Minor issue in DxPlay sample

There is, I think, a minor bug in the DxPlay sample (in capture.cs).
In the function IPToBmp() you (re-)calculate the buffer size for the bitmap data through the formula:

int iBufSize = m_videoWidth * m_videoHeight * 3;

Now, AFAIK in Windows bitmaps each indiviual row is DWORD-aligned (i.e.the rows are padded, if necessary, so that the stride is always a multiple of 4).
So, in a 24bpp bitmap the stride equals (<width>3) only, if <width> is a multiple of 4. ( In general, it is: ((((<width></width></width><bpp>) + 31) & ~31) >> 3) )
Yet not all codecs require the frame width to be a multiple of 4. If it isn't, an access violation (due to the incorrect value for <scan0> parameter in the Bitmap constructor) occurs upon trying to save the bitmap. (You can verify this by cropping a video appropriately and saving it e.g. as Uncompressed RGB.)</scan0></bpp></width>

Why not simple use the stored value for the size of the bitmap data <m_imagesize> obtained in SaveSizeInfo()?</m_imagesize>

Regards

uli

(I would like to conclude - I should have started with this - by expressing my most heartfelt gratutude for all your efforts in this immensely valuable project!)