Width and Height of the Image in RenderTargetBitma

Using the below code I'm drawing on DrawingVisual then rendering it to an Image using RenderTargetBitmap. The final Image is later added to a Canvas and displayed on the screen.

My problem is with the pixelWidth and pixelHeight arguments the RenderTargetBitmap method wants. What valued should I give to it? I have noticed that if I give it lower numbers parts of the image is not rendered. On what basis should I choose these? I have given it 1000 in the code below.

public class ModelBeamSectionNamesInPlan : Image
{
    private readonly VisualCollection _visuals;
    public ModelBeamSectionNamesInPlan(BaseWorkspace space)
    {
        var typeface = Settings.BeamTextTypeface;
        var cultureinfo = Settings.CultureInfo;
        var flowdirection = Settings.FlowDirection;
        var beamtextsize = Settings.BeamTextSize;
        var beamtextcolor = Settings.InPlanBeamTextColor;

        beamtextcolor.Freeze();
        const double scale = 0.6;

        var drawingVisual = new DrawingVisual();
        using (var dc = drawingVisual.RenderOpen())
        {
            foreach (var beam in Building.ModelBeamsInTheElevation)
            {
                var text = beam.Section.Id;
                var ft = new FormattedText(text, cultureinfo, flowdirection,
                                           typeface, beamtextsize, beamtextcolor,
                                           null, TextFormattingMode.Display)
                {
                    TextAlignment = TextAlignment.Center
                };

                // Draw Text
                dc.DrawText(ft, space.FlipYAxis(x, y));
            }
        }

        var bmp = new RenderTargetBitmap(1000, 1000, 120, 96, PixelFormats.Pbgra32);
        bmp.Render(drawingVisual);
        Source = bmp;
    }
}

You can query the DrawingVisual's ContentBounds property, which

gets the bounding box for the contents of the ContainerVisual

or the DescendantBounds property which

gets the union of all the content bounding boxes for all of the descendants of the ContainerVisual, but not including the contents of the ContainerVisual.

Something like this should work:

var bounds = drawingVisual.DescendantBounds;
var bmp = new RenderTargetBitmap(
    (int)Math.Ceiling(bounds.Width), (int)Math.Ceiling(bounds.Height),
    96, 96, PixelFormats.Pbgra32);