light and shade in artificial scenery

[Lazur]

Sneek peak to the process of building up my poster for digital graphics subject at the university.

From left to right:
original concept, permanent markers and wood finish on packaging paper
pencil drawing on plywood
permanent marker on glossy print foil
two rendered images with blender
a composite of the previous two
the raster image traced with inkscape

Original size is 350 mm / 1000 mm, had cut off the header part on the top -two of the photographs taken had them.
Will have to print out the final result, still I'd like to add in some details and end up with a grubgier, more industrial look as on the first image.
Have to figure out how to manage it, since filtering may be a bit processor melting at this scale.

[ha1flosse]

very nice work and pretty cool concept drawings! i like the antagonism of the grasshopper as a "natural" element" and the technical structures. so the result is nice and the alteration of the grasshopper surely fits into the concept but the antagonism is fading looking at the result. so what was your original intention regarding the composition and the elements you've been using?

[Lazur]

Thank you!

Tried to make the composition somewhat dynamic looking, emphasizing on the image's given width/height ratio.
Looking up a small corridor can be quite dramatic, which was a basic idea to me to incorporate. Light shows from the top, and only a small ray of light reaches the bottom. With all the complex shapes, only one pole standing in the distance as a straight axis, connected to two other supporting towers in the foreground of the composition.
Needless to say it didn't come out so well. Once modelled in 3D, the right side's supporting column turned to be steeper, blocking the continuity of the thin negative space on top. Now the composition falls apart to two sides, with a bit bland right side.
To overcome this, now the three poles on the left are representing the "light hitting the ground",
and for the final image I'll have to add in some fancy stuff crossing the right side as well.
Was thinking of a glitch looking decompositing and of growing glowing cables from the grasshopper. Will consult with the teacher a week later and then will decide what would be more (time)effective.

[Lazur]

An update on this project:

[Moini]

In my non-artist view, I like the third one best, looks very futuristic and technical. I'd maybe add the structure on the yellow shapes from the fourth to it.
But I guess you've moved to the fourth?

[Lazur]

Didn't felt like those swashes were blending in to the background so went with the next one for this time.

Tried to grunge up it a bit in gimp, to have a similar feel as the original:





Unfortunately that's the size it was created in so probably I'll go with the clean one for the printing.

[Lazur]

[Moini]

How big will it be printed?

[Lazur]

350 mm / 1000 mm.
(@ 600 dpi that would be 8268 px / 23623 px.)

[Moini]

That's huge Do you have some kind of printing service at your university, or do you need to go via some online print service?

[Lazur]

There is a print service at the university just in case, but planning to print it with a more quality oriented print shop.
They are making giclée prints with 12 different kind of inks, and have no problems printing from rgb format -however their ability to print colours "accurately" I can hardly match on my side.
For colour matching I'd need my screen calibrated -doesn't make too much sense to me. No matter how I set, it can only produce 256 luminance levels and Mach banding is present (probably on the operating system side). So all in all not sure if pdf-s produced with inkscape's default colour profile means no banding in print or not.

[Moini]

I kind of 'inherited' a (spyder) colorimeter, and actually once did calibrate my screen. It definitely did make a difference (but atm. it doesn't work, the profile isn't applied - I haven't looked into the cause, yet, though). Even if for me, it *really* doesn't matter. I rarely need any accurate colors. I'm already happy when my printer actually prints something that doesn't have any stripes from dried ink in the jets.

I think the pdfs (if there are no filters and all objects are rasterized, that is) should contain the gradients directly, in their mathematical form, and their banding would depend upon the software that displays them. If you expect to need this more often and it's affordable, maybe order a test print, that contains all sorts of critical stuff?