|
ASI224MC IMAGE PROCESSING TUTE |
Part 4 PHOTOSHOP CS4 PROCESSING |
|
|
|
|
|
This is a simple "drag & drop" loading & I like to have all the images I may choose to open/process in CS4 etc open in separate windows by enabling the option in the 2nd image above. |
|
|
|
|
|
In the above image we have chosen the eye-dropper tool from the lhs of CS4 & used it to sample the background around the planet. You will notice that over on the right in the "Color" display it shows a black colour after we click/sample the background - do not sample right next to the planet's edge as you can see that it is lighter there (we will attend to that shortly!) & of course do not click in that blue artefact that we have created when we did the "RGB Align" in Registax6. |
|
|
|
|
|
We've now brushed out that blue to create a fairly uniform black background around the planet except for the area immediately adjacent to the planet's edge - keep your brush well away from this region & do NOT be tempted to try & "fix" these areas close to the disk edges using the brush - I've seen many ugly attempts using this bad ploy & it is definitely NOT the way to handle those areas..! |
|
|
|
|
|
|
|
By selecting "Image>Adjust>Levels" we bring up the histogram displayed in the 3 images above. In the first of the histograms/graphs you notice a peak that rises at the very lh side of the graph then drops back to almost zero before we see the main body of the graph which represents the histogram of the planet itself. This initial peak represents to all intents & purposes the brighter region around the planet's edge/rim. By shifting the lh (black) slider to the position indicated you will see in the 2nd image that we have now lowered the brightness there to an equivalent of that in the general sky background: do NOT move it too far, only to where that initial peak on the graph finishes! The 3rd image above shows the results of moving the rh (white) slider to the left - this brightens the image. (compare each of the 3 images above, particularly the first & last) We don't move it all the way to meet the rh side of the graph because we want to leave some room for any of the other applications that we will do - & we want to avoid clipping the histogram as described earlier. |
|
|
|
|
|
Fairly self-evident steps above: we record with gamma set at unity ("1.0") but lowering the gamma does assist with contrast/detail clarity) so it is quite acceptable to lower it to around "0.90" here as shown... |
|
|
|
|
|
As with everything in this tute these apply to this specific capture data: they form a basis for your own data's processing but you will not only need to experiment & alter specific values throughout these steps but also adjust for your own personal perspectives: we find that these later cmos-based cameras such as the ASI224MC colour & the ASI290MM mono benefit from an increase in colour saturation, but also an increase in vibrancy, which for these explanations is a broader "span" of colour or dynamic range. The above adjustments are par for the course for us with our captures, but you are always advised to experiment for your own understandings! :) |
|
|
|
|
|
|
|
These 3 images show the results of adjusting curve values for the image - compare the top image of these 3 with the bottom image. By "pulling down" the bottom half of the curves graph in the direction indicated & doing the opposite for the top half we create a slightly more "3 dimensional" appearance to the disk of the planet by inceasing the contrast as we look from the middle of the disk towards the outer edges. Once you've done this a few times you will become very good at arriving at the values of "195/185" that I find quite effective most of the time - but don't get hung up fiddling about...just make the 2 "pulls" then type those values into the boxes showing the values! :) |
|
|
|
|
|
Self-evident steps here above - remembering that the image is still 200% of capture scale so that if we reduce it to 50% now, we arrive at an image the same size/scale as when we captured it. |
|
|
|
|
|
|
|
|
|
|
|
More self-evident steps: I like to reduce the background canvass slightly first before enlarging the background canvass because sometimes you can see the division between any new enlarged canvass & your original background's canvass sizing, reducing it a tad first seems to remove that likelihood. |
|
|
|
|
|
Here we've saved a file for the web as well as one for any alterations as well as a more permanent record of the full file image. I always save as "png-24" for web images as this is the best format for fidelity when posting any images on the internet. |
|
|
|
|
|
|
|
|
|
|
|
In these examples I display the steps in sharpening your
image: NOTE that the sharpening is
done BEFORE the size reduction stage shown earlier - we do this for a
significant reason in that if we sharpen the image, then reduce it to the
size/scale we wish to finish up with, we do NOT need to apply so much
denoising (in this case gaussian blur) to the final image. Obviously the last image above has not only been reduced in size before we apply the gaussian blur, but we have also made an enlarged canvass in the manner we did before... |
|
|
|
Here we have our smoother image next to our sharper image - in these 2 examples I would fairly obviously opt for the sharper image because although a slight edge artefact can be seen, it is quite subliminal & the clearer details of the sharper image would influence my choice... ;) A marathon amount of typing & screenshots & as much again for Pat to upload: I hope this tute is of some benefits & I will be creating more shortly, where I go into more involved (but just as easy) processing applications to (hopefully!) allow people to develop their image-outcomes further! :) |
|
|
|
|