The Computer Chronicles - Speech Synthesis (1984)

Special thanks to archive.org for hosting these episodes. Downloads of all these episodes and more can be found at: archive.org/details/computerch...

[@TheHabitman]

I love old television. You just can't get that charm anymore.

[@lenovovo]

+TheHabitman I could not have said that any better. You put it perfectly. Have a great day! :-)

[@iNCoMpeTeNtplAyS]

The charm comes from limited budget and time. They didn't have infrastructure or technology to make shows as easily as we do today so they had to nail it the first time or risk being g late.

[@rabidbigdog]

Heaps of people yelling at their PCs in an office didn't really catch on. I can't imagine why.

[@rabidbigdog]

@@patrickglaser1560 The fascination with speech synthesis in the early 1980s was so weird.

[@straightpipediesel]

@@rabidbigdog From the 1880's through the 1970's typing was a specialized skill that women took classes for in high school and did as a career. The ordinary businessman didn't know how to type, merely dictating letters. Voice recognition and voice interaction was a way to make computers accessible to the masses. (Manual typewriters require both physical finger strength, and also lack of automatic correcting typewriters meant you had to have very good accuracy) Even in the 90's, people got overwhelmed with GUIs with words everywhere. I remember old people thinking they had to read every single thing on screen, not familiar with say a dialog box taking attention. We still have this stupid fascination today, with Alexa and Google Assistant and Siri, that doesn't make sense outside of specialized situations (basically driving).

[@cyberyogicowindler2448]

​@@rabidbigdog I collect sound toys and research the hardware. In early 1980th they made a talking version of almost everything (e.g. watches, alarm clocks, calculators, bathroom scales) not only for the blind, but as a luxury/novelty hitech gadget. Later it was considered annoying, so for a long time most cars or appliances didn't speak anymore. (But e.g. in a washing machine it would make sense if an unusual fault prevents it from working. E.g. "I get no water. Please open the water tap and check the hose." or "Unbalance. Please untangle your laundry and restart spindry.") I am only happy that my talking alarm clock (unlike Alexa) is not designed to spy the mankind.

[@retro-nostalgia-vintage9600]

I love the computer chronicles.

[@raven4k998]

you gotta love there suits it made the show very unique no other computer shows on youtube even to day have to same look with them

[@Tristinfate]

40 years later and Alexa and Google still incorrectly pronounce some words and make people laugh.

[@SWRadioConcepts]

I asked Alexa to give me updates on my stock portfolio and all she replied was "Sorry, I can't do that." :(

[@Trusteft]

You watch this and you take what they say as accurate and objective. Even if they from time to time fall way off in their predictions. Even when they have guests from companies showing off their products, you know they are trying to get to the truth. Which quite interestingly enough, is given by the guests, without lying or over exaggeration. You try to watch tech programs today, youtube or TV, and you just know it, it's nothing but part of the marketing department of companies.

[@Jwdude123]

Trusteft truth

[@SWRadioConcepts]

you wouldn't get the actual engineers and designers as guests today, you get marketing and *high-energy* sales people overselling their product.

[@ArumesYT]

I guess you missed the episode in 1983 where Alan Shugart from Seagate did nothing but defend his own products in his "predictions". Same with Morrow defending his 5.25" drive in a laptop when the entire industry had already moved on to 3.5". And there are many more examples like that. Don't idealize the past, it doesn't get you anywhere.

[@nikw3026]

@@ArumesYT Hes right tho - don't look down on the past. The 80s-90s were peak America

[@jgood005]

Marketing knows how to sell. While I'd love to hear an engineer talk in-depth about their new product, it won't cause the masses to get excited to buy it. You need a marketer to do that. Which is unfortunate, because now nobody actually talks about the technical aspects. It's all quick, flashy marketing for the ADHD crowd. Presenters today can barely let someone complete an entire sentence, let alone two, without cutting in with some flashy graphic or something.

[@leonjones7120]

Amazing technology for its time. the sampling rate is well-chosen for speech.

[@raven4k998]

well Speech Synthesis didn't really catch on cause we don't use it on modern computers it must have been a fad🤣

[@oubrioko]

20:56 The clunk and stutter of the those original IBM PC full-height floppy drives sure brings back memories.

[@raven4k998]

imagine if they brought it back modernized the floppy drive and floppy disc to have more capacity it would be bought up more for nostalgia then anything else

[@louisloizides7488]

Thanks to these guys I have to deal with those horrible voice prompt menus now whenever I call customer service 😂

[@marctronixx]

lol was kinda on the same thought path and came here looking for this post. ::)

[@joecat4892]

the voice rec is incredible... lets not forget.. this is 84!!! .... I had problems with voice rec doing my uni work in 2004!!

[@raven4k998]

yeah and better yet is the voice recognition I swear it's just some guy talking on voice chat it's that realistic

[@oldtwinsna8347]

@@raven4k998 it's just prerecorded speech. this thing they are demo'ing has very little permutations going on so they could put together some phrases and digitize it for replay. would never fly for a real speech synthesis system.

[@UncleFeedle]

I wish Alexa had an option to sound like a vintage speech synth. Having the voice of Gorf or Wizard of Wor answering my questions would make it would seem more like I'm living in the future. 😄

[@raven4k998]

yeah well it's shocking at how fast the voice recognition works on such old slow hardware of that time when you think about it now days the hardware power is seriously overkill compared to that

[@davidt8087]

@@raven4k998they did it on a slow 286 or 386 SINGLE computer with one core one thread and only a couple mhz processor with less than a million transistors. Today, Alexa, Siri, Google , ALL use MASSIVE servers and THOUSANDS OF SERVERS to process speech. You couldn’t fit the code of Alexa or Siri for example in anything less than MANY MANY TERABYTES

[@cyberyogicowindler2448]

"Surrrendr the Blizzard of Blor..."

[@unclefreddy2009]

I miss these kind of shows and I remember when this was on the air.

[@RonJohn63]

I really like the intro music. And MicroFocus' definition of "visual computing" is pretty amusing (though understanding in the context of COBOL programming in 1984). 14:27 en.wikipedia.org/wiki/Epson_HX-20 "The Epson HX-20 (also known as the HC-20) was the first "true" laptop computer."

[@captainkeyboard1007]

I love my microcomputer and watching The Computer Chronicles. If the shows could exude strength, they would keep my fingers and right thumb on the keyboard.

[@8BitNaptime]

Fun! I had a great time as a kid in the '80s hooking up the SP0256 speech synthesizer chip to my VIC-20.

[@bastardtubeuser]

Herb rules its great when he is on, TCC had great recurring guests.

[@newworlddisorder156]

What a charming series

[@halfsourlizard9319]

Taking something that you could just read ... and turning it into a thing that you have to access over a voice call is the most-boomer tech imaginable.

[@Petr75661]

Stephen Hawking texted - he wants his synthesizer back.

[@patricknelson]

Just googled that company; turns out Speech Plus Inc. was actually the creator of the speech synthesizer for Stephen Hawking.

[@Petr75661]

@@patricknelson wow!

[@ens8502]

Haha, handicappeddly funny! :)))

[@raven4k998]

to late he wants his teeth fixed first

[@aarongreenfield9038]

​@@raven4k998 Might not be very practical to fix them now, being that he is dead.

[@jefferee2002]

I love watching these old vids. I'd love to have walked in on their set and showed them my smart phone with speech recognition.

[@aarongreenfield9038]

​@@chaeobods It's ok to violate, as long as It's to get rid of tribbles.

[@tom7601]

I had "SAM" "The software mouth for the Com-e-door 64." When it counted, it pronounced each number except 64, "6-2, 6-3, 64, 6-5”. It was fun...

[@chriscannon8527]

There's a port of it for modern Windows that works in the command prompt, pretty fun little program :)

[@cyberyogicowindler2448]

The C64 game "Tales of Arabian Nights" had plenty of that speech.

[@UncleKennysPlace]

Digital speech, in an analog film camera.

[@blazed85]

11:40 the dead look on his face after answering the question is quite funny 😂

[@jaybird57]

On the road, stop at a gas station to get your email over payphone! No way!

[@AiOinc1]

4:51 Siri and Cortana fans, eat your hearts out! But seriously, the things people think are new these days...

[@AiOinc1]

+AIO inc. Oh neat 18:19 as well with the responses! More ammo against Apple users who think they have the latest and greatest!

[@AiOinc1]

+AIO inc. Dat accent tho Kansas much?

[@jfcash84]

AIO inc. Hmm. Appletalk was standard on 68030 and 68040 Macs in the early 1990s. PCs didn't include built in speech synthesis till Windows 95, both platforms though didn't include speech recognition until OS X adapted Siri and Cortana was taken to the Desktop PC. Neither of which are perfect at what they do.

[@nitramluap]

Not only that, but they whine about how 'bad' they are when they really have NFI just how *good* they are.

[@cmatthews718]

@@jfcash84 Appletalk has nothing to do with speech. That was a networking technology.

[@janruudschutrups9382]

"Phonemes are the basic elements of speech, for instance 'shu', 'thu', 'fuh'." 😝

[@wallacelang1374]

These Speech Synthesis programs are specific language dependent, so when the English language version is installed than it will only speak in the English language out loud. I remember that when the phone company had debuted a speech synthesis feature saying current time it was ended just a little bit later when the US government chose to break up AT&T from its baby Bell phone companies.

[@marctronixx]

1984 I was at the World of Wheels (Sheriff Roscoe P. Coltrane was there!!) and there was a kiosk with this system in it. you could type anything in it and it would do just as it did here. me, being silly, typed in a bunch of random letters and see how it sounded. well, it sounded just like the phonemes those letters represented -- a bunch of random sounds... :p

[@bastardtubeuser]

for all the Herb-a-maniacs out there 16:05 brilliant presentation. blooming fantastic.

[@bastardtubeuser]

and why would there be so many Herb fans ? well Herb knows what he is talking about. not pressing a button that automates centuries of research in man hours then laughing about how easy it is. big difference. Herb can do it in his head and presents himself well. Not mentioning his actual work.

[@calvinsaxon5822]

All hail Herb! All hail Herb! All hail Herb! Herb is our past, Herb is our present, Herb is our future salvation. Herb. Herb. Herb.

[@CB3ROB-CyberBunker]

this guy's voice is the reason why early speech synthesizers all sound that way :P he kinda speaks like one himself lol. according to him that sounds 'natural' lol.

[@CB3ROB-CyberBunker]

also he asked how it works. not 'a general description of what it does' lol :P the expected answer would be something like 'we put the pcm samples in some rom and hook an offset table to the start of each one then somehow link that to ascii values or combinations thereof' :P aaand then we simply ram that through a resistor array at the back of it to generate the voltages' :P or 'something like that' :P would be 'how it works' not 'what it does' lol.

[@curiousottman]

Wow that speech synthesis was disappointing in 1984 however still sounds like a scammer calling me from the IRS in 2020.

[@jimmybuffet4970]

The Votan system was AMAZING! Wow! 1984!!!

[@tarstarkusz]

Human speech is an area where computers haven't really progressed at the level of computing generally. Even the text to speech features on new computers aren't that good. Android and IOS are getting much better though.

[@PhoenixNL72-DEGA-]

Yeah, there are plenty of computer programs that read books aloud for blind people (I don't mean audiobooks but programs that read out epub, pdf, plain text etc) and they all still sound pretty bad.

[@tarstarkusz]

@@ghost_mall Nobody would ever mistake t2s as a person talking.

[@jr2904]

​@@tarstarkuszolder people do

[@tarstarkusz]

@@jr2904 What do you mean by older? I'm in my 50s and they don't fool me. If anything, I would think younger people are easier to fool because they've heard it all their lives.

[@cyberyogicowindler2448]

​@@PhoenixNL72-DEGA- Many blind people playback text2speech at 5x speech that sounds like completely gibber to untrained novices, but permits them to read at the same speed like seeing people.

[@CMDRScotty]

"Would you like to play a game?" "I can't do that dave." "Come with me if you want live." "Your all going to die down here." "Danger!!!Will Robinson, Danger!!!" A dream come true or a nightmare waiting to happen you decide.

[@retrogamer33]

They're all dead Dave (Holly - Red Dwarf)

[@777jones]

Number five is alive.

[@simonRTJ]

"I dont much pay attention to warning lights on my car" wow really? you should not be on the road sir.

[@dhpbear2]

Design flaw: "Buy a thousand shares of Votan" Should have echoed back a SYNTHESIZED version of Ron's speech, not simply a recording of it! Oops! After viewing the rest of the presentation, that WAS a synthesized version of his voice!

[@janruudschutrups9382]

David Perkins My suspicion, especially given the major difference in quality between the Call Text product and the Vodan demo, is that maybe the latter was all simply pre-recorded samples instead of actual computer voice synthesis.

[@TheGreatAtario]

That was a note for his broker, who, presumably, is a human

[@DavidJG242]

I like this episode

[@WaybackTECH]

The Votan V5000 in 1984 beats Siri, S-voice and Microsoft Product Activation Center in 2015.

[@ruthlessadmin]

WaybackTECH I was just about to comment: It's sad that voice recognition doesn't work even a tiny bit better in 2015...OTOH, the device I typically use it with does fit in my pocket, so there's that, I guess. Personally, I think they had someone in the back w/ a mic ;)

[@HikikomoriDev]

statikreg It`s staggering too because with all the data mining Google has, from Google voice, etc etc, it`s still kinda bad if we think about it. Not the best yet.

[@liamchilders9516]

WaybackTECH Microsoft Is really good go Microsoft 3 of 5 stars.

[@mubd1234]

I don't believe the voice was generated by a computer. I think he's lying!

[@Fygee]

It sounds like it's using precorded samples.

[@skyelarmurray467]

1:20 Siri on some hardcore shit 😂😂

[@tarstarkusz]

22:40 10kb/S (75,000 bytes per minute) is not bad for voice. Probably costs an arm and a leg at the time.

[@MrGarg10may]

This technology doesn't even exist today 😂 the computer voice was so human, today bots don't sound like humans

[@velimir5780]

Not true. Some really sound very human-like!)

[@lindaoffenbach]

@@velimir5780 I guess you just confirmed... 'human-like' does not equal Human.

[@JBuchmann]

In the last segment the voice seemed too realistic for 1984. And the back and forth conversation and processing seemed way too fast. I got the feeling the interaction was staged, like there as a guy in the back room pretending to be the computer voice.

[@oldtwinsna8347]

@@JBuchmann pre-recorded discrete samples. they had a bunch of that in the 80s which was basically just an analog sound track being activated by some digital logic circuitry. very limited use but the practice of it sounded futuristic and amazing. only thing is, the entire way this is done has zero application to the methods needed in real digital synthesis.

[@BdR76]

19:11 Speech recognition was clunky back in 1984, now 30 years later; still problematic.

[@RonJohn63]

And two to three years away!

[@doplere6364]

2021 and computer still has trouble understanding what I am saying lol

[@rzober89biologia]

he talks about RS232 but it seems to be some different port, more like LPT

[@blackstar2008]

It’s a standard DB25 connector. It then became smaller using DB9 connectors

[@alphonsocarioti512]

Hey Siri... What is Herb talking about?

[@jaybird57]

Is that a zilog80?

[@n.miller907]

Kraftwerk used a lot of these devices in their music.

[@infinitecanadian]

This must have been a tape from a college course or something, because the man called it a 'lesson' and referred to a textbook.

[@syferdet]

I believe he was a professor and the shows were for a college course he developed and taught then coordinated the curriculum with Stuart. That's my guess anyway. Computer Chronicles must have been required viewing, which wouldn't bother me at all.

[@infinitecanadian]

@@syferdet It would be great to have this as required viewing.

[@AbdiPianoChannel]

Speech recognition still not working.

[@NickKont]

Google/Samsung and Apple will strongly disagree with you and they have a strong case for that

[@AbdiPianoChannel]

@@NickKont Well, none of them can pronounce my names. I have to write my name different way for the speech tech.

[@NickKont]

@@AbdiPianoChannel well give the AI behind it time to figure out things and it will be solved that too.

[@patricknelson]

Wow! Computers. Aren’t they somethin’? That’s somethin’.

[@lindaoffenbach]

Still a big challenge today and still extremely annoying in all cases applied. Speech simply can not be isolated from situational feeling/emotion and instant intuition.

[@tommyeastwood4393]

Come on. It finally works well..

[@liamchilders9516]

Siri can't remember anything Crotona remembers everything.

[@melmaciandissenter2324]

I don't want my com pewter to talk to me !

[@oldtwins]

have to admit that voice synthesis and recognition is one of the poorest areas of computer development. graphics, sound, processing power, network speeds, all beyond the wildest dreams of where we are compared to 1984 but only incremental gains in voice

[@RobertLock1978]

I was thinking the same thing. Even speech synthesis hasn't improved much.

[@dorlaretz5901]

@@RobertLock1978 Bollocks

[@codeoptimizationware2803]

@@RobertLock1978 : Well, ya know, these things don't advance themselves! hehehehehehehehe

[@RobertLock1978]

@Code optimization Ware - Hahaha - good catch.... though the doers of the action upon speech synth. [i.e., engineers] were implied..... :P

[@thestarglider]

Alexa may wanna have a word with you now ;-)

[@iliadde]

Stephen Hawking voice! 😮

[@clemstevenson]

And that's how Stephen Hawking was constructed.

[@clemstevenson]

@@ghost_mall I was only joking.

[@murdaone261]

UNITED STATES CENTRAL INTELLIGENCE AGENCY DIRECTOR LAMONTE M. WARD

[@RonJohn63]

27:52 Lesson? Textbook?

[@HikikomoriDev]

RonJohn63 ...I wonder where those are. Seems like good material for historical reviewing..

[@jfcash84]

This show was used in computer science classes in the mid 1980s in community colleges. Therefore the reference to the textbook.

[@RonJohn63]

@@jfcash84 that's scary. (I went to a third-tier state school in the mid 1980s, and got a good education in programming.)

[@BBC600]

jfcash84 If that was the case I’d have thought they’d advertise the supplemental textbook (this was aired on PBS based on my knowledge). Based on a Google search it appears Herb Lechner wrote the book to accompany the show yet I can’t find any record of the text itself. EDIT: This made me dig deeper and here’s an Amazon link for the book 📖 www.amazon.com/Chronicles-Wadsworth-continuing-education-professional/dp/0534033962

[@jr2904]

​@@RonJohn63how is that scary?

[@retrogamer33]

Faux Neemes

[@kevinnivek8907]

Scarlett Johansson does not approve.

[@tarstarkusz]

Herb Lechner sucks. He sounds like he's reading lines. Gary Kildall is a MUCH better tech co-host.

[@RajeshKumar-mw6uo]

best computer engineers