Learning Task 9: Poster collage


So I like the concept of this but not the result. I don’t feel like the image is balanced correctly, the characters and sword area should be smaller to give the tree more of an impact, but when I did that it looked worse… The symbolism is also very heavy-handed and the antagonist is practically a walking trope.

There’s also only seven elements, but adding more made the image very cluttered, so I think that was the right choice.

Format wise I went with jpg on the 50 setting. Quality wise png 24 was a little better, but also four times the file size, so for normal non-art related sites this was the better option.

I’m Back!


It’s been an annoying few months, what with an inability to breathe and a couple of surgeries (nothing major 🙂 ) but I have returned to the net. Yay?

In other news, I’ve purchased a new monitor. While (whilst?) setting it up it occurred to me that I never really bothered adjusting my old one. I just upped the digital vibrance and had at it… So expect the colours and contrasts of future work to be a tad different (and hopefully better).

Learning Task 8: Image Compression Experiment

Given the “busy” requirement of one of the images for this task I decided to have a play with reflections and raytrace shadows through trees. The image has WAY too much going on, but it was a good learning experience nonetheless (as lame as that sounds, lol).


Compressed as a 60 quality jpg it comes in at ~78KB. The “plain” image comes in at a minuscule 9KB.


The reason for the size dependency is all due to the compression algorithm. The “Busy” image changes colour almost every other pixel and there’s no other discernible pattern. In the “Plain” image however the algorithm can simply sum up the giant area of white as “the next x pixels are white”.

Learning Task 7: Combining Images using Layer Masks

I didn’t quite follow the brief with this one. The task called for pictures of me and there was zero chance of that happening. As far as the learning goes though it’s all the same, so hopefully that’s acceptable.


I’m not happy with this one to be honest. The landing pose is wrong, but more importantly I just don’t like the feel of it.

Also, full disclosure, I purchased the architecture model for around a buck. It’s quite nice from a distance, though I need to work on the shadows some more, but up close the textures aren’t great and there’s some clipping between walls and floors. I’m going to try and edit it myself at a later date to see if it can be salvaged for closer shots.

I also gave this pic a once over with the previously used HDR highpass method.

Below are the (drastically reduced) original renders.


Learn Task Six: Making Selections & Combining Images



Lego cars these days look so real…

This task was possibly THE most tedious thing I’ve done in recent memory. Not entirely sold on her hand pose, but I spent so much time on this task that I’d rather stab out my eyes than adjust it. It was very educational though, I’d never used the stamp or patch tool before. The tedious repetition also got me into the habit of using hotkeys, another plus.

Additionally, one of the tutorial videos had an easymode HDR tip, so I gave that a go as well.


Not the best piece to test the technique on, but it was still quite interesting. The skin looks great near the eye glow, but the car has a sort of over-sharpened effect going on.

Learning Task Five: Adjusting Layers

I was slightly worried about how this task would go. With no background and lightning controlled completely by me I was concerned not much would change between images without making them over or under exposed. That sort of held true, especially with the darker image, but the bright one turned out OK.

Dark Image OriginalNENightOrig50

Dark Image Adjusted
NENightAdjusted50Now the second image does look under-exposed, but she is meant to be wandering around in the dead of night, so I think it looks more accurate this way. It’s hard to tell, but it is also more vibrant.
One annoyance with this image is the eye glow. Because it’s hard to make out the hair that blocks her left eye the glow comes across an unbalanced, something I’ll have to be wary of in future.

Bright Image OriginalNEBrightOrig50

Bright Image AdjustedNEBrightAdjusted50After uploading this I’m now convinced it’s over exposed. I may have to resubmit this later when I get some time. This is why you always back-up your original images, lol.

I also have to say that I’m not a huge fan of the levels tool. Perhaps it was just my image choice or that I’m overlooking something but levels felt a little redundant when I could accomplish the same or better with the curves or vibrancy tools.

It was also interesting to see the Elf’s facial expression change slightly as the jpg compression kicked in, albeit minutely. Another thing I’ll have to watch out for in the future.

Now with 100% more armour!

It may have taken me a while to get all of the details right, but the Night Elf now has clothing, just.

In all honesty though, I took the easy way out for this, beginner that I am and all. None of the armour really needs to conform to movement, they’re mainly just static models attached to rigid body parts. Also, full disclosure, the metallic items are freeware that I’ve modified slightly.

I’ve also uploaded this as a png, rather than the usual web saved jpg used for learning tasks to compare quality. The file size is still only 315KB, so don’t worry, it won’t murder Aussie download limits.



Added clothes! Yay?
Increased hair length.
Changed hue of hair texture.
Corrected majority of hair clipping issues, pauldrons still an issue.
Adjusted highlight colour of nails slightly, still not happy.
Increased eye size and adjusted eye slant slightly.
Added one small light source to each eye because Night Elf eyes glow for some reason…
Adjusted default lighting to remove random blue highlight issue.
Fixed hair/pupil interaction problem.