Category: 3D


In Caromble! it is possible to charge the paddle and give the ball a boosting power such that it will not bounce off objects, but goes right through them. Adding the charging mechanic would make the gameplay more interesting and gives high-score hunters more possibilities. However, we have been struggling with the implementation. How can we strike the right balance between a powerful powerup and not making the game too easy?

In our first implementation, the paddle could be charged in a continuous manner; the more the paddle was charged, the more boosting power the ball would get. If the power was maxed out, we would not only boost the ball, but also give it the ability to crash through objects, instead of bouncing off them. As a trade-off for using this power, we reduced the movement speed of the paddle significantly, making it far more difficult to reach the ball.

For some reason, this did not feel right. Blocking the paddle was too annoying, and it seemed impossible to find the right speed for charging. Either players would charge all balls, or none at all.

Also it felt a bit gimmicky as it was not needed to complete any levels.  To solve this we introduced hitable blocks that could only be destroyed with a charged ball.

Now, after several tries, we finally found an implementation for the charge mechanic that feels in place. Instead of boosting the ball proportionally with the amount charged, we only have two possibilities now; the paddle is either fully-charged, or it is not charged at all.

Pressing the charge button (LMB or SPACE) fills the charge meter. While  charging, the paddle cannot move at all. When the button is released, the meter empties quicker than it was filled. If the ball hits the paddle when the charge meter is not full, nothing  happens.

However, if the charge meter is full, the paddle is charged, and stays charged until the ball is hit. The charge button can be released, and the paddle can move freely. Having the possibility of aiming the ball whilst being charged, seemed to be the missing piece of having a good charge mechanic. It works and feels good. Now all we need is an awesome graphics effects for charging the paddle and boosting the ball.

Because pictures tell more than a thousand words, here is also a video of the Charge mechanic (charge meter in upper left corner):

Tickets for Beyoncé …

… is the reason why the team had to work without internet for over an hour. This Friday it was my turn to provide a working location (indie to the max), so we were all at my place. However, as a bigger brother, I sometimes sacrifice my dignity for the goodwill of my little sister, meaning I was about to stand in the digital queue for tickets for Beyoncé (singer / actress / model / idol for my sister). As my router sometimes refuses work when too many machines are connected, I had to politely ask the team to disconnect, because disconnection for me would mean re-entering the queue. What I expected to last for 10 minutes, took more than an hour. And I didn’t even get the right tickets! One for sitting in the back, and for standing in the front (sorry sis).

After all this nuisance, we could finally do some business, and on the menu for today was: the new Trailer!  Our first trailer was very simple: a few seconds of gameplay, from different levels. We didn’t put much effort in editing slickly or adding juicyness. Why would we need that, the game speaks for itself doesn’t it. Wrong! The youtube statistics showed that many people already clicked away whilst showing the logo’s. They hadn’t even seen anything from our game and they gave up on Caromble!. So our GameTrailer 101 was: start with the action. Show immediately what your game is about, because the attention span of your viewers is… well, probably less than you’d hope or expect.

We think our next trailer is much juicier than the previous one. Also we show some more of the content from Caromble! We hope you like it. Expect it to be online somewhere in this week.

New Paddle

paddle

Santa came early this year and look what he brought: a brand new paddle! This paddle matches our background story much better. And it looks better too. On top of that the shape has been updated too, it is a bit more flat at the top, so players will have greater control over the ball.

We also would like to share some concept art previously posted on our Greenlight page. It shows you the direction we’re headed with Caromble!, as we are nearing the final stages of development.

We are eager to hear any thoughts and feedback!


 

Orderly Rubble

Another new addition: Pieces break into rubble (just for the looks). So here are some amazing rubble models!

Rubble 01

New Asset: The Oil Silo

This new object is a bit simpler than the old batch, because the rushed assets usually turned out nicer in the game than the detailed ones. It’s one of the toughest parts in game art to keep the bigger picture in mind. When an artist works on something, then it is easy to optimize detail/contrast/colors for this specific item. But all those properties work completely different in a scene it has to share with many other objects.

So here it is: thicker lines on the texture. Less and softer details. And less color.

Oilsilo

Asset: The Bus

One of the assets rushed in for Indigo. Will get some brush-up and normal maps later.

That was by the way some of the interesting feedback I got: the most rushed assets got repeatedly positive reactions. My perfectionism makes me cringe a bit – but of course I should be happy because I can work much faster than I did with some of the older assets. They turned out too detailed – it does not only not show up but also makes their impact lower.

So maybe even this one is a bit more than needed.

level_block_bus01

Final Asset: The Farmhouse

One of the first bigger assets in it’s final form: the farmhouse. Which is right next to the start on or prototype level. I made a little turnaround movie – one with the diffuse map, and one with just the normal map (Normal maps are basically just showing the shape of a form, instead of the color).

I think I’ll keep this style throughout the project – after all the testing I’ve written about earlier, I’ve got it working: It’s quick, and unique enough. With some after effects we’re planning for this should work well in the final game (again, those asset shots are not done in the real engine yet, but close to it).


assetshot_farmouse01_diff01
assetshot_farmouse01_diff03
assetshot_farmouse01_norm01
assetshot_farmouse01_norm03

More Steps in 3D

On request to add desktop sizes of the last images, I’ve made the caromble wallpaper page. Hope this works for everyone.

And the following images are some more asset-shots to test the final 3D look.

cow01 cans01

As pointed out in the last post, this is not in-game, but close to it. If we get features like the black outline to work, it will help us also with the gameplay. We could change the color to highlight important objects.

cow01_outline

While working on the cow screenshot, there was a little accident. I put the wrong texture on all objects. But I think it actually has something to it – it looks more gritty, evil and in despair. Wile the overall technique for the other assets is ok, I will try to incorporate some of the effects from this mistake. Maybe the cow will get a gasmask one day.

cow01_error

Next Step in 3D

After the 3D tests I showed earlier I had a pretty clear path to go. Inked lineart did a neat job: it is clearly visible, gives a unique look and is well workable. The next step was coloring to create some final assets so I know it works for all the remaining pieces. From the concepts I had a good idea of how the coloring should look like – and I wanted to take the loose approach very directly into 3D.

level_idea03_2_colorstart

Here are some 3D tests that are further along. Note though, that while these are real-time shots, they are not from the actual game. Certain features, like the black outline are still missing. But I think we’re close – the depth of field (blurring in the distance) is working already albeit not as perfect yet.

boxes01 chicken01

Here the texture of an oil barrel. Each object will have also a damaged version – thats what the cracks are for. For who is not into 3D usually: The light blue part is a normal map, which helps the game with the lighting. It adds little details like the rims, without having to change the mesh.
It was a guideline from the team to create these, because the Ardor3D engine is able to handle modern shaders quite nicely and we want to show that off (and it looked nicer anyways). While in the game the lights is not set up well yet, these normal maps will look better than in those test shots.

level_base_wallrock01_diff

From Concept to 3D

When we went for the city design – I had to start thinking about how to transfer this look into 3D. It’s great that the team gave me the OK to go for a pencil experiment. I am used to work on huge game projects with hundreds of people – there is not much flexibility there, and not much trying of new things. The switch to single artist is a very nice experience.

I started with a simple container as first test object. And I did the texture in a pencil style. Hoping to get as much of the drawings in to the final version.

3dtest_container_outline level_block_cargo_container01

There are some problems with this approach. As you can see in the texture-scan, it’s all a bit noisy. And when adding a color overlay (since the final version would surely be more colorful), the lines are hard to read.

In 3D this all got even worse. In the screenshot above you can also see a version with a normal map added. While this looks nicer to me since it seems much more detailed, it makes the line-effect nearly disappear.

Maybe adding contrast on the lineart would fix this, but it tends to look “dirty” with scanned pencil art. Which also creates another issue: Whenever I would add something in Photoshop, I would have to emulate the noise effect. Keeping in mind that I will have to do all assets on my own for now (the downside of not having a big team), I had to find an easier way.

So I made several dummy-objects with just ink-lines. It is faster, and way easier to fix in Photoshop.

3dtest_several

The lines are much stronger, and won’t disappear, even if I add color later. I decided to go for this – and made one more detailed object to try the workflow. Some ambient shading definitely helps to sell the shapes.

I think with this I have a good balance of “easy to work”, while still having a somewhat analog look to it.

3dtest_house