Messages

Scavenger, I agree about every point you make there and I'd personally all for going with that route, the problem here is that I know the rest of the Pixel Hunt people are much more into a spontaneous, chaotic Let's Play style, while you are thinking of a planned out, almost scripted and very educational style.

You can still make something appear more chaotic and spontaneous even if it is scripted, it just takes longer to put it together. If you remember MST3k/Rifftrax, that appears like the commentary track is spontaneous and everyone is ad libbing and ribbing on the film. But it's actually meticulously put together, with jokes just in the right place, and recorded such that noone flubs their lines or talks over one another unless it's a deliberate joke.

You need to be insanely talented in order to pull off a purely ad-libbed let's play with a group of people you can't see the reactions to in one take. I'd wager that it is an impossible task. I mean, sure, you'll have fun doing it, and it'll be fun for the people involved, but producing something for an audience? That's a different kettle of fish. You need to think about making it approachable for an audience. Spontaneous group blind run isn't working. I've seen too many let's plays that have been ruined by the insistence that "doing it live" is the thing to do. You can still be funny, you can still be "spontaneous", and you can still be "wacky" and still create something worth watching, without relying on your first take. It'll take a bit longer, it might even require reshoots of footage and rehearsals where you brainstorm funny jokes, but it's totally possible to make your videos more watchable while still achieving the same feeling you're going for.

The audience, more often than not, wants to see the game being played, and be shown around it because otherwise, they'd just play the game themselves. Don't lose that in a cacophony of conflicting personalities trying to make jokes. Embrace the format you're working in, don't try and fight against it.

I just watched it. I'm going to be harsh, but I believe in you guys.

I didn't really like it. It was very overwhelming and scattershot, and I think it could have done with a lot more structuring. The blind run doesn't really suit groups, as people tend to react and talk over each other. Group lets plays also don't work well with reading out the dialogue - it confuses the listener as someone who is commentating on the footage is also presenting themselves as part of the footage (as a side note, if you are going to read out the dialogue, don't put on a voice to do it. Just use your normal voice.)

I'd set out a script or at least a list of talking points for each part of the video, to give you some direction. A let's play is about informing the audience as much as it is entertaining them. You must know more about the game than the viewer, and present it to them in an informative manner. You missed out the good ending? Tell the audience how to get the good ending. Get all the points. Find all the secrets. Present the game in a manner that is easy for the viewer to digest.

The intro was a little bloated, and since you have who the people are explained in the accompanying text, you don't really need to have an elaborate intro for a let's play. Just say "Hello, and welcome to  ~Name of Show~. I'm DKH, and this is Noa (Hi! ~Witty line~) and Wyz (Hello! ~Witty line~)..." etc, "... and today we're playing Errand, by Radiant.". This is the internet we're producing for, and as it's on demand we don't need long intros to make sure everyone is sitting comfortably. Just jump right in there. Be quick. Be snappy. Grab the audience's attention, but don't overwhelm them.

Perhaps the slow intro is partially the reason for the lack of response? After all, if you dither on getting to the meat of the video, people will shut it off and forget about it quicker than you think.

Also, don't be afraid of a second take. If something goes wrong in your video, don't ignore it, take a second take instead. The final product will be a lot more polished and well recieved. This isn't live television, you can edit and reedit however much you like.

So, in brief:
- Structure! Bring some new information to the table, and show the viewer interesting things about the game. Be on top of your material!
- Impact! Don't fill the void with chatter, make the things you say meaningful and interesting! Brevity is the soul of wit! Get into your videos with the minimum of introduction!
- Polish! If something goes wrong, fix it, streamline it. If too much goes wrong in a final video, it turns the audience off since you're literally breaking their immersion with issues. This isn't going out live, it's prerecorded. Take some time and make every episode gleam!

You guys have a lot of potential, but you need to focus your efforts and coordinate to create something great. It's a lot of work, I know, but I know you can do it. Once you have the right system, this could be an entertaining and informative series. Without that focus, you're not providing the audience with anything they need more than providing videos of people talking over a game, which is something the internet has in excess.

A minor thing, by the way, I think this calculation is a bit wrong:

Code: ags

(System.ScreenWidth - width - border_size) / 2, (System.ScreenHeight - height - border_size) / 2,

You are adding border_size (multiplied by 2) to width and height just before this, so there's no reason to subtract that even more.

That's for setting the coordinates, not the height/width of the box, without it the background box is just drawn out of sync with the border.

Sorry for double post. I was curious and looked in the engine code. DisplayAt function does not use the width you gave her to break text into lines, but (width-6) for some reason.
I can't get the sense of all that is going on there atm, but I wonder, will that be a working hack to give width + 6 to DisplayAt?

Adding 6 does nothing to the background box shooting off to the side - it merely messes up the already working ones. Tried both adding 6 to only DisplayAt, and to every width calculation after the initial one.

This is a little frustrating. I don't want to have to build my own text window function, I don't think I'm a good enough programmer to do so. There must be a reason why AGS is choking on this function, and I hope it's actually my code and not something within the engine. I mean, I could work around it by not using two long words in a row every time I write the function, but that's silly.

You guys have been an invaluable help so far, I don't think I could have gotten even this far without your insights. But this last bit is just stumping me to death.

EDIT: Oh, duh. I can just put a corrective variable in there to shorten the width if I need to. A dirty fix, but it'll do.

Okay, trying with THIS code:

Code: AGS

function DisplayText (String text)
{
  DynamicSprite* sGuiBG = DynamicSprite.Create (320, 200, false);
  int width = GetTextWidth (text, 0);
  if (width > 256) width = 256; // So the image doesn't fill the screen and look ugly, also stops ColouriseArea freaking out.
  int height = GetTextHeight(text, 0, width);
  int border_size = 6; // Each piece of the text window is 6 pixels in width/height.
  width += border_size*2;
  height += border_size*2;
  sGuiBG.ColouriseArea ((System.ScreenWidth - width - border_size) / 2, (System.ScreenHeight - height - border_size) / 2, width, height);
  Overlay* oGuiBG = Overlay.CreateGraphical ((System.ScreenWidth - width - border_size) / 2, (System.ScreenHeight - height - border_size) / 2, sGuiBG.Graphic, true);
  DisplayAt((System.ScreenWidth - width) / 2, (System.ScreenHeight - height) / 2, width, text);
  oGuiBG.Remove ();
  sGuiBG.Delete ();
}

Gets me....



Adding in Crimson Wizard's fix solves the vertical problem (which is friggen awesome!), but not the horizontal. What I'm suspecting is happening is AGS is reporting a larger than max number (256) because of the two long words in a row. This truncates the Colourised overlay down to 256 pixels, but by the time it's done a DisplayAt, AGS has put the offending word on a newline instead.

How do I take THAT into account?

I'm trying to put a semitransparent image behind my text windows, but it's not working out. I stole the code from my previous game, which appeared to work at least a little, but this time, no luck at all. GetTextWidth seems to choke on really short or multiline images.

Code: AGS

function DisplayText (String text)
{
  DynamicSprite* sGuiBG = DynamicSprite.Create (320, 200, false);
  int width = 10+ GetTextWidth (text, 0);
  if (width > 256) width = 256; // the fixed width. change this value as preferred
  int height = GetTextHeight(text, 0, width);
  sGuiBG.ColouriseArea ((System.ScreenWidth - width) / 2, (System.ScreenHeight - height) / 2, width, height);
  Overlay* oGuiBG = Overlay.CreateGraphical ((System.ScreenWidth - width) / 2, (System.ScreenHeight - height) / 2, sGuiBG.Graphic, true);
  DisplayAt((System.ScreenWidth - width) / 2, (System.ScreenHeight - height) / 2, width, text);
  oGuiBG.Remove ();
  sGuiBG.Delete ();
}

Produces:

(I know it's a little unreadable, I'm working on that too, but that's out of the scope of this question)

I actually don't know how to make a background plate work for a vanilla display command, but I'd like to, as well as one for a DisplayAt, so it's more flexible. I'm a little over my head here, I think.

Okay, ProgZmax, you have blown my spriting out of the water in one deft hit. That hair looks absolutely gorgeous.

Rather than try to imitate it on the same sprite (and fail miserably, seriously that is way above my competency right now), I instead took the pointers you showed in this edit and applied it to a couple of other sprites I did:

Old Version
[imgzoom]https://dl.dropbox.com/u/50882197/temp/jake_talksprite_wip2.PNG[/imgzoom]

Old Version
[imgzoom]https://dl.dropbox.com/u/50882197/temp/desiree_talksprite_wip2.PNG[/imgzoom]

Am I anywhere close to getting it? Any other pointers you could give to take advantage of this palette?

Also, would it be wrong to ask if I could use that edit in my game? It's seriously very nice.

[imgzoom]https://dl.dropbox.com/u/50882197/temp/nell_talksprite_wip2.PNG[/imgzoom]

Fixed up the hair and made it feather a bit, turned the loose tentacle of hair into a braid.

Still not sure about the placement of the highlights, though.

[imgzoom]https://dl.dropbox.com/u/50882197/temp/nell_talksprite_wip.PNG[/imgzoom]

I'm currently spriting some online avatars for some of the characters, and I'm having a bit of trouble with this girl's hair. It's supposed to be crazy punk hair, and it looked good on paper, but I just can't get it to look right. It looks kind of flat, and I don't like that. But since it's crazy hair, I'm not sure where to put the highlights and lowlights - my usual tack for it isn't working as well as I'd like it to.

Any suggestions?

First, let me say that that's some really cute concept art. It's clear and easy to read.

[imgzoom]https://dl.dropbox.com/u/50882197/temp/boy_paintover.PNG[/imgzoom]
I'm not too good at explaining this, but:
All I did here was clean up the double pixels (the stair-case effect, it makes sprites look less refined than they could be), fixed the shading in places, and tried to make him look more like he is in his concept art. You're quite a good artist already, but I think you've fallen into some of the pit traps that everyone falls into at one point or another. The hair, for instance, is pillow shaded, when it would be better served shading across the hair rather than along it (I don't know the technical term - anisotropic, maybe?). Another thing is contrast - don't be afraid of too much contrast between shades, you can always add a median shade later. It's easier to have too much and tone it down, than have too little and spruce it up. You can see it especially on the hair and the eyes. I merely created a darker eye shade colour and hair lowlight, and removed the old dark middle shade.

The sprite itself is not that bad, it just needed some refinement. Keep at it, and you'll go far! :3

Basic concepts that have been explained exhaustively elsewhere, a lack of respect for pure pixel art (their pure pixel art, to be quite honest, is horrible. It shows none of the skill their digital painting has. It's primitive and amateurish), and... dress up dolls?

This would be a waste of money if it was free. Now a pixel art course that teaches you all the stuff you can't get elsewhere, that sounds interesting. But this is not that.

I'd better show some of my work that I've done.

As part of the huge update, the entire metaverse section of the game is being reimagined. This is part of it!


Doing spritework in the c64 palette is really quite fun.

And things can only get better from here.

I use this thing:

http://www.compuphase.com/software_palettemaker.htm

It's a little tricky to set up, but it's way good for making suitable colour palettes. Just make sure you increase the canvas of your background over 200% so you can crop out the watermark. I also tend to reduce the amount of colours in photoshop first to 256 BEFORE doing the AGS palette reduction, with diffusion dithering, or pattern depending on what kind of picture it is.

By importing your AGS palette into PaletteMaker and locking those entries, you can have a background that uses all 256 colours while still taking advantage of the other slots!
Once you've done that, just paint over all the weird dithering colours (sometimes they creep into a picture) using your favorite 256 colour painting program.

What do you want to use these 8 bit images for?

I managed to do a similar thing with MIDI files once, making them loop before the song ended to a point after the intro, and transition between different files - but I'm not sure the same thing can be done with OGG files simply because the Get/SeekMusicPos function was last I checked, more inaccurate than the MIDI/MOD get/seek functions. I'm not sure if this has been fixed.

But if that's been solved, then I don't see why it couldn't be done. No metadata, though, AGS doesn't grant access to the files themselves.

[HUGE UPDATE]

!!HUGE UPDATE!!

Game is back in production on an official capacity. Check first post for prettier screenshots and less outdated information. More screenshots forthcoming as they get tightened up to my satisfaction.

Taking both of yers advice, I fixed up the Jakob sprite and redesigned two more.

[imgzoom]https://dl.dropbox.com/u/50882197/spriteupdate.PNG[/imgzoom]

Buster is supposed to look more like a free runner than a chav now, so I lithed him up and drew him better. I also redesigned his mask a bit so it looks like a mask. And generally made it not a horrible abberation.

The other one is:

[imgzoom]https://dl.dropbox.com/u/50882197/desfixedface.PNG[/imgzoom]

Does anyone know how to make that sideview look anything but absolutely terrible?

Coded a "module": https://www.dropbox.com/s/4nv8xi59wc4qh2w/Bounce0.1.scm

That's perfect! I already use dummy characters for discarded inventory items in the room, so this'll be relatively painless to implement - practically the only thing left to implement is the movement itself. Thankyou so much!

Oh, duh, I knew I was missing something. That was a really stupid mistake of me to make. x3 Thanks!

I've run into an odd problem when working with my game. In every room, my walking code runs fine, but on this one room, the player character will walk towards a certain spot and not get themselves out of it, no matter how fervently I click
Here is the walkable area map with the position with which the character sticks itself. Why is the character gravitating towards this spot and walking away from it's target? It's all very mysterious, but again, doesn't happen in any room other than this one. I tried my best to break the pathfinding in other rooms but it's just this one that has it. Details could be: I've turned off automatic walking, and done "player.Walk (mouse.x,mouse.y)" in it's stead on on_mouse_click. It's frustrating. Will I have to scrap this room and rebuild it?

Secondly, I've been wracking my brains for a way to implement a certain Legend of Kyrandiaesque inventory system, and so far have had a lot of luck with it. Everything works well, but I'm not satisfied with the way the character drops items. I would like to make it so that it plays something like this:
[imgzoom]https://dl.dropbox.com/u/50882197/throw_object.PNG[/imgzoom]
In which the player clicks on empty space and the character throws it, how far depending on how far away the character is from the target click. Cyan being a walkable area, so that the player doesn't throw items into ravines and stuff. It's wholly unnecessary, but very in-character. I'm just not sure how to make a thrown object arc like that so that it comes to rest roughly around the mouse pointer. I'm not a very good programmer at all!

Tap if you're using a parser, press otherwise.

Reasoning:
When you're using the tap, you are freeing your hand to type on the keyboard. If you don't, you are limited to typing one handed, which is less than efficient in time sensitive scenarios. You want to be able to type in your command as fast as possible. When you're walking across the screen, the steady pace of the character allows you to input your command as they get to their destination, as opposed to stopping and inputting the command. And if you're going to be going long distances, you don't want to have to rely on holding down the button to move, since it is fatiguing on the hand.

Pressing, on the other hand, is more precise, and if combined with a swift and active protagonist and a simple mouse interface. You reduce the amount of clicks for walking to 0, and since you don't have to type, it leaves your other hand free to use the mouse to control their actual actions. This is commonly used in 3D games, and in situations where precise positioning is more important.

Most of the reason why people hate the tap interface, I think, is that Sierra tried to use it in places where it was inappropriate. For example, in a situation where you could die if you walked off of something (usually while typing a command and not looking where you're going). And in a system where dying was slow and required saving manually, it breeds contempt of the mechanic. But the tap mechanic isn't bad in and of itself, it just was implemented in a way that wasn't playing to it's strengths at crucial points. You don't want your hands off movement mechanism mixing with twitch reflexes. That sort of thing is bad mojo.

In short: It depends on the rest of the game.

[mystery]

So here I am, painting a background, and I notice that my protagonist's sprite hasn't been updated since I released the last version of my game in 2009. Obviously, it now looks rubbish compared to the other sprites, so I have to fix that.

[imgzoom]https://dl.dropbox.com/u/50882197/newjakobsprites.PNG[/imgzoom]
New on top, old on bottom.

I'm trying to get him to look more consistent between poses, so that he doesn't suddenly look like a completely different character if he turns around. Also, I noticed his earfins were way too big in proportion to his head, and his head was awful. So I fixed it, and gave him the suggestion of buck teeth.

Am I going in the right direction with this? Are there some anatomical or balance errors that I'm totally missing? How do I shape his snout in the front view? It is complete mystery to me.

For reference, some concept art I made of him. Ignore the hair, it's the way it is on the sprite for ease of animation:


As always, thank you in advance.