I’m happy to introduce an article by Nicolas Richelet, offering insight into his visualization workflow relying mostly on 2D techniques in Photoshop. Fascinated by the works of Luxigon, Mir & Labtop, Nicolas started a personal exercise in exploring their style of visualization which he shared on the forums under the thread – #ap00t Tower. Since I was fascinated with this style myself, I started the Black & White Tower miniMAX challenge on the forums (which Nicolas took part in too) as a way to further explore this style with others. Since then I’ve asked Nicolas if he could share more insight about his workflow and he kindly distilled it into this article. Enjoy!
These images were made with a goal of studying the style of Mir, Luxigon & Labtop. Its a personal project that allowed me to explore a bit of this kind of rendering styles. In this making-of you will find some tips on how I made some specific parts of the images and an insight about my 2D workflow. I hope you will find some interesting stuff to help improving your own skills.
You might like to read the following articles before or after this article to get in the mood even more…
Before beginning the project I made some research on websites for renderings and architectural stuff and here is what I found that was useful to me
In the image above you can see:
- OMA, Jeddah Airport – Artefactory Lab
- ADPI, Biejing – Labtop
- Moonhyun Tower MIR
- Ateliers Jean Nouvel, Le Louvre, Abu Dhabi – Artefactory Lab
The major elements I get from these images were Composition, 3 points perspective, Wrapping and mood.
The 3D Tower
Here you can see how all the different part of the project are designed.
Here are the different parts of the towers: The floors, Pillars, glass and railing parts, and the wrapping.
I used a 3dsmax procedural map called cellular to make the wrapping element of the towers. This type of texture can be set with different shapes, and I choose the CHIPS one. There is a bit of reflection in the texture to catch light and environment, but thats all very simple.
Lighting Main Tower
For lighting the towers interiors, I was inspired by a Vyonyx tutorial and populated all the floors with textured lighted plates. I made a multi-sub-objet material with slightly different V-Ray Light shaders. The plates were placed with the help of the Multiscatter plug-in and concentrated on the borders of each floor, so we can see them from the point of view.
You can see below the shader of the office-lighting-plates. This is a Multi-Sub material with two slightly variations, one blue and one orange color. You can add more color variants if you want too. Using V-Ray light material allows me to light the office ceilings at the same time, giving the feel that they are occupied and busy.
Lighting Background Towers
The transparent background towers were textured in Photoshop but the textures was made in max. I made a render of some tower elevations, without the wrapping element, and this made me some 2D textures I used in post production. You can download these textures if you like
Im using them in Photoshop in screen mode most of time, and duplicate them when needed to increase their power.
2D Post Production
Heres the Brute 3D Render before any post production done on it
Im using different type of adjustment layers for color grading and contrast settings. You can see below the ones I used on the tower image. The most important for me is the Black & White adjustment layer. Here, the layer is in Overlay mode at 95%. This gives more punch to the rendering. On my composition, the Group containing the adjustment layers (called SETTINGS) is usually on top of the other groups, so it affects all the elements under it.
Wire Color & Object Selection
When I made this image a few months ago, I was using the V-Ray Wire Color render element to select different parts of the image. This can be useful if you want to be able to color correct or paint everything on the image, or if you want to create masks when you add people or foliage behind 3D elements.
Now Im using PSD Manager 3 from Cebas. it allows me to have more controls on the masks I want to get from my 3D rendering and outputs a PSD file directly with render, render elements and masks inside. Its a real-time saver.
- I have three options to make light glows, sparks and lens flares:
Paint with a soft round brush, in screen or overlay mode, usually to add glow to an existing light source.
- Finding light source images on the web (google images) , putting them in screen mode and cleaning them a bit to fit with my scene. This way is interesting because there can be unexpected and interesting result mixing some other source elements. This method is very good for experimenting.
- Using Photoshop plugins like Knoll Light Factory to create lights (see Lens flare example below, made with this plugin).
Here you can see the benefit of the same Lens Flare sample on the image. It really adds the atmosphere to this part of the scene.
Adding some car light streaks works the same as light sources and flares. On this image I took a night photo with light streaks (found on Google), and made it fitting the perspective with the Photoshop Transform tools. It was composited in Pin Light fusion mode. I added some styled painted streaks to add a bit of chaos in the middle of the straight red original streaks (see white and orange ones on the close shot below).
The look of your cut-out people and the way you are inserting them in the composition contribute to the overall quality of the image. Their clothes, color and attitude must be chosen with great care.
People must fit with the identity of the building(s) surroundings. In this image, even if the camera is away from the people, it was important for me to choose good people, because, even if you cant see the details in the distance, you can always feel them.
When I put cut-out people in post-production, Im always trying to choose people lighted in the same way the scene is, and if I dont find enough that fits, Im choosing uniform lighted peoples and paint light and shadows on them. In Photoshop, Dodge, Burn and Sponge tools are useful to me for this task.
Putting a lot of people can be a very repetitive and boring task. To help me saving time on this, I made some people boards. This is big PSD files where I have put many cutout peoples in there; each file has people sorted by category casual people (backlight, front light, uniform light, office, shopping, children, seated, and so on). This way, I dont lose time opening individual people cutout file, selecting the alpha channel, copying & pasting. Im just dragging people from their board to the composition.
When I need some kind of people, I run a script via Photoshop that simply open the file I need (according to the script buttons below).
And here is one of my people board
My (Actual) Photoshop Workflow
In my professional or personal work, I always try to make my Photoshop files as clean as possible, naming and sorting layers in groups. These groups are always the same and are always on the same level. This allow me to update my rendering quickly when I need to make changes requested by me or by the client.
On this project, I was not as organized as usual, I experienced ton of things and this leads me to a nice image and a big mess in my PSD File. Thats why I dont show too much screenshots of it, I dont think this would be of any help.
See below a screenshot of how the structure of a typical PSD file Im working on looks, on another quick test perspective (no 3D involved!).
Thats all for this making-of. You can see both final images below and If you have any questions, ask away and Ill do my best to answers them all.
Following your requests, in the comments, I’ve provided the original PSD file for the “CHIPS” image, reduced to manageable size of 352*640. The layer structure of the file is a bit messy, and don’t reflect my actual Photoshop workflow.[wpsharely][/wpsharely]
After several years as Jazz Pianist, Nicolas decided to dive into the graphic world by learning design & painting using digital tools and making visualizations. Nicolas did an internship in “Architecture Studio” Company in Paris, following later with a course at the Beaux-Arts School in Marseille to further develop his artistic skills. He then worked 2 years in Atelier 4, an architectural company in Clermont-Ferrand.
Nicolas joined the V-Pictures studio in Bordeaux in 2006 and after four years decided to leave it and go about his own way working mainly on images for marketing and communication, national or international architectural competitions, always exploring other creative fields, matte paintings, or another challenging works.