlimage.h
inline HRESULT CImage::Load( LPCTSTR pszFileName ) throw()
{
if( !InitGDIPlus() )
{
return( E_FAIL );
}

Gdiplus::Bitmap bmSrc( (CT2W)pszFileName );   　　　----　★
if( bmSrc.GetLastStatus() != Gdiplus::Ok )
{
	return( E\_FAIL );
}

return( CreateFromGdiplusBitmap( bmSrc ) );

}

I steped into ★, because all the BMP data it's used below is from here, and I get:

inline
Bitmap::Bitmap(
IN const WCHAR *filename,
IN BOOL useEmbeddedColorManagement
)
{
GpBitmap *bitmap = NULL;

if(useEmbeddedColorManagement) 
{
    lastResult = DllExports::GdipCreateBitmapFromFileICM(filename, &bitmap);
}
else
{
    lastResult = DllExports::GdipCreateBitmapFromFile(filename, &bitmap);   ---　★★
}

SetNativeImage(bitmap);

}

:-D ★★：this is what I posted in last time, sorry for confusing message.

[Mark Salsbery]

Oh it was Gdiplus::Bitmap :) Anyway, CreateFromGdiplusBitmap() is the method you want to step into. That's the one that may change the format to 32bpp, but it doesn't for me. Mark

Mark Salsbery Microsoft MVP - Visual C++ :java:

[Mark Salsbery]

Also, if after stepping into the CImage loader code you find it does't do what you want, it's pretty easy to load BMP files yourself - much easier than other file formats with complicated compression. Mark

Mark Salsbery Microsoft MVP - Visual C++ :java: