Sunday, May 23, 2010

Experiment 2 General Feedback

Hey guys, please read this. It's a summary of the issues I found when looking at your submissions. Issues you can address quite easily to make for a stronger final assignment. It may not be evident in the ranty nature of these points, but I was quite pleased with the general standard of Experiment 2 - most of you ticked all the boxes and had good arguments for what you were doing.

1) Constantly scrutinise your work. Pretend you're marking it, and that it's a piece of work you're looking at for the first time and don't know anything about. Are your captures informative? Do they show the labs, ramps and meeting places? I know that you upload your models and environments, but if your captures show me everything I need to know about how your model is situated, how it is occupied, and how the parts relate to each other, I don't need to open these files. It is best if you include just one image of the whole complex from a distance showing where it is on the island (or other landscape) and one close up that shows the textures, among your 5. Include a person on at least one shot too, to indicate scale. Don't put in repeated images of the same view but from very slightly different angles, it's redundant.

part b) Make sure your model works. The submission requirement is that you make your model available to me, so if it is impossible to walk up your ramp I'm going to find out. If your link doesn't work, I will also find out. If I think something in your model is a little iffy I will want to check. If I can't because your Filefront link is corrupted, you will not get the benefit of the doubt.

2) Constantly scrutinise your work. Are there glaring inconsistencies between what you say you're doing in your text, and what you're actually doing in your model? If you say that shadows are important to you, turn the shadows on in the model. If you say the point of something is to be solid, make sure it looks unequivocally solid. The clients you're designing for are paramount, keep their requirements clear in your mind as you design. Test every assertion you make against the internal robustness of the program you have devised for yourself based closely on the brief.

part b) I have been encouraging you to provide some text or narrative of your thoughts alongside your images, and I continue to do so. I understand that you make some decisions for reasons that may not be obvious to me as I look at your images, so it's good if you alert me to subtleties I may miss. Do not, however, use text as a substitute for design. If you write something, make sure that it reinforces what is actually in the model. The text is there only as supporting material.

3) The reason that all of the weekly exercises emphasise the technical procedures (texturing, uploading, importing, exporting, saving) is so that you experience no glitches in the last few days when getting your final model ready. It is hard to accept excuses about technical problems for issues that should have been ironed out 3 weeks before, when the due date wasn't stressfully imminent. Please make sure you take these steps when the independent study schedule tells you to.

4) Proofread. Proofread. Even if your first language is English. Get someone to read your text who knows about the correct placement of apostrophe's (I did that deliberately...). You shouldn't feel ashamed about asking someone, the text you're writing is so minimal I'm sure anyone would be happy to help out. It's a real shame to look at a strong submission that is diminished by bad grammar.... (I'm not asking anyone to proofread my rant here, so if you find some typos you get a prize).

5) When you put things on your blog, make sure you view your blog before you sign out and wander off. Sometimes I see empty posts.

6) Just a little point about lighting: put in spotlights with restraint and taste. Don't go crazy with this, one well-placed light is often more evocative than five bright white lights cranked up to maximum intensity. The more lights you put in, the messier the cacophony of shadows, too, and the less impressive each one. Put in coloured lights if you have a very good reason, otherwise it looks like a kitsch disco.

7) Please render your images at 1920 x 1080 pixels. That way they will be the same aspect ratio as most computer screens. For Experiment 3, please also remove the text that is automatically on the screen (check Andrew's blog for how to do this.) Remember that you will soon collate a portfolio of the design work you do at uni. If, after each semester, you have a selection of really excellent images ready to show to a potential employer, your life will be much easier.

finally) ATTENDANCE. You are required to attend a minimum 80% of contact hours in order to pass. Most of you have no problems here. Some of you are on thin ice indeed. Also, the 10 marks allocated to class participation do not automatically become yours if you have physically come along to each tutorial. They hinge very heavily on you having done enough work to talk to me about each week. I don't mind if you upload your process work to your blog, or if you show me printouts so long as it facilitates a constructive discussion. This will be important for Experiment 3, which is due two weeks after your final tutorial.


All the best with the rest of semester, here and with your other subjects.

Friday, May 14, 2010

Things about Experiment 3

1) Know about this helpful site: http://www.crymod.com

2) You will be assessed on 18 perspectives. You have been asked to draw 36. This means you need to upload your best 18, demonstrating that you know how to draw both 1-point and 2-point perspectives - ensure you have a fine assortment of both.

3) The Es you are drawing are for inspiration purposes. You will not be required to translate them directly into any part of your bridge.

4) The quality of your drawings play a part in your assessment. Do not draw and scan them as though they don't matter.

5) You make holes and caves in your Crysis model with the 'voxal painting' tool. They let you push/pull things sideways.