Analogue tones of a ZX Spectrum Load set to ride again via podcast project The glory days of audio-cassette loading are set to return in the coming weeks, with retro fans to be treated to a broadcast for them to hit Play and Record to. Audio cassettes were the medium of choice for software back when Sinclair and Commodore's 8-bit hardware ruled the roost. The floppy disk seemed impossibly glamorous …
> We struggled with making tape-to-tape copying work
With a twin tape stereo for the copying and a decent tape player connected to the Speccy almost every tape I copied worked perfectly. For personal backup reasons of course, ignore that box of 50 C90s full of games...
Now, where did I put the colour code grid for Jet Set Willy?
"Of course I would never tell you to look over the web for some kind of archive site that has all those games, cause that would be illegal."
Actually, a lot of the original games developers or software houses have given formal permission for their games to be included in Spectrum software archives, as they're not exactly selling them any more, and many of the software houses are no more in any case. They got paid when we bought them in the first place, and so that's a pretty fair and decent stance for them to take, and we thank them again for (too) many hours of youthful mis-spent fun.
You might be interested in the work of 4am on twitter. They're cracking every single piece of Apple II software that was ever released in as non destructive a way as possible, so that they can be run on emulators and preserved for posterity.
The write-ups of the cracks are surprisingly interesting, even for me who's never used an Apple II.
For example, circumventing the mutilayered, homebrew protection on "The Quarter Mile".
The earliest piece of software I wrote in exchange for money was precisely to discourage this behaviour in Commodore owners. As I recall, it involved disassembling the system ROM and working out how to execute code in the tape buffer to unscramble lightly-obfuscated code (about all you could manage with a 6502 and 128 bytes). I have no idea whether it was ever deployed or, if so, succeeded in its aims.
But I did spend an inordinate amount of time loading files from tape and, frankly, it was painful even though I knew I was being paid. Can't see why anyone would repeat the experience voluntarily.
But I did spend an inordinate amount of time loading files from tape and, frankly, it was painful even though I knew I was being paid. Can't see why anyone would repeat the experience voluntarily.
A lot of old men remember the utter shit produced by Sinclair with remarkable fondness. I think it's a form of Stockholm Syndrome crossed with Survivor Bias.
Shit computer, or no computer? Which was the choice for most. The speccy did, if nothing else, lower the barrier to entry.
I started out on a sharp MZ80; a hand me down. At boot, you had only a very, very basic interface to start the tape or disk loader. 10 mins to load basic, then 10 more to load an actual program.
So the zx spectrum+ was quite the upgrade all things considered. I have little nostalgia for the speccy beyond it being a starter for ten. The c64 on the other hand, I have a lovely upgraded example sat here with SD card adapter and Epyx fast loader. Use mostly for Games obviously, but that SID synth is distinctive and useful.
There was a Tron-like Spectrum game called Blind Alley which had the most recognisable sound on tape. I could identify that from about 3 seconds of audio, when fast-forwarding through my tape player.
One thing I did manage once was to use 2 tape recorders and a phone call to transfer a program to my mate round the corner, with a simultaneous SAVE and LOAD and some silence it worked like a non-flow-control modem.
I liked the custom loaders. One of the best was Fairlight. It had an initial short load to get its custom code into RAM then silence.
..
..
Until suddenly the data came screaming in with no discernible lead-in tones. The custom loader also read data into screen RAM in the correct order to make the image appear from top down to the bottom.
And of course who can forget the intro music that demonstrated what a piezoelectric speaker could do even though the code could only send a 1 or a 0 to it.
"And of course who can forget the intro music"
I was kinda expecting the legendary Fairlight crack-tro with those words...!
Who remembers the ability to download software via teletext?
It was promoted on "Micro Live", presented by Ian McNaughton-Davies
I seem to recall the software was only available for BBC Model B owners, and you had to buy a separate teletext adaptor, in order to "download" the code.
You could actually goto (not the BASIC command) the Ceefax page on a suitably equipped teletext TV, but the code could only be seen.
https://en.wikipedia.org/wiki/Micro_Live
We had one of those adapters at school. It worked very well, but there was quite a limited selection of software from memory - some worthy education stuff and the odd game equivalent to the things you would type in from a magazine. Beat doing the actual typing in of course, and as it was all in BASIC it gave you extra time to examine the code and "adapt" it :-)
It also picked up normal teletext (it had four tuning pots and could switch between channels in software) and could save the pages (1k each). It was the first teletext system I'd seen with "fasttext" - the coloured buttons and pre-loading of pages.
M.
"It also picked up normal teletext (it had four tuning pots and could switch between channels in software) and could save the pages (1k each). It was the first teletext system I'd seen with "fasttext" - the coloured buttons and pre-loading of pages."
We had one too where I used to teach. As a project for the learners, we wrote something to download and link every page so we had an "instant" offline CeeFax :-) (Skipped the "live" pages that were constantly updating though)
When I was 15 I applied for funding to buy one of them and a BBC plus a specific model of video recorder that was easy to hack. The idea / purpose? To code a method of setting a VCR programmer using Ceefax/Teltext program listings and adjust the start and stop times automatically if the program start was delayed by say an over-running sporting event. I had a rudimentary 6502 machine code program already written to handle the user interface and control the tape transport system - I just needed real teletext data, a BBC to work on at home instead of writing it all at school during lunchtimes and a VCR with full solenoid control instead of just a prototyping board stuffed with optoisolators and LEDs.
Funding request was rejected but two years later the first PDC VCRs were launched followed shortly after by VideoPlus+ program code machines. Im not sure if this was connected or not!
>a method of setting a VCR programmer using Ceefax/Teltext program listings
My first VCR (can't remember make or model) did just that. It also gave me Teletext on my non-teletext TV; you had to do it all via the video channel though. It also had a neat trick of using the teletext data to superimpose the TV show's name and broadcast time & date onto the first ten or so seconds of the recording which made searching a tape much easier.
Double thumbs up for the ICU ref, although I think Nik Turner might have been a bit pissed off at describing them as a Hawkwind spinoff, even if the two were not entirely unrelated.
For those unfamiliar with ICU - check 'em out - especially the 2nd LP "The Maximum Effect".
Boring bit of info but I lived in a house 3 doors away from a Hawkwind member. We grew up on that street in the late 70s to early 00s before we moved out of London for good. They were still there when left or at least the mum was. His son Sam would come down their drive and sell his toys at the end of it (this was in the 80s). Went round their house a few times growing up. He played drums. They had a big tree in the middle of the garden with, what I remember to be a sort of treehouse wrapped round the bottom of it.
Some magazines even provided a short(ish) BASIC programme and put checksums at the end of each line of hex so using the programme to enter the data would check each line as you completed it before eventually writing out the hex data as a "save" file.
Software Hidden in Records | Nostalgia Nerd (31 mins)
The title is right, shame the video is littered with the word "vinyl" all the way through. Humbug.
I remember fondly counting out numbers of parentheses in complicated formulas, you only needed to be one out and the logic didn’t work work, and the program would generate an error. Was a bugger going through the whole page working out which formula had the missing (or indeed extra) bracket!
There's been plenty of people over at comp.sys.sinclair and World Of Spectrum who've experimented with various ways to play recordings into a ZX Spectrum - MP3 players, iPods, even (back in the day) minidiscs and the like.
There's even been programs tucked away on vinyl albums; I've even got vague memories of some magazine doing a flexidisc, and about a decade ago, I helped to slap together a loading screen for a "hidden" program on a CD album!
So yeah, it'll probably work, so long as they don't overcompress the audio.
I used to work on the BBC Wales off-shoot of the Micro Show in the early 80s. (I still have a Jupiter Ace and Oric Atmos that was sent into the show for review all those years ago).
One of the tasks at the end of the show was to hang around in the transmission continuity for Radio Wales and Radio Cymru and wait until the end of the normal schedule of programmes, which was normally around midnight. I would then load up a ¼ inch tape machine with the tape containing a BBC Micro or ZX Spectrum listing (being the BBC we had to be fair to both sets of enthusiasts and alternate every week). Then turn off the normal transmitter signal processing (an early form of Optimod) and press the play button.
Hopefully, someone was at home listening and loading the game off-air. The only problems we faced was that the system was very fussy about the tape head azimuth and quite often some pass band filtering improved the chances of success. However, we did get quite a few letters from Norway claiming it got that far.
Better luck on FM Radio Cymru than AM Radio Wales I'd imagine. I did try getting software off-air once or twice, with limited success. I was in Caerphilly where the surrounding hills made FM reception from Wenvoe rather hit-and-miss. The BBC Micro's format worked best in my experience, particularly as it was sent in 256 byte chunks and you could re-wind the tape a short distance and try one chunk again if it failed the first time, rather than every other computer which needed to start from the beginning.
At one point the off-air software was in some "agnostic" format which I never managed to decode.
I worked at Red Dragon / Touch Radio in the 1990s and we had AM reception reports from pretty much the whole of Europe once the sun had gone down, even though local after-dark reception was limited to about ten miles from each 200W transmitter!
One bloke in Germany sent us a photo of his girlfriend sat next to his radio (girlfriend? yes, we were surprised too), and one bloke (it was always a bloke) sent us a cassette recording of our FM station, picked up in Sweden, complete with a transcript of the RDS as proof it was us.
Fell to me as junior "engineer" to reply to all these oddballs.
M.
FM gave us the fidelity, and you could always drop one channel if there was an azimuth problem; often caused by a cassette duplication problems at source from the game companies.
The AM gave us the range as long as the upper frequency in the FSK (frequency shift keying) pair was clean (about 2kHz or so), which was OK for the 4.5kHz bandwidth of a AM Medium Wave transmission if slightly under-modulated. Also, we did have at least 100kW to play with.
As for the mystery format, could be anything we had in the cupboard at the time. As well as the usual Commodore and Tandy suspects there was an Oric 1, Oric Atmos and Jupiter Ace as mentioned, but also my favourite, a Memotech MTX512. The programme presenter was a guy by the name of Chris Jones.
Radio reception in Caerphilly was always bad due to the 'mountain', I used to live north of Ystrad Mynach in Penpedairheol.
"BASICODE" was an "agnostic" format. A very limited dialect of BASIC to run on many home computers. The BASICODE 2 "tape" sounded like BBC data without the (?) 256 byte chunks, and I could save a listing from BBC Micro, but ZX Spectrum would only load from the tape - and of course you had to load a BASICODE loader first.
Like many ZX owners I also had a tape to tape box and a bunch of C90s to play my 3rd or 4th generation "backups". Many hours spent fiddling with bass and treble trying to persuade the computer to load. Hard enough for normal recordings but later when there were various fast loaders and snapshot devices using weird baud rates it was even more glorious when the game actually loaded.
As a young spotty Herbert, I and my fellow Herberts would yell “Bagpipes!” when the loading noise changed on the loading screen. At that point, you knew that the monochrome image was about to be coloured. It was competitive. Call it too early and you’d receive a dead arm.
Like biblical numerologists looking for decipherable patterns, I would listen to the loading noise trying to discern the “meaning”.
One pre-pubescent Herbert mimicked the loading screeches and patterns into a cassette recorder to see if he could fool the machine. I recall the sum total was intermittent rolling blue and yellow lines around the screen edge.
Happy naïve days.
Essentially audio tape formats in the home computer era work by storing pulses on the tape. This was done in order to keep the hardware simple and to be able to deal with extremely horrible tape decks running at the wrong speed.
If you manage to keep a pulse vaguely coherent, you will be able to get the data through. This is also called a linear phase response, or a constant group delay. This isn't as important for general audio, so back in the analogue days people didn't take the effort of doing so.
Now since everything is digital, it's easy to have filters with linear phase response.
Also any general purpose audio codec has very little reason to mess with the phase, so the data will likely go through given sufficiently high bit rates.
BTW back in the analogue days there was a variety of ways to deal with non-linear phase responses which you could have when the audio was transmitted via television or the radio. The WDR Computer Club (West Germany) had a device which would cup off the overtones of the tape signal. That way they didn't need to care as much about what the overtones contributed to the signal. In (communist) East Germany, the solution was to broadcast the signal in stereo, loud on one channel and with 10dB attenuation on the other one.
audio tape formats in the home computer era work by storing pulses on the tape
Erm... no. The sound you heard from a cassette was a rapid switching between "high" and "low" tones (frequencies) with one used to represent a "1" and the other a "0". Filters in the computer would discriminate between the two to produce a serial stream of 1s and 0s. The switching point between the two might be subject to some phase errors, but that didn't matter because several cycles were recorded.
The two frequencies were chosen to be fairly far apart (the Kansas City Standard linked above used 1.2kHz and 2.4kHz) which meant that the filters could be reasonably forgiving about the exact frequency which resulted in a "1" or a "0", but a badly over- or under-speed machine, or one with too much flutter could still break the transmission. Phase, less so as in a single-channel system (mono) recording single - relatively low - frequencies it doesn't really have that much meaning.
Didn't stop some machines being far, far more sensitive to variations (particularly in level - volume) than others. I seem to remember a reasonable number of problems with my Spectrum, but not enough to be utterly frustrating, while my later BBC Micro only really failed on tapes with large amounts of noise, not fully erased previous recordings, or dropouts. Friends with other machines (naming no names) fared a lot worse.
M.
I built a home brewed Z80 machine around the time of the Spectrum. Programs were stored on audio tape with a 1 represented by 5V for 1 tick followed by 0V for 1 tick. Zero was 5V for 2 ticks followed by 0V for 2 ticks. This ensured that there was no DC component in the signal. Data was stored in 1k blocks with a checksum. It ran reliably for years.
I also vaguely remember a system of transmission of Spectrum code with a flashing square on the TV screen. It was picked up by a photo detector. Does anyone know anything about it.
If you plonked 5V into the "MIC" socket of a standard cassette recorder of that time you'd get a click on the tape, probably oversaturate the tape briefly too, though that depends on the auto level control. Following it by 0V doesn't guarantee "no dc component" it guarantees an average level of 2.5V. That said, there was probably a capacitor somewhere in the input circuitry (if not in the output of your board) so no dc would actually get through to the recording circuit.
What you would record would probably be a crude sine wave - a square wave shaped by the response of the capacitor(s) inline and how the FM circuitry dealt with such abuse - and playing it back through a speaker would result in a nasty two-tone buzz; a "high frequency" and "low frequency".
If your "ticks" were (let's say) at 2kHz, a string of "1"s would result in a 1kHz buzz while a string of "0"s would buzz at 500Hz.
The circuitry for reading it wouldn't be reading "5V" or "0V" directly - the output circuitry of the cassette recorder wouldn't allow that, and anyway it would be affected by the level of the volume control - the input circuitry on your Z80 board would possibly have been some kind of Schmidt-triggered "AND" gate or S/R flip flop, though I'd need to think about how to deal with "1"s arriving at twice the rate of "0"s. Was it actually "HLHL" for "1" and "HHLL" for "0"?
If you read down the article I linked about the Kansas City Standard it notes that while the original standard used was four cycles of 1,200Hz for "0" and eight cycles of 2,400Hz for "1", faster developments such as the version used on the BBC Micro used one cycle of 1,200Hz for "0" and two cycles of 2,400Hz for "1" which is conceptually similar to your scheme.
Yes, I remember the flashing square too. I don't think it was Spectrum code, but another kind of "agnostic" format that could be loaded into any one of several supported machines. I seem to remember that the sensor was a simple phototransistor connected to a data line, 5V and 0V. Simple and safe enough on a machine such as the BBC Micro which had a "proper" user port (or was the joystick port or maybe even the lightpen input used?) but dangerous on a Spectrum where the edge connector was the unbuffered lines direct from the micro inside. Of course, for a relative youngster such as I, sending off to Maplin for a single phototransistor, resistor, wire and maybe one of those connectors was a bit silly when you added the P&P, and I didn't get into Cardiff often enough to visit Mr. Marks in Wyndham Arcade in time to build something.
Dare say it would have survived video recording too, but we didn't have a VHS player until much, much later.
M.
I had a UK101 micro (UK version of the Ohio Superboard) that was terribly sensitive with loading tapes.
It turned out that the original design was based on 7400-series chips but at some point they had switched to the low current 74LS00-series chips for the later boards but hadn't compensated for this in the timing circuits for tape i/o. It was struggling to lock on to the frequencies because it was looking in slightly the wrong place. The fix was to work out the correct values for a resistor/capacitor pair to get it looking for the correct frequencies, it then worked as like a champ... until I found my next 'toy' (a BBC Model B)
This post has been deleted by its author
"Right up until the all-too-common R Tape Loading Error (which usually seemed to come right at the end of a lengthy period staring at a loading screen)."
There was a CRC check (or maybe just a bit count, I can't remember) on the whole file. If there was some glitch that didn't cause a dropout, it would only be detected once the whole thing has loaded .
I recall from the 1980s that one rock/pop band had a BBC program on at least part of one side of their vinyl record LP album release.
This was designed to be played into the BBC micro tape input just like the usual tape machine, to load the program in. As such this would have therefore been the same variant, CUTS, of the Kansas City Standard for digital storage on analogue tapes - so I would understand, from: https://en.wikipedia.org/wiki/Kansas_City_standard#Home_and_personal_computers
Back in the mid-late 1980s, I saw that rock/pop band on Number 73, Saturday Morning Kids/Youth ITV show, talking about it in an interview. I can't remember the band name. However some searching suggests it could have been one of these: Mainframe - Talk to Me ( ref: https://stardot.org.uk/forums/viewtopic.php?t=15405 ) or Kissing the Pink - The otherside of heaven ( ref: https://orchardoo.com/histsingle2.htm ).
Such a concept of combining music audio and digital data on physical analogue media resembles the concept with the CD: where CD mixed mode and CD plus CD extended standards. CD audio was the red book standard, CD-ROM was yellow book and CD plus (separate audio and CD-ROM parts on same CD) was the blue book standard - https://en.wikipedia.org/wiki/Rainbow_Books Some of those latter standards appeared to come after the aforementioned here vinyl with computer program concept - wonder if that inspired those later CD standards?
The spectrum was a huge improvement over the ZX81 that I had.
Reading and writing to audio tape only worked about 1 time in 10 and only then if you used the expensive Chrome Oxide tapes, had the knobs on the tape deck set just right and performed some kind of satanic ritual.
So realistically, if I wanted to wanted to run any programs, I had to type them in every time (on the crappy touch keyboard, while trying not to wobble the 16K RAM pack of fail).
My friends all had Spectrums. They learned how to load and play games. I learned how to type in programs and make them work. Which eventually led to a long career typing in programs and making them work.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
