Wavelet Noise Reduction
This is the core of PureImage. Unlike all conventional methods of Noise reduction, our wavelet NR produce much higher quality of noise reduction while retaining image sharpness and details.
Unlike some frequency domain profile based NR, PureImage will automatically auto-tune the noise function to different noise type across the image, making the interface much simpler and universal for all digital or film cameras.
In addition PureImage includes advanced Color Correction, Color Matching and Crop functionsThe main screen has four display modes:
You can compare the noise reduction of the original image. If you make any change to the parameters, the bottom part will need to be recalculated. This is relatively quick and should take a second.
Since this mode display the image in pixel size you will see only small part of the image. You can drag the image around to see and to compare how the noise reduction affect various parts of the image.
On the Preview window you will
see marked the part that is
displayed. You can drag the red rectangle also there.
It is important to realize that to see the true effect of the Noise Reduction function, the image must be displayed in pixel size (1:1). Zooming out will distort our perception of noise so this mode does not allow any zoom changes. You can also adjust Colors in this mode, but you don't really get the true feel of the image in tis mode.
This is the main setting of the Wavelet Noise Reduction algorithm. Higher level will remove more noise, but also may remove some fine details from image making it look unnatural smooth. Therefore it is important to set the level only to the amount needed to remove noise.
Remove Block Noise This will remove a blue or red noise chunks often seen in high ISO images in dark areas. Also this removes JPEG artifacts.
Advanced Processing Our goal is to create image that will have removed noise but still retain its details and sharpness. This is the task of Advanced processing. First setting is the Mode.
There are few modes:
Cropping image is one of the
most common trick of professionals.
you take the image, the subject may not be always framed correctly. By
enlarging and cropping we can make the whole composition look better.
There are many rules for composition, but the simplest one is a rule of
thirds.You can overlay a thirds grid over the image in this mode.
The image is displayed in a Crop Format similar way that an enlarger in photo-lab works. This us let see the whole composition as it will appear on photo paper or screen. For example our camera takes 4:3 images, but we want to print them on 6 x 4 photo. Obviously some parts would need to be always cropped.. You should choose the format depending on your desired destination.
To zoom-in the image you can use the enlarge slider, mouse wheel or the +/- buttons on toolbar. Similarly to the Details view, you can move the crop position by clicking and dragging the image or the rectangle in preview.
When you select the best crop of your image, press Apply Crop button. This will return to the Details mode and crop the image. You can also click the Details tab in which case you will be asked if you want to apply the crop.
This mode can be also used to check the overall Color correction impact. If you don't change the crop settings (Enlarge or position) then you can freely switch between these modes without actually cropping the image.
Note: In Full/Crop mode the NR is not applied to the displayed image on screen (It would simply take too long calculate NR on whole image just for preview and you won't really see the correct impact because of the resized image.)
There are few sliders to adjust the color and overall impact of the image. Color correction changes will shows also in changes of Curves and Histogram screen.
These two sliders will stretch the dynamic range of dull image separate for each shadows or highlights. You can see the impact directly on the image or in the Curves and Histogram screens. Moving Shadows will make more of the shadows and mid-shadows deeper black, moving the Highlights will make more of the highlights bright. The result will be more dynamic image with deep shadows and bright highlights.
Under-exposed image will have lost details in its shadows and overexposed image will have overblown highlights. A typical example is taking picture of people against strong light. Without correct compensation the faces will be dark because the camera set exposure on the bright background. Our image will be under-exposed. We can use the Exposure compensation (+) to bring up some of the lost details in shadows back.
It is common that digital
cameras underexpose the images just very
little, only to avoid over-burning highlights.
On some cameras flash images tends to be bluish and also bright outdoor images may have a slight blue cast. On the other hand, no-flash indoor images with slow shutter will appear yellowish. The temperature correction can add or remove the warmth.
This will boost color of the images to make them look vivid. Unlike normal Saturation, this process doesn't work on HSL space so there is no danger of clipping during RGB to HSL conversion and back.
Color Match uses special matching and optimizing algorithm that can match colors and tone of the image to the image loaded as reference. In addition you can save the color profile of reference image for later use.
There is no restriction what can be the reference image. It can be any size or color. The matching and searching algorithm will make sure that the result image will look as much as the reference image in terms of color and tone while preserving smooth transitions between color. However a match between two totally unrelated images may not always produce best looking result especially if our destination image is in poor dynamic range or color.
You can choose to load the reference image from disc or use saved reference profile. In addition you can save any reference image to a new profile with the small save button.
You can check the translation function built by the color match in curves display.
This will display the translation color curves that are applied to the image.
These parameters contribute to the curves:
This will display a histogram of the output image.
All of the color Correction functions and Color Match will change the Histogram graph.