In the past, many companies were quoting animations based on a dollar per second rate for finished footage and there appears to be a trend away from this method of pricing. Part of the reason is that the amount of work involved in producing one second of animation can vary enormously depending on what is required and the level of realism. The advances in computer visualization technology available today are mainly responsible for the widening gap between what was possible and what is possible today.
The best thing one can do is educate themselves on what parts of the animation process are the most time intensive and where they can expect to pay the most (or least) for the requirements.
Below is a breakdown of several cost factors that are "tangible". As with any outsourced service, you also pay for reputation, experience and overhead costs, but these are more difficult to quantify. The greatest cost in a forensic animation is the number of man-hours required to prepare, assemble, edit and finalize the animation. Since vehicle collisions are the most common forensic animations, most of the examples used below will reflect this particular example, however, the process and cost factors are more or less the same regardless if the animation being produced is of a crime scene, personal injury case, or medical procedure.
1. Information Gathering and Preparation
During the initial stages of a forensic animation, it is imperative that the animator be brought up to speed on the details of the case. This often means that all photos, video, drawings and reports must be provided to the animator and they must subsequently go through all the materials of the case. Initially, it is helpful to bring the animator into the initial discussions about the trial strategy and what is the objective of the animation. Further, it is useful to have the animator contact the accident Reconstructionist as applicable to go over details of the accident report and if there will be any transfer of other data such as a digital site survey or simulation data.
Where applicable, the animator may be required to travel to the location of where an accident or crime may have occurred. This is to obtain subsequent images and information that may not have been obtained or was not available in the existing scene materials or reports.
2. 3D Models - Recreating the scene assets.
Normally, it is very rare that a forensic animator can reuse the main 3D models in a scene. In the case of an automobile collision, there could be some "standard" 3D models reused such as a stop sign, traffic lights or electrical poles, but there are often times when even these smaller objects must also be built from scratch in order to obtain the highest level of realism.
There are typically three types of 3D models in a forensics animation. These are classified by their level of importance, accuracy and necessary level of detail. Primary objects are those that are directly involved in the animation. Think of them as your primary characters. Secondary objects are those which may not be directly involved, but their movement or relative position plays some part in making the animation accurate or credible. Tertiary objects are those which are strictly for the benefit of visualization and do not play a direct role in the animation. An example could be a tree off in the distance which would not affect the animation if it were removed except for the level of realism.
Creating accurate and realistic primary 3D models is still a time intensive task which has not changed very much over the years. The quality of the models and options associated with the level of realism are greater and therefore, there is an equivalent amount of time required to create the models. Where possible, it may be possible to purchase an existing 3D model and tailor it to suit the needs of the animation. This should only be done when the accuracy of the model is not a requirement and it may be a secondary or tertiary object in the scene.
In the case of a vehicle collision, the ground terrain is a good example of a large scale model which is required to be accurate and a true likeness of the actually area where the collision occurred.
Often, modeling may come from other sources such as CAD files, 3d scanning or photogrammetry. In each case, the model normally requires subsequent work to get the correct look or to adjust the surface topology of the object.
3. Mapping - Preparing and adjusting images
Although this could be classified as part of modeling, it is in fact a separate part of the 3D modeling process which involves the use of photo imaging software. If you were to consider a simple example of a stop sign, there is the geometrical portion of the model and there is the "textured" or "mapped" portion of the object. The red colour of the sign with the letters S-T-O-P is normally not modeled but need to be mapped. Mapping can be accomplished by taking an existing photo of the stop sign and extracting only the portion of the image that relates to the sign itself and not the background. In some cases, photos need not be used at all and the entire image map of the stop sign could be made with a program like Photoshop.
Since photos are often taken at various angles and depth of fields, the image needs to be rectified (i.e. any depth removed such that the image becomes orthographic or "flat") so that it can be pasted like a sticker onto the model.
However, in each case, there is a considerable amount of time committed to the process of editing photo images to extract usable maps for models and the more accurate, and realistic one wants their models to look, the more it will cost.
4. Assembling the Scene
Once all the scene objects have been created, there comes the task of assembling them accurately in the scene. Since accuracy is the single most differentiating point between any animation and a forensic animation, there is a considerable emphasis on the placement of objects and positioning of all the scene elements. In some cases, there are automated utilities to help the animator reduce the amount of time to place objects, but unfortunately, since each forensic animation is different, this is normally done through manual means. Hence, if there are many objects to be accurately placed in a scene, the amount of labour involved increases.
5. Animating the Scene
There are two ways to animate a scene. One is through the import of simulation data and the other is through keyframing. Keyframing is a manual technique where an object is fixed at a certain location at a certain time of the animation. At each point in the animation where the object is "fixed", this is called a keyframe. The 3D animation software interpolates between keyframes to distribute the motion.
A complex scene involving 5 or 6 vehicles would be more involved than a simple animation of 2 vehicles crashing since there are many more vehicles and positions to keep track of. Since there are many clients who request different collision scenarios, there may be several iterations or versions of the animation.
6. Lighting and Special Effects
Once the entire scene has been animated, this is normally the time to adjust the lighting and prepare any effects which might be required in the animation. When trying to replicate the exact lighting of a scene based on reference photos it can often be very difficult to achieve. Consider the range of lighting and weather conditions such as fog, rain, snow or night scenes.
When appropriate, it may be necessary to show breaking glass or skid marks. Creating accurate looking effects is a time consuming process. Normally, it is difficult to quantify something as "accurate" unless it is verified by an expert. For this reason, there could be several iterations of the lighting and special effects in order to match the expert's report/testimony.
7. Rendering
The act of rendering is how the 3D animation software converts the entire scene into a set of images which run (typically) at 30 frames per second. Depending on the complexity of the scene (i.e. complexity of items 1-6 above), a computer can render a single image at anywhere from a few seconds to an hour or more for each frame. Considering that a 1 minute animation has 1,800 frames, the time involved can be as little as 5 hours to 60 hours on a single computer.
Additionally, if there are 5 or 6 different views to be rendered and not taking any edits into account, the rendering portion can take a considerable amount of time.
Fortunately, many animators make use of either render farms (i.e. network of computers strung together that divide the work of rendering) or invest in high powered workstations. There are firms which specialize in providing rendering resources to animators and when a quick turnaround is required, this is an additional expense which is normally translated to the client.
Normally, rendering only makes up less than ten percent of the total cost of an animation. The actual time is dependent on the type of hardware used and this is part of the reason why companies have been moving away from quoting animations based on their length or dollar per second of completed footage.
8. Video Editing and Final Packaging
As the animation approaches the final stages of completion, the last few steps are the editing and compositing of the animation. The addition of any text, title screens, overlaying images and making any adjustments to the colour take additional time which is normally associated with how the animation is to be packaged and presented.
Most people today are finding that downloads and CD's are quite useful for viewing on a computer, however, if a client requests a self running DVD with menus and sound then this is obviously another involved step.
9. Changes and Edits
One of the most overlooked parts of creating a quality animation and avoiding constant changes is to ensure that clear objectives are understood by all parties up front.
Changing the animation half way through a project can substantially affect the total cost and may delay the delivery considerably. Knowing exactly what needs to be shown and limiting the number of edits keeps things on track and limits the risk of cost overruns.
The basic rule of thumb for the cost of a forensic animation is that as the level of detail, realism, accuracy and number of scenarios increases , so does the cost. A complex animation with many details means more preparation, production and verification hours for the animator.
A forensic animation can range widely from $3,000 to $30,000 depending on each of the factors explained above. There are no hard rules to pin down the price of an animation since each one is "custom built".
Be sure to get a clear idea of the costs by providing the initial case information to the animator. Most reputable forensic animators will supply an initial consultation at no cost in order to provide an accurate quotation or to tell you if an animation is even feasible. Also, ensure that clear objectives are set prior to embarking on a forensic animation. This will avoid unnecessary changes, increased cost and will keep projects on track.
Eugene Liscio is the owner of AI2- 3D Animations which specializes in forensic animations for litigation support. AI2 actively promotes the use of Forensic Animations, 3D Virtual Models, photogrammetry and other visual strategies for the courtroom. Eugene is a registered engineer in the province of Ontario, Canada.
Forums: http://www.ai2-3d.com/Forums
3dfree downloads
