Hello everyone.

I'm nearing the completion of our first game called "DDD Pool"
(pronounced "three D pool"), and I would appreciate comments from fellow indies before I release the game to the public.

One of the useful information I would like to get is what are the specs of your machines and what framerate you get (key "F" during play). I only have a few different machines at my disposal, and they are all pretty new and run the game flawlessly. I want to know what are minimum hardware requirements for game to run. Software requirements are Win 98 and up, DirectX 9.0. I'm not using any of the modern DX9 features like pixel shaders or stuff, it's just developed using DX9 SDK.

No help file for now, I hope you will manage around (This I believe is also a good test for a game of how intuitive it is). Anyway, when you choose "Practice" (the only available mode in the trial version) and click "Play" you can get quick help on "F1".

No instalation necessary, just unzip and run "DDDPool.exe" from anywhere you want (even CD).
The file size is about 4.2M.

Download DDD Pool V0.9 beta (www.paprikari.com/download/DDDPool.zip)

Once again, since we are new in the business, every comment is highly appreciated.

I got 60.64 fps on my 3Ghz P4, but I asked for 85.

The display-mode selection dialog seemed pointless. I could only choose full-screen at 1024x768. I could choose refresh rate but that didn't seem to make any difference.

There were some white unlit polygons in the pool environment (wall banister and trim where wall meets floor).

I didn't like the cue for a mouse-pointer, it felt odd, too long and skinny. Personally, I'd prefer a more traditional mouse-pointer. If you really want a themed pointer, maybe a white gloved hand would work.

On the main menu, the context-sensitive help appears over the 'play' button, obscuring it. I tried to just double-click 'Practice' hoping that would work.

The option buttons on the main menu don't have context-sensitive help, and it's a complete mystery what will happen if they are clicked.

Otherwise, it's very polished. Nice and simple controls.

Originally posted by Wayward
I got 60.64 fps on my 3Ghz P4, but I asked for 85.

The display-mode selection dialog seemed pointless. I could only choose full-screen at 1024x768. I could choose refresh rate but that didn't seem to make any difference.

There were some white unlit polygons in the pool environment (wall banister and trim where wall meets floor).

I didn't like the cue for a mouse-pointer, it felt odd, too long and skinny. Personally, I'd prefer a more traditional mouse-pointer. If you really want a themed pointer, maybe a white gloved hand would work.

On the main menu, the context-sensitive help appears over the 'play' button, obscuring it. I tried to just double-click 'Practice' hoping that would work.

The option buttons on the main menu don't have context-sensitive help, and it's a complete mystery what will happen if they are clicked.

Otherwise, it's very polished. Nice and simple controls.

1) (60 FPS) Nice figure. I suppose you chose "best" ball quality.
I only get about 34fps on my Athlon 2500+ and ATI 9100 (you didn't state your VGA type). I don't quite understand what you meant by "I asked for 85". You probably have chosen the monitor refresh rate of 85 Hz. this is in no relation to how fast your system can draw a 3D scene. For this beta release I chose to let the system draw as meny frames per second it can without waiting for vsync. This is to better esimate the fps on specific machine. In final release I plan to set it to wait for vsync because the animation is then more steady and consistent. This way, if you chose a monitor refresh rate of, say, 85Hz, you couldn't get any more fps than 85 (which is more than enough for smooth animation).
2) (1024x768) Three possible screen resolutions are planned for final release: 640x480, 800x600 and 1024x768. In order to have better 2D image quality (by not letting the 3D hardware stretch it for different resolutions) I will have separate 2D images for every resolution. For now, only the images for 1024x768 are (almost) done, and also I wanted to make this test release as smaller to download as possible.
3) (white polygons) Yes, it's not finished yet. You may also see that the room is in the early stage (there are shadows of picture frames on the walls, but not pictures, ...). I decided to post this test release as it is now to be able to get more feedback.
4) (mouse pointer) That was going to be changed too.
5) (help over play button) This was intentional. There is no space elsewhere in the picture, and I really wanted to make one screen main menu in order to avoid wandering through multiple level menu. I think this makes it simpler and faster to set up game and play. Anyway, while the mouse pointer is over the game type picture and the context sensitive help is displayed over the play button, you don't really need the play button displayed, since you can't click on it while the mouse pointer isn't over it. The moment you move towards the play button and leave the game type area, the play button is accessible again and ready to be clicked on.
Double-clicking Practice (or any other game type button) is excellent idea. I will include this feature in the final release. Thank you.

6) (Context sensitive help for option buttons) Hmm. Worth thinking about it. Do you think if I explained it well in the manual, I would still need it?
Could you please be more specific what was "mysterious" about those buttons.

Thanks for your feedback.

Waiting for more.

Duke

6) (Context sensitive help for option buttons) Hmm. Worth thinking about it. Do you think if I explained it well in the manual, I would still need it?


I haven't played the game yet, but let me disillusion you of this right here - nobody ever reads manuals or instruction screens. And if they do, all they see is "instructions: blah blah blah". You'll still need it.

In fact, you'll need to use every last trick in the book to help people understand better. And even then, they still won't understand.

[edit] Jeez, who typed this? Typos corrected...

I suppose you chose "best" ball quality.
Yes.

you didn't state your VGA type
QuadroFX 1000.

Could you please be more specific what was "mysterious" about those buttons.
There are very few clues as to what the options buttons would do when I clicked them. Will they open another menu? Will they enable/disable? As it turned out, they acted like spinners, going through a list of options, but I couldn't see what all the options were without clicking through them all, and I couldn't go step back to an option I clicked past without going around again. Perhaps left-right, or up-down arrows would make it clearer as to what kind of buttons they are, but personally I like to see all the options available.

AMD Duron 1Ghz
Nvidia Geforce 2 mx

1024 x 768, best details: 9.10 fps
1024 x 768, low details: 68 fps
1024 x 768, high details: 41 fps

Originally posted by Wayward

There are very few clues as to what the options buttons would do when I clicked them. Will they open another menu? Will they enable/disable? As it turned out, they acted like spinners, going through a list of options, but I couldn't see what all the options were without clicking through them all, and I couldn't go step back to an option I clicked past without going around again. Perhaps left-right, or up-down arrows would make it clearer as to what kind of buttons they are, but personally I like to see all the options available.

I was thinking the buttons were selfexplanatory:
1) Easy - Medium - Hard - this could hardly be anything but CPU level of play
2) Best of 1 - Best of 3 - Best of 5 - match duration in frames
3) The button entitled "ROOM" with a thumbnail of a room with a pool table cycles through different choices of environment, including no environment
4) The button entitled "BALLS" with two choices of balls for 8-ball pool: solids-stripes and yellows-reds

We left out the more lenghty titles because they would turn the screen into a pile of text, and our graphics designer Alexander wouldn't allow it :)
Anyway, there is no more than 3 possible choices and you can cycle through them in a matter of 2 seconds, so I'm not sure about the purpose of arrows (which would also be one detail too many on the screen).
And for the "all options available", I can't do this without opening a new window, which I was trying to avoid, especially for 2-3 choices.

I don't know, maybe I'm wrong about this. I'd like to hear more comments on this.

Come on, people. Say something, please.
I was so excited to see the first feedback just three hours after I posted initial message. I hoped they'll keep coming at that rate.
Maybe people are tired of trying other people's stuff and writing critics, I don't know. As I said, I'm new in this.
Or is the theme for my game so uninteresting?
I saw from the log files that at least dozen people downloaded the game. Why are they saying nothing? Is it not worth spending couple of words?
I would do it for you if my opinion meant anything to you.
And your opinion does mean something to me.

I just hope potential customers will not have the same (un)interest in my game when I release it.

People ask for opinions and feedback all the time. Nobody here is paid for their time, so you should really be grateful to get anything at all.

Just my opinion...

Originally posted by EpicBoy
People ask for opinions and feedback all the time. Nobody here is paid for their time, so you should really be grateful to get anything at all.

Just my opinion...

OK, no hard feelings, I understand that.
I was just hoping...

Hi duke,

I briefly tested it out but it was totally unplayable on my laptop. Screen res was 1024x768, 32 bit, 60 Hz (I couldn't change any of this). My specs: CPU is P3 1.2GHz, 256MB RAM, NVidia Geforce 2Go (32 MB). I don't know exactly what the framerate was, but it couldn't have been faster than 1 frame per 2 seconds. Your scenes did not look too complicated, so I wonder what could be the reason for this slowness.

You asked why so few people gave feedback. I suggest you write more enticing headlines. "New indie in town. Feedback appreciated" doesn't really spur me into taking a look as much as "Please test my new 3D Pool Game" or some such. Consider the subject of your post a small sales pitch :).

Hi Matthijs.
Thanks for your feedback.

Your laptop should run this game smoothly. The problems might be following:
1) you've chosen detail level that is too much demanding for your graphics card. When you start the game, change settings for "Ball detail". Try 'High' or 'Medum' settings
2) If you choose lesser ball details settings and the game still has low framerate, than the problem is probably graphics card drivers
3) If that's not the case, maybe antialiasing is enabled in nVidia control panel. That can decrease performance.

You can get Framerate number by pressing "F" key during game.
Anyway, the game in it's final release will run in 640x480, 800x600 and 1024x768 screen res.

Please, let me know if you still have problems.

Originally posted by Matthijs Hollemans
Your scenes did not look too complicated, so I wonder what could be the reason for this slowness.

The scene itself is not so complicated (not sure, but around 12000 - 15000 polys for room and table, plus 1100 polys per ball = total of around 30000 polys), the slowness comes from the way I render balls. That's why you are given the "Ball detail" choice in the initial dialog.
"Best" ball detail first renders the balls with texture and lighting (one directional light from above). Then it renders the room, table and all the other balls into a cube reflection map (128x128 pixels per cube side) for each ball. That means it renders the scene 16balls x 6cubesides = 96 times. THAT's a demanding task.
The app then alpha blends the ball with this cube reflection map over the initially rendered ball.
"Higher" ball quality (I hate the name for this level of detail, but couldn't come up with anything smarter so far, english not being my native language) is everything like the "Best", except for 64x64 cubemap resolution. About 10% gain in framerate.
"High" quality is {texture + lighting} plus fixed environment cube map for all balls, calculated from the middle of the table and not including other balls in reflection. About 200% gain in performance, compared to "Best".
"Medium" is {texture + lighting} plus a fixed Env/Specular texture map. 5x framerate of "Best".
"Low" is only texture + lighting. 6.5x framerate of "Best".

This is all experimental, but currently it seems that the final version will only have one or two levels of detail. "Medium" with proper map is the most promising candidate. Our graphics designer is not satisfied with the lighting of "High" - "Best" methods. I'm currently working on this.

You asked why so few people gave feedback. I suggest you write more enticing headlines. "New indie in town. Feedback appreciated" doesn't really spur me into taking a look as much as "Please test my new 3D Pool Game" or some such. Consider the subject of your post a small sales pitch :).

Thank you for this. I am aware that my lack of experience might do me something like that. I was just wandering why people don't reply because the topic has significant number of views, and the download count is much more than a couple of replies I got. Never mind this, I'm moving on with the game.

Could you amortize the updating of the cube maps? Like, only update 2 per frame or something? Sort them based on distance from the camera and update the closest ones first?

Just some random thoughts...

And also consider limiting the camera range on the cube maps. The reflection fades into nothingness relatively quickly on a small sphere, so you wouldn't have to render the whole room necessarily. Cut back your far plane maybe...

Oh, and I'm no marketing expert, but "DDD Pool" isn't the best name. I know it's a play on "3D Pool" but it's not going to click for most people and it looks like a typo...

It would also help if you posted screen shots.

I usually don't download anything unless I can see what I'm getting into.

Hi Duke,

I downloaded your game a few days ago, but haven't even unzipped it yet. So lack of feedback may just be that people are busy and not a comment on the quality of your game.

Originally posted by EpicBoy
Could you amortize the updating of the cube maps? Like, only update 2 per frame or something? Sort them based on distance from the camera and update the closest ones first?

Just some random thoughts...

And also consider limiting the camera range on the cube maps. The reflection fades into nothingness relatively quickly on a small sphere, so you wouldn't have to render the whole room necessarily. Cut back your far plane maybe...

1) The only time you see reflections clearly (or at all) is when you are preparing your shot (aiming). So I thought I calculate cube maps once before I enter the prep shot loop. That would be a huge gain in performance.
Unfortunately, the cube maps are dependant on the view point, so I HAVE to calculate them every frame if I want them to be correct while you move the "camera"(view source) around.
2) I can't cut the far plane either, because the walls wouldn't show up in the reflection map. Not rendering far away balls doesn't help much too. The bottleneck, as far as I see it so far, is the number of cube maps calculated per frame, not so much what is rendered into those maps (the balls seen in reflection are not the same models as "big" balls on the table - they have 5 times less polygons)

Thanks for your suggestions anyway. If you come up with more, I'd like to hear them.

Oh, and one more thing.
I also tried to optimize calculation by trying to have one fixed reflection map of room and table for all balls (calculate it once per frame) and then copy it into cube map for each ball and calculate ball reflections separately. I'm not sure how much this would help (I think it should) because I was not able to copy cube map texture from one to another (I tried UpdateTexture()). The system only copied it for the first ball (cue ball) and some scrambled lines and dots appeared on all other balls.
Does anyone know how to copy a cubemap texture into another cubemap texture? (Admitedly, I'm a beginner in DirectX programming, but since I was able to figure out some more complex operations of DirectX from DX SDK help, why is such a simple operation like copying a texture made so "complicated" and "foggy"? I praise my days of assembler programming on Amiga! Everything was so simpler back then)

Originally posted by Chris_Evans
It would also help if you posted screen shots.

I usually don't download anything unless I can see what I'm getting into.

Thanks for the advice. Will be corrected ASAP.

Originally posted by kevryan
Hi Duke,

I downloaded your game a few days ago, but haven't even unzipped it yet. So lack of feedback may just be that people are busy and not a comment on the quality of your game.

Thank God it's that!

Originally posted by EpicBoy
Oh, and I'm no marketing expert, but "DDD Pool" isn't the best name. I know it's a play on "3D Pool" but it's not going to click for most people and it looks like a typo...

I was afraid someone might say that:(
It's the first name that came to my mind after 2 minutes of thinking of a working title for the game to be.
Recently, when the things neared it's end, we tried to think of a more suitable name, but couldn't come up with anything.
There are already 3d pool, arcade pool, 3d live pool, virtual pool, super pool and all kinds of "XYZ pool" on the market.
Maybe I should have consulted with you guys a bit earlier on this, but now it's all done. The logo is done, the domain is taken (www.dddpool.com) and I'm kind of not in the mood of changing name at this stage.
Let's just hope that you may be wrong about this:D

Best regards,

Hiya Duke,

It took me longer than I'd like to download the demo on my 56k modem, but because it's a pool game, I went for it.

I am a big fan of pool games and haven't found anything to beat Virtual Pool 3 yet. That game plays so wonderfully with excellent control and input. VP3 just lacks polish, graphically.

So, I tried your demo and it does look nicer than VP3, but I'm afraid I am still waiting to see a pool game that beats VP3.

I did enjoy testing DDDPool though and it IS promising indeed. My major gripe is with the control\input. I think the very best shot taking\cue control is in VP3. Nothing comes close.

The power of the cue hit should always (in my opinion) be controlled by the gesturing (sliding back and forth) of the mouse. Ideally, without having to click a mouse button too.

In VP3, the player depresses the "S" key (shoot key) and then gestures with the mouse to hit the cue ball. This feels the most natural way to me with the best accuracy and control for the player. It's a very satisfying way to play. I have never liked the power bar method. Just my opinion.

Okay, some stats:

I have a P4 1.8Ghz with GeForce 2, running with 768 meg ram with Windows XP.

I chose 60hz, fullscreen, 32 bit colour and got these frame rates: (average fps)

BEST - 19.4 fps

HIGHER - 23.5 fps

HIGH - 61 fps

MEDIUM - 112 fps

Hope this is useful.

Please accept my comments (above) as constructive and not an attack on your control methods. I did like DDDPool, but still prefer my old fave, VP3.

I am always looking to buy a VP3 beater, but I think some changes to DDDPool would need to be made to do that. DDDPool looks nicer, I just dislike the control method.

I did buy another Pool game last week called "Pool Paradise" for the PC. This also fell well short of VP3. I think DDDPool has the potential to beat Pool Paradise. It's already slicker. It could also beat VP3 if the control method was improved.

Good luck with DDDPool, anyway. I will check out it's progress in future.

I've not downloaded the game myself as I just don't like pool games, but I have a couple of points that should help you, if you're interested.

1) For your full-screen options, if you're not willing to allow the user to pick just anything, you should add support for 1280x1024. That's the native resolution of most (if not all) 17" flat panel LCD monitors and anything other than native resolution looks utter shite. I won't even buy commercial games unless I can confirm beforehand that they support this res because it really is that bad if they don't. I can't be the only one.

2) Forget all that cube-mapping stuff. For mapping onto spheres you can do a really good reflection map with one render per ball instead of six. When divided by a further 15 (for each ball) this will get you an fps to die for.

Put a 'camera' on the target ball center and orient it to face the players viewing position. Set the FOV to 179 degrees and render what you can see from there into a standard 128x128 rendertarget.

Map that rendertarget image back onto the ball with spherical mapping and you'll find little noticeable quality difference and your game will be 6 times faster. If you're not using DX9 you'll have to setup the spherical mapping with a matrix but it's really simple and theirs plenty on the net about it. (it's the same matrix you used to position the camera for doing the rendertarget (possibly inverted, would have to check))

Hope that helps...

I have a personal preference for a screen resolution of 1600x1200 for two reasons:

1) It's the native resolution of my LCD.

2) It's my default desktop resolution. I have a dual monitor setup, so when a fullscreen application changes the screen resolution it moves the windows around on the second monitor, which is ugly and irritating. This problem is solved by switching to a fullscreen resolution that is the same as the desktop resolution.

But this is just me.

Supporting the default desktop resolution would seem a safe thing to do too; you can be sure that that resolution works.

I understand your difficulty in supporting more resolutions because of your 2D menu, but I'm sure that can be made more flexible without having to create a custom image for every resolution.

Originally posted by Sausage
...I did enjoy testing DDDPool though and it IS promising indeed. My major gripe is with the control\input. I think the very best shot taking\cue control is in VP3. Nothing comes close.
The power of the cue hit should always (in my opinion) be controlled by the gesturing (sliding back and forth) of the mouse. Ideally, without having to click a mouse button too.
In VP3, the player depresses the "S" key (shoot key) and then gestures with the mouse to hit the cue ball. This feels the most natural way to me with the best accuracy and control for the player. It's a very satisfying way to play. I have never liked the power bar method. Just my opinion.
...
Please accept my comments (above) as constructive and not an attack on your control methods. I did like DDDPool, but still prefer my old fave, VP3.
I am always looking to buy a VP3 beater, but I think some changes to DDDPool would need to be made to do that. DDDPool looks nicer, I just dislike the control method.
I did buy another Pool game last week called "Pool Paradise" for the PC. This also fell well short of VP3. I think DDDPool has the potential to beat Pool Paradise. It's already slicker. It could also beat VP3 if the control method was improved.
Good luck with DDDPool, anyway. I will check out it's progress in future.
Thanks! It's so nice to hear this from a pool fan.:D
One of my main goals for this game was simple, easy and comfortable controls (Wayward thinks I achieved that: "...Nice and simple controls...").
I respect the fact that different people like different things. I personally find gesturing method (often called "True swing" in golf games) unprecise and inaccurate. Admitedly, I never took the time to try to get used to it.
But given the fact that a pool fan likes this method best (and presumably there are more of you out there), it's well worth considering putting this as an alternative method in the game. I will do my best to implement this feature. When (and if) I do it, I'd like you to test it and tell me if I got it right.

Originally posted by PalmTree
...you should add support for 1280x1024...
Originally posted by Wayward
...I have a personal preference for a screen resolution of 1600x1200...
The 3d part of the game is ignorant of screen resolution (it can run in any res your VGA supports).
The problem, as I explained, is the 2d part (menu and some other 2d images). You know that already.
Every developer wants to please as many different pesonal preferences he can, and so do I.
Originaly, 1280x1024 was planned along with current 3 resolutions, BUT...
In contrast to previous 3, its aspect isn't 4:3! :( That means the menu image can't just be scaled, it must be expanded.
Also, separate images for 1280x1024 would (probably) make the download file significantly larger, since it's the biggest resolution of them all.
Btw., I can't figure out what should be the top file size for download acceptable by most (almost all) potential customers?
I set the ceiling for final version at about 10MB. Is this too much? Or am I allowed a bit more?
If only I would be publishing the game on CD. Number of different resolutions would then be no problem (nobody on our team is paid "per kilo" of his work).
Anyway, I just consulted Alexander about this, and we'll definitely do the 1280x1024. Our guess is that there are many people who would like that, so it's worth the extra effort.
As for 1600x1200, ... (where is the ceiling?):rolleyes:
My cousin has a notebook who's natural resolution is something in the aspect of 16:9 (can't remember exact figures now).
The game concept is such that I just can't adapt it to all possible resolutions there are.
I kind of don't like it what they did in WCS 2003: you can play in 1600x1200 (maybe even higher), but the menu and all other 2d images are the same size in pixels (taken from 640x480) for all resolutions, so when you select 1600x1200, the menu occupies just a small portion of the screen and it gets, well not unreadable, but kind of ugly.
Wayward, I'll try to stretch 1024x768 to 1600x1200 in-game and see how it looks. If it's acceptable, it's on. If not, sorry, but making separate images for 1600x1200 is just out of the question right now.

PalmTree, about that alternative reflection method:
Something just like that struck me yesterday, although I didn't think of sphere-mapping the images on the balls. I was trying to figure out what kind of mapping to use.
Yes, I'm using DX9. I don't quite understand you how to sphere-map it. Did you mean I should use automatically generated texture coordinates with the flag D3DTSS_TCI_SPHERICALMAP?
Thank you for your valuable advice. I'll try it and we'll see what comes out. That would definitelly be the winner if it works!

Once again, thank you all guys!

Exactly right on the TCI_SPHERE stuff. So many people were doing this kludge I spoke of that they included it in the source, so to speak, and is why that new texture transform flag has been added. It should just work right out the box. You'll still need a texture-transform matrix as its a TCI thing, but it just needs to be an identity with the translation stuck in as camera.pos - object.pos, I think. If not, it needs to be aligned as I suggested previously (been a while since I set this up). Have a tinker

As to the aspect ratio issues, I just don't see the problem, sorry. You get to specify the aspect ratio as width/height in the projection matrix setup already.

As to the gui/title screens etc, just blank the screen and draw your images at 1024x768 or whatever slap bang in the middle. A lot of games do this (even commercial ones) so you'll lose no brownie points. Games that don't (ie quake) look pants imo. I'd rather have a small image at 1:1 pixels than some crappy stretched thing that goes all blurry.

Or another trick that I'm a big fan of is to simply render a set-piece scene as the title screen, with the table, some balls, etc/whatever, then stick some billboard textures over the top with the title and other iconic stuff that wouldn't work well as part of the 3D.

HUD elements can be snapped to the relvevant screen edge and maybe filled in with a repeating texture if you want to 'go around corners' with the look. I'm fully aware that this is all extra work but I think it really will be paid back for you. LCD owners may not buy your game just because of this, and LCD is becoming very popular these days.

Hope that helps - just my opinion of course.

PS. I just had a thought as I was typing. It might be better for you to snap your game to a maximum of 1024x768 but still allow greater resolutions to be picked. Just setup the main rendering viewport to be 1024x768 centered on whatever the display is and that should do it. Personally speaking, I'd much rather play with a small amount of letterboxing on my 1280x1024, than having 1024x768 set on my monitor as at least then all the pixels would still be 1:1 and my second monitor desktop area wouldn't have a heart attack.

PPS. 1600x1200 is probably the maximum limit. Sure, some people may have a monitor that can show bigger than this, but there won't be many and certainly not many at all with something like this in the home where the games get played.

PPPS. It's about time AOC or someone released a widescreen LCD that could do 2048x1024 with square pixels. Sure would save all this dicking around with multimonitor!

Hiya Duke,

Well, I'll be happy to test the gesture control if you do decide to add it. If it plays well, I'm sure I'd buy a copy!

It's important though, with the gesture method, that the player doesn't need to depress a mouse key when sliding the mouse back and forth. The pressure from pressing the mouse key hinders the finesse and accuracy.

This is how VP3 works (you depress "S" while sliding the mouse). In Pool Paradise, you still have the mouse gesture method, but you also need to press the mouse key and it really makes a big difference....it reduces fluidity.

As a pool\snooker fan, I will buy any game that I enjoy playing....and some that I don't (like Pool Paradise). Maybe I should take it back and ask for my money back! It was only £20, but I just won't be playing it cos it falls short of fun.

:)