Perhaps you would return them and continue to use V brakes?

That’s not how I roll.

The ideal solution (obviously) is to painstakingly measure the frame and caliper mounts, and then lovingly model a pair of adapters to lower the brakes to the rim! Huzzah!

The last question is where can I get these milled? I’ve used shapeways.com for 3D printing before, but their tolerances in steel are too rough for this job. Do you know a good place? Answers on a postcard.

The camera in question is a Flip Mino HD.

Handsome, no?

Not so much. This portable and useful camera is built around a tiny, tiny sensor. Whilst capable of shooting 1280 x 720 progressive video, it has a darker side. It suffers from rolling shutter. Badly.

A close approximation of my reaction to this news.

Some explanation.

There are two types of sensor used in digital video cameras, the older and cheaper CCD, and the newer, and more expensive, CMOS. CMOS sensors have the great advantage of all the pixels being read all at once. CCDs, however, are read line by line, meaning the top of the image is captured before the bottom of the image.

‘How does this affect me?’, you ask. It means that any objects with vertical lines in the image will become objects with diagonal lines, if the camera, or the object moves.

Here is an excellent video explaining this

If you need to camera track the footage, to generate a 3D camera track for Maya or Nuke, rolling shutter can compromise the quality of that track. Another area in which you will have difficulty, is if you try and composite rendered 3D elements into that footage. If the footage is skewed, and your elements aren’t, it’s going to look bad.

This distortion needs to be corrected, and to do so, we use RollingShutter, an excellent plugin from The Foundry for their compositing tool, Nuke.

For the Flip, I have had favourable results with a Correction setting of 0.7, and Shutter Sync set to 0.5.

The footage I’m working with suffers from this so badly, that I’ve opted to use RollingShutter to skew my 3D elements in line with the footage, rather than correcting the plate. This is because the grainy nature of the Flip video causes a significant number of artefacts to appear after processing.

All in all, this has been a learning experience, and it has taught me a valuable lesson. Rent a good camera, rather than buying a shoddy one.

Well, I better have something to show for all that time away, and the answer is, ‘Maybe?’.

I have been soldiering on with my university degree. My current modules are Special Effects, Research Seminar, and Screenwriting.

The research seminar is by far the most interesting, as I have been given free reign to investigate my chosen topic of Render Optimisation. The (likely wrong) report I made for the module is available here. Please sound off in the comments if I’ve made mistakes.

Screenwriting is a different matter, and I intend to hide that from human eyes until the end of time itself!

Special effects though… That has more than a few tricks up it’s sleeves, and I’m expecting a good laugh from that one before the end of the semester.

This will likely eschew a take-down letter, but here is a pre-vis or two of the SFX project.

Enjoy, and there is certainly more to come!

This was my first test of GoZ, although I only used it to transfer the base-mesh to ZBrush. It’s refreshing to not have to stop and save out a .obj of whatever I’m working on. That small break in the workflow always felt incongruous.

I rapidly block in forms with the move brush and the standard brush

The sculpt begins to take shape, with the nose, ears and the brow line becoming apparent.

I finally add the artificial eyes and strengthen the chin considerably.

I was stumped as to how the lips should look and, admittedly, it stalled the whole piece. I’m happy with this as a study, and I’m really motivated to work on some more!

Here is the piece of concept art I created for my Character Design proposal.

This is a quick paint over of a ZBrush sculpt, and as such, is not animatable.

I’ve been fortunate to have access to ZBrush 3.2, and I have no doubts that GoZ is going to completely change my workflow. It’s all looking very exciting!

The work is all very simple and was made in quite a rush, however, I am pleased with the way this piece has turned out

It is a mashup between a whale and a turtle. I was very interested by the skeletons of terrestrial whales and how the flukes of the tail have no supporting bone.

A quick concept sketch is laid down. Other areas are studied in closer detail on other sheets of paper.

In Maya, I rapidly build the body of the Whale from a cube. I choose to drop the hind flipper as it makes the creature resemble a turtle too closely.

Detail is sculpted into the mesh in ZBrush, with steadily increasing levels of subdivision to accommodate the increasingly complex details.

While building this I tried to reduce the time it took to reach completion. The entire process, from concept to final sculpting took a little over one and a half hours.

This coral is the first of almost 30 individual models for the interior. That number is for coral alone and does not include crabs, seaspiders, fish, detritus or the structure itself.

The base mesh built in Maya, shown smoothed.

A rapid sculpt from ZBrush. There is a minor issue with ZBrush interpreting elements of the mesh from Maya incorrectly.

I am more excited about the Lighting & Rendering module now than ever before, mostly because a very abstract and far-reaching idea is starting to become a reality. 

It is becoming readily apparent, however, that in order to deliver this scene in the next 11 weeks I’m going to have to dramatically step up the pace.

The three characters are a Herder, a Herd Animal, and a Parasite.

Below is the development of the herd animal, affectionately (and I think quite aptly) named, The Space-Pig.

In an effort to make a more intriguing and original creature I began with simple ink-blots. Here I can see the bony head of a creature with a very unusual jaw.

Next  I draw a fast and cartoony sketch. This will act as a guide during the modelling process.

In order to progress quickly with the creature, I lay out Zspheres in Zbrush. This allows me to interactively work with the proportions in 3D and experiment with different forms.

The Adaptive Skin mesh is imported into Maya. This mesh isn’t ideal and will pose challenges later in the modelling process.

The finished mesh in Maya. I have added accoutrements such as the start of a saddle and a set of reigns with a bit.

The partly finished sculpt in Zbrush.

I have not proceeded to add intricate detail to this model as I am unsatisfied with the proportions, and I will likely remodel it.

The reason for this not meeting my expectations is due to me losing objectivity and not taking the time to stand back and assess the model against the concept, and against what I wanted it to become.

I may still use this as a variation on the Space-Pig concept, the reason for the reigns and saddle, as it would make a pleasing hero-shot with a Herder astride it.