There are less than two months left until the beginning of Siggraph, where some of the best minds and most innovative companies in the CG industry will be on display. This will be the 40^th International conference and exhibition on computer graphics and interactive techniques and throughout the four-day event will feature everything from student work to blockbuster VFX.  The festival is recognised by the Academy of Motion Picture Arts and Sciences as a qualifying festival, and several works originally showcased at Siggraph have been nominated for or have received a Best Animated Short Academy Award.

The art gallery will explore CG work related to this year’s theme: XYZN:Scale; or rather projects which portray an innovative idea or story through scaling and transformation techniques. This year’s impressive preliminary lineup of production sessions will be open to the public and include the techniques behind the creation of the visual effects behind Life of Pi and Pacific Rim, as well as a presentation on the fusion of stop-motion and visual effects technologies used in blockbusters such as ParaNorman and Coraline.

Major software companies will also have booths at the festival. The Pixologic booth’s line up alone includes studio panel discussions by Blizzard Entertainment and Disney  (amongst others) focusing on how they use ZBrush, as well as individual artists such as Joseph Drust (from Ubisoft), Jared Krichevsky (from Aaron Sims Company) and Josh Herman (from Marvel Studios) giving live demonstrations of their work.

Check out the official website  for more information

Can you please tell us a little about Set Visions and the history of the studio?

Set Visions was established over 25 years as a commercial photographic studio, combining photographic and creative skills with emerging technology. Having transitioned from film to digital, and more recently (approximately five years ago) launching a unique photorealistic CGI brand named ‘PIX’, Set Visions has advanced, and can now successfully deliver dual discipline image solutions to a much wider audience.

Why did you decide to go from a photography-only studio to a photography and CGI studio? 

A key element of our success is timing, and a renewed awareness of CGI caught our studio’s attention at the right time! As the economic downturn began, particularly within the architectural area, we noticed a slight increase of CGI awareness within the home interiors market. This new prospect for CGI visuals in our market resulted in an acceleration of CGI technology, and so an opportunity arose. Supplying duel disciplines made us more competitive, utilising assets and adding value when budgets needed it most.

What tools/software do you mostly use at the studio for your CG work?

Software used are 3DS Max, Rhino, Vray & Photoshop.

You’ve mentioned that your team of CG Artists include photographers and stylists – in what way do they work together? How are the disciplines of photography and CGI combined?

Each image created is a collaboration of skills; CG artists mixed with professional photographers and stylists balance the aesthetics of every image. Our traditional creative studio process remains, uniting such skills as lighting, camera angles, decor and the use of CG technology to produce the final image.

What truly excites you about CGI today?

How it thrives through the ever-changing boundaries provided by developing technology. Passionate and obsessive, we enjoy the challenges this brings when creating photorealistic visuals. New tools open new avenues, enabling growth from 2D stills to movies, 360s, design visualisers and interactive web tools for use across multimedia.

What’s next for the studio?

The Set Visions team relishes the opportunity to constantly develop new skills and evolve. Through resource, talent and technology available, we aim to showcase our progression with each image created. Our CGI team is expanding, helping us to get more of our work recognised and viewed by a wider audience, beyond the UK and Europe…worldwide!

To learn more about Set Visions, please visit setvisionspix.co.uk

For the past 25 or so years I’ve been working in the videogame, TV and film industries, creating graphics and animations, directing commercials and leading art teams. While working at various studios I have hired many artists and seen various showreels drop through the letterbox – some very memorable, if only for all the wrong reasons. However, there were a few that have certainly made me sit up and want more.

This feature aims to guide you through creating and preparing your showreel to the standard needed to wow your potential clients or employers. Make them want to choose you for their next project. We will be looking at industry-standard Non-Linear Editing (NLE) tools and I will outline tips, tricks and the hard-and-fast rules that will help you get the best out of your work.

We’ll look at the importance of choosing the right music for your reel and how to create a soundbed using freely downloadable audio tools. We’ll explore how you can edit tracks using the timeline to accentuate your reel. I’ll also be giving my advice on how to promote your showreel; who you should be showing it to and which online websites you should use to get the maximum exposure for your self-promotion.

If all this is done well, your reel will become an extension of your talent, rather than just a list of clips edited together to a tedious techno track.

Start strong

The opening of your showreel has the biggest impact on the client, so it should be the most impressive part. Show the work you are most proud of and that represents your talents. Starting off strong will captivate your audience, draw them in and set the tone for the rest of what’s to come.

As an employer, when I press Play I expect to see a title, the name of the artist and what the showreel represents. Mine, for example, opens with ‘Chris Hill, Character Animation Showreel 2013’. Keep it short and use a simple, clear font with a backplate that won’t detract from the information you’re giving. Using a short, punchy title sequence can also give your showreel the edge, as it’s this sort of attention to detail that will get you noticed.

Create your showreel at the highest possible resolution with the largest bit-rate, so you can flexibly downsize if needed. I use full HD 1,920 x 1,080 pixels at 25fps wherever possible. If you have to scale your clips, make sure you maintain the aspect ratio or your clips will become stretched.

Keep it short

Keeping your showreel brief will help to focus the viewer on your particular strengths. Make it no longer than three minutes and no shorter that 30 seconds. A client will be able to get a feel for your skills quickly, so show diversity but don’t feel you need to explain your life story. Around one minute and 30 seconds is generally considered the optimum length for a successful showreel.

If you have an extensive amount of work over a number of fields, break the reels up and have one for each discipline; for example animation, effects and so on. This will enable you to apply for specific jobs.

Never repeat clips in your reel and if you only have one or two projects to show, it could be worth showing them in full, rather than cutting two together. Showing narrative rather than chopping back and forth in a jumbled mess would be the better option in a situation such as this.

Good pacing

Keeping people interested in your work is the main goal of a showreel. If you pace the reel too slow your viewer will very likely get bored – they may be inclined to skip forward or even hit Stop. If you time it too fast you will fail to show off your talents – another recipe for a quick exit. These rules can be pushed a little in either direction; if you’re editing a motion-graphics reel it can be pushed a little quicker, whereas character animation reels might be a little slower to enable a better read for your viewers.

Caption where needed

Captions will help the viewer understand what they are looking at. Keep captions simple and use a bare minimum number of words. I recommend captioning a shot with the role you played in creating it; for example ‘Character animation’, ‘Effects animation’ and so on. When you create a caption, keep the font simple and the size large enough to read, but not so large it overpowers the clip. Use a complementary colour – usually a simple white or a muted yellow that’s easy on the eyes. Be sure to keep your captions positioned consistently throughout the reel so your viewer learns where they are placed and can therefore read them quickly.

Stay in tune

Most people create a reel for personal promotion and think its okay to use copyrighted music. It’s really not okay. If the music is copyrighted then you should get explicit permission to use it. It’s much easier to find music that’s copyright-free from websites like www.jamendo.com, which hosts thousands of tracks to choose from. Just remember to credit any music you use and read any terms and conditions before adding it to your reel.

Using the right soundtrack will set the pace and speed of your cuts. You can cut on a beat or at key moments in the musical piece. Choose your track wisely and don’t pick a tune that’s offensive or difficult for the uninitiated to listen to. Using a track that’s too isolating will guarantee the soundtrack is turned down. You may think it’s the coolest music on Earth, but if it’s muted it will fail to serve its purpose.

Choose something that enhances your particular style of reel. If you’re creating an effects reel with particles and explosions, think about using music that’s punchy to help accentuate events. A fluid simulation reel might benefit from music that has a serene feel to it – something that flows along and complements the beauty of the simulations. Also, select music that has many facets to it rather than just a monotonous beat, as this will enable you to cut at interesting musical events. For example, if we have a clip where a large creature jumps off a tall building and lands heavily on the ground, we could align the landing with a key moment in the music or edit the clip so the creature lands on a beat to accentuate the animation.

Use sound effects

If adding sound effects seems like your kind of thing, check out www.soungle.com, www.freesound.org, or even www. soundjay.com for tons of free effects that can help add atmosphere to your reel. If you’re feeling really creative, why not write your own track that can be tailored specifically to your needs? You can find a collection of royalty-free tracks supplied with this issue’s disc (unfortunately unavailable for digital readers).

If applicable, lip-syncing should be demonstrated, but this is something you will have to consider when choosing your audio track. Put the speech into the edit, if at all possible, then add a new audio layer for the dialogue to the audio track and lower the music levels so the client can check your lip-syncing skills. It’s at this point you may find you haven’t used a soundtrack with vocals running throughout, so also consider this when choosing your music and placing your lip-sync clips.

There are lots of tools out there for editing audio, but one excellent piece of software that will help you edit sound effects or tracks is Audacity, which can be downloaded for free at http://audacity. sourceforge.net. Try to work with WAV files rather than MP3 if possible. MP3 files are compressed and suffer a loss of quality. Also use files that are stereo 16-bit 44.1kHz (CD quality). When you finally come to encode your reel, the audio and video will get compressed.

Once you’ve edited your audio clips, bring them into your editing software and layer them up in the timeline. You can then sync them to key events in the showreel. This process may take some time, but the impact it will have on your reel is huge.

Link your clips for the best effect

During the edit you will transition from clip to clip, but you shouldn’t just dump one clip after another. While this gets the work in front of the client, it pays to give every transition some thought. Good transitions create a nice, smooth flow throughout.

Simple five-frame blends can soften between clips and reduce the jolt from an abrupt cut. Quick fades also soften edits: simple take your first clip down to black for three frames and then fade your second clip for two frames. Try to use the shots to drive wipes and fades; for example if you have a camera move from left to right, create a five-frame wipe from left to right that follows the speed of the camera. This five-frame wipe adds detail to show you’re thinking on a creative level about every shot. Try not to use too many standard wipes and transitions often found in editing software, as they can be very cheesy and dull.

When putting together a reel it’s also good to show shot progression. A modelling reel should start with the wire cage to show how clean the model is, then transition to a shaded view and again to a textured model. If you have a compositing reel, then show the clean plate, add each layer and ultimately the complete shot.

Go out with a bang

The start and finish of your showreel will be the most memorable to your audience, so it’s vital it opens and closes on strong pieces. You need to make an impact and you want to leave them desiring more, so pick your very best piece to end of a high.

After the final clip has played out, add your credits and contact details – primarily your name, email address and telephone number – again in a clean, legible font. Leave contact information up long enough for it to be read slowly and then end the reel. Try not to fade out at the end, as some players will leave the last frame up on the screen.

Be flexible

You should always be prepared to refresh your reel on a regular basis, so consider updating it and sending it out at least every six months. Remove weaker clips and reorder your reel to cover new and better work. This is where you may need to adjust transitions and possibly tweak the cuts to fit in the new clips. Good luck!

Thanks go out to Filippo Veniero for this Blender tutorial. See more of his work at ifilgood.it

01 Begin modelling

We’ll use simple objects for this tutorial: a cup to fracture, a bullet to simulate the impact and a drinking straw in the cup to make the scene more interesting. Add a circle (Shift+A>Circles), then in Edit Mode Extrude (E) to model the cup. You can model a bullet in the same way (with a circle and extrusion) as well as the straw (Shift+A>Cylinder). Simply place the straw inside the cup and proceed with your lighting setup. Add a set of white light emitter planes above the scene and two lights at the side (one warm and one cold). Don’t forget to set Ambient Light in the World panel to 0

02 Textures & materials

We’ll use the Cycles render engine now, so choose a nice texture for the cup (here we’re using the Blender logo), select the cup and in Edit Mode unwrap the mesh. Don’t forget to use UV as the texture co-ordinate. Add a second material to the cup to be used as cell add-ons for the sides of the shards

03 Begin fracturing

In Object Mode, select the cup and move it to another level (M). Drawing with the Grease Pencil, break the lines of entry and the exit point of the bullet, then add some extra points that will form the other chips. Each point is the centre of a chip. Enable the Fracture add-on and select Cell Fracture. Set the Point Source to Grease Pencil, Recursion to 2, Material to 1, Margin to 0.000001 and Group Name as ‘fract’.

04 Set up the game engine

We should have the cup divided into fragments on a new layer. Select all the chips, move them to the main layer, pick Blender Game as your render engine and Game Logic as the layout of the windows. We need to freeze the fall of the fragments before impact with the bullet and enable the impact to its passage. Select a chip and set the logic nodes as shown, then grab all the chips (Shift+G>Group), hit the spacebar and search for ‘copy logic brick to select’. Use the same settings for the straw.

05 Adjust your settings

Select the bullet, then in the Physics panel tick Actor and in the Game Logic window add a new property: ‘trigg’ (it’s a variable used by logic nodes to unfreeze the animation). Add a motion node, as shown, to move the bullet along the X-axis against the cup. In the Game menu, tick Record Animation and then click Start Game Engine. Blender will simulate the impact and record all movements as IPO curves.

06 Rendering & post work

Select the Cycles renderer and go back to the window’s default layout. Choose a good frame from the animation, set the camera and enable the new motion blur for Cycles. Press F12 and wait until the end of the render. Save the image and open it in GIMP for post-production (Levels and Curves)

Simulation troubles

Here are some problems that may arise:

• If the add-on behaves strangely, and instead of the cup it shows a cube, the normals of the mesh are inverted.
• If the object has visible cracks after the add-on, close the gap between the various pieces using the Margin parameter.
• If the bullet doesn’t move all the chips, simply replace it with a sphere for recording the animation and then undo the replacement for rendering.

Thanks to James Morris, director of t-zero communications for this review

Essential info:

Price: Prices are available on the official website

Specifications:

• 3.2GHz Intel Core i7-3930K running at 4.4GHz
• 32GB 1,600MHz ECC DDR3 SDRAM
• 3GB NVIDIA Quadro K4000
• 240GB PNY PREVAIL ELITE Enterprise Solid State Disk
• 2TB Western Digital RED 7,200rpm SATA hard disk
• 18x Sony NEC Optiarc DVD rewriter
• Windows 7 Professional 64-bit
• Three-year warranty: the first year on-site; second and third years RTB

The Good

• Incredible graphics performance
• Good rendering power
• Decent storage selection
• Reasonable price

The Bad:

• None I can think of!

NVIDIA’s first Quadro card based on its latest Kepler GPU generation, the K5000, arrived in October last year. However, we’ve had to wait until now to see the full range appear. With the Magnetar S16-AW750 Armari has brought us the high-end K4000, which promises to be as much of a leap in performance as the previous Fermi generation 4000, but without a high-end price to go with it.

The S16-AW750 is based around Intel’s Core i7-3930K processor, which runs at a nominal 3.2GHz. Armari has permanently raised the clock speed, so this CPU will run at up to 4.4GHz across all cores and has a custom water-cooling system installed to ensure that temperatures still remain under control. The 3930K is the top non-Extreme processor from Intel’s Sandy Bridge generation, so offffers six physical cores, which hyper-threading presents as 12 virtual cores, giving parallel tasks like rendering a boost. The processor has also been partnered with a healthy 32GB of 1,600MHz DDR3 memory in four 8GB DIMMs, leaving four slots free for upgrading to the maximum of 64GB.

This brings us neatly to the star of the show: the new NVIDIA Quadro K4000 graphics card. Numerically, this is the successor to the Quadro 4000, but its specification even surpasses the Quadro 5000 and 6000 in some areas. For a start, it sports a whopping 768 CUDA cores, where the 4000 had just 256, the 5000 had 352 and the 6000 had 448. This would imply a threefold leap in performance over the 4000, but it’s not quite as simple as this.

The previous few Quadros had double- pumped CUDA units, where the Kepler generation keeps these at the same frequency as the main GPU core. This is sure to make the performance increase a little more muted and Kepler Quadros have reduced double-precision performance too, although this won’t affect most mainstream 3D software. Memory bandwidth has also gone up from the Quadro 4000’s 89.6GB/ sec to 134GB/sec and the quantity from 2GB to 3GB of GDDR5.

Armari has provided an impressive storage provision, too. There’s a 240GB PNY Prevail Elite Enterprise solid state disk, for super-fast main operating system and application loading, as well as a 2TB conventional Western Digital RED 7,200rpm SATA hard disk for general data. There’s an 18x DVD rewriter, but this more budget-conscious Armari system doesn’t include the company’s usual multi-format card reader.

The K4000 doesn’t disappoint when it comes to performance testing. The score of 96.69 in Maxon CINEBENCH R11.5’s OpenGL test is one of the highest I have ever seen. SPECviewperf 11 results were similarly impressive, with 91.78 in the lightwave-01 viewset, 113.41 in maya-03 and 67.14 in the SolidWorks-based sw-02. All are among the best test scores I’ve recorded. The Armari system also managed a very healthy 12.48 in the CINEBENCH R11.5 render test, which is what I would expect for a frequency- enhanced six-core Intel Core i7.

Despite these phenomenal test results, the Armari S16-AW750 still comes in at £2,295 ex VAT, which is a decidedly mid-range price for a workstation. This sums up NVIDIA’s new Quadro K4000 nicely: high-end performance for a mid-range price. It’s a very exciting proposition for independent 3D content creators or studios looking to expand their seat count without breaking the bank.

NVIDIA’s Kepler Quadro range

NVIDIA has taken an unusual route with its Kepler generation of Quadros

The high-end K5000 was released alongside the consumer-grade models based on the same generation about six months ago. We’ve seen a number of workstations using the card, with very impressive results. Sporting 1,536 CUDA cores and 4GB of GDDR5 memory, the K5000 is a formidable professional 3D accelerator. However, at over £1,300 it’s also pricey. Those looking for a more cost-conscious workstation have had to wait until now for the more reasonably priced K4000, mid-range K2000 and entry-level K600.

All three new cards are pretty exceptional. The K4000 is a little over half the price of the K5000, but offers much more than half the performance, as shown in this month’s review. The K2000 has an even more compelling value proposition for the professional 3D content creator on a tight budget. It offers 384 CUDA processing cores and 2GB GDDR5 memory, so has a better specification than the previous high-end Quadro 4000’s 256 CUDA cores, although memory bandwidth is only 64GB/ sec compared to the Quadro 4000’s 89.6GB/sec.

The Quadro K600 similarly offers specifications to compete with the next class up from the previous generation. It sports 192 CUDA cores, twice what the Quadro 600 had, the same as the Quadro 2000 and 1GB frame buffer. However, this is DDR3 memory, so bandwidth is just 29GB/sec, whereas the Quadro 2000 offered 41.6GB/sec. The K2000’s power consumption drops a little, too, drawing just 51W compared to the 2000’s 62W, although the K600 remains similar at 41W compared to the 600’s 40W. With twice as many CUDA cores, you’re getting quite a bit more performance per Watt.

Overall, NVIDIA’s Kepler Quadros bring unprecedented performance for the money, with the K2000 in particular providing high-end performance to low- and mid-range workstations.

Our verdict:

Features: 9/10
Performance: 10/10
Build quality: 9/10
Value for money: 10/10

The Quadro K4000 provides awesome performance, decent rendering and can be bought on a budget

Final score: 9/10

For more information, head on over to the Armari website.

“At Weta Digital, our focus is feature film visual effects. In order to fulfill the director’s vision for the shot we often create digital doubles of characters in the film. These digital versions of the characters are used in place of real actors in situations where it would be too dangerous to use a real person or it was not possible to shoot the scene with a live actor.

These digital doubles need to move and act like the live action character; indistinguishable from the actor. We use dynamic simulations on their hair, skin and clothing to add a high level of realistic movement. These replicate the effect a real environment would have on an actor and his/her clothing on a live set.

For The Hobbit we wanted to create a larger number of clothes that would be indistinguishable from the behaviour of real clothes when simulated. To do this we chose a pattern-based approach at the modelling stage which could accurately create the volume of the real clothing. Marvelous Designer 2 was chosen because it had a fast and stable solver for draping the clothes when modelling and a simple, user-friendly interface.

The patterns for the live action costumes used-on set were available for us to use as templates for our digital cloth patterns. We then created accurate digital copies of the characters’ bodies from reference pictures and scans and used these as our models.

Although we did not render the geometry from Marvelous Designer 2 , the wrinkles and draping the software produced were an invaluable starting point. This geometry was used as a template and the topology rebuilt in Maya with seams, buttons and other details added and used as the hero digital costume for our Models Department.

The Marvelous Designer 2 geometry was also rebuilt in our Creatures department with topology that fed into their cloth simulation approach. The hero geometry was wrapped or attached to the simulation geometry, which was invisible in the final renders.

For The Hobbit we had to produce a digital double of nearly all of the characters seen in the movie. This amounted to hundreds clothing elements – most of which used Marvelous Designer 2 in their creation. The most intensive use was the dwarf costumes that had up to five layers of clothing. The ability to manage that many layers in Marvelous Designer 2 was a real bonus and made the set up for cloth solves simpler and more stable.

Even though Marvelous Designer 2 was created as a tool for the fashion industry it was adaptable enough to meet our needs for film making and we were very happy with the results.”

You can learn more about Marvelous Designer 2 over at the official website.

Within Can Pekdemir’s world, bones disappear from faces while whales are dropped from great heights, deforming beautifully at the shock of landing. On first discovering his work, you can’t help but take notice of the oddly incredible works that Pekdemir presents, while on further investigation you’ll be glad that you wanted to know more. Though Pekdemir’s images are seamless and bizarrely natural, his execution is calculated. Nothing is done without reason, so should you spy melting bovine hind legs among his breathtaking portfolio of artwork, you can rest assured that there’s a very good explanation.

Pekdemir was born in Istanbul in 1982 and he is still based in the hustle and bustle of the city. In 2008 he took the leap of faith that so many fine artists wish they could bring themselves to take: he quit working for agencies altogether.

Swapping creative, corporate projects for a new environment where ideas and learning take centre stage, his move paid off. Pekdemir went to work at Istanbul Bigli University as a full-time instructor at the Department of Visual Communication Design. Here he feels the benefit and influence of working with people from different disciplines, giving him an excellent grounding to start introducing justification into his work. “Working in a school [provided me with the] great advantage of being able to reach both theoretical sources and technical equipment at the same time,” he explains.

Theory and methodology are an imperative part of Pekdemir’s approach to his art. While he’s stepped away from academia for the moment to focus on his work, his academic experiences still inform a lot of his creative decisions. Pigs can fly in Pekdemir’s world and when they do, he ensures they fly with great artistic merit.

Learning 3D for art

3D is the tool of Pekdemir’s trade, using programs such as modo and ZBrush to bring his ideas to life. “I’ve been into 3D for around 15 years,” Pekdemir offers. But his interest in computer graphics stretches right back to the demo-scene time when real-time 3D graphics were created by code. He later moved to using 3D software to create artistic forms. “For the last four or five years I’ve been using it not only to form shapes but as a platform for deforming them – sometimes by misusing tools, letting mistakes happen as I work and doing methodological pieces,” he says.

Digital 3D is a fascinating medium to work with, because it’s very nature can be misleading. “Even though we work in a 3D environment on our computer screens, it is actually a two-dimensional representation of a three-dimensional space,” Pekdemir explains. “Although we deal with all angles of a model, most of the time we end up with a two-dimensional image looking from a specific angle.” Because of this, the artist ultimately chooses for the viewer which side of the multifaceted creation should be observed. “With a real sculpture, the viewer has the freedom to control his or her point of view and walk around the work. In 2D images [this is limited by] the decisions of the artist,” he adds.

Defining methodology

3D may be Pekdemir’s method of choice, but the artistic decisions and theoretical nuances that form the final piece mean that his work takes on a three-dimensional quality quite separate from what a software package can offer. “I joined an exhibition in Istanbul’s 2010 European Culture Conference,” he explains. “At the time I was interested in the semiologic analysis of facial movements. I decided to analyse how a news speaker’s face moves and changes during different kinds of news titles, like politics and sports. By 2D-tracking the eyes and mouth visually, and using the audio form as a different visual element, I was then able to visualise this data and draw alternative news-speaker portraits. Following this, I continued working on methods to deform anatomy.”

While his ideas have developed and refined over time, Pekdemir points to a single statement that had a great impact on how he approaches his work: “Once I was reading an article titled ‘Art Common Sense and Photography’ by Victor Burgin, who is a British conceptual artist,” he begins. “I found this sentence in the article: ‘There is no content without a form and no form which does not shape content’.” Reading this made him revise his old work and try to analyse the decisions he had made throughout the creative process. After some analysis, the decisions and justifications of his work became a bigger part of the final outcome. “At the beginning of the ‘Procedurals’ series I made, I composited models into photographs by deforming the 3D shape affected by objects in the environment, or by placing them in deserted areas where no living thing is seen. These decisions were made very personally,” he offers.

Pekdemir’s ‘7 Sculptures’ series is perhaps one of his most methodological, linking concepts with images in both surprising and meaningful ways. “In the ‘7 Sculptures’ series, I tried to connect sculpture and photographs by using the time and location numerical data of the photographs, to control algorithmic image parameters,” he says. “This meant using specific numbers as input data to create a procedural in modo. This results in a unique virtual sculpture to be (de)formed for every location and time.”

Weathering the creative process

To begin any work of art is a tremendous act of violence, if the Bohemian-Austrian poet Rainer Maria Rilke is to be believed. In Pekdemir’s creative process, he finds the hardest part comes just one step before, while the final image is just a distant dream and the initial thought is still embryonic. “The biggest struggle is to be satisfied with the idea you are coming up with, not with the final visual output,” he says.

Pekdemir begins with a question and lets the answer lead him to the result. “Most of the time, I start my works as though they are experiments by asking: ‘How would a form react to a certain impact/force or to an extraordinary condition?’” Whether that form is a human body or an animal drawing of a theorem, Pekdemir follows through in his own way to break up the delicate membrane between idea and creation.

His images have sprung from questions such as, ‘How would we look if we turned into boneless structures in a second or if the gravity on Earth was different?’. He also challenges the boundaries set within 3D programs, questioning the use of certain tools and applying them beyond their original purpose.

These key ideas, deforming images and misusing tools, run consistently through Pekdemir‘s work. A series like ‘Soft Body’ illustrates the haunting results of morphing a recognisable shape. “‘Soft Body’ is an ongoing project to experiment with how a body reacts and deforms when it suddenly becomes boneless, or the rigidness of the bones change, or how we would look if gravity was different,” Pekdemir explains.

A glance over his other works shows that the act of distorting reveals just as much as it hides. The premise of his creations may be fictional, but the results offer the viewer a unique perspective into what is real. When asked how he wants people to react to his work and how he knows when he’s got that reaction, Pekdemir doesn’t have a clear answer. He does have an observation at one way his images have been perceived, though: “When I see my work featured on different websites, I notice the writers often use keywords like isolation, depression, loneliness and amorphous,” he notes.

Surround yourself with inspiration

While Pekdemir draws his ideas from posing himself questions and then answering them visually, he also believes that social environments play a big role in how you act and behave.

When he began at Bilgi University, Pekdemir found inspiration in the company of like-minded people, but he also thinks that the place where you live plays a role in the way you think. Born and bred in Istanbul, he feels that the rapidly changing capital is inductive to complex thought. “Social environments surely change people’s ways of thinking; their approach to handling problems and creating solutions,” he says. “Istanbul is a chaotic, crowded city and the centre of migration in Turkey… It’s home to over 13.5 million people and ranks as one of the largest cities in the world. You have to deal with traffic, catch buses, be at work and work harder to make a living here. This affects the way a person thinks and gives him or her a more complex point of view, rather than a simple one. You may lose your ability to think, in clean and simple terms.”

A 3D future

Looking to the future, Pekdemir has plans to overcome the limitations of 3D work in a 2D space, by using a 3D printer to take his images from the digital realm to the physical one. As such, he is planning to print works such as Clash (right) to a life-size scale, to give viewers the ability to walk around it and appreciate it from every angle, rather than stick to the angle that he offers them.

3D printing is achieved when a machine reads a digital design and lays down layers of liquid, powder or sheet material and then builds up from a series of cross sections. Since layers are added, printers have the potential to create almost any geometric shape – a must-have feature if they are to attempt to re-create Pekdemir’s artworks!

There are also plans to take the dynamic series ‘7 Sculptures’ one step further, as Pekdemir explains: “My future projection is to turn [the series] into an interactive project, where the user can be a part of [the art] by determining the location, time, space and exhibit it as a dynamic holographic sculpture.” For now, though, Pekdemir is focusing on his current work and developing images around his new fascination with cloud and plant forms.

]]> http://www.3dartistonline.com/news/2013/05/can-pekdemir-interview-the-fine-art-of-3d/feed/ 0 http://www.3dartistonline.com/news/2013/04/how-to-achieve-better-renders-with-keyshot-4/ http://www.3dartistonline.com/news/2013/04/how-to-achieve-better-renders-with-keyshot-4/#comments Fri, 26 Apr 2013 09:00:25 +0000 Chris McMahon http://www.3dartistonline.com/news/?p=9413

Thanks go out to Craig A. Clark for this LightWave and KeyShot tutorial. You can see more of his work at www.scorpiocgi.co.uk

 

01 Compose the shot

The first step is to set up the shot, so here I have the mechanical dragonfly modelled in LightWave (exported as OBJs) imported and all the materials assigned accordingly.

Next I check that I have all the material designations I need before going too far (especially if importing OBJs). I’ve selected an environment HDR image to be used. In this case a serene country road works well, as this subject should definitely be outdoors and the basic lighting it provides looks pretty decent.

 

02 Alter the environment

My chosen HDRI map is quite good, but is slightly flat and cold at this stage. So let’s warm things up a bit. I open up the new Edit option of the Environments tab and add a Pin (light source). From here I can colour, intensify and position the HDRI any way I like. The new Pin light is orange to replicate the warmth of the sun and I also reduce the brightness of the underlying original image to boost the contrast.

 

03 Let there be light

I’ve warmed up the tone of the image and things are looking much nicer, but I still need to bring greater definition to the shadows. Back in LightWave Modeller I load up the dragonfly body and wings for reference, then create a light for boosting the Sun, a fill light, and up-and-down lighting fills. I position these as I see fit, but they can always be moved around in KeyShot.

 

04 Configure the highlights

I can now export the lights as an OBJ file from LightWave and import it into my dragonfly scene. Once imported, these are in the correct positions compared to the dragonfly. I also set the Sun up as a Diffuse Point Light to provide some crisp shadows. My fill lights are Area Lights, which you’ll notice remain visible and provide options to hide certain attributes. Point Lights are never visible physically because they only have a single point of origin.

 

05 Reposition the lights

Optimise the position of the lights and place them where you think is best for the shot. You can also model them at world zero and position after importing into KeyShot. The lights can always be selected by the name of the material and then moved as required at any time (I always use the Move tool for this, as the Transform handles make it much easier). The Sun light does actually need to be moved to get the result I’m looking for.

 

06 Begin to render

Now it’s finally time to render out the scene. I always opt for 32-bit TIFF files, including Alpha Transparency, as I find it provides the most flexibility with the output file. Quality options enable you to specify maximum samples, while the Advanced setting defines the individual control of depth of field, sampled, aliasing and so on. I personally find the best solution is to specify the maximum render time. This leaves KeyShot to produce the optimal render it can within the set time.

 

You can see more tutorials like this in issue 54 of 3D Artist. Get your copy today through the Imagine Shop, or digitally through greatdigitalmags.com. Alternatively, why not make big savings on a subscription?

]]> http://www.3dartistonline.com/news/2013/04/how-to-achieve-better-renders-with-keyshot-4/feed/ 0 http://www.3dartistonline.com/news/2013/04/mari-2-0-review/ http://www.3dartistonline.com/news/2013/04/mari-2-0-review/#comments Thu, 25 Apr 2013 14:14:47 +0000 Chris McMahon http://www.3dartistonline.com/news/?p=9474

Thanks to Daniele Orsetti, texture artist and modeller at MPC for this review

Essential info:

Price: £1,210/$1,995 US
Operating Systems: 

• Windows, Linux

Optimal System Requirements:

• Quad-core processor
•  250GB disk space available for caching and temporary files (or a minimum of 50GB if you’re working on a small project)
• At least 4GB RAM
• Display with 1,680 x 1,050-pixel resolution
• AMD/NVIDIA graphics card with at least 1GB of RAM and OpenGL 3.2 support

The Good

• The Layers system is so similar to Photoshop that you can start working with it straight away
• Shared layers will make your life much easier
• Easily handle hundreds of high-res textures on large models featuring multiple UV tiles
• Fully integrated with Photoshop and NUKE

The Bad:

• System requirements are high so MARI won’t be right for everyone
• You can’t go back once you open an old MARI project in 2.0 – be sure to archive a backup first
• Workflow for creating a shared single mask across multiple channels could be easier. Perhaps soon?

MARI is a 3D texture-painting tool originally developed at Weta for Avatar. The Foundry acquired it around three years ago, and since then it has become the new industry-standard for 3D texturing, being notably used in nine of the ten feature films shortlisted in the ‘Visual Effects’ category at the Oscars this year.

The real beauty of MARI is that artists can paint high-resolution texture maps directly onto 3D models without ever having to worry about technical issues. The software can handle millions of polygons and more UV tiles than they’ll ever need. Most of MARI’s modules are also similar to the ones of other 2D applications, for example Photoshop, so new users generally won’t get lost when they first start learning how to use it.

The overall feeling when working with this software is that MARI has been created to take the best of what was out there and make it even better: from small tools like the multi-resolution Warp (Photoshop users are still waiting for this), to bigger features like the fully renewed Layers system of this latest version. MARI has also introduced brand-new concepts for the texturing workflow. One of these tools, for example, is the Paint Buffer, which is a plane that lets you edit your paintings before baking your work onto the 3D geometry.

While the core feature set artists are used to hasn’t changed in this new release of MARI, the big update is the introduction of the new 100 per-cent non-destructive layer system. In previous releases, artists had to create channels for every single image they needed to apply to a model, and then mix those channels inside shaders through shader modules. This could cause some confusion – particularly to users approaching the software for the first time. In version 2.0, The Foundry changes this workflow: channels are not the basic components of the texture any more and all of the work is done at a layer level, in a similar fashion to Photoshop.

Channels are now like folders, with each of them able to contain as many layers as needed. Shaders on the other hand are no longer used to mix layers and are now closer to the shaders found in other 3D programs, with the new ability to plug each channel with slots – COL, SPEC and BUMP, for example – and preview the result in the viewport.

The majority of the work now is done inside the Layer Stack palette. Here, users can create different types of layers. The most common is the Paint layer, which is an empty canvas that artists can use to paint or project images onto the 3D model (similar to the Channels of previous releases). The second-most important layer types are the GPU-accelerated procedurals, which are non-destructive editable layers such as fractal noises, tri-planar projections and tiled textures (previously in the Shader modules).

All these layers can be blended together using the common Photoshop blending modes – and also advanced blending – to key and remove only a part of the image. Integration with Photoshop is further enhanced by the possibility of importing PSD files as layers in the Layers stack, as well as exporting layered files out as PSDs.

MARI 2.0 doesn’t stop there, either. New and unique concepts have been introduced to make the layers experience even better than in rival packages. Each layer mask can be converted in a Mask Group, which is a layer stack of its own where artists can use blending, groups, adjustments and procedurals for a more flexible workflow. Layers can also be shared by Ctrl+Shift-dragging them into different stacks (or the same one) so that they automatically and instantly update wherever it’s used.

If you’re afraid of leaving Photoshop but curious to give MARI a try, I recommend you test it out. You won’t regret it!

Be prepared

MARI 2.0 is a beast, so you’re going to need a good system to run it on 

Due to the very high system requirements of MARI 2.0, in order to test it for 3D Artist magazine I had to get my hands on some sturdy hardware. Saving the day was the HP Z820 Workstation running Windows 7 (64-bit). The system featured dual graphics processor support, 32GB of memory, a 300GB Intel SSD and the AMD FirePro W7000 4GB PCI-e graphics card. The awesome graphics card coupled with the HP workstation gave me ample power to test-drive MARI 2.0 and give it the spin it deserves. Learn more about HP workstations at www.hp.com and check out the FirePro W7000 at www.amd.com.

Our verdict:

Features: 9/10
Ease of use: 8/10
Value for money: 9/10
Quality of results: 9/10

MARI has never been this powerful and flexible before. Try it!

Final score: 9/10

If you’re interested in trying out MARI, head over to The Foundry website

]]> http://www.3dartistonline.com/news/2013/04/mari-2-0-review/feed/ 0 http://www.3dartistonline.com/news/2013/04/mpc-digital-doubles-interview/ http://www.3dartistonline.com/news/2013/04/mpc-digital-doubles-interview/#comments Thu, 25 Apr 2013 09:00:03 +0000 Chris McMahon http://www.3dartistonline.com/news/?p=9426

Can you give us a ten-step process for creating digital doubles in today’s films?

1. Reference Photo Shoot – The actor is shot from multiple angles on a turntable using a polarised filter, and again without a filter. This helps to give a good base for texturing, specular reference for lighting, and will also be used as part of the modelling process.
2. Head and Body Scan – The actor is scanned to give a detailed 3D representation, which is used as the starting point for modelling.
3. Facial Capture Shoot – If there is a close up performance involved we capture this performance in 3D. In the past we have used techniques such as Videogrammetry to reconstruct point data in-house, or 3rd party capture vendors who provide an animated mesh of the performance. An advantage of Videogrammetry for digi-doubles is that we also construct an animated texture of the performance, which helps at the lighting and rendering stage, giving accurate colour pigmentation as the blood flow changes in the face.
4. Body Mocap Shoot – Quite often, the actions of the doubles need to be keyframed – since they are most likely performing the impossible! We capture body motions for the animators to mix with keyframe animation to help create a more believable performance.
5. Modelling and Topology – Once the scans and photographs are acquired, 3D cameras can be created to help modelling realise the model in 3D. The scan data needs to be recreated in a form more usable for animating and rendering, creating a neat mesh. We call this process ‘topologisation’. There will be rounds of feedback with the Supervisors to refine the topology and match the reference.
6. Texturing – The polarised and non-polarised photographs are stitched and cleaned up to create textures on the finished topology. Maps for specular, displacement and other passes are also created. The 3D scan is used to extract fine details.
7. Rigging – The bones, muscles, and skin setup are applied to the model, and puppet controls are created for the animators.
8. Animation – The rigged puppet is used together with mocap and keyframe animation to bring the digi double to life in the shots.
9. Look Development and Rendering – The model and textures are brought together in a lighting setup matching the shots to be rendered, and the shaders and textures are balanced to create a photo-realistic image.
10. Compositing – The rendered images are blended in with the filmed footage (referred to as the ‘plate’), and polished to perfection!

How do you go about accurately capturing the face of an actor and making sure it is as believable as possible when viewed on screen?

The team at MPC would aim to acquire as much reference as possible: photographic reference, filmed reference, static scanned body parts and dynamically scanned 3D facial capture. To make it as believable possible we need as much reference material as possible. We then ingest and understand what makes the actor individual.

What are the most challenging features to get just right in a digital double? What will audiences notice if done poorly?

The face! More specifically, the eyes can be difficult to get right. Not just the response to light, but the subtle animation of blinks and fleshiness surrounding the eyes. The rendered skin quality is also a big challenge. There are so many subtleties in an individual’s skin that solving it for one actor will not necessarily work so easily on another.

How do you go about seamlessly transposing from a real human actor in a shot to a digital actor, and vice versa?

The digi double will be matchmove animated to match the plate. Then we can project the 2D plate onto the 3D geoemetry, and work with that as an animated texture in lighting and compositing. Quite often we find that the full CG digi double, with dynamic CG hair and costume, looks more convincing than the blended transition, and so we decide to go with full CG all the way!

What new techniques/tools are making the creation of digital doubles easier and resulting in more believable on screen performances?

It depends on the type of shot. For regular mid-screen height, the advancements have been better photo reference acquisition, improvements in cloth simulation tools and more advanced rendering techniques. For close up shots the improvements in scanning dynamic, high-resolution facial data has helped a lot to create more believable facial performances.

Where can you see the practice of creating digital doubles heading in the future? How long, for example, do you think it will be before we see films that revive deceased actors in a leading role?

The trend has been for dangerous or physically impossible actions to be realised through digi doubles. This allows directors to be more imaginative with the actions, and to have the flexibility to add additional scenes or change shots after the filming has finished. I expect this trend to continue, with the film-making process continuing well after the shoot has wrapped.

However, I’d say it’s a long time before we are reviving dead actors! The best digi doubles take advantage of detailed scanning, photography and performance capture. With dead actors, this data doesn’t exist, and we have to rely on reference of existing performances to extrapolate to a new performance. This extrapolation will have assumptions and inaccuracies that I expect to fall short of creating a truly believable performance. It’s all about great reference.

If you want to read our full feature on digital doubles, then pick up 3D Artist issue 54 today! Get your copy today through the Imagine Shop, or digitally through greatdigitalmags.com. Alternatively, why not make big savings on a subscription?

]]> http://www.3dartistonline.com/news/2013/04/mpc-digital-doubles-interview/feed/ 0