×
Namespaces

Variants
Actions

Instagram X-Pro II Filter Recipe by Miracle Machine (Nokia Imaging SDK)

From Nokia Developer Wiki
Jump to: navigation, search

This recipe replicates Instagram's X-Pro II Filter using Nokia Imaging SDK inbuilt filters.

See Also
Article Metadata
Tested with
SDK: Windows Phone 8.0 SDK
Devices(s): Nokia Lumia 920
Dependencies: Nokia Imaging SDK 1.0
Article
Created: hamishwillee (27 Feb 2014)
Last edited: hamishwillee (03 Mar 2014)

Contents

Introduction

The Instagram X-Pro II filter makes colours more vibrant and gives a warm appearance (sometimes described as "vintage"). This filter is used in about 0.32% of all photos shared on Instagram (source /filterfakers.com).

The recipe in this article shows how to recreate X-Pro II by combining some of the 50+ inbuilt filters from the Nokia Imaging SDK. The recipe is loosely based on How to Make Instagram X-PRO II Effect Using Photoshop (Miracle Machine blog).

The article includes the recipe source code, instructions on how it can be used and tested, performance measurements and a gallery.

Source code

The code for creating the effect, the source, renderer and target is described in #How to use the filter and Core concepts (Nokia Developer Library). The fragment below shows only the "recipe-specific" code for adding the filters to the List<IFilter>.

Warning.pngWarning: The code below uses a Rect based on the image size (m_FrameSize) to define the vignetting region. Depending on the architecture of your app this may be available directly where you're defining the filter, or indirectly from the source. This is discussed further in #How to use the filter.

var filters = new List<IFilter>();
//Instagram Filter: X-Pro II
 
//Vignetting
Color black;
black.R = 0;
black.G = 0;
black.B = 0;
 
// NOTE: How m_FrameSize is obtained depends on the architecture
Rect region = new Rect();
region.X = 0.1 * m_FrameSize.Width;
region.Y = 0.1 * m_FrameSize.Height;
region.Width = m_FrameSize.Width - (2 * region.X);
region.Height = m_FrameSize.Height - (2 * region.Y);
 
filters.Add(new VignettingFilter(1.0, black, region));
 
//Adjust colors
double a = -0.00002746151084310468;
double b = 0.010566561680750;
double c = 0.091211513981563;
 
Curve curve_blue = new Curve(CurveInterpolation.Linear);
curve_blue.SetPoint(0, 25);
curve_blue.SetPoint(255, 230);
 
Curve curve_red_green = new Curve(CurveInterpolation.Linear);
for (int i = 1; i < 255; i++)
{
int yvalue = (int)(a * Math.Pow(i, 3) + b * Math.Pow(i, 2) + c * i);
curve_red_green.SetPoint(i, yvalue);
}
 
filters.Add(new CurvesFilter(curve_red_green, curve_red_green, curve_blue));

How to use the filter

This recipe uses only filters based on IFilter. These are combined by simply adding them to an IFilter list, which is then passed to a FilterEffect along with the source. The code fragment below shows the general strategy (this is explained in more detail in Nokia Developer SDK - Core concepts).

var filters = new List<IFilter>();
//The recipe filters are added here
...
...
 
using (var source = new StreamImageSource(stream))
using (var filterEffect = new FilterEffect(source) { Filters = filters })
using (var renderer = new WriteableBitmapRenderer(filterEffect, writeableBitmap))
{
await renderer.RenderAsync();
}

The recipe uses a Rect based on the image size (m_FrameSize) to define the vignetting region. In the code fragment above we have access to the source as a StreamImage, and so can get the ImageSize using the following asynchronous function call:

//Get image size from stream to use in vignetting. Note is asynchronous call.
ImageProviderInfo imageInformation = await StreamImage.GetInfoAsync();
Size m_FrameSize = imageInformation.ImageSize;

Depending on the application architecture it may be possible to get hold of frames directly. It is also possible to define vignetting without a region, and this might be an area for further exploration.


Testing the filter

The easiest way to test the recipes is to add them to the Test Apps for Viewing Custom Filters (Nokia Imaging SDK) (a fork of the Real Time Filter Demo). Simply to add the filter source above to both RealtimeFilterDemo and StaticFilterViewer test apps in their respective switch statements. You may also need to add the line:

using Windows.UI;

Note that the code for both test apps makes the frame size available as m_FrameSize (as used in the recipe snippets).


License

The code has been released with the standard MIT License.

Performance

This section lists the approximate frame rate range (in frames-per-second) obtained by using the recipe in the Test framework.

Device FPS (Frames Per Second)
Lumia 1020 12-14 FPS
- -

Gallery

Note, original (unmodified) image is here.

Summary

This is a fairly accurate reproduction of the Instagram X-Pro II filter using Nokia Imaging SDK recipes.

Note that it is usually possible to reproduce Instagram filters using different recipes, and to trade-off off "accuracy" and "performance". If this recipe does not meet your needs, see if we have any others in the category, or create your own (with associated article!)

This page was last modified on 3 March 2014, at 06:28.
175 page views in the last 30 days.
×