Jump to content
patrick99e99

Trying to generate some TI99 speech

Recommended Posts

I have tried Qbox in Linux using Virtual Machines for Win 3.1, Win 95, Win 98 and Win XP, all with similar results: No coding tables found...

I know I had previously run Qbox on my old (now defunct) Win 98 desktop with the same files, so I have no clue why this is no longer working...

 

 

UPDATE: I figured it out. After checking the Install.log file that comes with the distribution, it looked like the install program placed the Qboxpro.ini file in the WIndows directory. Copying that file to C:\Windows fixed all issues and now Qbox runs in all versions of Windows. I still have some issues with sound on my WIn 3.1 Virtual Machine, but that's not Qbox's fault.

Edited by Vorticon

Share this post


Link to post
Share on other sites

Just a thought ... have you tried adding the C:\Qbox directory to the Path environment variable?

Yeah, I added C:\QBOX to the path, and got the same results...

Share this post


Link to post
Share on other sites

UPDATE: I figured it out. After checking the Install.log file that comes with the distribution, it looked like the install program placed the Qboxpro.ini file in the WIndows directory. Copying that file to C:\Windows fixed all issues and now Qbox runs in all versions of Windows. I still have some issues with sound on my WIn 3.1 Virtual Machine, but that's not Qbox's fault.

 

Ahhhh!!! Hooray. Yes that indeed fixed it for me too!!! Thank you so much! Now onto the next step-- actually making some speech!

Share this post


Link to post
Share on other sites

Ok-- now I am stuck again. I have gotten to the assembler phase, and attempted to just compile the war of the worlds demo you had posted:

http://www.planet-99.net/wotw.htm

 

I am using the Asm994a assembler, and have turned on the options "Def Regs (R0- R15), "Produce Listing File", "Produce HEX Obj file", and "Compressed Obj".

 

When I attempt to start assembly, the console reads:

 

>> Assembly Started - 11/10/13, 08:02:21PM
Pass 1: wow.txt
Pass 2: wow.txt
Error #3: Line #547: Label must begin with a-z, $, _ or .: 234567890123456789012 Phase error - symbol undefined 2nd Pass: 234567890123456789012 Multiple definition of symbol: 234567890123456789012
Assembly Complete - Errors: 3, Warnings: 0
>> Processing Complete - 11/10/13, 08:02:21PM
Line 547 is: text 'century, that human affairs were'
Which makes no sense to me, as nothing *appears* wrong with that. Does anyone have any idea why this is breaking?
To continue making progress on this project, I removed the offending code pertaining to displaying text, and just made it so that it would only be streaming the speech bytes (that's all I'm interested in anyway)-- my modified version is here:
Anyway, after that I was able to assemble the code.... However, now I am hung up on actually getting the Win994a simulator running. I got the editor/assembler disks from here:
Their extensions were .dsk, which I changed to .TIDisk as the Win944a emulator said that's what it expects? But when I insert these two disks into the simulator, I can't get it to show an option for "Editor/Assembler".
I can catalog the disks and I see:
ti99ea1.png
ti99ea2.png
What am I missing here?

Share this post


Link to post
Share on other sites

... What am I missing here?

 

Unless you want to edit, assemble or debug in the simulator or mess with the Tombstone City program, you do not need the E/A disks. What you need is to "insert" the E/A cartridge, which should have been supplied with Win994a. You select the "Cartridge" tab and if E/A is not "inserted", remove it and "insert" E/A.

 

If you want to run your assembled program, you need to copy it to a TIDisk image that you then insert in DSK1, DSK2 or DSK3. You can get Asm994a to copy it at assembly time or use Fred's TI99Dir for the copying.

 

...lee

Share this post


Link to post
Share on other sites

Error #3: Line #547: Label must begin with a-z, $, _ or .: 234567890123456789012 Phase error - symbol undefined 2nd Pass: 234567890123456789012 Multiple definition of symbol: 234567890123456789012

Assembly Complete - Errors: 3, Warnings: 0

>> Processing Complete - 11/10/13, 08:02:21PM

 

Line 547 is: text 'century, that human affairs were'

I had another error. It was fixed with just removing a space in front of a label (line 544).

 

Looking at the HTML source, there indeed seem to be spaces (rendered as one space here) in front of that label.

 

I think that our systems might copy different things due to OS, Browser etc.

Edited by sometimes99er

Share this post


Link to post
Share on other sites

What you need is to "insert" the E/A cartridge, which should have been supplied with Win994a. You select the "Cartridge" tab and if E/A is not "inserted", remove it and "insert" E/A.

 

If you want to run your assembled program, you need to copy it to a TIDisk image that you then insert in DSK1, DSK2 or DSK3. You can get Asm994a to copy it at assembly time or use Fred's TI99Dir for the copying.

 

Hmm.. so the win994a that I installed seems to only have "Demonstration", "Disk Manager 3", and "TI Extended Basic" cartridge options. I've been doing a lot of googling and can't seem to find much info on these E/A cartridges/disks...

 

I was able to create a blank TIDisk, and have the Asm994a assembler copy the speech files to that disk, but I can't get to the "Load & Run" option as demonstrated in the video...

 

I'm so close!!

Share this post


Link to post
Share on other sites

 

Hmm.. so the win994a that I installed seems to only have "Demonstration", "Disk Manager 3", and "TI Extended Basic" cartridge options. I've been doing a lot of googling and can't seem to find much info on these E/A cartridges/disks...

 

I was able to create a blank TIDisk, and have the Asm994a assembler copy the speech files to that disk, but I can't get to the "Load & Run" option as demonstrated in the video...

 

I'm so close!!

 

When you follow the "Win994a Simulator Download" link at http://www.99er.net/win994a.shtml, follow the "Cartridges" link to get the Cartridges.zip file containing 198 cartridges. Also, the Disks.zip file (following the "Disks" link) will net you 81 disks images.

 

...lee

Share this post


Link to post
Share on other sites

HOORAY!!! Total success! I was able to make my own 8khz wav file and hear it as TI-99 speech!

 

Thank you everyone for helping make this possible!!

  • Like 1

Share this post


Link to post
Share on other sites

Hey that's brilliant news! And thanks to vorticon for figuring it out! I guess we should update the archive in the resources thread with a game to do this problem from happening again!

 

Patrick, I'm glad the demo was (eventually! ) of user to you!

Share this post


Link to post
Share on other sites

So--- After making several audio files, I am curious how big of a sound quality difference there would be between using the emulator and an actual TI-99/4a. If I go the route of purchasing a TI-99, I have a couple of questions:

 

1. Am I going to be able to find an editor/assembler cartridge? I've been looking on ebay and don't see any currently.

2. Is there an easy way to transfer my LPC source files from my desktop computer to the TI-99? Has anyone made a usb interface?

 

Thanks.

Share this post


Link to post
Share on other sites

The MESS emulator produces speech that is *identical* to the real synth. This is because it is actually simulating the chip down to the level of the internals of the chip. Incredible.

Share this post


Link to post
Share on other sites

The authors of the tms5220 emulation in MAME discovered that the description in the patent is incorrect in some places. This was found by "decapping", meaning that the chip was opened, photographed, and then the logic arrays and mask ROMs were "read" from the photograph. No, really. :)

  • Like 3

Share this post


Link to post
Share on other sites

The authors of the tms5220 emulation in MAME discovered that the description in the patent is incorrect in some places. This was found by "decapping", meaning that the chip was opened, photographed, and then the logic arrays and mask ROMs were "read" from the photograph. No, really. :)

 

Like these guys: http://visual6502.org/

Share this post


Link to post
Share on other sites

The authors of the tms5220 emulation in MAME discovered that the description in the patent is incorrect in some places. This was found by "decapping", meaning that the chip was opened, photographed, and then the logic arrays and mask ROMs were "read" from the photograph. No, really. :)

Now that's effing amazing!

Share this post


Link to post
Share on other sites

The authors of the tms5220 emulation in MAME discovered that the description in the patent is incorrect in some places. This was found by "decapping", meaning that the chip was opened, photographed, and then the logic arrays and mask ROMs were "read" from the photograph. No, really. :)

 

Are the photo's of the chips guts available online anywhere?

Share this post


Link to post
Share on other sites

The MESS emulator produces speech that is *identical* to the real synth. This is because it is actually simulating the chip down to the level of the internals of the chip. Incredible.

 

So you think MESS will sound way better than what I currently am getting from win994a?

Share this post


Link to post
Share on other sites

Also-- do you happen to have that "CES" wav file handy still? I have been getting a lot of pitch errors when trying to convert my own samples, and so I would like to try comparing my audio to the CES one and see if EQing it similarly helps at all.

Share this post


Link to post
Share on other sites

So you think MESS will sound way better than what I currently am getting from win994a?

Yes. Both sound chip and speech are far better.

Share this post


Link to post
Share on other sites

Also-- do you happen to have that "CES" wav file handy still? I have been getting a lot of pitch errors when trying to convert my own samples, and so I would like to try comparing my audio to the CES one and see if EQing it similarly helps at all.

 

I think so, yes. Please see attached.

 

When I make my own WAV files, I do the following:

 

  • Apply compression
  • Correct bias (the cheap microphone I use seems to have a slightly positive bias)
  • Remove breath - the compression in step 1 raises the volume of the intakes of breath between sentances/paragraphs, which can fool the QBOX algorithms, producing noise. I remove the breaths by hand (just across them in the WAV editor and hit delete).
  • Down sample to 8khz

This seems to produce quite good results, though it is still a bit hit and miss. If you have a boom baffle on your microphone that might help - it helps to remove the boomy sounds on p's and b's etc, and reduces the effects of "esses" etc. I have actually put socks over the microphone in the past to produce the same effect!

 

Finally, note that you can correct pitch errors in the QBOX editor itself, and do things like force 0 energy between sentances etc. The likely hood that QBOX will produce a perfect translation without requiring hand editing is quite small in my experience!

test.wav

  • Like 1

Share this post


Link to post
Share on other sites

 

Finally, note that you can correct pitch errors in the QBOX editor itself, and do things like force 0 energy between sentances etc. The likely hood that QBOX will produce a perfect translation without requiring hand editing is quite small in my experience!

 

Did you happen do any of this pitch error correction on the LPC you used in your demo video? Do you have any tips for how to go about manually fixing pitches?

Share this post


Link to post
Share on other sites

I have tried Qbox in Linux using Virtual Machines for Win 3.1, Win 95, Win 98 and Win XP, all with similar results: No coding tables found...

I know I had previously run Qbox on my old (now defunct) Win 98 desktop with the same files, so I have no clue why this is no longer working...

 

 

UPDATE: I figured it out. After checking the Install.log file that comes with the distribution, it looked like the install program placed the Qboxpro.ini file in the WIndows directory. Copying that file to C:\Windows fixed all issues and now Qbox runs in all versions of Windows. I still have some issues with sound on my WIn 3.1 Virtual Machine, but that's not Qbox's fault.

 

Bumping this thread. Vorticon, above, you say the following:

 

 

 

After checking the Install.log file that comes with the distribution, it looked like the install program placed the Qboxpro.ini file in the WIndows directory. Copying that file to C:\Windows fixed all issues and now Qbox runs in all versions of Windows.

 

Did you mean: After checking the Install.log file that comes with the distribution, it looked like the install program placed the Qboxpro.ini file in the QBOX directory...

 

I'm helping another AA user to get QBOX up and running.

 

Thanks

 

Mark

Share this post


Link to post
Share on other sites

 

Bumping this thread. Vorticon, above, you say the following:

 

 

 

 

Did you mean: After checking the Install.log file that comes with the distribution, it looked like the install program placed the Qboxpro.ini file in the QBOX directory...

 

I'm helping another AA user to get QBOX up and running.

 

Thanks

 

Mark

 

No, I meant the c:\windows directory. The QBOX directory already has the .ini file in it.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...