Just a quick translation since the question arised again in an exchange of emails.
So, instead of using a background pic, using a UVmapped box. I began doing that last year, when I had only a TNT2 M64 for graphic card, since my computer, a 750 MHz Thunderbird AMD, was crawling when I was using Background pics.
This tutorial was first meant as a companion to the head modelling tutorial from WizardNx, aka Nicolas The Corsican.
Hope this helps.
Jean
Great tutorial IamInnocent !!
Keep blending and making (so good) tutes !! 8)
bye
WizardNx 
Nice Tut, this is the way I have always done it too. I like it better than using background images because you can actually move the image through your model.
Nice work.
I’m intrigued : why moving the image through the model. If it’s a real advantage I’ll modify the tut to have 4 to 6 intependant planes instead of a box.
Funny, I have used both background images and uv textured planes methods and I haven’t noticed any speed increment… actually I find it harder to be forced to model while in Alt-Z texturing mode.
By the way, my system is a 2.4 P4 with 1/5 Gbyte DDR Ram.
I don’t see much speed difference myself since I upgraded to an ATI 8500.
Yet I like the box better for it is easier to line up and scale the different view.
Did you try to set your model to wire rendering even when in shaded mode.
Anyway, this was an answer to a diredt question : the rest is up to you.
I’m actually trying right now to find a new presentation for my tutorial so I can give the information and still leave them open on every possibilities. It should be ready for the roller coaster tutorial.
I forgot to mention one important thing about my machine… it is powered by a NVidia Titanium 4 with 8X AGP (the 8x is just marketing stuff anyways)
Now, what is that rollercoaster tutorial you are talking about?
Something I’ve been building to with the whole serie of tutorials of the last 2-3 weeks. It came about during a discussion trying to find out where the famous sudden twists in the tilt of 3D curves came from. This problem I already pretty much solved in the first tutorials. Now it is time to put it together, by describing an application of all the workarounds and, better, the means to avoid having to use them.
I have a version of the tutorial written : 9 big pages, 100+ images. This is too long for a blow by blow type of tutorial (you know : “do this/do that…”) and I have to come up with something better, a mean to put the tools in the hands of the visitor, let him work and yet give him the chance to get some feedback from his efforts.