MPEG Version (larger file, lower quality)
Much of what happens at Who3D (www.geocities.com/who3d but soon opening our new site at www.who3d.com) is behind the scenes. Perhaps too much. Now that we've got a bit more web space, perhaps it's time to show a little bit more of what's involved in working on this reconstruction effort... hence this Video Diary which will show the history of production of (at least) one shot from December 25th 2000 onwards.
Knowing in advance that this project is going to take a lot of time, we're working on more than just the modelling and we've got an eye on some software which isn't ready yet - but should make for improved results once we reach "final render" stage. Along with the history of this scene as it's developed, I'll also give a few signposts to future work to be done which will further enhance the final result.
Throughout this "diary" I'll use smallish .JPG thumbnails to display a single frame of the current state, and highly encoded .AVI movie clips at twice the resolution. You may need to update your CODECs to view these, as they use what is a comparatively new and efficient video CODEC by Microsoft.
Below some of the images there is also a text link to an MPEG version of the file. These are slightly larger files (about 100KB larger) and slightly poorer quality, but should provide a fair idea of the original animation without eating up huge chunks of web space. They're hosted at tripod.com, so I'd recommend you use a download manager to retrieve the files and watch them later. Click on the link once to go to the Tripod file download page, then use your download manager to retrieve the file. Press the "Back" button on your browser to return to this page.
Picture it. It's Christmas Day 2000 and I've got time on my hands. Gareth and Dave have complained that we don't show people enough of what we're doing (including them - the 3D artists tend to concentrate on *doing* rather than *showing*). But we've been told we can have some serious web space at who3d.com... So I (Cliff) set about producing some animation to display what we're up to...
Noel's 1st Doctor model, hair, and texture have, so far, tended to wear either a business suit or the standard Poser 4 tuxedo. This is largely because getting the characters themselves to look and work right has been the first priority, with showing them off as a far lower priority task. Putting together some slightly cleaned-up audio from the missing episode of The Tenth Planet, Noel's 1st Doctor model + texture, the tuxedo... and a program called "MIMIC" to perform automated lip-synching for us, I quickly come up with the animation above. Now although no-one is currently working on a "proper" set of clothing for the first Doctor yet, I've used the tuxedo rather than the business-suited Poser figure because to my mind the tux is closer to what William Hartnell wore. Oh, the tie is terrible, and the collar far too small - but these are things I think I can work on... so this animation is simply displaying the Doctor as he stands on 25th December 2000, and gives a brief demonstration of how easily we can get some life into the models using the episode's audio, MIMIC, and a little effort.
So far, so good - but however good the model and texture, the clothing and environment make the model *obviously* a model. What we're displaying will be harshly criticised if the surrounding area doesn't help to provide the right "feel". So the clothing needs to be closer to what William Hartnell actually wore in the episode.
Right, it's not perfect - it's not even meant to be. But the modifications to the clothing (hiding the integral bow of the tie and the shirt buttons, reshaping the collar, repainting the shirt front to include a rough tie and waistcoat) help to provide more of a "feel" of the character we're after. To give you an idea of how simple some of this can be, the tie and waistcoat are *really* simple modifications - painted on a texture map with Ms Paint (the free "bitmap editor" included as part of Windows). Hiding the bow tie and reshaping the collar were a little trickier within Poser 4, but still a matter more of persistance than skill.
So far, so good - but there are still numerous steps to go through before we could hope to call this animation "final" - or even close. The remaining steps include:
Bearing in mind that however good the model and texture, the clothing and environment make the model *obviously* a model, I decided to slap in a quick "who-ish" background. In this instance, since it was readily available, a panel of roundels from Danny's Doc1 TARDIS Console Room.
Importing the Console Room panel into Poser displayed some software limitaitons with Posers rendering ability. To be blunt, the consoles look great rendered in most programs but in Poser they don't look smooth - they're slightly jagged. this is the kind of issue that will be solved both by surface subdivision and/or by rendering in a more sophisticated environment - so the "fix" is already planned. For now - especially as it isn't the right setting, but is just being used as a backdrop to give flavour - it'll do. To reduce the harshness of the flaws in the render I've applied some additional blurring in post-processing.
Bearing in mind my request for some animations to show off what we're up to, Danny has put together this animation with a festive theme.
Sadly I can't manage to upload the better quality .AVI or a viewable-quality MPEG file to tripod... I get timeouts every time. We'll improve the quality of the AVI and add an MPEG version when we move to who3d.com :)
OK, it was a little rushed - but it's a pretty good little animation to have been built in 2 weeks of Danny's spare time, no?
Although I'd not finished my next step, I thought a new render might be a nice way to celebrate the very beginning of what looks to be a promising year...
Chris Sutor has a much better 60-fying technique than that used here - which basically consists of greyscaling and adding in some noise. The result isn't quite as good as I'd like, due in part to the discovery of a bug in the "logo" plug-in for VirtualDub that I'm using to overlay noise. But even so, it's a bit spooky.
So far, so good - but there are still numerous steps to go through before we could hope to call this animation "final" - or even close. The remaining steps include:
There are hundreds of shots to produce, and we're working on all of the characters, clothing, and sets as well as producing the shots with those characters. Ignoring (of course) other projects that Who3D has in hand, and solo projects that we each are working on. Working on this reconstruction is still going to be a higher priority than showing off the results, but I'd expect to see, on average, at least 1 update to this video diary per month (December/January being a particularly good time for updates because of the holidays).