Thanks guys - don't mind at all. It's a pretty logical process to be honest. The key to these really is breaking it down into stages and not trying to do too much in one go, instead relying on merging frames together later on. So generally, my workflow goes something along the lines of:
- Ambient layer, no lightpainting. Allows you to more effectively control your exposure so you know how long you will have to physically lightpaint the surfaces of the car. Generally it's best to keep your ISO down and your aperture fairly narrow so you have longer to move, but obviously this is dependant on the brightness of your location and your light-source. There's not much point giving you example figures here because it will vary between locations, but I usually try to stick to a shutter speed of around 8 seconds and juggle the other two parts of the exposure where possible.
- The actual light-painting itself is the difficult bit, but the most important thing is to stay smooth. I tend to split the image into four and then work my way through lighting them in stages. So maybe start with the shadow on the floor, then the side facing the camera, then the far side, then the roof/bonnet. Often you can combine these, it all depends on what you want out of the image and how bothered you are about getting it in-camera. Ultimately though it means you end up with something like this:
Where the green lines are passes with the light-source (and green squiggles are the areas on the car where the highlight will show) and the blue areas are from the ambient layer, so light as a result of the long-exposure.
Once you've got your base-frames, it's then a case of layering them up in photoshop and creating a single image out of them, then editing that. Hope that helps!