3D Experiments and Modelling
Ongoing screenshots of the 3D development and defining processes can be seen in the Appendix of the State of the Art Review and on Behance (unfortunately they aren't as aesthetically pleasing as other behance galleries because of this, but blogging sites are insanely slow and to understand these ramblings you would definitely need reference photos).
I began by modelling on and off with the assets which you can see in presentations two and three however for the purposes of chronological understanding i will explain the modelling from the building from the ground up exterior to interior and hopefully then onto UV maps, materials, lighting, cameras etc.
Just to note, the 75TH Eleanor Roosevelt Exhibit happened during this process, so What was exhibited at the show was not the final model or even the model i'm going discuss in this post.
Unlike previous 3D projects that although ticked the boxes i wanted this to be a show piece in my final year portfolio, i also had a level of responsibility to the client and Base One Europe and the past to recreate this bunker to the best of my ability. unfortunately this meant my usual attitude of 'its not great but it will do' wasn't goo enough.
Every time i asked 'Why Why Why' during this project i always over analysed if it was finished or if it was done to the best of my abilities where it mattered.
This meant stopping and starting and constantly redoing something that yes could've taken mere minutes but instead took me hours.
The thing to know about this project is it took me an obscene amount of time compared to how long i had intended for it to take, and i can honestly say i only moved on to the next stage if i was confident in the result.
I begun the model in several different ways, but like recording measurements i will only bore you with the way it worked for me for the best end result i could create at this point.
Incase it wasn't obvious from my calculations and previous measurement discussions, i set my scene with metric units in 3Ds Max.
I edited the high res images in Adobe Photoshop where only necessary information was shown, i did this by cropping the necessary views and deleting all measurements and notes (i had the physical copy to go by if these were needed). I made a note of the image size and created a plane of the same size (with extra centimetres of wiggle room) and projected the image on it.
I did this for each blue print as i went along. this wasn't for measurement reasons as you probably think, this was actually just as a reference to cross check my measurements. I measured by snapping the grid to an appropriate unit, and aligning the walls of the model with the measurement models.
For example i create a box for the the wall and use the create section to modify the measurements to the ones i've previously calculated. i then created measurement boxes, aligned them exactly between walls etc, this was cross checked with the the reference plane and periodically with Autocad for precise measurements. Understandably i couldn't check with Autocad every single move unlike the measurement boxes and reference plane but all the measurements were checked in the end, just not as ongoing as the other measuring methods. This type of good practice saved mis-measurement issues down the line.
Including details that might never be seen or graded was all necessary in the big picture for everything to mesh so perfectly and to fit not just digitally but in my own vision. This includes fillers between ceilings, wooden blocks in concrete footings and panelling within the walls. time consuming but worth it as i go along.
Certain things might have been obvious to a construction crew with minor details and no follow on details but unfortunately i hadn't a clue what that meant. for example i know the size of the bowstrings in the roof but only a 2 dimensional view of its structure and at that only for one row, how many is needed to hold up the roofing? i used my best judgement and by online resources guessed about two per panel in a V shape (most common in bowstring roofing which is what the plans roof structure most resembled.)
Or the concrete footings boarding the outside walls. i assumed at the start these resembled the concrete footings in the centre of the building, but i was wrong and it was one of the major set backs during the modelling stage. they appear in early drawings along with all centre footings, but since the centre footings didn't appear from the floor panelling upwards i assumed the wall footings were the same, i was in this case wrong, and although it doesn't say for certain i believe full length concrete pillars would be the most likely if going by similar building interior looks.
This required me to redo and modify each individual pillar, and measure all of again. it set me back by at least a full day of modelling. which doesn't sound like alot but i only got a few hours here and there uninterrupted along side other commitments and then a few straight dawn to dusk days modelling, this set me back one of those valued long days.
These guesses hiccups can all be seen in the modelling screenshots.
Another set back was somewhere between the first floor walls and the roof panelling i had unconscionably stopped grouping objects, i still named them appropriately but there was one point i had skipped a group if not two large groups and it took three times the amount of time back tracking and grouping them accordingly as it would've if i had just grouped them as i had gone along just as i had been doing until that point.
A handy tip/technique i picked up during early days of modelling and experimentation was using the clay shader. this had a cleaner and easier to analyse look about it, especially for presentations and screenshots. I mostly worked with the default shaders and wireframes unless modelling furnishings. clay shaders were only used when i was near finished with a model and wanted to inspect it before moving on. It was simple new techniques like this that pushed this project along and helped create a sense of professional practice.
At this point the exterior from roof to basement is done as I created the exterior walls and then interior walls before the windows and doors as its modelling was similar in structure and technique, whereas the windows and doors required finer modelling and booleaning and attempted spline modelling as opposed to mostly creating and modifying standard primitives.
I began by modelling on and off with the assets which you can see in presentations two and three however for the purposes of chronological understanding i will explain the modelling from the building from the ground up exterior to interior and hopefully then onto UV maps, materials, lighting, cameras etc.
Just to note, the 75TH Eleanor Roosevelt Exhibit happened during this process, so What was exhibited at the show was not the final model or even the model i'm going discuss in this post.
Unlike previous 3D projects that although ticked the boxes i wanted this to be a show piece in my final year portfolio, i also had a level of responsibility to the client and Base One Europe and the past to recreate this bunker to the best of my ability. unfortunately this meant my usual attitude of 'its not great but it will do' wasn't goo enough.
Every time i asked 'Why Why Why' during this project i always over analysed if it was finished or if it was done to the best of my abilities where it mattered.
This meant stopping and starting and constantly redoing something that yes could've taken mere minutes but instead took me hours.
The thing to know about this project is it took me an obscene amount of time compared to how long i had intended for it to take, and i can honestly say i only moved on to the next stage if i was confident in the result.
I begun the model in several different ways, but like recording measurements i will only bore you with the way it worked for me for the best end result i could create at this point.
Incase it wasn't obvious from my calculations and previous measurement discussions, i set my scene with metric units in 3Ds Max.
I edited the high res images in Adobe Photoshop where only necessary information was shown, i did this by cropping the necessary views and deleting all measurements and notes (i had the physical copy to go by if these were needed). I made a note of the image size and created a plane of the same size (with extra centimetres of wiggle room) and projected the image on it.
I did this for each blue print as i went along. this wasn't for measurement reasons as you probably think, this was actually just as a reference to cross check my measurements. I measured by snapping the grid to an appropriate unit, and aligning the walls of the model with the measurement models.
For example i create a box for the the wall and use the create section to modify the measurements to the ones i've previously calculated. i then created measurement boxes, aligned them exactly between walls etc, this was cross checked with the the reference plane and periodically with Autocad for precise measurements. Understandably i couldn't check with Autocad every single move unlike the measurement boxes and reference plane but all the measurements were checked in the end, just not as ongoing as the other measuring methods. This type of good practice saved mis-measurement issues down the line.
Including details that might never be seen or graded was all necessary in the big picture for everything to mesh so perfectly and to fit not just digitally but in my own vision. This includes fillers between ceilings, wooden blocks in concrete footings and panelling within the walls. time consuming but worth it as i go along.
Certain things might have been obvious to a construction crew with minor details and no follow on details but unfortunately i hadn't a clue what that meant. for example i know the size of the bowstrings in the roof but only a 2 dimensional view of its structure and at that only for one row, how many is needed to hold up the roofing? i used my best judgement and by online resources guessed about two per panel in a V shape (most common in bowstring roofing which is what the plans roof structure most resembled.)
Or the concrete footings boarding the outside walls. i assumed at the start these resembled the concrete footings in the centre of the building, but i was wrong and it was one of the major set backs during the modelling stage. they appear in early drawings along with all centre footings, but since the centre footings didn't appear from the floor panelling upwards i assumed the wall footings were the same, i was in this case wrong, and although it doesn't say for certain i believe full length concrete pillars would be the most likely if going by similar building interior looks.
This required me to redo and modify each individual pillar, and measure all of again. it set me back by at least a full day of modelling. which doesn't sound like alot but i only got a few hours here and there uninterrupted along side other commitments and then a few straight dawn to dusk days modelling, this set me back one of those valued long days.
These guesses hiccups can all be seen in the modelling screenshots.
Another set back was somewhere between the first floor walls and the roof panelling i had unconscionably stopped grouping objects, i still named them appropriately but there was one point i had skipped a group if not two large groups and it took three times the amount of time back tracking and grouping them accordingly as it would've if i had just grouped them as i had gone along just as i had been doing until that point.
A handy tip/technique i picked up during early days of modelling and experimentation was using the clay shader. this had a cleaner and easier to analyse look about it, especially for presentations and screenshots. I mostly worked with the default shaders and wireframes unless modelling furnishings. clay shaders were only used when i was near finished with a model and wanted to inspect it before moving on. It was simple new techniques like this that pushed this project along and helped create a sense of professional practice.
At this point the exterior from roof to basement is done as I created the exterior walls and then interior walls before the windows and doors as its modelling was similar in structure and technique, whereas the windows and doors required finer modelling and booleaning and attempted spline modelling as opposed to mostly creating and modifying standard primitives.
Comments
Post a Comment