by date
My orange Quark flow.
My orange Quark flow.
| Name: | Affectworks | |
|---|---|---|
| Date Posted: | Dec 19, 2006 | |
| Rating: | 4.3 out of 5 | |
| Public: | YES | |
| Comments: | YES | |
| RSS Feed: | or Subscribe with . | |
| Profile Page: | View profile page for Affectworks |
Blog post
It's been awhile since we posted anything regarding Switching Gears, so I thought a blog post would be in order. After we went to Oregon in October we have been busy with taking all the feedback we got from the nice people who played our test we brought along and have been implementing changes both in style and in ways of thinking.
Besides being a blog just about how awesome Switching Gears is going to be, I thought I'd actually try to put down something useful I have learned about working with Quark and levels.
First off, I can't wait for Constructor for one simple reason: Edge smoothing. It will make all of our .dif work look about 1000 times better. We have applied to be in the beta program a couple of times, but no luck so far. But the work I put into Quark now should be easily ported over to Constructor once it's released, though.
First thing I do is sketch levels. I am trying to come up with as many level sketches as I possible can. I can't recommend enough getting a cheap sketch book and just carry it around. If you see something cool, do a rough sketch of it. No one will see these sketches but you, so if you only have time to get the rough outline down, that's fine.
Sketches are cheap and easy to do, so just churn them out. For every level I come up with a "hook" for the level. It might be "fight in arenas" or "sewer waters" or something like "giant statue" and I build my ideas and maps around those hooks. This gives each level a flavor of its own even though the central theme continues. (I took this idea after having played Ratchet and Clank over and over and over again. It's in my opinion the best action-platformer out there.)
Here's a first draft of an idea which has "water obstacles" as its main theme:

As I sketch I try to envision what the player will see as he runs through the level and put in obstacles and rewards at equal distances. This usually doesn't work well, which brings me to the next phase: Orange!
Next I sit down with Quark and go all orange. What is "orange" you ask? Well, it's something I picked up from reading "Half Life 2 - Raising the bar". Instead of worrying about textures, lighting, exact shapes and such I use crude blocks and two textures, one black and one orange, to block out the path(s) that the player will be following.
I have created a little library of these building blocks using Quark's prefab system (ToolBoxes / New map items... / create a new main folder and just drag your objects in there). This way I can quickly drag and drop to assemble a quick facsimile of what I drew.
At this point, I really don't bother with the "coolness" factor of the level. All I worry about is the flow. How will the player get from point A to point B .. Z and is the run there interesting and fun? When I know a more detailed object will be placed somewhere I put a black box there to mark the location of where it will be placed. As I work I try to keep things organized with folders so I can easily find brushes.
There are 5 things I consider during the orange work flow:
1. How far will the player have to run to get to the next encounter? If it's too far, then it needs to be broken up. No one likes to just move through a level doing nothing.
2. Is the destination easily accessible? (that is: Can the player make the jumps. Are there objects in the way?)
3. Can the player find loopholes? Such as... can the player squeeze himself through somewhere where I don't want him to go?
4. Visibility. Can the player see where he is supposed to go next?
5. Is it fun to get there? This is the biggie at this point. Very often I find that the sketches I did were insufficient and the map needs more love. This is when I iterate the hell out of the map. I play it over and over again, moving the crude building blocks around and adding / removing items / cutting the floor up, putting in stairs and so forth. Anything I can think of to keep the run to the next encounter interesting.
At this point the map is just a bunch of squares sitting on top of a giant square. Maybe there's some detail here and there to mark things, but basically it's blockland.
After I am satisfied with how the level flows, I go back and start putting in the second level of detail (no relation to LOD). This includes rough pillars, arches, stairs, diggers and so forth. Everything is still orange, but it's starting to resemble a level.

The final step I do before Ryan (programmer, co designer) gets his mitts on the level is that I place little dts word bubbles around the level so he can understand the flow of it.

After he gets back to me with feedback (some general, some specific), I go back into the orange level and fixes things / change the path and iterate, iterate, iterate until fun appears. Nowhere around this time do I even think about lighting, textures, dts decorations or anything other than the flow of the level.
The end results is this:

A completely orange level that is actually fun to run around in even though it's bare to the bones. As your library of prefab items grow, the faster and faster you'll be able to create these orange levels.
Whoa.. this turned out to be a giant blog. I didn't even get to the tips and tricks with Quark. Maybe next time.
Cheers!
Fredrik S

Besides being a blog just about how awesome Switching Gears is going to be, I thought I'd actually try to put down something useful I have learned about working with Quark and levels.
First off, I can't wait for Constructor for one simple reason: Edge smoothing. It will make all of our .dif work look about 1000 times better. We have applied to be in the beta program a couple of times, but no luck so far. But the work I put into Quark now should be easily ported over to Constructor once it's released, though.
First thing I do is sketch levels. I am trying to come up with as many level sketches as I possible can. I can't recommend enough getting a cheap sketch book and just carry it around. If you see something cool, do a rough sketch of it. No one will see these sketches but you, so if you only have time to get the rough outline down, that's fine.
Sketches are cheap and easy to do, so just churn them out. For every level I come up with a "hook" for the level. It might be "fight in arenas" or "sewer waters" or something like "giant statue" and I build my ideas and maps around those hooks. This gives each level a flavor of its own even though the central theme continues. (I took this idea after having played Ratchet and Clank over and over and over again. It's in my opinion the best action-platformer out there.)
Here's a first draft of an idea which has "water obstacles" as its main theme:

As I sketch I try to envision what the player will see as he runs through the level and put in obstacles and rewards at equal distances. This usually doesn't work well, which brings me to the next phase: Orange!
Next I sit down with Quark and go all orange. What is "orange" you ask? Well, it's something I picked up from reading "Half Life 2 - Raising the bar". Instead of worrying about textures, lighting, exact shapes and such I use crude blocks and two textures, one black and one orange, to block out the path(s) that the player will be following.
I have created a little library of these building blocks using Quark's prefab system (ToolBoxes / New map items... / create a new main folder and just drag your objects in there). This way I can quickly drag and drop to assemble a quick facsimile of what I drew.
At this point, I really don't bother with the "coolness" factor of the level. All I worry about is the flow. How will the player get from point A to point B .. Z and is the run there interesting and fun? When I know a more detailed object will be placed somewhere I put a black box there to mark the location of where it will be placed. As I work I try to keep things organized with folders so I can easily find brushes.
There are 5 things I consider during the orange work flow:
1. How far will the player have to run to get to the next encounter? If it's too far, then it needs to be broken up. No one likes to just move through a level doing nothing.
2. Is the destination easily accessible? (that is: Can the player make the jumps. Are there objects in the way?)
3. Can the player find loopholes? Such as... can the player squeeze himself through somewhere where I don't want him to go?
4. Visibility. Can the player see where he is supposed to go next?
5. Is it fun to get there? This is the biggie at this point. Very often I find that the sketches I did were insufficient and the map needs more love. This is when I iterate the hell out of the map. I play it over and over again, moving the crude building blocks around and adding / removing items / cutting the floor up, putting in stairs and so forth. Anything I can think of to keep the run to the next encounter interesting.
At this point the map is just a bunch of squares sitting on top of a giant square. Maybe there's some detail here and there to mark things, but basically it's blockland.
After I am satisfied with how the level flows, I go back and start putting in the second level of detail (no relation to LOD). This includes rough pillars, arches, stairs, diggers and so forth. Everything is still orange, but it's starting to resemble a level.

The final step I do before Ryan (programmer, co designer) gets his mitts on the level is that I place little dts word bubbles around the level so he can understand the flow of it.

After he gets back to me with feedback (some general, some specific), I go back into the orange level and fixes things / change the path and iterate, iterate, iterate until fun appears. Nowhere around this time do I even think about lighting, textures, dts decorations or anything other than the flow of the level.
The end results is this:

A completely orange level that is actually fun to run around in even though it's bare to the bones. As your library of prefab items grow, the faster and faster you'll be able to create these orange levels.
Whoa.. this turned out to be a giant blog. I didn't even get to the tips and tricks with Quark. Maybe next time.
Cheers!
Fredrik S

Recent Blog Posts
| List: | 04/24/08 - Still Alive 02/19/08 - Back again with Survival 10/17/07 - Z 08/03/07 - Switching Gears Demo Available 07/24/07 - Maya tutorial set driven keys 06/25/07 - Switching Gears 04/17/07 - Switching Gears 02/08/07 - Switching Gears - New bad guy |
|---|
Submit your own resources!| Stephan (viKKing) Bondier (Dec 19, 2006 at 21:05 GMT) Resource Rating: 4 |
Thanks for sharing.
STef
| Allyn "Mr_Bloodworth" Mcelrath (Dec 19, 2006 at 21:11 GMT) |
| Andrew Hull (Dec 19, 2006 at 21:25 GMT) |
| Tom Eastman (Eastbeast314) (Dec 19, 2006 at 21:31 GMT) |
| Aaron e (Dec 19, 2006 at 21:53 GMT) |
Edited on Dec 19, 2006 23:46 GMT
| Dustin Sims (Dec 19, 2006 at 22:13 GMT) |
| Phil Carlisle (Dec 19, 2006 at 22:22 GMT) |
I'll have to have a go with that myself. Usually I use a "scaling" texture, which is a given size tiled (a one unit cube essentially with "1 unit" texture mapped on it).
Good one!
Looking forward to more SG progress too!
| Russell Fincher (Dec 19, 2006 at 23:05 GMT) |
| David Montgomery-Blake (Dec 19, 2006 at 23:37 GMT) |
I like it!
| Justin Kovac (Dec 19, 2006 at 23:53 GMT) |
| Florian (Dec 20, 2006 at 20:06 GMT) |
btw, have you got the link of the interview/article about HL2?
| Kano (Feb 07, 2008 at 00:54 GMT) |
Is there a way to position an object exactly where you want it or do you have to keep dragging it around with the mouse until you get it in the right spot?
The orange world looks pretty good!
You must be a member and be logged in to either append comments or rate this resource.


4.3 out of 5


