Papa Intellivision Purchase

[+Lathe26]

 

I purchased a modest portion of the Papa Intellivision collection, as was mentioned on the "Intellivision Purchases and Deals" thread (starting here).  A separate, fresh thread for this was requested so here we are.    For completeness, the previous posts will be reposted here so folks can reply to individual items for any questions.  However, some of the reposts will be edited to also mention what their current analysis status is.  These will be followed by new posts of the remaining items.

 

Current mood remains the same:

 

[+Lathe26]

I got my hands on an ORIGINAL Keyboard Component manual.  Exciting!!!

 

This is not the recent reproduction of the manual.  How do you tell the difference?  In the last photo, look at the crispness of the lettering between the original (slightly yellowed paper) and the reproduction (bright white paper).  Especially take note that the ® and ™ are less legible on the reproduction.

 

 

 

 

[+Lathe26]

  

Here's the new Printer Interface that came with the manual!

 

Just so folks aren't confused, the 3rd photo shows my older Printer Interface to the right of the new one.  Long lost sisters reunited at last.  👭

 

Analysis: case opened up and the circuit board was identical to the other Printer Interface.

 

 

 

[+Lathe26]

This time the rare KC Printer came with the Printer Interface.  Much Wow!

 

Analysis: tested and still prints!

 

 

 

[+Lathe26]

... and check out this early version of the  KC BASIC Cartridge that was part of the deal.  It has a white label with black writing instead of the typical blue label with white lettering.

 

Analysis: I think checked this one out and found that it works and has identical data to a typical BASIC cartridge (not 100% sure I did this).

 

[+Lathe26]

The set also included a Diagnostic Cartridge for the Keyboard Component!  I wonder what problems it was designed to detect.

 

Analysis: Dump it.  Circuit board is same as BASIC cartridge but with 2 different EEPROMs (3rd socket is empty, as is typical).  It kind of runs but this was written for the ultra-rare Computer II Assembly version of the KC, not the "more common" Computer III Assembly (think of these as motherboard versions).  Thus, it briefly runs and then says the the keyboard fails.

 

 

 

[+Lathe26]

Onward to the tapes...

 

Here is the BASIC tape for Household Inventory.

 

Analysis: I've already dumped this tape and will share the results some time in the future.

 

 

[+Lathe26]

... above are the previously posted items.

 

Below are the newly posted items...

[+Lathe26]

Here is the BASIC tape for Address List.

 

Analysis: I've already dumped this tape.  I will share the most results in the future.  However, I will censor out the actual people, home addresses, phone numbers, and other identifiable information.  Just as a test, I briefly googled for one of the people and was able to find out an unnerving amount of info so this portion of the data definitely should not be shared publicly.

 

 

[atari5200dude82]

Very impressive. Most impressive 

[+Lathe26]

Here is the General Instrument Speech Synthesis Demo tape.  The label is no longer attached since the glue dried out (might restore it later with some adhesive).

 

Analysis: none so far.  Someone else on AA has stated that this most likely just contains normal audio recordings of various compression levels for the SP0256 (i.e., the Intellivoice's speech chip).  In other words, just audio and no data.  I suspect this is correct but still need to confirm.

 

Analysis 2.0: This has been recorded.  It only contains "Keyboard Keys" audio (i.e., lines 3 and 4 on the label) that lasts 40 seconds.  This consists of counting words from "zero" to "nine" followed by other random words like "dollars", "percent", "equals", etc.  The "National Weather" audio (i.e. lines 1 and 2) was likely erased and recorded over.  Even using spectral analysis was not able to find anything other than 2 seconds where there is faint audio of something intelligible, but using noise reduction, filters, and normalization was not able to recover anything usable.

 

 

[+DZ-Jay]

5 hours ago, Lathe26 said:

Here is the General Instrument Speech Synthesis Demo tape.  The label is no longer attached since the glue dried out (might restore it later with some adhesive).

 

Analysis: none so far.  Someone else on AA has stated that this most likely just contains normal audio recordings of various compression levels for the SP0256 (i.e., the Intellivoice's speech chip).  In other words, just audio and no data.  I suspect this is correct but still need to confirm.

 

 

I find very interesting the notes scribbled on the label of that tape.  It implies two distinct modes of operating the speech synthesis chip.

 

First, it mentions "National Weather" encoded (presumably) using LPC, which suggests an example of the sampling capabilities later employed in the talking games.  These would be actual voice recordings sampled digitally, and then encoded using Linear Predictive Encoding.  The tape also includes the original (presumably human-voiced) recording for comparison.

 

Second, it mentions "Keyboard Keys" generated (presumably) using formant synthesis, which is an algorithmic synthesis of voice, such as what was used in the Speak & Spell and early microcomputers.  It is not based on voice samples or allophones, but on a set of rules that attempt to reproduce the human vocal tract (think the old Apple II SAM).  It then includes two versions: pitched and unpitched.

 

I imagine that this formant "Keyboard Keys" is a precursor to the "RESROM" voice samples we got built into the Intellivoice ROM, that produce human-sounding but clearly synthesized speech calling out letters, numbers, and a few other general words.  I suppose that the "unpitched" ones would be close to what we get with the basic allophones today in our own home-brews, which sound rather robotic.

 

Fascinating!  I would love to hear these tapes.

 

    dZ.

[+Lathe26]

In the long haul, I do plan to fully analyze all the items.  That will take time since I have 1 project in the queue ahead of these.  However, I plan to record the tapes soon to at least get an early look.

[+Lathe26]

Here is BI Tape 2.0.  I have no idea what the "BI" stands for.

 

Analysis: I have already recorded this one and it is quite interesting.  It appears to be an early or alternative version of the Keyboard Component Demo tape.  The voice actor is different and some of the spoken sections have a different script (the words spoken) while others have the same script as the normal Demo tape.  The data tracks are similar but not identical to the Demo tape as well.  This tape hasn't been fully analyzed, but I'll start a separate thread for it.

 

 

[+Lathe26]

This is the Diagnostic Test #1 - American Sound Tape.

 

Analysis: None so far.  Someone else on AA has stated that this most likely just contains normal audio recordings because the tape was used to test the tape duplication machine or facilities that would later duplicate all the KC tapes to be sold to customers.  Sounds likely based on the label, but the contents are still yet to be confirmed.

 

Analysis 2.0: This has been recorded but not yet decoded.  In audio software, the audio looks visually identical to Diagnostic Test #1 - Mattel Tape so it is strongly suspected to contain the same data.  To be confirmed.

 

[+Lathe26]

I have done some early analysis and here are the updates.  The earlier posts have been updated.

1. The General Instrument Speech Synthesis Demo only contains "Keyboard Keys" audio (i.e., lines 3 and 4 on the label) that lasts 40 seconds.  This consists of counting words from "zero" to "nine" followed by other random words like "dollars", "percent", "equals", etc.  The "National Weather" audio (i.e. lines 1 and 2) was likely erased and recorded over.  Even using spectral analysis was not able to find anything other than 2 seconds where there is faint audio of something intelligible, but using noise reduction, filters, and normalization was not able to recover anything usable.
2. The Diagnostic Test #1 - American Sound Tape has now been analyzed.  It's sister Mattel Tape is closely related so the analysis is posted there.

[+Lathe26]

This is the Diagnostic Test #1 - Mattel Tape.

 

Analysis: This has been recorded and early decoding has been done.  The audio tracks each have a single, though different, sine wave tone.  There is 1 record on the Read-Write data track (not extracted yet).  The Read-Only data track contain a 5KB program and a single 100KB blob of test patterns.  The program contains the text of "TAPE DIAGNOSTICS" so this appears to be a previously unknown Keyboard Component tape!

 

Theory: At a glance, the Diagnostic Test #1 - Mattel Tape and the Diagnostic Test #1 - American Sound Tape appear to contain the same data.  Since there is currently a company called American Sound & Electronics, it is possible that the Mattel Tape is the original tape and the American Sound is the professionally made copy made by a company that specializes in high quality cassette tape duplication as a test before going into tape production for the KC software.

 

 

[+Lathe26]

Here is a Demonstration Tape from Dec 31st, 1980.

 

Analysis: This has been recorded but not yet decoded.  In audio software, the audio looks visually identical to other production Demonstration Tapes so it is strongly suspected to contain the same data.  To be confirmed.  It is different than the BI Tape 2.0 mentioned above.

 

 

Edited August 30, 2022 by Lathe26
Fixed typo in the date

[+DZ-Jay]

7 minutes ago, Lathe26 said:

Here is a Demonstration Tape from Dec 13st, 1980.

 

Hey! That's my birthday!  I was 10 years old that day, and probably still hoping for an Atari VCS that Christmas.  (I think I got my Intellivision the next year.)

 

UPDATE:  I see that you meant to say "December Thirty-first," and not "December Thirteenst." 😄

 

  dZ.

[+Lathe26]

3 minutes ago, DZ-Jay said:

Hey! That's my birthday!  I was 10 years old that day, and probably still hoping for an Atari VCS that Christmas.  (I think I got my Intellivision the next year.)

 

UPDATE:  I see that you meant to say "December Thirty-first," and not "December Thirteenst." 😄

 

  dZ.

[+Lathe26]

Here is the other Demonstration Tape from an unknown date.

 

Analysis: This has been recorded partially decoded.  It is definitely a Demo tape of some kind since the first decoded records contain code with the expected text... but it is different than both the production Demonstration Tapes and BI Tape 2.0.  Additionally, it does not have anyone speaking in the audio tracks but instead just has a continual (and annoying) tone.  My current hypothesis, yet to be confirmed, is that these tapes were created in the following order:

1. "Demo" tape (this one) - earliest known pre-production version
2. "BI Tape 2.0" - later pre-production verson
3. Typical Demonstration tapes - the final production version

[+DZ-Jay]

So, if your theory is correct, what would you expect that annoying tone to represent?  Is it a bias line or additional data?

 

   dZ.

[+Lathe26]

1 hour ago, DZ-Jay said:

So, if your theory is correct, what would you expect that annoying tone to represent?  Is it a bias line or additional data?

 

   dZ.

Definitely not data since it is on the analog track.  My guess is that it was simply a placeholder before they got someone to speak the lines.  This way, the dev could here if/when analog audio playback was activated.  This is still to be confirmed.

[+Lathe26]

We're done with the tapes.  Now for the BASIC Wash Cartridge!

 

Analysis: This cartridge has been tested and it works.  It has also been dumped.  The BASIC code it the same as all other BASIC cartridges.  The 3rd chip, the WASH chip, adds some extra functionality that will be posted about in the future...

 

 

[+Lathe26]

I have 1 more item in the mail.  Until it arrives, the below item will be last one posted for a short bit.

 

This is Dave Chandler's Business Card!