IRender Tutorials - Realistic Human Figures - Quick Tips
Realistic Human Figures in IRender nXt and Photoshop – Quick Tips
Creating photorealistic figures in SketchUp and IRender nXt can be quite tricky. I personally think a good quality model is essential to any interior design presentation and adds realism and a sense of scale.
The simplest way to create a photorealistic model is to insert a 2D photo based cut out of a figure , however the resultant render tends to be flat and washed out due to the complexity of the light configuration usually required for most interior visuals. I have found in my experience the best way to achieve realistic human representation is to insert good quality models in Photoshop afterwards. This tutorial assumes you have a basic knowledge of Photoshop and SketchUp.
The YouTube link below was posted by a fellow forum member Al Naggar Shows the 2d model process:
Contents
- 1 Tip 1 - Model Scale and Perspective
- 2 Tip 2 – Is There Life After The Internet !
- 3 Tip 3 – Save those Files !
- 4 Tip 4 – Make those models work harder !
- 5 Tip 5 – Improving 2D Cutouts
- 6 Tip 6 – Using Props for realistic integration of 2d cutout models
- 7 Tip 7 – Walking Figures
- 8 Tip 8 – Speeding up Model Editing process
- 9 Summary
- 10 See also
- 11 Video Tutorials
- 12 Videos
Tip 1 - Model Scale and Perspective
To make a figure look correct in a visual it is essential the perspective and relative scale of the object is correct to make the person appear integrated and part of the scene. The quickest way is to produce a second render with figures only and overlay this in Photoshop to give a guide to position and scale. When I initially started using IRender nXt I included the `scale figures in the original render spending most of my time removing \cloning out bits of model you do not want to see. This way you will have a separate layer you can turn off after use and not have to spend the majority of time cloning or removing unwanted image artefacts. If time gets short before that big presentation you can also reduce the number of figures inserted without having to spend time editing.
The final visual
I have positioned all the figures on a independent layer and render the figures on a quick one pass render separately.
This image shows figures only – please note I have also included a few `props` these are useful to help registration of overlay if you render in different file type or scales. Also I always explode the PNG based 2D cutout models as IRender nXt seems to sometimes have a little problem with them and ignores them on render – exploding seems to cure the problem most times.
Remember to turn off sky ,ground shadow and ground pane before render and save image has transparent PNG file (see tutorial 1).
Open both the `visual` and the `figure overlay` image in Photoshop and copy and paste figure overlay into main visual. If you have saved both files as .PNG the scale should be the same and match accordingly. This will now provide an easy template for inserting and editing accurately your model files.
Tip 2 – Is There Life After The Internet !
There has been many a time I have spent hours scouring the internet for the correct type of model only to nip into the kitchen with my compact digital camera and take a quick shot of `her indoors` or one of my kids – have the subsequent image background isolated and figure inserted into my visual in less than 30 minutes ! The moral to this quick tip is that in this digital editing age there is always something near to hand that will do the job. In the shot below I have quickly caught my son eating a snack just before I produced a visual for a student eating area.
Tip 3 – Save those Files !
I produce a lot of school interiors and therefore am very lucky to have some very `cheap` models to hand in the form of two teenagers – Every time I bribe then to do this I make sure I get a lot of different angle shots and have now saved a large reference set which speeds things up a lot .
Tip 4 – Make those models work harder !
Once you have created your model inserts – make them work harder by making them more versatile – in the example below I have used various Photoshop techniques to produce separate layers to change hair colour - both the head and right arm are separate to main body layer – this allows a subtle amount of articulation. The model will look much more realistic if he or she relates to the scene – making them look down at a screen or type a keyboard always adds to the realism.
Tip 5 – Improving 2D Cutouts
As already mentioned if you use rendered 2d cut outs in SketchUp\IRender nXt the finished image can sometimes be washed out and lacking in depth. To quickly fix this use your original file and overlay in Photoshop. If you do not have the original file – in SketchUp right-hand mouse click on material in material dialog and export texture image – overlay extracted image in Photoshop. Being an individual layer – you can lock the transparency of layer – allowing you to create much easier strokes\fills and shadows to improve realism of 2D image
Tip 6 – Using Props for realistic integration of 2d cutout models
However well the perceived perspective of a cut out model fits the visual - how it interpenetrates with the floor is all important.
There are several ways to help this by either `sinking` the model into the floor or editing the floor shadows in Photoshop – the easiest way is to place an object or prop in front of figure to hide convergence of model with floor.
Tip 7 – Walking Figures
However good the figures – walking poses tend to look very static – In Photoshop duplicate layer – reduce opacity and using `filters – blur and then motion blur apply a 2-10 pixel lateral motion blur and move the new layer a few pixels to the side (depending of direction of movement)to make the image appear that it is moving in front of camera.
Tip 8 – Speeding up Model Editing process
With all those saved model files you can speed up the process even more by pre-applying shadow and reflection on additional layers and grouping them together – a `generic` shadow cast directly below the model will probably fool most people to believe the figure is within the space. The reflections are basically a duplicate of original – vertically flipped and a gradient applied to a mask layer on reflection layer. The shadow is made by stroking with the burn tool on a hardlight layer – I may try and include these non-destructive editing techniques later.==Summary
Summary
I hope you find the above useful
Boothy
See also
Video Tutorials
Videos
We have created a number of Videos which will help you get started.