reflection

This post is a simple list of issues that upon reflecting back on the project are what I would have done to achieve a better end result if time was no object and if I had the knowledge I now have from completing this project:

•  I have established that the biggest flaw in my dissertation has been creating my high poly first instead of my low poly as this has caused most of the issues, but this is personal to myself as others find it easier to model the high poly first so it is dependent on personal experience and I believe I would personally more comfortable and achieve better results modelling this way.

• I think that if I had created the model all as one object the issue with Mudbox may have not been created as this would have made the poly's and geometry flow better within the model, instead of creating the different aspects of the model separately and adjusting/ adding to the geometry to attach the new parts of the amour whilst attempting to make the geometry flow fluidly. I believe if I did it this way that the poly count could have been even lower as well.

• If time was no object I would create the low poly first and carry the model across into Mudbox and add the details as originally planned to make the high poly and then try Xnormals again to see if the errors I came across were my own fault or the software. If that didn't work I would use projection mapping again but this time I would ensure that I research thoroughly into more depth about UVW creation and again into more depth with secular maps then finally add texture maps to give the model more aesthetic appeal and reduce the poly count even further, which will aid with all of the previous processes greatly. 

• I would also have more time to experiment and make mistakes due to creating low poly first.

• Do not get enticed by the Retopologize tool. 

• I have learnt not to use the Turbo Smooth modifier and if I have to use any sort of smoothing, I should use Smoothing groups instead, but if I can, I should create the geometry so that smoothing isn't needed.

• I have learnt not to use optimizer or pro-optimizer modifiers, unless it is for a last resort and not for this sort of work as it was near impossible to use for baking as my normal maps weren't created when I used it. I found this out the hard way as I mentioned previously due to the poly's being reduced but not uniformly which wasted a lot of time that could have been spent on my UVW creation/ research and adding more details into the model, normal and specular maps.  

• If is was to do this again I would have modeled the low poly better to aid in getting better results when modifying the cage when projection mapping.