Tuesday, 6 December 2011

3D Realism: Practical & Easy Workflows



Hi All,

The above image is part of a project I have finished for a client recently. The final 3ds files were produced in two versions: for mental ray and VRay.
However, the techniques described here can easily be implemented across a variety of 3d applications and rendering engines.


You can find the entire detailed Manual series by clicking here on Amazon.

It's worth noting that, the 98 page manual only has the interior 3ds Max scene seen on the cover; and was published by, CreateSpace Independent Publishing Platform (an Amazon.com publishing company).

The Pre-production and Post-production chapters do not require any rendering engine.
However, the screen grab settings for production chapters are in mental ray.

If it's any consolation, there are two 3ds max scene projects: VRay and mental ray.

Finally, I have new book coming out, just for VRay:

3D Photorealistic Rendering: Interiors & Exteriors with V-Ray and 3ds Max

There's a blog post about it here:

http://jamiecardoso-mentalray.blogspot.cz/2013/04/3d-photorealistic-rendering-interiors.html



The main brief was to create an image that could sell the space to potential clients.
These types of marketing images are quite popular within the professional visualisation industry.
As such, I took the usual path of following meticulously three main steps:

1-Pre-production: This vital preliminary step consisted in taking the brief from the client. In this brief, they came up with their drawings and sketchy ideas.
I then took all the information on board; did some heavy R & D based on their original brief and presented them with a mood board depicting my vision of the art direction.
This mood board led to another brainstorming session between me and the client, which culminated into a final mood board that we both liked (i.e. final collage).

It was vital for me to signoff this crucial stage of the project to prevent changing the course of the art direction at later stage, and to ensure that we both had a clear idea of the final result.





2-Production: Having gone through the pre-production process and considered some of the artistic and technical approaches available, I began applying the shaders and textures. All shaders and textures were based on real photo references supplied by the client and sourced by me during the R&D process. I have carefully ensured that the scene had interesting reflections (i.e. plain and diffused); highlights from the sunlight and from artificial lights; direct and diffused shadows; glossy highlights from materials (plain and diffused) and the correct bump values on materials.



The following process was to begin lighting the scene carefully.
By that, I first introduced the daylight system to set the sunlight direction and its shadows in the scene. As guide,I have used references from photos of the agreed mood board.



Next, I began adding the artificial ceiling lights. There was a quick test render of each artificial light created (i.e. with a plain non reflective white OVERRIDE material).This was to prevent the scene from quickly becoming "bleached out" with too many lights.
Each artificial light had soft shadows.

I had also ensured that, there was a clear definition between lit and dark areas in the scene,in order to create depth.



Once I was fairly satisfied with the overall lighting, I had disabled the material override for the final big size region renders.


The final image Size was rendered at 4000x3000 pixels.

I had produced the renders for mental ray & V-Ray.





3-Post production: This process is arguably the most important stage towards signing off a project. All previous stages and the technical decisions will culminate into this final process.
Effects such as vignetting, depth of field, glow/glare, grain, colour correction, etc, are often applied and tweaked in post. Mostly due to these effects being under more scrutiny by clients.


While it's possible to implement all these effects directly from 3Ds Max, applying these affects in post increases the work efficiency when matching the photo reference closely. It will also provide more flexibility to add or omit such effects if or when necessary.


The above mentioned workflow allowed me to produce the final renders within the allocated time and budget, while using different platforms and rendering engines to achieve one result.

For the next few weeks, I will continue posting only key sections of a Manual entitled: 3D Realism: Practical & easy Workflows.
This new and unique Manual is aimed at intermediate and advanced users who may or may not be familiar with rendering engines and wish to consistently/repeatedly produce high end marketing renders while implementing 3 simple steps: Pre-Production, Production and Post-Production.

*This Manual is lavishly illustrated with step-by-step tutorials and it is filled with useful tips & tricks throughout as well as theories and analyses of each technique implemented.


*The Project 3ds files are for VRay and mental ray.


*There is detailed process of preparing and importing external files; followed by modelling from 2D drawings.


*Robust Post-Production techniques to polish your renders.


*...and much, much more.




From the Manual :

Section 1

Pre-Production:

1.1 Introduction

It’s common practice to establish from the start: the type of project involved; the target audience and the media platform/s in which the final product is to be showcased on.

Projects can range from a simple design exercise for a PDF/PowerPoint presentation; to high end marketing images for magazines, newspaper, billboards, TV, etc.

Design exercises or renders for PDF/PowerPoint documents, do not require the same level of expertise and precision as marketing images.

Marketing image gurus have in fact concluded that, it’s far more difficult to produce compelling images without a clear art direction (i.e. photo references; etc.)than otherwise.

The pre-production phase is designed to help the Studio/Artist and the client to establish the art direction prior to entering the production stage.
When failing to sign off this crucial stage (i.e. pre-production), the final product will be more likely to be exposed to criticism and constant reviews by the client.

Furthermore, it may subsequently result in “stretched” budgets and unfulfilled expectations by both: the client and the Artist/Studio.

The following tutorial will take you through the entire process of consistently producing captivating marketing images.

1.2 Research & Development

Creating prestigious marketing images often starts with the initial brief from the client describing their ideas, concepts, textures/finishes, design drafts, vision, photo reference, etc.






With all the information supplied, one should begin making preliminary decisions with regards to finding the best photo references to use for art direction, design possibilities, lighting, finishes, etc.
This process is considered be one of the most important stages of the pre-production, as the artist/s/studio will liaise directly with the client about the following:

1. The most relevant features of the project/design.
This will later prove crucial when setting up the cameras and the composition.

2. Find out the story behind the design/project. This may also play an important role in deciding:
The composition, the overall lighting rig/mood, the types of people to be added in post, and how people should interact with the space/shot, and with each other.

3. The media platform that the marketing image will be displayed on (i.e. Magazines, Billboards, TV, Newspaper, etc).
This process is fundamental when selecting the types photo references, as it will have a direct impact on the choice of the 3d contents, colour, camera position, render output size, etc.

4. Other alternative art directions. This process involves having numerous photo references representing other dynamic camera positions, composition, lighting, colours, colour scheme, materials/finishes, etc, that may benefit the final image.
As mentioned earlier, this is the stage where the artist/studio should source for numerous striking photo references relevant to the criteria (i.e. similar spaces, structures, design, media platform and ideas) to help the client decide the art direction- Mood Boards.
Most of these striking photos can easily be sourced from websites such as Flickr; Books; etc.
Mood Boards





1.3 Camera & Composition

The following step is for the artist/studio to create draft/low poly 3d versions of the most relevant items of the 3d composition.
These should be based on sketches/drawings and photos supplied by the client. Alternatively, one can use the 3d model supplied, if available.
3D Model from Client





Next, one can start setting up the camera positions that encompass the relevant areas of the design, based on the previous information supplied, and the photo references sourced by the artist.

With help from the artist/studio, the client should sketch out the camera position/s, and the field of view (FOV); preferably on a 2d drawing print.
Printed Drawing with Camera Positions




This is a good technique to quickly assess where the camera/s should be placed.
The camera position and its settings play a crucial role in producing the final render.
It is very important at this early stage to decide how one wants the final image to be interpreted through the camera lens (i.e. in an artistic, cinematic or standard photography manner).
Even untrained eyes can quickly assess whether or not an image is realistic, based on how the camera is positioned or setup.
The camera’s FOV (i.e. field of view) should reflect the values commonly used in real cameras, coupled with appropriate render output size to help capture the scene’s essence (portrait/landscape). The final result should be based on how well the camera, the scene and its contents complement each other in a dramatic and effective way.

Having the camera at eye level (i.e. 1.60m) or another realistic position is important for two main reasons:

1. Our eyes can easily spot odd camera positions (i.e. unusual height/position).

2. Having the camera at eye level will facilitate integrating people, and other objects in post.
At times, even accurate scenes may look disproportionate, as result of not having the camera at eye level.
Creating a Target Camera



Professionals often create the original camera type as “Target” to set its direction. Once that is done, the camera is then immediately changed to “Free Camera” type, for more flexibility in moving/rotating the camera in the viewport.

Camera at Eye Level





Setting up the Camera Output Size





“Signing off” the shot/ composition also involves going backwards and forwards with the client, while deciding on the geometry that should be in the foreground and background.



1.4 Previewing the Art Direction

During the process of establishing the final art direction, you will be required to “Photoshop” elements such as photos, notes, sketches and other effects on top of the previously taken screen grab, or a draft render. The final result should be an artist’s impression, or a collage depicting the overall art direction of the final shot.

The entire exercise is designed to help the client quickly understand and preview the impact of their choices based on the design and the artistic decisions previously made.
Draft Render of the Camera Shot for Collage





Creating a Collage preview of the Final Result


It’s also worth mentioning that, while liaising with the client, one should be adventurous by trying to break the mould, and suggest new ideas and effects such as:

*Vignetting

*Subtle surface discrepancies

*Chromatic aberration (…this is the one effect that you should not mention to the client, but simply add it in a very subtle manner).

*Depth of field (it is mostly relevant when there are objects in the foreground).

*Glow

*Lens Flare

*Dynamic camera positions

*ETC

All the above mentioned steps will aid the client when reviewing the overall budget and the final art direction.
It will also help avoid the usual project constraints related to lighting, composition, textures/finishes; final quality, etc.

As mentioned earlier this article is just a small section of my latest Manual.


For this Manual, while a 3d software and a rendering engine had to be chosen as a platform; the majority of its content focuses mainly on workflows and general methodologies used by most reputable studios. These techniques can easily be implemented across a variety of 3d applications and rendering engines.

The image depicted below is one of many which proved to be remarkable. I have produced them while using 3Ds Max and VRay, for a client.
.




I hope you have found this post interesting.

Ta

Jamie

More tips and Tricks:

Post-production techniques

Tips & tricks for architectural Visualisation: Part 1


Essential tips & tricks for VRay & mental ray

Photorealistic Rendering

Creating Customised IES lights

Realistic materials

Creating a velvet/suede material

FoxRenderfarm

www.arroway-textures.com 

Renderpeople

Gobotree




.

Wednesday, 30 November 2011

3D Realism Practical & Easy Workflows (First Manual) Patreon





















 More chapters will be serialized soon...


It has always been my ambition, to be able to dedicate my Full Time helping and sharing my knowledge/experience with up-and-coming 3d artists, as well as Veterans.   

Due to pressing work commitments, it has been difficult to fulfil this long time passion of mine, wholly.  

However, my free time (...dwindling ever so much), is spent publishing free content on this blog. 

Recently, I have been presented with an opportunity to fully serialize many of my tutorials.   In order to do it diligently, I will be required to take some time off my normal work.   

By supporting this project, as my Patron, in patreon.com, YOU will enable me to dedicate more of my time publishing/serializing this content, and future ones!   

Without Your invaluable support this project will drag on for a lot longer than necessary! 

For more information about becoming my Patron, please visit my Patreon page on: 




Thank you ever so much for your support 

 J

.

Sunday, 12 June 2011

RichDirt: The most natural solution



The idea started years ago when I first pleaded with Matthias (then, working as a Director of product integration, at mental images) to develop a plug-in that could aid professional 3D Visualisers to integrate their 3D elements into Photomontages; verified views, and general external shots!

My desperate plea spurred him to single handily develop a ground breaking plug-in that is now known as RichDirt (now available for VRay also).


In addition, check my New Book: 3D Photorealistic Rendering: Interiors & Exteriors with V-Ray and 3ds Max


Most professional 3D Visualisers can relate to the laborious process of making a pristine/clinical external daylight render to look believable/photorealistic.

While the above mentioned is the general consensus amongst 3D professionals, there might be the odd project that may involve great buildings and designs (i.e. numerous modelling details; superior material finishes; nice metallic surfaces; glass reflections, appealing colour schemes; great compositions; etc.) that can make one’s life a bliss when creating photorealistic CG images.

Since the above mentioned projects are very few, one still has to be faced with the reality that most projects are riddled with challenges related to integrating CG elements into photographs, or even just making a full CG external daylight shot to look realistic!

One of the most popular techniques amongst CG professionals is to add very subtle surface discrepancies such as dirt, etc.

This popular technique had been borrowed/ copied from the film industry.

While Film/production companies make discrepancies very prominent in their shots, professional 3D Visualisers tend to make it as subtle as possible; as clients often expect their buildings to appear relatively New (i.e. Architects, Developers; etc.).

We are all too aware of the fact that, this perception from most clients does not correlate to reality:

The fact remains that during the construction of any building (i.e. 1, 2, or more years); surfaces are constantly exposed to adverse weather conditions and human activities (i.e. rain, wind, dust, sun rays, scratches, splotches, etc.) that will unequivocally affect the finishes/surfaces over a period of time.

While the natural building changes caused by some of the factors mentioned earlier are visible, they still look subtle and acceptable after just 2, 3 years in construction; therefore still considered/perceived to be relatively “New”.

If it was humanly possible to omit all discrepancies accumulated during the period of construction of a building, it would probably look CG (i.e. unreal).

Human eyes can inexplicably notice the absence of these natural and subtle discrepancies that we often take for granted!

The Job of a professional 3D Visualiser is to try to emulate the same type/s of discrepancy subtleties that often go “unnoticed”.

In doing so one is essentially making a CG render as realistic as any other professional architectural photograph!

This is one of the many reasons Film companies and highly specialized 3D Visualisation firms use captivating real photographs as reference, to achieve the ultimate result!

The images below depict some of my numerous 3D Visuals that involved having to manually paint extensively subtle discrepancies (i.e. dirt, etc.), in order to fully integrate the shots!

At the time, one of my biggest challenges was to be able to individually paint and “churn out” thirty different shots of each project, in a very short period of time!









RichDirt is currently the best suited Plug-in to address the technical “nightmares” highlighted earlier, especially due to its powerful/superior multiple render elements’ functionalities.

In addition to being ultra-fast to render, RichDirt also provides users with numerous options to plug/ mix with other proprietary shaders, and a variety of Max’s procedural maps such as cellular, smoke; Mix; Noise; Speckle; etc.

Furthermore, while 3D Visualisers will benefit immensely from its powerful functionalities, users from Games and Film industry may also find it utterly indispensable in their everyday projects.

RichDirt is currently designed for Max/ mental ray and V-Ray. However, since users can efficiently output its effects as multiple elements (i.e. rendered passes); it is possible to overlay the elements in post, whilst using other rendering engines (i.e. Maxwell; FPrime; etc).



The following tutorial (for mental ray) will hopefully help users understand some of its amazing key functionalities:

First ensure that the mental ray renderer is loaded.



Also, it is utterly crucial to set the display unit scale, as this function will help determine whether or not one is working in real world scale (i.e. door height: 2.13m; chair seat height: 0.4m; etc.).




Prior to beginning to apply the RichDirt, it is commendable to select few key objects in the 3Dscene, followed by isolating them; as opposed to test rendering the entire scene…which could be time consuming.

Finally, disable all lights; exposure/environment settings; reduce the rendering output size along with the general FG and global rendering settings.

This workflow will speed up the test renders toward the final desired results.

Isolate key objects




4-There are multiple ways of using RichDirt:

A-Directly onto a 3D object

B-As a rendered element/pass in post.

This exercise will focus mainly in its functionalities as a rendered element.

We will start by loading RichDirt into Max first.

RichDirt is only accessible through the main diffuse material parameters toggle, and can only be found under the “mental ray rollout” group, in the “material/ map browser” dialog.




5-To quickly fine-tune its settings and speed up the process, we are going to primarily use RichDirt “globally” as an override:

A-Open the “render setup” dialog box by clicking the F10 key, or by simply choosing it from the “rendering” tab dropdown list.

B-In the “processing tab”, enable the “material override” function.

C-Back on the “material editor” dialog, return to the main material parameters, by clicking on the “go to parent” button.

D-On the main material parameters, drag and drop the shader from the “material editor” dialog onto the” material override” toggle.
Choose the “instance (copy) method” option.




6-The RichDirt from the material editor is now linked with the material override.
Do a quick test render (i.e. Shift+Q) to see the results.



7-The test render is looking ok. However, it is looking slightly grey and reflective.

On the “main material parameters”, decrease the “reflectivity”; “glossiness” and the “glossy samples” values to 0.0.

Also, change the RichDirt colour to plain white.
Test render again to see the results.




8-The render now looks slightly better.

Due to the viewport that the render was taken from (i.e. orthographic view), the image is looking slightly faceted, and with artifacts.
To correct this, simply change the viewport to perspective/ or camera.

Also, to begin tinkering with some of its pre-sets, choose the “rich building +streaks” from the dropdown list; and test render it again.




9-With the “Perspective/camera” viewport selected, the renders look smoother and cleaner.

Most RichDirt settings are self-explanatory, especially when hovering with the mouse over most parameters and functions.



10-To quickly go through some of its core parameters, we are going to start with the “Clean” function.

Clean: This function determines the main colour of the geometry.

To change its default colour, simply left click on its colour swatch and choose the desired colour from the “color selector” dialog, as previously done.

Alternatively, users can also plug a procedural map, texture, or a proprietary shader onto its toggle!


Dirt: This function determines the main colour of the dirt areas of the geometry.

To change its default colour simply left click on its colour swatch and choose the desired colour.

Alternatively, users can also plug a procedural map, texture, or a proprietary shader onto its toggle!

Accessing procedural maps; etc.



The “Spread” value generally helps to feather the prominence of the dirt effect on the surface.

These values can be positive or negative, depending on the effect intended.




11-The beauty of this plug-in is that one can save out multiple variation effects as a separate rendered element (i.e. passes), to later compose in post…without any render compromises.

Personally, I like both results (i.e. value of 20.0 & -20.0) of the “spread” value.

It is also worth noting that, RichDirt rendered elements (i.e. passes) have the perfect contrast between bright and dark areas; therefore excellent to compose in post.

This process will be later explored!

Furthermore, the inner occlusion (IO) is one of the best and most unique features of this plug-in:

It essentially computes realistically the occlusion from the edges of an object. The images below depict its superior effect on surfaces.

"Inner Occlusion(IO)", disabled

"Inner Occlusion(IO)",enabled


12-The “Radius” function sets the distance of the dirt from its point of origin.

High values may turn the renders a bit dark, and make the dirt to lose its definition.

Users can also plug a procedural map, texture or a proprietary shader into its toggle.

"Radius" value set to 1.5m

"Radius" value set to 5.0m



13-The “Falloff” sets the amount of decay as the dirt expands away from its main point of origin.

This function is quite good to set the concentration of dirt on certain areas of a surface.

Users can also plug a procedural map, texture, or a proprietary shader into its toggle.

"Falloff" value set to 1.0


"Falloff" value set to 5.0



"Falloff" value set to 0.05



14- The “Bias” function helps to reinvigorate the appearance of dirt in a variety of directions (X; Y; Z).

Users can also plug a procedural map, texture, or a proprietary shader into its toggle.



15- In the “Extreme” group, there a number of functions to help RichDirt pick up tiny details/objects in the scene.

By enabling the “optimize thin” function, the plug-in will help accentuate the dirt visibility in areas that would have been “overlooked” otherwise.

The “Horizontal” ; “Upper Edge” and “Lower Edge” function sets the direction of the accentuation.

Users can also plug a procedural map, texture, or a proprietary shader into its toggle.

The “Up Slope” and “Dwn” function sets the degrees at which the “streaks” should stop following at.

The default value of 0.0 is set for the “streaks” to follow the surfaces in all degrees.

"Optimize thin", disabled

"Optimize thin", enabled

"Lower edge" set to 5.0



16-Under the “streaks” parameters rollout, we have the “streak map ” toggle and its parameters.

A-The “Streak map” toggle allow users to plug a procedural map, texture or a proprietary shader into its toggle.

By default, this function is overridden by the “Builtin” function.

The “use builtin Noise instead of Streak map” function overrides the “Streak map” toggle.

The “Builtin” function tightens up the streaks’ appearance.

"Builtin" value set to 3.0

"Builtin" value set to 13.0



The “For AO” function enables streaks on ambient occlusion (AO); and the “For IO” enables the streaks on inner occlusion (IO).

"For AO", disabled

"For AO",enabled



The “Amount” values determine the visibility on the streaks on the surface.

"For AO" and "For IO", set to 100


"For AO" and "For IO", set to 1



17- After having gone through most of its relevant parameters, it is commendable for users to save out multiple pre-sets as render elements (i.e. passes) for more control over its effects in post!

Let us start by first saving out the first RichDirt settings under the name of “strong Streaks”, in a new material library.

To create a new material library, simply select the relevant material slot on the material editor.

Next, click on the “get material” button to open the “material/map browser” dialog; followed by clicking on the “material/map browser options” button, and choosing the “new material library…” option from the dropdown list.

Name the material library “RichDirt passes.mat”.






18-Now that a new material library had been created, one needs to add the RichDirt material to the library.

While in the “material/map browser “ dialog, scroll down to the “sample slots” rollout. Select the relevant sample slot, and right click on it.

On the pop up list, choose the “copy to” option, followed by selecting the “RichDirt passes.mat” material library.




19-The next phase is to rename the saved material as “strong Streaks”.
Scroll back up to the newly created material library rollout.

Select the pre-saved material; right click; choose the “rename” option from list and type in the name in the dialog box.





20-To save the most recent changes; simply select the “RichDirt passes. mat” rollout, from the “material/map browser “ dialog, and right click.

On the dropdown list, choose the “ save” option from the pop up list.





21-To add the RichDirt as a render element (i.e. pass), simply open the “render setup” dialog box(F10).

Next, click open the “render elements” tab, and hit the “add” button; followed by choosing the “mr shader element” from the dialog list.




22- Once the “mr shader element” is loaded, scroll down to its “parameters” rollout, and hit on the “shader” toggle to load up the pre-saved RichDirt material.

Pick the relevant RichDirt material, and close the dialog.





23-Once loaded, drag and drop it onto one of the material editor slots. Choose the” instance (copy) method”, to close the dialog.





24-By default the pre-saved material is under the name of “map #66 (Rich Dirt). This is mainly due to the render element toggle only accepting the RichDirt map itself, as opposed to the entire shader.

Rename it “Strong Streaks”, by typing it on the material name field.





25-To create new variations and effects of RichRirt, simply drag and drop the current slot onto the adjacent material slot, and rename it .

Next, load up a new RichDirt pre-set from the dropdown list.





26-To test render the results of the newly created pre-set; simply select the original material slot from the top left slot (i.e. render override shader); followed by dragging the newly created material slot and dropping it onto the “diffuse” toggle of the top left slot.

The newly created material is now part of the material override toggle.
It is worth test rendering to see the results of the new RichDirt settings.

Prior to test rendering, disable the render element functions, to prevent Max from automatically saving it/them in a folder directory.







27-We now have multiple RichDirt parameter variations to use as rendered elements. If not satisfied with the current results, one could alternatively tweak the parameters further by plugging /mixing some of its toggles with Max’s procedural maps; shaders; textures; etc.

Once satisfied with the results, simply put this sample slot onto the“ RichDirt passes. mat” library, and save it over again, as previously done.






28-If there’s no desire to create new RichDirt variation elements, simply enable the render elements functions again, and add the newly created material to the list of elements, as previously done.

Also, it is worth renaming (i.e. type in) this newly created element from the “selected element parameters” name field.

This is mainly for one to be able to clearly distinguish the attributes of each listed element.






29-Now that everything is set, one can exit the isolation mode; disable the material override if desired; increase the sampling quality ; reset the render output size; set the t file type to TGA; etc.

One of the great aspects of the previously discussed methodology is that, one can access this pre-saved material library from a variety of different 3d scenes, on demand.





30-It is worth noting that one can also add other render elements such as ZDepth; Object ID; mr A&D Output: Diffuse Direct Illumination; etc; if desired.

As previously mentioned, the rendered elements’ contrast between the dark a bright areas are very apparent.

This is excellent for blending in post!





31- Once everything is rendered; in Photoshop, bring in your main rendered image and your elements.

To begin overlaying document layers, select one of the rendered elements document.

Next, open the channels tab, hold down the "CtrL" key and left click on the “Alpha” channel layer to select the main layer.

There should be a selection marquee around the Alpha area. Copy (Ctrl+C) the selected layer from the main document.

Now select the main render document, and paste it (Ctrl+V) the previously copied layer onto it.





32-Once the render element is on top of the main render, and adjusted, rename it and choose a different colour, to make a clear distinction between the two layers.

To edit layers, simply select it; right click and choose the “Layer properties” from the list.






33-To begin blending the render element, simply choose the “multiply” blending mode from the dropdown list, as it worked best for this exercise.

Next, create the “add a vector mask” tool. The “add a vector mask” tool is excellent to partially or fully omit elements of any given layer.

On the “tools” dialog box, select the brush tool (B); set the opacity to 30%, and begin brushing around the undesired areas.

The “add a vector mask” tool works best with white and/or black colours, when masking with the brush tool.

The black brush colour omits areas of the document; and the white brush colour does the opposite.

It is worth noting that, in addition to using the “multiply” blending mode, one can also reduce its overall appearance by simply decreasing its layer opacity value.

Bring in the other render element/s and repeat the previous steps.







34-The 3D renders below, depict the before and after shots using RichDirt.

As previously discussed, most 3D Visualisation clients would have preferred the RichDirt effects to be more subtle.

However, for the sole purpose of this exercise the RichDirt effect/s were made very visible...for obvious reasons.






I hope you have found the article interesting.

For more in-depth information and download of this Plug-in please visit:

http://www.enrichpro.com/en/richdirt/index.html


So far, I would like to give a especial “thank you” to the following group of experts; talented individuals and companies that have taken the time to look; test and recommend this product:

Cyrille Fauvel: Worldwide ADN Sparks program Manager; Autodesk Developer Network


Dano Battista: Director @ DB3D.co.uk


Iain Banks: Senior 3D Visuliser and owner of iainbanks.com


Graham Macfarlane: Visualisation specialist and Co-owner of Elyarch Ltd

Maher Zebian: Senior 3D Visualiser @ Glass Canvas Productions


Thorsten Hartmann - Infinity Vision



I hope you have found this article somehow useful!



My 3D Portfolio:

New Book: 3D Photorealistic Rendering: Interiors & Exteriors with V-Ray and 3ds Max


More tips and Tricks:

Post-production techniques

Tips & tricks for architectural Visualisation: Part 1

Essential tips & tricks for VRay & mental ray

Photorealistic Rendering

Creating Customised IES lights

Realistic materials