When NOT to use a render farm


You may say that the title goes against our interest. Yet it is the truth – using a render farm is not always worth it. When faced with a perspective of a long wait for the rendered files one should consider whether using a render farm pays off. Not that render farms are expensive (sometimes the electricity bill is bigger than the fee for rendering). It all boils down to estimating the costs of rendering on our own and the costs of using a render farm. To begin with, having all the farm’s computers waiting for our files and instantly at our disposal is a rare occurence. Usually, before we begin rendering we need to wait for the previous renders to be finnished, or – on farms which assignes priorities to renders such as Garagefarm.net or Copernicus Computing, where the cheap priorities are counted last. In the case of simple scenes and short animations it may occur that rendering at home is faster than waiting in the queue. On the other hand, when we finally manage to get through we can obtain fast rendering at a great price. Another thing that makes using a render farm unprofitable is the matter of managing them – determining how are files become accessible to the render farm servers. Usually, a scene has to be packed and sent to the farm. This causes a risk of an occurance that I, as a farmer, call missings – missing files. Each and every element of the scene needs to be present in the right place on a farm’s computer. When something is missing we may experience gray surfaces instead of our textures, or some objects missing. This may require uploading the missing files, often together with the scene file. If the farm uses uploading software (our does) the scenes are automatically packed along with all the necessary elements. In such cases we do not waste time on re-uploading the files. When the render time lenghtens for unknown reasons the trendering time lenghtens we should reconsider using a render farm. Of course, sending the scene to the farm lets us obtain a rendering at a reasonable price, yet it will not solve our problem. We probably made a mistake while componing the scene or made a mistake while setting the quality of the rendering too high. In this case one should try to optimize the scene on his own. All right, in order not to dig my own grave I will now tell you when it is definetly a good idea to use a render farm. The first case – when the rendering is to bring in profits. If the rendering is to earn money, get us a lucrative job or help us with an graphic desing assingment it is worth it to invest in our peace of mind. Let us not forget, that the farm allows us to upgrade the quality of the renderings a bit due to the fact, that time is not limiting us all that much. The second case is having a deadline and not being able to meet it using home methods. Nobody likes a inconclusive subcontractor and sometimes we may lose more than it seems. The customer will pay, but he will never again use our service. The third case – when it is necessary to modify the existing rendering. If there is risk that the scene, composition, camera movement, and so on will change we are unable to calculate the cost and duration of the task. A render farm is a perfect solution.