LinkoVitch

Perhaps you could start a RNG at the start of the game, and let it run until the player starts the game, use whatever value it has as the seed for the procedural content? That way you are adding a small amount of entropy to the mix (based on how long they wait until they start the game) ? Or perhaps have that prefill the values and give them the option to randomize or enter their own level code?

 

Cunning way of implementing a save system and randomizer though squire

I thought also in the past that the joystick bus was directly connected to the DSP, but it's not the case.

In fact, the data read from the joystick port is shared with the cartridge bus that is shared with the main bus.

If you read from DSP : the DSP will take the bus to read joysticks bits

If you read from another CPU : the CPU joystick addr will be decoded from the DSP and the DSP will enable the joystick input transceivers.

 

The advantage to read joystick from the DSP is that it has the highest priority on the bus (after the OP) so it can read with accuracy rotary and mouse controller.

 

Yeah, that, exactly that, that's why I did it

 

honest...

 

What.. wait why you all looking at me like that?? :D

 

Glad it polls both pads simultaneously now, was worth the extra OCD effort Explains why writing storew to the JOYSTICK address actually writes a word and not a long too! Pesky Jaguar internals, might expose a mechanism to reduce the resolution in future. (It's only 60Hz at the mo, so nothing epic)

If you are procedurally generating environments, then you could probably store pre-defined set of 'locations' in ROM easily by just storing the seed value for your RNG used for generating the environments...

Now includes joypad reading feature.

Are you sure it's still a sound engine, or are you trying to compete with Raptor?

 

 

I am hoping there is some sanity in the Jag design in that the DSP reading the joystick regs won't touch the main bus (as they exist within the DSP address space), a small part of my mind suspects that it probably grabs hold of the main bus anyway for this.. but I hope not

I'd second what CJ said. In the process of working on your game you may find it ends up smaller than you planned for one thing, or you may discover some techniques to reduce the storage footprint. Make the best you can, you can then edit it as required.

 

I am sure once you are at the point of nearing completing and have a good understanding of what you actually have you will be able to find someone to help with the publishing of it.

Version 0.21 released (changelog), Small update. Now includes joypad reading feature.

 

U235 SoundEngine Homepage

I imagine a great deal of the files are available elsewhere anyway. Anything particular you were looking for?

Version 0.20 released (changelog), Fixed bug with looped samples making small sample loops sound horribly off-key, Vibrato & vol slide added, added a pseudo random number generator...

 

U235 SoundEngine Homepage

 

 

well .. I don't really understand why using a tool if you need to download 20Mbytes of DLL first...

 

Use lodepng for png loading, or bmp is easy to read.

 

For me 20MB isn't much, especially in this day and age. My small (by today's standards) HDD is 160GB, the OS takes several GB of that, so, for me at least 20MB is nothing. I just wanted something that with very little effort can load any image format, rather than just the ones I could think of. Also it gives me access to a variety of filters and image processing methods too, so I can add those as well if I wish.

 

I wrote this tool for myself primarily part as a need, part because I wanted to play with C++, I decided to release it because I find it useful, and others who have seen it have expressed an interest. It may not be for everyone, but the more options available I am sure you can agree is better than fewer choices.

Can't start it in windows mode on my computer : CORE_RL_wand_.dll is missing

 

That is a library from ImageMagicK.. have you installed the ImageMagicK libraries? if so which version?

Just finished fixing up a couple of outstanding bugs in a graphics converter tool I started writing a couple of years ago. Just taken a while for me to bother fixing it It's by no means complete (or bug free I imagine) and I have a few features to add in and de-kink yet. But Ithought I'd put it out here and get some initial early feedback from it.

 

It uses ImageMagicK to allow it to read pretty much ALL file formats (JPG, BMP, PNG, GIF, TIFF etc etc), and supports Jag formats down as low as 1bit images and up to 16 bit images (dont' think I have added 32bit yet )

 

Anyway, enough waffle, I need my sleep. You can get the file from here: http://www.u-235.co.uk/gfxconv/gfxconv-0.02.zip

 

With a very very small changelog here: http://www.u-235.co.uk/gfxconv/changelog-0.02.txt

 

Please don't host these files anywhere else, by all means link to these files.

 

Thanks

AC is going to be like

 

 

Ah cool! someone made a video of the Battle Barge driving to the event last year

For those that would like to enter the speed-coding contest, the rules are there : http://www.yaronet.com/posts.php?sl=&s=155018

 

So for the English the rules are simply "Bottom!" HUR HUR HUR

 

I'll go and dig out the DVDs and get swotting up

YAY! I think I managed to complete the registration form challenge!

I tried reshoot.mod from 505 and it's working

but the sfx aren't :/

I tried your code and there is still no change, the best I got is a sample from the module bank playing, but not my actual sfx data

do you have u235 dev email ? the one on his website doesn't work.

 

 

Oops! sorry not been paying attention too much to the scene.. My email address on my website should work though.. linkovitch@u-235.co.uk ??

 

Did you solve all the U235SE problems?

 

(just been pointed at this thread)

 

Just checked, and the config for that mailbox was wiped out in a server crash! DOH! sorry (the info@ one!) my bad! Have fixed that now, sorry for cockup

I tried reshoot.mod from 505 and it's working

but the sfx aren't :/

I tried your code and there is still no change, the best I got is a sample from the module bank playing, but not my actual sfx data

do you have u235 dev email ? the one on his website doesn't work.

 

 

Oops! sorry not been paying attention too much to the scene.. My email address on my website should work though.. linkovitch@u-235.co.uk ??

 

Did you solve all the U235SE problems?

 

(just been pointed at this thread)

Version 0.19 released (changelog), Better music timing, finetune, slides, & vibrato...

 

But if French translation is done by Gt Turbo will it not be full of bugs? :D

 

Excellent idea.. perhaps I should look into getting my SE manual translated? Or as this is only really of use for French jag devs is there sufficient demand for me to give someone that task?

 

 

Version 0.18 released (changelog), Major rewrite of the sond rendering code. Smaller, faster and now with greater tollerance of bus latency.

Does it make your laundry whiter than white, too ?

 

YES! *

 

* actual claims maybe 100% through the teeth fibs

 

Version 0.18 released (changelog), Major rewrite of the sond rendering code. Smaller, faster and now with greater tollerance of bus latency.

OK, more on this.. I haven't stuffed the release of 0.17! YAY! It seems I have possibly tickled a bug in SLN!

 

For my test I have assembled main.o with madmac.

 

I have then linked with ALN, and also linked a version with SLN using the same binary blobs (main.o and dsp.o) which were produced by MAC.

 

The output from ALN works as intended, the output from SLN doesn't. I think SLN is introducing an alignment issue within the DSP code.

OK, looks like I may have stuffed up the release of 0.17 somehow :/ I will confirm and rectify as needed when I get home. Sorry about this.. I'll post again when I have an update..

Version 0.17 released (changelog), Major changes to the API adding simpler calls and greater flexibility as well as improved module accuracy.

I am making some significant changes to the SoundEngine interface, these should make for more intuitive API calls without you having to do any linkomaths before playing a sample! (I am porting the LinkoMaths into the RISC code).

 

What will this mean?

- The dropping of the periodXX lookup tables,

- The removal of some internal hackery,

- More accurate module pitch bends and slides,

- Removing the dependancy of mod parsing code on specific SE commands

- The removal of those SE commands completely

- Being able to simply specify a sample frequency to play at to the API and it doing the hard work for you!

 

The future? These are just ideas I have that this change will potentially make possible, how feasible they are I won't know right now, but

- The ability to switch to a new playback rate whilst the engine is running without (hopefully) any/much distortion to active audio

- Dynamic adaptive playback adjustments! (this one is likely a way off and possibly a pipe dream, but I think possible...)

 

Once I have completed this chunk of work I should hopefully be able to complete more of the missing mod parser effects. 2012 is going to be busy

Merry xmas guys!