Painting with Satori

Extract from PC Plus magazine January 1998

Canvas & RIR files mentioned here are not supplied on the Satori web site.  All references to screen shots refer to the printed tutorial.   Some, but not all, screen shots will be included here

Part 1 – Introduction to photo enhancement
Part 2 – Layers and masks
Part 3 – Advanced Tools
Part 4 – Putting it together


"In this series we’ll be completing a number of exercises to build your technique and explore different aspects of the program. This month we kick off with some photo image editing. Desktop scanners have been around for years and digital cameras are beginning to make their mark. The ever-increasing use of photo-based images on the Web and in self-published document is assured, and so the need for skills to correct, enhance and be creative with photographs can only increase. Not only that, but it’s whopping good fun too, if like me you are artistically inclined but haven’t got much in the way of conventional skills.


No doubt you are familiar with paint and photo-manipulation packages, and the idea that the bitmap files used by these programs are basically grids and dots, or pixels, each of which can take on a range of colours. The pixels are small, and are viewed at normal magnification are almost unnoticeable. Like the dots that form the TV picture, they blend together well. But if you zoom in on a bitmap, the pixels become larger and more noticeable. Previously smooth lines and fills look more and more like chunky coloured squares.

Another disadvantage of traditional bitmap files is that the entire image is one entity. Once you’ve made a change, unless the program has made a copy of the image in its previous state, you can’t undo or revise an earlier editing operation. Bitmap editors generally have an undo feature, but common bitmap file formats don’t allow a list of these changes to be embedded in them. Any editing you did in a previous session is part of the fabric of the image. Contrast this with the files produced by vector drawing programs such as CorelDraw in which the image is composed of many objects. Any object or group of them can be edited at any time, so whatever you do can be revised, even in a later editing session.


Although it’s a bitmap processor, things aren’t so crude with Satori. It has important refinements you need to appreciate before getting stuck into the program. Satori is resolution-independent. When you paint a brushstroke, for example, Satori doesn’t just change pixels in the bitmaps file. Rather like a vector drawing program, it stores a description of the brushstroke. When you zoom into the picture, the brushstrokes is redrawn in the same shape but at an increased resolution to reserve fineness of details.

To cater for this and other special features, Satori has its own file format. Only when your picture is finished do you convert the Satori file into a conventional format, at which point everything is transformed into a single conventional bitmap of the required resolution and colour depth. You can, of course, keep the original file and edit and re-output it should changes later prove necessary.

Satori’s own files end in the extension CVS, and are known as canvas files. There is also an RIR format which makes large imported bitmap files quicker to work on. Read up on RIR and other Satori-specific file types in the help system.


Our first exercise is to perform plastic surgery on the picture you can see on the left – a shot of Bath’s Royal Crescent. I wanted to remove the lamppost and tidy up the rubbish and graffiti. The way to do this in any photo editor is to copy details from elsewhere in the picture and paste it over the object to be removed. A fake but plausible-looking background is thereby constructed. Other tools are then used to smooth out any visible joins.

The copy and paste method can be very laborious work in some programs. If you’ve ever tried it, you will be looking at the complex background, thinking that it’ll be quite a job to patch it up convincingly. Not so in Satori. Start the program and follow me through. Surprisingly, it only takes a few minutes.

Because the image is larger than the editing window, it’s best to convert it (ROYLCSNT.TIF) into RIR format for faster working. Go to File, Convert to RIR, and pick the TIF file. In the next dialog box, if you haven’t copied the file on to your hard disk, be sure to choose something other that the CD as a destination for the conversion.

Now open ROYLCSNT.RIR. Just click OK in the Load to Layer dialog. Satori will read the RIR file, incorporating it into a CVS format file for editing. The help system explains the relationship between RIR and CVS. If this is the first time you have used the program, you may want to dock the floating toolbar at the top of the screen. The lower part of the Satori window is given over to drawing tools and associated controls. It’s a bit if an eyeful at first, but you’ll soon be whizzing round it like an old hand.

The tools are logically arranged in pages, selected by the upper row of buttons. Click on Paint, to display the painting tools. To the right of this page are the painting styles on offer. Some mimic conventional tools, such as airbrush and pencil. Others are for special effects. The one we’re interested in is the clone brush, so click on Clone now.

With this in operation, when you paint on the screen (click and drag with the mouse or press and drag with a tablet pen) the pixels under the tip are made into copies of pixels in another part of the picture. If you carefully choose the offset between source and destination, you can effortlessly paint over an unwanted object with a convincing background detail from elsewhere.


Setting the correct offset and exercising a reasonably delicate hand with the pointing device makes the removal of unwanted detail much easier than the more conventional way of copying and pasting parts of the image.

Zoom into the picture; press the magnifying glass in the bar you just docked and drag a box over the area you want magnified – the lower part of the lamppost. Don’t worry if you get a warning dialog saying that the screen area is too small to load the image at one-to-one scale – just click OK.

Now we can give the clone tool a try – we’ll eliminate part of the post’s shadow. First we must set the relationship between the copy and paste elements of the brush. Right-click on the picture and select Set Clone Offset. The pointer turns into a rubber stamp, waiting for you to set the source position. Click on clear sunlit tarmac midway between the post’s shadow and the shadowed area above it. Now click in the post’s shadow.

The second click has set the offset between the brush tip you’re going to paint with and the place where Satori picks up the colour the brush will apply. There’s one further thing to do before erasing the shadow. If the brush is too large, it will pick up too big an area and probably include unwanted detail. The brush size was denoted by the green circle you saw when setting the offset, but now it’s gone. It’s useful to turn the circle back on by right-clicking and selecting Brush Outline.

Remember that the source will be the same size as the destination. Is it too big? Then choose a new size by clicking in the selection of brushes on the far left. Go for the smallest yellow circle of all. If it’s still too large, enter a value smaller than 12 in the edit box under the black square displaying the brush tip. A value of 6 works well for the operation you’re about to undertake. See how the sliders move when you set the value? They are another way of fine-tuning brush size.


Now place the brush tip over the posts shadow at it’s left-hand end, and drag along it. Two or three passes should see most of it off, but some shadow still remains. Drag the editing windows to one side for a moment, so that you can see the hint of shadow remaining in the unscaled view. The slowness of the brush is due to its action having a degree of transparency. This is good for careful blending, but in this case we really want a stronger action.

The shape of the brush, shown on the lower left, gives a clue as to why – its edges are very blurred, giving a transparent effect. Satori is highly customisable, and the profile of this brush can be changed. Go to View, Brush Setup, Profile. In the Brush Setup dialog, to the right of the image of the brush is a graph. Try dragging the handles around to change the brush characteristics. With the middle handle pulled towards the upper left you get a more solid brush that will have a stronger action.

If the painted-over area is too light, you can tone it down by resetting the clone offset to take in the slightly darker area below the lamp-post’s shadow. A couple of strokes blends it in perfectly. The smudge tool, listed in the left-hand column, can be used to smooth areas that don’t look quite right, but don’t over do-it. You lose fine texture, giving a synthetic look to grainy-looking surfaces like the path.

Zoom in very close until you can see individual pixels. Select the smudge tool and try it on a small area. See how the effect builds up the longer you press, and how the grainy look is smoothed away? It’s worth playing with other brush effects at very high magnification. Eradication of blemishes can be achieved with the Feathers brush, for example – give this one and others like Feathers a try at high zoom, so you get a feel for what they do. Also try adjusting their characteristics in the brush setup dialog. While some brushes apply a straight colour, some manipulate what’s already there to give various effects. Don’t forget that [Ctrl][Z] undoes the last operation.

Now have a go at the part of the shadow that stretches in to the gravel, and the line of broken tarmac where the lamp’s cable has been buried. If you want to scroll the zoomed view, click on Pan in the Zoom Controls window, drag the picture around within the editing window, then click on Pan again.


With this sort of touch-up work you need to keep an eye on the brush offset. For instance when you move on to erasing the post, the job is easy if you get the correct offset, and a mess if you don’t. The first point at which you need to be precise is where the background changes from light to shadow. The angled boundary must be reproduced so that it joins up either side without a kink.

All you have to do is set the offset at the same angle. Click on the shadow boundary for the source, and further along the boundary for the destination. If you’re careful to line up the offset so its parallel with this line, when you paint over it the gradient will continue straight over the present position of the lamp base.

When you get on to the railings, the steelwork must be put in the right place. The way to do this is to make the offset horizontal, and to use the two railings to the left to set the distance apart. When you erase the lamppost, the clone brush will copy the railing from the left and paint it at the same distance away as the two railings you used during setting.

Further up, set the clone offset to match the spacing of the columns. Once again, the column to the left will copy over the post at the correct spacing. Be careful on the roof – it’s easy to mistakenly copy branches, chimneys and the lines of brickwork. Careful and frequent changes in offset distance and angle will avoid these problems.

In the case of the tree, a short offset inclined at the prevailing angle of the branches gives a convincing effect but a bigger brush works better (say 8). Regularly adjust the clone offset angle and work with slow strokes along the line of the branches. Total time spent on erasing the lamppost on the fast PC is under ten minutes once you’ve had a little practice, which is impressive for such a complex background.


Correcting blemishes is only one part of photo editing work. The thing you’ve most likely to want to do is correct problems with exposure and colour balance. Load up the file BATH.CVS which was originally a PhotoCD image. This picture of Bath was taken on a cloudless January day. It looks a little washed out and has a blue cast.


Before successfully adjusting the colour balance of a picture you need to be aware of the way a computer handles colour. Each pixel on the monitor is lit up by a signal composed of red, green and blue components. Combinations of these colours give all possible hues, from black to white and all shades in between. This is known as the RGB colour model. When the proportion of red, green and blue are all the same, you get a shade of grey. Unequal proportions give a colour. In a 24-bit file each colour channel can range in value from 0 to 255. Since eight binary are required to represent each of these values, every pixel requires 24 bits to store its colour.

You can use Satori to familiarise yourself with the way the values of the RGB channels affect the overall hue. Go to File, New, click OK and push the Colour button to bring up the Colours dialog. Make sure the RGB radio button is checked and click on the Slider tab. Here you can mix your own colours by dragging the Red, Green and Blue sliders, watching the result at the bottom. You can also press the eye dropper button and pick a colour from any image you have loaded, to find out how it was mixed and, if necessary, re-mix it.

Screen colours are made by illumination. Printed colours are reflected, so they use different primaries. CMY (Cyan, Magenta and Yellow) is the scheme used by printers. If your output is to be printed professionally, it’s best to work with a CMY-based model. The films used for printing will separate the picture into its cyan, magenta and red components, so it’s best to work with these rather than RGB.

A full mix of all these inks is supposed to give black, which in practice is not as black as it should be. So in professional printing the black areas are handled separately as a forth channel, so they can be printed with black ink. This is the CYMK model, and is what you should use in professional printing applications.

Satori can deal with this quite easily.

Bring up the Colour Correction dialog by selecting the only option on the Colour Correction menu. On the slider tab the brightness and Contrast controls can improve the picture quite noticeably. There is a flatness of tone across the sky, hills and city. Contrast reduces or increases the range of intensities in a picture. So if we increase contrast, the lighter areas get lighter while the darker ones get darker.

Winding up the contrast to about 40 per cent certainly improves the look of the buildings, because they are a lot lighter than anything else. The hills and sky still don’t look right, however. Pull the brightness slider down to about –25 per cent. This richens the green in the hills without harming the look of the buildings too much, though you can push the contrast up a bit further to compensate if you want to. I ended up with –30 on the brightness scale and 60 on the contrast. This made a dramatic improvement, though blue from the bright sky still gave a slight cast to the green areas.

Below the Contrast and Brightness sliders are controls to shift the colour balance. Bearing in mind that green is a combination of blue and yellow, it seems sensible to push the bottom slider to the left, increasing the proportion of the yellow in relation to blue. A setting of in the range of –20 to –30 reduces the blue cast without affecting the colour of the buildings beyond what you might see on a picture postcard. Press OK to check out the full picture. Not bad – and some fine-tuning of the brightness and the contrast settings will finish things off nicely.

Before finishing, it is worth digressing into the object list. Press OK in the Colour Correction dialog, and press the Object List button on the right of the screen. The tree view that comes up next shows everything in the file. The Polygon item is the image itself. Your colour correction is shown too; deleting it from the list will undo the change. Go to the Edit menu and see the options under the Remove Top... item. They all refer to objects in the Object List.

That’s all for now. In the next issue we’ll be digging deeper into Satori’s powerful toolset with particular emphasis on layers and masks."