Wider New Logo tests:
Thanks to some of the feedback I got from TSML I was able
to reinforce some of my own misgivings about some aspects of first tests (in the
paragraph after this one.) These new ones are only 120 pixels tall 50k each so I
decided not to use thumbnails but the actual images so you could compare them
more easily. The first is using rounded shading to give the letters a 3D shape.
The second is the same but with a slight WFMM bloom. I am interested in whether
you think the first or second look better i.e. with or without the bloom effect.
The 3rd flattens the text only. This causes the letters to appear larger but not
3D. I am not sure which effect has more impact. What do you think? The 4th and
last image carries the flattening over to the treble clef and notes. In all
cases I had to adjust the shaders shine down to or near zero to avoid a white
having broad white faces devoid of color. Though it may not make much difference
I used Darktree shaders on all but the slabs and frame. On something this small
it may not matter but I may want to do some sort of animations, close-ups or do
things with individual elements in the future so I wanted to have everything set
up in advance. Just in case. I don't know if it is my imagination but the treble
clef in the last one looks kind of like velvet. If I use the last version I will
probably apply this to the notes too.
Added #5,6 & 7:
I just added numbers 5, 6, and 7 below. They all use flat
letters and symbols and add the velvet black look of the treble clef to the
notes as well. Their frames also are all using a Darktree, thanks to Jane Lin,
so I was able to cycle the hues and rotate them until they started and ended
where I prefer. #5 uses is a straight forward version of this. #6 uses a
"painterly" mottling effect to mess with the borders between the hues
in the frame. #7 is the same as #6 but the mottling is at a smaller scale.
I would REALLY like to know which of these seven
you prefer. I would also like to know which of the last three (5,6, and 7) you
prefer. If you prefer one of them but with an element from another please let me
know that too. Thanks. dgriffin@computer-music.com
#1
#2
#3
#4
#5
#6
#7
New Logo tests: (These are thumbnails, click
them.)
I am working on a new logo for my company using my newer
graphics abilities and, hopefully not as tall. This is my first possibility. The
first two thumbnails are for the same image. The first at approximately the size
I would use on the web page. The second twice as large so you can see more
detail. The third image is the previous one done several years ago in Raydream.
The first two need to have the white area trimmed away. I am also not yet happy
with how the upper left edge of the frame does not show a groove because of the
light angle. I may do a new frame that is deeper so it will still have a shadow
in spite of the nearly straight-on lighting. I used WFMM Bloom. I am not sure if
I like it. I used flat faced letters then applied their textures rounded so I
have some weird uneven surfaces on the letters. I was hoping the bloom
would soften the problem. I am still deciding if I like the results. I tried to
keep the height down with a narrower frame but I think the older frame is more
striking so I may make the new one a little larger.
Etched Glass Test:
I think this could look more like etched glass if I had
used a larger map than 512x512 and possibly a different fill algorithm in the
letters. This one was done using a texture map bump at a very low setting like
0.01. The same image was used with layered colors. I could have used a single
texture map for color but wanted to be sure this would work in other, more
complex situations. The first layer was plain color white and the second was
off-white using the texture map at 20 percent transparent. No filtering if the
image since it made the glass look less rough, more like it melted. I know I
could have used a black and white image to mask the etching area then use white
noise image or a noise filter to have more immediate control over the etching
surface texture. I tried it but got quicker results by just filling the letters
with a colored noise pattern in Picture Publisher.
Xylo-Bot story animation:
A xylophone playing robot. Design improved for low poly
and range of motion. Story animated. Motion to be fine tuned. Sound will be added
soon.
381k 128x96 Indeo
5.11 AVI 270 frames (9 seconds). Recommend NON-looping. Final pose is
supposed to stay on screen.
The story: The robot is playing the xylophone. A very fast 1 octave
chromatic scale then a 1, 5, 8 to finish the concert. The final notes are hit
stronger and stronger. Then he bows to the cheering audience. To the left, to
the middle, then as he is about to bow to the right he passes out from
exhaustion. The audio will include the xylophone notes, crowd noises and
cheering, and a nice thump as his head hits the ground, possibly accompanied by
a little microphone feedback for comedic effect. This version has the default
textures from Rhino for fast test animation. The final version will probably
have to be in this small a window or smaller to make room for the audio. But it
will probably include a moving camera perspective or cuts from one angle to
another.
Xylo-Bot rough animation:
A xylophone playing robot. Design and animation changes
planned. Sound will be added eventually.
211k 160x120 Indeo
5.11 AVI 101 frames LOOPING.
I plan on adjusting the motion curves for the down-swing to be faster than
the up-swing. I also designed the robot with too little range of motion (again)
so I am deciding whether to make a newer one on the same skeleton or start over
with the robot and skeleton and the puppeteer key frames. I am deciding if
the whole idea is worth the trouble. In theory he could play almost anything and
multiple copies could be produced to play different parts of the tune. But
showing both of them would make them each smaller in the view so you would see
less. The more elaborate the show the less there would be to appreciate in terms
of detail and accuracy. I also am thinking a simpler animation may be more
educational for examples on my web site. At least one of you will suggest having
the bars move as they are struck. It may be cartooney but in real life you are
rarely close enough and at the right angle to see the bars move and even then
the player has to be striking the bars pretty hard. I used Puppeteer to key
frame several positions and it allows me to create almost any tune by loading
key frames. That is how I did this example. I had two problems while setting
things up. First The joints would show themselves to be set at 0 degrees and
puppeteer would show them to be set to -1 AND the Min and/or Max would be off by
1 or 2. Eventually I hammered one joint into submission. I am not sure how. The
other one never capitulated. The other problem is that I included target cubes,
nulls, that, when the spherical hammer centered on them, would cause it to be in
the exact right spot to strike the relevant bar. But I could find no way to
align the hammer with the location once it was boned. I even lengthened the head
bone so its center coincided with the center of the hammer sphere but since I
could not individually tell that bone to go to that point while in IK mode
(requiring me to select the full skeleton instead) I had no way to use the
targets. Does anybody know how to do this? The best I could do was use them as
visual cues. And since I had to set each target's axes to center they were
visible which produced a nice axis in the center of each target I could use to
visually align the center of my hammer sphere. So, while I was hoping for an
easy way to exactly strike the bars I had to settle for a visual cue and manual
adjustment.
Spinning
Spheres & Lights animation
(no
thumbnail) 108k 80x60 Indeo 5.11 AVI 1.33 seconds LOOPING.
IK Robot animation
This is my first IK bones animation. Without anything but
the sky for motion reference it may not be obvious. Your perspective is that of
a camera person, maybe in a jeep, panning up at this towering robot only to see
that it is about to step on you so you start to drive out of danger. You avoid
being stepped on but see that the robot has turned its attention to you as you
drive away. I made the sky as my first custom sky using Michael Gallo's Infinity
1.2. I found using the fog foreground shader (this one is ground fog) softens
the seam at the horizon where the plane and environment sphere meet. The other
textures are tweaked darktrees except for the shiny part of the cockpit which is
my own. I envisioned it being blacked out so you could not see in but a pilot
could see out. But I also wanted it shiny enough to be reflective. I erred on
the side of reflective but it creates some neat reflections in the first few
frames. I made the model in Rhino so I could export at any resolution I wanted
but chose a low one for easier animation. This one started at 13,570 polygons
and I reduced it to 3,408 by boolean subtracting a cube from each part. This
project really taught me the dire need for at least two nails in trueSpace. I
would have also liked a constraint system so the turret would always stay level
unless I wanted otherwise.
Click here to see the movie.
Indeo 5.11 AVI 318k.
240x180 2.33 seconds.
It may be more interesting to step through the frames or grab the frame
slider and drag back and forth. I tried to keep it short so things happen
fast. |
|
|
Click here for
stills from the movie.
640x480 JPEG about 30k each. |
Test pose image
This is my first ever attempt at a robot or even a vehicle of any kind. I
imported a lower resolution than I would have liked for ease of animation later.
The top section is meant to be improved either by trying different pods or
adding weapons stations or even turrets. It has been successfully boned for IK
and that is how I posed it this way. I made a few mistakes in the design. The
knee joints can not flex more than 90 degrees because of their geometry. So he
can not squat down very far. One future idea is for two joints close together so
the legs could fold all the way against each other. I also did not allow a
plausible way for him to turn. I am cheating with the assumption that the feet
rotate around their center. A bad design but its just for fun anyway. I thought
it would be fun to build a pod that looks overly large for the feet. It should
make the statement that the technology is so good that it is very strong and can
keep balance very well. A longer pod would allow turrets on the underside front
and back so it could "scratch its legs."