The Nearest-Neighbor algorithm is the fastest of the three, but produces relatively low-quality thumbnails. The Bilinear algorithm offers a much better quality but is about twice as slow as Nearest-Neighbor. And finally, the Bicubic algorithm provides the highest quality but is approximately twice as slow as Bilinear and, therefore, 4 times as slow as Nearest-Neighbor.

A resizing algorithm for your thumbnails can be specified via the property Jpeg.Interpolation. Valid values for this property are: 0 (Nearest-Neighbor), 1 (Bilinear), and 2 (Bicubic). The default value is 1 (Bilinear). For this property to take effect, it must be set before calling Save, SendBinary or Binary.

The following table illustrates the effect of setting the Interpolation property. Notice how much smoother the thumbnails produced by the Bilinear and Bicubic algorithms are in comparison to Nearest-Neighbor. Notice also that in many cases (such as this one) Bicubic provides little, if any, improvement over Bilinear.

No sharpening	Sharpen(1, 120)	Sharpen(1, 250)

The Sharpen method uses two Double arguments: Radius and Amount. Radius controls the size (in pixels) of an area around every pixel that the sharpening algorithm examines. This argument should normally be set to 1 or 2. Amount (expressed in %) specifies the degree of sharpness. This argument must be greater than 100.

If one or more coordinates passeed to the Crop method are outside the coordinate space of the image, this will actually expand the "canvas" around the image. This is useful, for example, if you need to create margins around the image. The following code creates a 10-pixel margin around an image:

jpeg.Crop -10, -10, jpeg.Width + 10, jpeg.Height + 10

IMPORTANT: Before version 1.7, the color of the margins created by "negative" cropping was always white. Starting with version 1.7, it is determined by the color specified via Canvas.Brush.Color, and is black by default. To create a white 10-pixel margin with version 1.7+, the following code should be used:

The Grayscale method sets the three color components (R, G, B) of each pixel to the same value L using the following formulas:

Method	Formula
0	L = 0.3333 R + 0.3333 G + 0.3333 B
1	L = 0.2990 R + 0.5870 G + 0.1140 B
2	L = 0.2125 R + 0.7154 G + 0.0721 B

The effect of the Method argument is demonstrated by the chart below:

Method	Effect
Original Image
0
1
2

The effect of Method 0 is what Photoshop calls desaturation (Image/Adjust/Desaturate), while Method 1 is similar to Photoshop's conversion from RGB to grayscale (Image/Mode/Grayscale).

Note that Grayscale is a method, not a property, so the '=' sign should not be used:

The Hue parameter controls the brownish hue of the output image, and should usually be in the range of 25 to 60 for good results. The Contrast parameter controls the contrast of the image. The value of 1 means no contrast adjustment. Values between 1.2 and 1.5 usually produce good results.

Original Image	Hue=50, Contrast=1.4

The sample Sepia conversion shown here was achieved as follows:

VB Script:

<% Set Jpeg = Server.CreateObject("Persits.Jpeg") Jpeg.Open Server.MapPath("clock.jpg") Jpeg.Width = Jpeg.OriginalWidth * .8 Jpeg.Height = Jpeg.OriginalHeight * .8 If Request("Grayscale") = "1" Then    Jpeg.Grayscale 1 End If If Request("Sharpen") = "1" Then    Jpeg.Sharpen 1, 250 End If If Request("Horflip") = "1" Then    Jpeg.FlipH End If If Request("Verflip") = "1" Then    Jpeg.FlipV End If Jpeg.Quality = Request("Quality") Jpeg.Interpolation = Request("Interpolation") If Request("Crop") = 1 Then    Jpeg.Crop 30, 30, 470, 320 End If Jpeg.SendBinary %>

C#:

<%@ Import Namespace="System.Web" %> <%@ Import Namespace="System.Reflection" %> <%@ Import Namespace="ASPJPEGLib" %> <%@ Page aspCompat="True" Language="C#" Debug="true" %> <script runat="server" LANGUAGE="C#"> void Page_Load(Object Source, EventArgs E) { ASPJPEGLib.IASPJpeg objJpeg; objJpeg = new ASPJPEGLib.ASPJpeg(); objJpeg.Open( Server.MapPath("clock.jpg") ); objJpeg.Width = (int)(objJpeg.OriginalWidth * 0.8); objJpeg.Height = (int)(objJpeg.OriginalHeight * 0.8); if( Request["Grayscale"] == "1" ) objJpeg.Grayscale( 1 ); if( Request["Sharpen"] == "1" ) objJpeg.Sharpen( 1, 250 ); if( Request["Horflip"] == "1" ) objJpeg.FlipH(); if( Request["Verflip"] == "1" ) objJpeg.FlipV(); objJpeg.Quality = int.Parse(Request["Quality"]); objJpeg.Interpolation = int.Parse(Request["Interpolation"]); if( Request["Crop"] == "1" ) objJpeg.Crop( 30, 30, 470, 320 ); objJpeg.SendBinary(Missing.Value); } </script>

Click the links below to run this code sample: