Dev Diary-Bird Of Light

9th September 2015

So close. It’s been a while since I’ve updated the diary, and a LOT has happened.

We have Music!!

We have Art!!

Those things are kinda important, right? Right. Well, what happened was that we were kinda unhappy with our art, since it was pretty rustic and basic at best. I had botched some really high poly models together in Blender that would make a pro cringe (it did). The music was unconvincing and the SFX consisted of whatever we found for free on the Internet. It wasn’t terrible, but it wasn’t good either.

I put together a trailer that was basically an Alpha trailer, and we showed it around. Just when we had lost hope, we found a musician/composer from Romania who was willing to make us custom music for next to nothing. We also managed to hook up with an awesome art team that agreed to partner with us on the project and replace ALL the art. 2D, 3D, character, the works.

We thought we were at Beta, but it turns out it was Alpha. Oh well.

I suddenly found that I had an art TEAM! Someone who could put my mental visualizations into real art. The first thing was to change the ‘cells’, which were basically boxes with grooves for the running track.

We decided to make the new ‘cells’ more organic and uneven, and rather than a simple cube floating in water as we had done before, make them floating islands.

Block in MaxNewArtShot

BEFORE                                                                                     AFTER

Some more updates:

  • We decided to release with twenty-one levels, seven down from the earlier plan of twenty-eight.
  • There’s a (kind of) interactive story now, the animals are the narrators of the story in an unconventional sense.
  • At some point, the realization dawned upon us that we had a game that was not for the 12-20 year old gamer market that we had thought it was for. The game was far too complex to target towards that particular audience, but the art style (very friendly, bright, cartoony) was pointed to them. this needed to change.

To this end, we decided on a low-poly look for the environments and a sober, stylized look for the UI. Drawing inspiration from native art, we redesigned all the 2D art of the game. Here’s a comparison of the old and new speedrun badges. Turns out a professional artist can make quite a difference. 😛

Medal speedrunBadge_Speed_runExplorerMedalBadge_explorer

Above, you can see some of our old badges alongside the new ones.

Here are some old vs new screenshots:

IMG_2817 1

As you can see, we’ve changed the character model as well. In keeping with the target of an older audience of close to core gamers, we’ve made Tara a bit older and much less cartoony. Below, you can see the difference in the new and old level setup screens.


1 May 2015

The game has been in production for one year and one month now, and we’re (kind of) at Beta. It turns out that my estimation of the scope of the game was out by 50%, which is not too bad.

Without further ado, here’s the latest trailer:

Here’s a GIF from the game as it looks now:


There’s way too much to write about since the last post, so I’m going to try and summarize it in not too many words. Time spent typing this is time I could be making levels.

So the major changes to the game are:

a) The game Name: has changed three times from “PumP” to “Jump Like A Girl” to “Jumpology” to (hopefully finally) “Super Veg Girl”. More about that later.

Update: Nope. It is now “Bird of Light”

 b) Implementation of the Map system: This has been among the most challenging features to design and implement. The game world, due to the frequent 90 degree turns and the uniform of the tiles, is easy to get lost and disoriented in. We needed to implement a way for the player to figure out where she/he is in the level and where to go next.

It would be impractical to try and use a mini-map in a corner like many RPGs / FPSs do, because of how small it would be on a mobile screen, so we decided to use the level setup screen as the map. The player would, by tapping the ‘Map button’ while running, be able to bring up the map as a 30-40% opaque overlay, showing the current position as well as the positions of all the eggs and other pickups (see the picture below)


The problem with this was that the player was running AND trying to look at the map at the same time; it was difficult to change the focus of one’s eyes so rapidly and ended up in the player dying most of the time.


We changed this so that now, the game pauses while the map overlay shows up, like this screenshot below. Whenever the player is ready, she/he can press the ‘resume’ button and start running again after a three-second countdown.


This feature took a big bite off the difficulty of the game, and that’s a good thing. By tapping the Pause button, players can take a breather and assess the situation without losing out on time (Where am I/where do I need to go/where are those eggs-key-feather)?

c) This map system also happens to be our primary mode of monetization. Information is provided to the player about the location of the various pickups with labels that can be activated by the map panel.


There are maps that indicate the location of the Key, Eggs, Feather and also the solutions to the SpeedRun and BossRun.

To unlock, the player must spend tokens; the blue and white ones. There are tokens to be collected in every level, and the player can also get tokens by watching videos, sharing achievements and rating the game/liking the FB page.

d) Removal of the timer during the puzzle phase: Initially, the timer started as soon as the puzzle/level setup screen showed up. The idea was to start putting pressure on the player to solve the puzzle as soon as possible. We removed this based on feedback from early user testing. The game was hard enough as it was, and we decided to allow the player to take their time over the puzzle.


Tick, tock!! This was before we removed the timer

e) Controls: This is the big one. This is a twitch action game, after all- and we need to get the controls right. We’ve iterated and iterated and brainstormed and tried lots of control schemes before coming up with the final one.

Initially, the jump was performed by a (Tap) on the screen and the leap was a (Swipe Up). After user tests, we changed the jump to (Swipe Up), as this was the most intuitive jump, established by conventional runners on mobile devices such as Temple Run and Subway Surfers. The leap was much more tricky to map. Here are the options we tried out:

  • Swipe Up- now mapped to jump
  • Long Press and hold-way too unresponsive and clunky
  • Long Swipe up (double the length of the jump swipe)- this was confusing and error-prone
  • Swipe Down and then Up-this was much better but still too error-prone. Error rate of 20-25%

Final scheme: Double Tap!!

We have finally settled upon the double (rapid) tap for activating the leap. While it’s not as intuitive as the long swipe, the error rate is way lesser. There’s nothing more irritating than error-prone controls in an action game, so this is a very important decision. After stage 2, the leap is all-important and the player has to activate it very frequently.

1 Feb 2015

The difficulty of creating difficulty

As I have mentioned somewhere else on this blog, I tend to make games that are hard to play. While that’s not a bad thing at all, I realize that I need to do it with more…finesse. I am very aware of this while I design levels for SuperVeg Girl, and it is a little overwhelming to try and achieve that perfect balance of advancement and frustration.

I’ve created about sixteen levels up to this point and and I’ve ended up revisiting the first twelve levels a total of (at least) thirty times in total. Most of the times, I remove and rearrange obstacles to make the level easier. I make it pretty hard to begin with, almost like I’m punishing the player. Don’t ask me why I do this- I really don’t know.

One of the issues I face is-who’s playing? We decided to keep the game targeted towards a 12+ casual/mid-core audience, but also include the harder-core players who don’t mind a bit of punishment. To this end, there are three “Badges” or objectives in each level, and each Badge targets a particular kind of player behavior. The player can progress through the game by just collecting one single badge in each level


 Explorer Badge

1. The Explorer badge: This badge is awarded to players that want to run about the level unhindered by the timer. It is awarded to players that pick up all the eggs in each level. If your thing is running around exploring game worlds, this one is for you. The tile arrangement does not have to be optimal, neither does the run need to be perfectly efficient.

Medal speedrun

 SpeedRun Badge

2. The SpeedRun badge: Unlocked by reaching the castle in the fastest possible time. Of course, the player does have to find and pick up the key. The tiles need to be placed in a way that the player chooses the optimum path from the start position to the key and then to the castle. This is for the kind of player that likes to burn through the game, unlocking levels and progressing just for the fun of it. It’s the fastest way to play through the game, and the player doesn’t need to pick up any eggs. It is not necessarily the easiest of the objectives, but frequently is.


BossRun Badge

3. The BossRun badge: To get this badge, the player needs to place the tiles in such a way that she/he chooses the optimum path to collect all the eggs as well as the key, and runs through it (almost) flawlessly. There’s very little room for error here; one wrong turn and you’ll run out of time. I’m not gonna lie, it’s hard. I have a really hard time getting the BossRun badge in my own levels most of the time. This one obviously targets the hardcore players.

 6 Oct 2014

I’m off to GDC Pune this month, and I’m hoping to have a few people play my game and give me feedback. Here’s some gameplay footage:

 1 August 2014

So we have a month left to go for Beta. Are we gonna make it?????

No, not really.

But it’s not too bad. The Grid view is ready; that’s a very important 30% of the gameplay done. Sushil made a pretty smooth and intuitive mechanic to place the “Bridges”. I’m calling it that in place of “Placeable Tiles” because of, well, obvious reasons. Tap on the screen to place the tile and it flies into place from the stack at the bottom, and tap it again to remove it. I’ve gone through two iterations of the art for the tiles, and the new one looks pretty OK for a Beta build.

This is the first level; the shape of a question mark was not intentional, in case anyone was wondering.


The Grid View

1 July 2014

The awesome news is that we have a new team member; S is a programmer whom I worked with briefly while he was doing the Diploma program at Backstage Pass. He’s got a full time job but takes out a few hours after work every day to work on PumP. It’s awesome as he has three years of experience with Unity and is a great guy to work with.

I’ve been creating the game world and putting together levels, besides lots else. I realize with some trepidation that this is quite an ambitious project; because of the fact that the budget for the game production is $0, I will have to make pretty much all the art. To this end, I have started making and integrating quite a few environment assets.

The idea is to figure out how it looks in-game; it’s going to be a circular process for making and integrating the art. Make it, put it in, see how it looks, and change the gameplay to suit the art and make art that suits the gameplay.

Here are a few assets:





Also, I have here a mockup of the puzzle phase of the game; I call it the “Grid View”. I had put it together in Photoshop to kinda get an idea of what the view would look like. Initially, the idea was to just have one lane per block/cell, that was placed like a cross in the middle. That does not show here, but there’s only one egg per cell and it’s in the middle. I was pretty proud of this mockup; took me a day or more of hand-painting. I’m not much of an artist, but I really love to paint assets digitally.

puzzle phase

At some point, this changed at A’s advice to a cell with a “Hash” pattern of paths, like this later mockup of a level:


This is what the level looks like


This is what a player would do


This is what a player will NEED to do


It’s at this point that I have now introduced a new element into the gameplay; the ‘Prime collectible’. It’s an object that needs to be picked up before the player can complete the level. The eggs are optional; the player would get a reward for picking them all up, but she could progress to the next level even if she missed one. I’m aiming to introduce some serious replay ability into the game; this device is geared towards that.

The reason for introducing the prime collectible is that I realized the player would happily go straight to the end block/objective without picking up a single egg and achieve the time objective. I had to introduce an element that she would HAVE to run around the grid to collect; now I had to place it wisely in a position with some separation from the end block.

1 June 2014

It has been one month now-one month of development. I have Shanghaiied a programmer, A,  and he has now created the core of the game PumP. I have tied him up to the window grille, allowing thirty minutes a day of exercise and a single bathroom break.

The core of the game is the character controller. It is a very specific kind of movement, one that is completely limited to a world created solely from these.

Block in Max

The ‘Cell’ is the building block of the PumP game universe. I’m creating the environment myself; the last time I used 3D software for production was three years ago so I had a bit of catching up to do. I’m pretty comfortable creating environment meshes now, and texturing them- as long as no professional 3D artist EVER sees my unwrapped UVs. This is version 1;no doubt, I shall be making several iterations.

The prototype, as one may call it, was ready a while ago- The character simply ran and turned at the junctions.


Yes, that is Old faithful Unity Builder guy with spanner. He’s awesome.

I was a bit worried about how disorienting it would be to rapidly turn 90 degrees rapidly; would the player be completely lost, direction-wise? Whaaa, whatwherehow?? It turns out that it was not the case. Temple Run, for example, has 90 degree turns, but they aren’t that frequent. It is going to be a challenge to create a rich, varied game universe from these simple blocks-but I have some ideas about that.

I just have the green spaces (see the 3D model f the cell?) to place stuff. The game world is going to be very geometric and organized by virtue of its building blocks, and I will need to create some chaos and disorder within that world. I can see houses, trees, barrels, cows(??). Windmills. I was taught at VFS that windmills in games bring luck.

I should have put a windmill into StarTrail.

1 May 2014

So it begins again.

StarTrail released in Nov 2012; It is now May 2014. I haven’t been idle, though it may seem like it; even to me. Design wise, I mean. A while ago, I became fixated on a Rubik’s cube as a Muse for my next game. I wanted to create a game that allowed players to manipulate the game universe to create/clear a path; allow some degree of creativity or at least offer real choice in the physical environment.

The platform would be mobile.

I planned, drew, sketched, thought. Walking Flor was when I had the opportunity to be out and about, and I chewed it over for months of walks and falling-asleep times. The game idea iterated, still in my mind, from 3D to 2D, and back to 3D. The environment manipulation would have to be in two dimensions; I didn’t think that mobiles or their average users could handle the disorientation issues in three dimensions.

These childhood games came to mind.

img_4005 numbertilesOne had to slide the interlocked tiles around to arrange numbers and solve the puzzle. Kinda like a 2D Rubik’s cube.

That lead to the conclusion that the game would have to be grid-based. Now I needed to decide what the size of the grid would be, and the orientation. This decision had far-reaching design consequences, so I had to think about it for a while. A mobile puzzle game is more convenient to play in portrait; you can play it using one hand and your thumb. Portrait it was, and the grid would be 6×8. A fair compromise between usability by fat fingers like mine, and space for good level design.

So the grid would be the top-town view of the game world, and a player character would have to navigate this game world. Now I set about designing the elements. There would be fixed blocks that the PC would be able to traverse on, and empty blocks that she would not (the protagonist would be female). The player would need to place ‘Place-able’ blocks in the empty spaces to make the grid traverse-able, so to speak. A few of my initial sketches to try to flesh out the concept:


First iteration of the grid, and game name


This was pretty close to the final concept

Now I needed to figure out the camera and controls; the game would have to be played in two different phases. It was a 2D puzzle that morphed into a 3D action game; there would be two game modes- the Puzzle mode and the action mode. The camera needed to be third person, because the field of view had to let the player survey the environment while running. Looking at the environment and taking decisions would be an important part of gameplay.


So the player would need to decide a path to follow during the puzzle mode, and then run that path in the action mode-but how would she/he remember when to turn right or left?? Potentially, during the later part of the game, the grid would have lots of fixed and place-able blocks. I initially planned to let the player place some indication of turning during the puzzle phase, and later dropped the idea as it would the make the puzzle phase too complex and clunky. In addition to logic and skill, the game would test the player’s memory as well.


This was another tricky choice. How big was the game world? How many steps would the PC need to take to traverse one block? I left this to be answered during prototyping. Here’s a mockup of the scale and the HUD:


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s