Talk:Amstrad CPC

From Wikipedia, the free encyclopedia
Jump to: navigation, search


Suggestion - Can we add a bit on the addition of added Firmware. My 6128 was always using PROTEXT? from Arnor? and a bank of other ROM based software Victuallers 11:23, 18 January 2007 (UTC)

Comparison section moved from article[edit]

The following text was submitted to the article by an anonymous contributor, but was subsequently moved over here due to its rather "unencyclopedic" form---as it stands now, it smacks somewhat like a newsgroup posting (nothing wrong with that per se; it just doesn't fit in a wkp article, looking like that).

CPC and other small Computers of this time

trying a subjective ;-) & short run on that item you got the following:

  • it was simple said better then most of Sinclair's computers, lacking even the microdrives, having a good keyboard and you haven't so much hardware bugs. Well, they are earlier in the market.
  • it compares well with the Commodore C64, better BASIC on the CPC, but there're no sprites
  • it was quite cheaper than the early IBM PC, but lacking it's modularity and future.
  • it was cheaper than the early Apple Macintosh, which have a GUI, but to less memory
  • it was worse than the later Atari ST and Amiga, in speed and 3.5" drives

In those days you've to wait for hours calculating a single mandelbrot, hoping to have a Gray some day ... :-) As the Commodore, the CPC was a well done homecomputer, made for working and writing programs, for playing, sometimes. It was brave one, surviving nowadays in emulators and old, but working electronical history.

Regarding the first of the items above, I observe that the "better than"-argument, as far as it's relevant, follows partly from the difference in release dates* of the machines (as the contributor indeed indicates), and that the keyboard should be given more attention in the article than is now present.

(* due to the time interval's progress of electronics, falling prices of components, and Amstrad's advantage of not having to do very pioneering work in home computers)

As for the second point above, which is generally correct, the main features of Locomotive BASIC on the CPC, compared to the Spectrum's Sinclair BASIC and the C64's Commodore BASIC should perhaps be more clearly noted in the article.

The three remaining items is not particularly relevant, I think, since the 16-bit machines listed were not direct competitors to the CPC (or the Spectrum or the C64, for that matter, although some CBM adverts of the early 80s actually listed the original IBM PC as an expensive alternative to the C64).

All in all, however, I think the section offered a reminder of several points to be improved in the present article. --Wernher 21:25, 8 Mar 2005 (UTC)

Gosh. What is it about the Amster that attracts people who can't write grammatical English? 8¬)

How good's your German? There's something about the style of the bulleted list and the paragraph after it that suggests a native German-speaker, to me. PeteVerdon 01:55, 27 January 2006 (UTC)

You raise some good points. I also see that since I went thru' that article a week or so back, several other [badly-written] additions have been made. I shall earmark it for another rewrite as soon as I have the spare time! -- Liam Proven 17:14, 11 Mar 2005 (UTC)

OK guys...I added tons of stuff and corrected some of the older ones..."Amstrad floppy disks were "hard" on the inside" like Zip Drives ??? Who wrote that ? :-) Try and slide a 3" disk open: it has a normal "soft" magnetic disk, unfortunately...

Also, no mention was made on the possibility of using 5¼ disks and (external) drives, which many power users did, for the sheer cost of 3" disks... EpiVictor 11:41, 29 Mar 2005 (UTC)

Someone should add how and why the CPC supports some C64 games, I experienced several C64 games that loaded normally. Most likely due to the cassettes just stroing BASIC commands not actually assembly.. but thats pure speculation. Info from someone who knows would be good. Also it would be good if this document could be changed to British english as well.... color -> Colour etc (If you need justification, its a British company that sold its product mainly to Europeans and Australians (who also use British Spelling)). UnlimitedAccess 06:57, 6 May 2005 (UTC)

It is just possible that the CPC may read DATA out of some C64 tapes, but this depends on how the data was encoded to an audio signal. E.g. the frequency for "1"'s and "0"'s might be different, or even by the whole encoding scheme, which would make "speedload" or "turbotapes" essentially incompatible with different computers' datacorders.

Even if the C64 and CPC (and ZX Spectrum, for that matter...) tapes or disks were or could be made compatible under certain circumstances, "normal" binary programs wouldn't load at all, and BASIC programs would only load if they were saved in ASCII (and not in the usual space-saving format), unless it was common between the machines.

Even then, they would only be parsed and able to run only if they were very "clean" basic programs, e.g. with only simple commands and structures common to all BASIC dialects like 'PRINT', 'IF'..'THEN'...'ELSE' etc. while they would miserably fail to run at the first 'POKE' instruction (which made 'BASIC' programs, de-facto, machine dependant) and at any specialized 'C64' BASIC instruction or expression (just remember how screen colors were mapped to keyboard keys in the C64, or how the C64 lacked 'BASIC' instructions for sound and graphics). EpiVictor 17:52, 6 May 2005 (UTC)

Well it happened, Garfield the game, Dynamic Duo and Kentilla for C64 all worked on my CPC, no other game I found worked though.. :), but I didnt have that big a collection of C64 games to try. UnlimitedAccess 21:26, 6 May 2005 (UTC)
Are you sure those games weren't actually stored in "dual format" tapes?

Some developers/publishers had, in fact, marketed "multi-format" games with the same game for e.g. C64 and Amstrad CPC on one tape, but on different sides, with little success. I mean, even the BASIC dialects of the two machines aren't compatible, let alone the hardware... EpiVictor 23:35, 6 May 2005 (UTC)

Possitive, Kentilla was a text adventure game and I have it for both the C64 and the CPC464 version and both work on my CPC... If they were duel format, they are unlabeled as such.UnlimitedAccess 02:13, 7 May 2005 (UTC)
Garfield very well could of been duel format because it appeared tempremental and only worked some of the time.. perhaps it was luck when i picked the CPC side or not.UnlimitedAccess 02:21, 7 May 2005 (UTC)
Well, the cases can only be two: since Kentilla was a text adventure game, as long at didn't use any sound or ghraphics, it could probably have been written entirely in BASIC and rely on ASCII saving for loading and running on both machines (like some sort of "portable software", if you prefer). The other option would be of course making a small BASIC loader (written so its compatible with both machines) and then the loader discerns what machine its running on and executes the proper machine code located in an external file or in the DATA portion of the BASIC file (with POKE's, PEEK's, etc. ). This has of course the disadvantage of taking up double space for each game, so it can only wotk with simple programs.

Most probably, your games were either twin format or, in the case of the text adventure, written in very clean basic. This can be verified with an emultor, at any time...EpiVictor 17:14, 8 May 2005 (UTC)

RS232 printer connector[edit]

Was it? RS232 is an 8-pin connection: I remember the printer port to be an edge-connector style of thing, much like the expansion connector, but I may be wrong.

It looks to me as though it wasn't. The French variant of Wikipedia states for the 646 that "On regrettera notamment l'absence d'interface RS232" and the comp.sys.amstrad.8bit FAQ ( states in answer 3.4 "Neither the CPC nor the PCW have a RS 232"
RS232 is a serial standard. The CPC's printer port was a Centronics-ish parallel port, an edge connector on the mainboard. As I recall looking at the manual, it actually only provided 7-bit support (for some reason the MSB was connected to ground). This caused me great bother when I tried to get it to provide graphical print-out on an old EPSON 9-pin printer because it couldn't send a complete ESC/P command set.
Amstrad released an RS-232 add-on, which plugged into the back. As said, the printer port was a parallel one with the 8th bit zeroed. I think the 8th bit of that IO port was used for something else. — Preceding unsigned comment added by Lovingboth (talkcontribs) 16:55, 22 March 2013 (UTC)

3 1/2" drives[edit]

There seems to be an argument as to whether these were more or less popular than the standard 3" drives. To my mind, the 3" drives were certainly more prevalent among Amstrad owners and a fair amount of commercial software was released on this format. Of course major hobbyists would have used 3 1/2" drives but these are in the minority. Cpc464 11:56, 18 October 2005 (UTC)

Commercial software (especially games) were always released in 3" format. However, the so-called "power users" as well as people needing to keep a lot of data quickly turned to the cheaper 5 1/4" disks for storing their games, and (partly) for exchanging data with a PC, which was sort of possible in the 5 1/4" format. There were even "certified" 5 1/4" drives issued for the CPC series. The 3 1/2" format was POSSIBLE to use with a sort of hardware hack, but I think that "official" 3 1/2" drives were never made for the CPC, so that's something only power users could do. Besides, not many other machines used 360KB double-sided in 3 1/2" format. EpiVictor 22:50, 19 January 2006 (UTC)
AMSDOS was on ROM and provided with the proprietary 3" disk drive. You could use a 3.5" drive for a _second_ drive, third party models being available.
A 5.25" drive may have been briefly available for the CPC, but they were never even remotely popular (3.5" becoming standard for Mac and PS/2 way back in 1984). (talk) 16:00, 10 September 2011 (UTC)

Exceptional games[edit]

As well as Gryzor and Turrican, there's a game I remember which was only available on the CPC and from a programming point of view it blew everything away. I think it was called Helter Skelter and the guy who developed it also wrote for CPC User as "The Hairy Hacker". The game was so complex that when run from tape, it actually had to overwrite some of the base O/S to give it enough space to run.

I'm not adding it as I'm not 100% sure this is the correct name, but if someone can confirm then this definitely deserves to go in as an example.

IainP (talk) 20:47, 25 December 2005 (UTC)

[1] Here is a review and screenshot of the game, but seriously neither the screenshot nor the review look very promising :-( It looks like it's yelling "I'M A SPECTRUM PORT!", "quite a contrast to the gorgeous loading screen and the cute and catchy music" as the review says. This other review is in german, but the CPC screenshots from the game look spectrumesque :-( EpiVictor 22:37, 19 January 2006 (UTC)
Just played the game...well it's nothing exceptional, and nothing seems to indicate a "complex programming". The graphics are basic, the music is catchy and the frame rate suffers a I don't have the "sk1llz" to check if it does indeed overwrite the tape OS while loading (BTW, the basic program modules are within the 64K limit) but else it's NOT an "exceptional game", under any point of view EpiVictor 22:45, 19 January 2006 (UTC)

Criticism of CPC software (games)[edit]

The second and third paragraphs in this section contradict one another. Either the CPC had hardware scrolling, or it didn't... but on reading the rest of the section, I suspect it only had horizontal scrolling. Is it saying that the display memory has a complicated, non-linear arrangement? Or that to scroll quickly, you'd have to selectively copy memory around?

Also, it does a bit of a disservice to the ZX Spectrum and C64 display attributes, which allow 16-colour, 'high-resolution' graphics by limiting the number of colours in each 8x8 pixel square to 2. (The C64 had hardware sprites, and afaics many C64 users were not even aware of this limitation.) Other systems e.g. the NES and MSX1 used similar schemes. --StuartBrady 15:01, 22 January 2006 (UTC)

-- I'm pulling this from my head, and I was a teenager at the time, but there was definately a few whacky little hacks you could use to get fairly convincing scrolling involving tweaking some registers around. There was a mess of strange undocumented stuff around , including something that could get it to adopt a higher res, but then swiftly shit itself due to it mapping into memory all wrong. God it was a fun unit to tweak, especially for a precocious Z80 hacking teen (talk) 11:06, 16 November 2008 (UTC)

Systems like the C64, Atari 800 and NES had hardware sprites and so simply scrolled in low resolution.
You couldn't do that on the Amstrad or Spectrum because you could only have character block by character block animation in low resolution with software sprites as the video processor only addressed character blocks not pixels (hardware sprites could work in any mode). So they used high resolution which chewed up heaps of processor time.
"Scrolling" is simply where you animate a character block moving across a block. By synchronising the different blocks, you get movement. (talk) 06:51, 24 October 2011 (UTC)
You are properly confusing several things here. Hardware sprites, hardware scrolling and resolution are not co-dependent. Of course the CPC could address single pixels. But contrary to hardware sprites it was "destructive" to paint over a pixel, which meant the software would have to keep track of what it painted over. Which screen resolution to use was an aesthetic decision since lower resolution usually meant a higher number of available colours. And hardware scrolling is a hardware function that shifts the screen by single pixels (after 8 pixels, you would then have the software shift the screen content by one character row). Simply put, the shortcoming of the CPC was that it had to do sprites and scrolling in the CPU, and that means that things like single-pixel scrolling or a big number of sprites are forbiddingly expensive. --Takimata (talk) 10:01, 24 October 2011 (UTC)
"Screen resolution" choice is not aesthetic. "Low resolution" had _precisely_ the same number of pixels as high resolution. The difference is the adressable space. In low resolution, only character blocks can be addressed (Note : resolution is not the same as "mode". Each mode has a low and high resolution). So software sprites can only work in high resolution if you want them to move pixel by pixel. Eaxctly the same is the case on the C64.
In fact, no 8 bit machines had "hardware scrolling". The scrolling registers on the C64 were scan registers, to synchronise the CRT scan with the drawing (eg raster bars), which the CPC also had. In fact smooth text scrolling was common on the CPC. (talk) 11:52, 5 November 2011 (UTC)

The NES had hardware scrolling, its a very well known fact — Preceding unsigned comment added by (talk) 03:52, 11 April 2012 (UTC)

7-bit printer port[edit]

I remember as a big deal the choice of supporting only 7 bits in the parallel Centronics printer port, i.e. the most significant bit was simply not wired. This meant that special characters such as umlauts were inaccesible, and numerous work-arounds were published then. Even CPC enthusiasts admitted that this was even more idiotic than the 3 inch floppy drive. I'm surprised not to find this fact mentioned here. Or did it apply only to the German Schneider models? Simon A. 13:33, 17 January 2006 (UTC)

It happened to absolutely all Amstrad computers, even for the first IBM PC compatibles (the later ones included 8 bit)
Improve the article correcting that fault :p
Claunia 21:24, 17 January 2006 (UTC)

The reason the printer port was only 7 bit was that the 8th bit output from the parallel interface chip was used to toggle the cassette motor relay on and off. This saved the cost of a dedicated chip to do motor-control on the cassette deck. One favourite hack was to perform surgery inside the machine and solder a wire across from the motor relay to the 8th pin on the edge connector (severing the track-to-ground connection at the same time), and then inserting some simple code in the firmware to use the 8th bit when required. (talk) 01:55, 15 April 2008 (UTC)

Ahh, yes, that's it. Anything to save an IO bit! It would have worked fine with Postscript, but anyone who could afford a Postscript printer wouldn't bother with a... well, I suppose they might not have any money left for anything else :) Lovingboth (talk) 10:06, 23 March 2013 (UTC)


From the article: "CPC titles rarely featured smooth scrolling or sprite handling due to programming complexities". Now, I suppose the smooth scrolling is a fair enough point, but sprite handling? Sprites are everywhere in CPC games. It may not be hardware sprite handling, but I don't think that makes a difference.Cpc464 13:16, 20 January 2006 (UTC)

The "smooth" adjective refers to both "scrolling" and "sprite handling", it doesn't mean that the Amstrad CPC titles didn't have ANY scrolling or sprites. If you find the phrase confusing as it is...well...rephrase it :-) EpiVictor 16:36, 20 January 2006 (UTC)

The c64 could have attribute clash too as it had a limited number of colours per pixel block, obviously not as servere as a speccy, thats why 64 games had chunky graphics to avoid it.

  • More colours (27 vs 16)
  • Faster disc drive
  • Much better BASIC and OS/firmware support
  • Better picture (CPC monitor vs TV) (Monitor included as standard)
  • Standard support for CP/M
  • Graphics modes are not attribute/colour map based (i.e. Any pixel can be any colour)
  • Better expansion ROM support
  • Easier to program CPU (Z80 vs 6502)
  • More reliable machines ( the return rate for CPC's was about 2% compared to about 10% for the C64 and 40% for the Amiga)
  • Faster processor (clock rate)

—The preceding unsigned comment was added by (talkcontribs) 2006-07-02 17:57:49.

I really don't think these sorts of comparisons are helpful. It's difficult to know whether they are biased... and sometimes bigger numbers aren't everything. (Consider, for example, the quality of a machine's BASIC implementation... how do you measure that fairly?) --StuartBrady (Talk) 22:34, 2 July 2006 (UTC)
As for the claim the CPC had a faster CPU, the clock rate isn't everything (we see that even today). A 1 MHz 6502 was very fast, a Z80 at the same speed would have been much slower. Anyways, I agree with StuartBrady about the comparsion stuff not being very helpful. 17:59, 5 February 2007 (UTC)

Aleste 520EX[edit]

I've cleaned up the text as much as I can, but I can't find anything out about this machine. If nobody can find any sources for this, I think it will have to go. --StuartBrady (Talk) 12:51, 2 August 2006 (UTC)

From its description it seems more like a "general purpose 8-bit computer" like the Sprinter, whose main purpose is ZX-Spectrum hardware emulation but is (should be?) flexible enough to emulate other Z80 based systems. EpiVictor 15:15, 3 August 2006 (UTC)
However, following and translating the provided link, it seems that this "Patisonic" firm has produced only a ZX Spectrum clone, and an unspecified 64-K machine with 4-channel digital sound output...but no mentions to an "Aleste 520 EX" machine or anything suggesting they designed and built a "Super CPC". EpiVictor 15:20, 3 August 2006 (UTC)

Is the Aleste 520EX just an Atari ST clone (or hoax)?[edit]

The link provided] seems more of a hoax or original research, both against wikipedia's policy, and does not contain any useful info, yet.

In any case, the alledged "Aleste 520 EX" specifications look a lot like those of the Atari ST and closer to the MSX2 than to the Amstrad CPC (512 KB Ram, nearly identical video modes with the MSX and Atari ST, same sound chip, same 3.5" floppy drive) and the CPU speed (8 MHz) is highly atypical for a Z80 (not impossible) but again, closer the the Atari ST's 8 Mhz (with a 68000).

Even the name...Aleste 520EX ...Compare to "Atari 520ST"...very suspicious, and "Aleste" was also a successful series of games on the MSX, if I'm not wrong. On the other hand the Patisonic firm, on their website, only claim to have made a Spectrum ZX clone called Patisonic 48ST, with no references to either Amstrad, CPC or MSX compatibles.

In any case, from the description alone the "Aleste 520EX" classifies more as an MSX clone than a CPC clone, perhaps as a counterpart to the Sprinter computer, which instead focuses on the ZX Spectrum. EpiVictor 23:15, 3 August 2006 (UTC)

No idea! There is an article on the Russian Wikipedia, ru:Алеста (компьютер) (English translation) but the article is rather new. It's an Article for Deletion, apparently because it was felt that the information belonged in the CPC article instead. It was created from, which also was used to edit the Aleste 520EX section in this article. --StuartBrady (Talk) 19:31, 4 August 2006 (UTC)
Looking at this in more detail:
  • Memory: I wouldn't take too much stock in the fact that this had 512 kB of RAM — it was quite common for these clones to add more memory.
  • Video modes: These don't match the modes of the MSX2 at all... so either the MSX and CPC software was modified, or the machine had more than just those two modes.
  • Sound chip: the MSX and CPC used the AY-3-891x, and the Atari ST used the YM2149 (which was basically the same). From the description given here, the Aleste had something quite different.
  • Disk drive: this is hardly a distinctive feature.
  • CPU speed: for the mid 90's, an 8 MHz Z80 doesn't sound at all suprising. The SAM Coupé had a 6 MHz Z80... but if it runs MSX-DOS and CPC software, it must surely have a Z80.
Quite believable actually: "Sinclair User" mentions the Z80H running at 7 MHz[2], and this was in 1986 — 8 years earlier! --StuartBrady (Talk) 21:05, 4 August 2006 (UTC)
I have to agree, it does sound more like an MSX clone... but perhaps with some CPC compatability? As for the name, they could have just wanted to pretend it was an ST. Some of the NES clones have Playstation and N64-like cases! OTOH, it could be a hoax, I've no real evidence one way or the other. --StuartBrady (Talk) 20:53, 4 August 2006 (UTC)
After some more digging: says (translated):
The computer Of "alesta" these are clone "Amstrad S.PS" but is working on by control MSXDOS. [...] Some technical special features were borrowed from MSX 2, "Atari ST" and "Amiga". "alesta" was popular among amateurs MSX since it had the analogous organization of memory, the size of disks, the operating system MSXDOS There was complete compatability with MSX not, but "alesta" computer was accessible, whereas To "yamashku" it was not purchase
It goes on to suggest that MSX compatability was done through emulation. I'm finding the whole thing rather confusing. --StuartBrady (Talk) 21:34, 4 August 2006 (UTC)
It sounds like a hybrid of various parts of 80s and 90s home computers. I meants the video modes match more those of the ATARI ST, not the MSX or MSX2:
  • Low resolution - 320×200 (16 color), palette of 512 colors
  • Medium resolution - 640×200 (4 color), palette of 512 colors
  • High resolution - 640×400 (mono), monochrome
So at least the first two video modes seem to match, as does the palette.
About the sound, it seems implied that this machine, if it's real, used the AY-3-891x chip (presumably built-in) as a means for MSX and CPC sound compatibility, and some kind of external "soundcard" for playing back MODs/ST3s. This would narrow it down to a DMA-fed 4 channel DAC or something equivalent, maybe even ripped Paula chips from Amigas, as it would be hard for a Z80 to play MODs by software mixing alone (well, there was a MODplayer called Galaxy that could play MODs at 22 KHz on an 8 Mhz XT, but that's another story).
No problems about the CPU...there's no reason why it shouldn't use a faster Z80 at the time. All in all however, the machine seems a lot like an MSX and CPC oriented Sprinter, which combines "advanced" hardware and software to act as a relatively expandable and versatile emulator. BTW, even the Sprinter, through heavy reprogramming, is capable of running CPC software but its primary architecture is optimized for ZX Spectrum emulation. If this Aleste 520EX thing is real, then most of its features suggest that its designed with MSX compatibility as its primary purpose (not least, because of the great numbers of MSX software available on 3.5" disks), and secondarily for Amstrad CPC emulation and acting as a low-end STM/ST3 player (if so, there could be a small demoscene for this machine). EpiVictor 10:45, 5 August 2006 (UTC)
In 1993, I really doubt it was using FPGAs. BTW, the S3M claim seems especially dubious, as S3M can use 32 channels. --StuartBrady (Talk) 13:16, 5 August 2006 (UTC)
Neither this section nor the Russian article mention the 640×400 mode for some reason. It's probably not a coincidence that both have been added at roughly the same time. --StuartBrady (Talk) 13:23, 5 August 2006 (UTC)
Yeah, the "russian page" seems to have been created recently, perhaps specifically for (misdirecting?) this article, and IMO there's a large chance that this "Aleste 520EX" machine is entirely imaginary and is a sort of "collage" between features of 8-bit and 16-bit home computers of the 80s and 90s: the video modes of the ATARI ST (and the name, too), the sound chip of the Amiga, the CPU of most 8-bit machines etc. etc. Also, on the few collectors pages there's nothing similar to this phantomatic Aleste 520EX. My proposal? Remove it from the article and at most make it have an article of its own, perhaps in the catetogory of "Rare hardware" or even "Imaginary hardware", if there is one ;-). Even if this "Aleste 520EX" was real, it is heavily implied by its description that CPC compatibility was secondary, and if so then the Sprinter should be mentioned as a potentially CPC compatible machine, too. In either case, there's little to keep it in this article. EpiVictor 15:13, 5 August 2006 (UTC)
It's best to assume good faith, but I would have to agree — it should go, so I'll remove it. --StuartBrady (Talk) 16:25, 5 August 2006 (UTC)
AMEN. Maybe You right there was reasons to remove info, but your doubts unreasonably. That computer was not imagination it was real. Look at it's site[3], there is emulator and you can play with it. So, you also wrong about other things: computer was was not clone MSX or ST, it was 100% clone of CPC. MSX DOS was used only cause it was one of real and popular OS at Russia witch support 720KB FAT12 disk format. Also MSX DOS is the successor of CPM. Sound board was real simple and had no any PLD (schematic, source code and example of sound on the site[4]). But sound quality was almost the same as AMIGA's sound. Today you can hear it. The site was madden recently you right, but computer and software was in production at 1993. --Via4any 15:37, 2 November 2007 (UTC)

An update, from 5 years further down the line, after a cursory google: Eh, it's just another one of those wonderful and wacky unofficial clone machines from Russia. Just, this time, instead of basing it on the Spectrum, they have for some reason gone the sort-of-MSX route (which the CPC also sort-of-was)... using the Amstrad as a base, then copying in their favourite (or maybe just easiest-to-copy) parts of other popular machines of the time. Probably not worthy of further mention in the main article, because, heck, I could go out and try to make something like that. But why would it be of interest to someone generally looking up CPC info on wikipedia? (talk) 17:27, 5 February 2013 (UTC)

Games quality[edit]

The games quality on the CPC could be quite variable but when it was good it was exceptional. Also, for certain styles of games the CPC was by far and away the best 8-bit mainstream platform, ie 3D isometric and vector orientated titles. The CPC even had a filled polygon game in the shape of Starstrike 2 and looked like an early ST or Amiga title. The disc version of Starglider also pushed the early 16-bit titles in frame rate and smoothness and was probably the most sophisticated wireframe game of the day.

The CPC versions of Knightlore, Alien 8, Head Over Heels & Batman ran at higher resolution and with more colour and no colour clash than compared to the Spectrum versions which are ususally used as the example platform for these particular titles. In particular the CPC version of Fairlight was significantly better graphically than the Spectrum version.

There was always a question mark on the scrolling ability of the CPC but when programmed intelligently it could be superb as shown in the port of Uridium which was arguably better than the C64 version. As previously mentioned the quality of the French software in perticular was very high with the likes of Get Dexter and Doomsday Blues. The audio side of these games as well also pushed the AY soundchip very hard into passable even for a C64 comparison. Though the SID chip in the C64 was a much better sound chip when coded for well the CPC sound chip could give a good account of itself and some excellent pieces of music were written for it.

Games that should not be forgotten as excellent (quality conversions from other formats or arcade conversions) and sometimes groundbreaking titles on the CPC included: Starglider, Rescue On Fractalus, The Eidolon, Yie Ar King Fu, Little Computer People, Starstrike 2, Mercenary, Elite, Kong Strikes Back, Bruce Lee, The Sentinel, Starquake, Dun Darach, Marpsort, Tir Na Nog, Sweevo's World, Knightlore, Alien 8, Head Over Heels, Batman, Bobsleigh, Combat Lynx, Project Future, Dan Dare, Uridium, Bounder, Trailblazer, Sorcery, Exolon, Frost Byte, Barry McGuigan's Boxing, Fairlight, Xeno, Revolution, Spindizzy, Tempest, Wizball, Cauldron, Cauldron 2, Barbarian, Get Dexter, Doomsday Blues, Starion, Highway Encounter, Alien Highway, Turbo Esprit, Deep Strike, Tau Ceti, The Academy. —The preceding unsigned comment was added by (talk) 11:54, 25 January 2007 (UTC).

Sure, none said the CPC didn't have good games (and the screenshots used are mostly from what would be considered "the good ones" for the CPC). On the other hand, as an ex-Amstrad user, I had my fair amount of Amsoft titles shoved into my face, and for each title I was proud (or at least not ashame) of showing to a C64 or Spectrum user, there were 3 or 4 that had no reason to exist, if not as a bad joke. EpiVictor 23:50, 25 January 2007 (UTC)
Of the games on the list, Sorcery in particular is worth looking at. The graphics will, of course, look extremely poor today, but for its time they were truly beautiful and lively.--Peter Knutsen 10:43, 1 November 2007 (UTC)

The graphics of Sorcery compares well to PC-games released in the early 90s that used VGA. —Preceding unsigned comment added by (talk) 19:00, 22 November 2007 (UTC)

Criticism section[edit]

This has now been removed twice. I believe it to be valid information - and hardly overly negative. Any comments? Tim (Xevious) 09:31, 24 April 2007 (UTC)

Removing it all is too drastic, but perhaps there should be some sourcing at least for the technical parts (e.g. scrolling, lack of attribute clash, perhaps a few comparative screenshots between similarly priced, same age games with extremely different coding quality). EpiVictor 10:13, 24 April 2007 (UTC)
I have no problem with sourced statements going back in. I'm very much opposed to a general philosophy that articles have to contain criticism sections, especially when said sections are populated by chatty commentary created on the spot by editors. That's not how we're meant to write articles. Chris Cunningham 10:49, 24 April 2007 (UTC)

Criticism of CPC software (games)[edit]

The quality of CPC games has been sometimes criticized due to the existence of releases that were simply a ZX Spectrum port, thus not measuring taking full advantage of the CPC capabilities.

The CPC shared the Z80a processor with the ZX Spectrum. Consequently many game manufacturers, seeking to cut costs, developed games for the two systems in parallel or ported older Spectrum games, yielding products that did not take advantage of hardware scrolling or the availability of 4 and 16 colour modes. Despite this, CPC versions would typically look better due to the lack of the attribute clash characteristic of the ZX Spectrum.

For the majority who targeted the CPC, challenges included the lack of hardware support for sprites, and difficulties in implementing smooth scrolling — particularly tricky at a rate of under 8 pixels per frame in the vertical direction. The complicated memory arrangement also made software sprite routines complex and comparatively slow-running, hindering the creation of smooth-running and colourful games.

Titles from the late 80s onwards tended to be coded more carefully than their mid 80s counterparts, making better use of the machine's graphics capabilities. When the CPC was programmed by an expert in the field, the smooth scrolling, colourful graphics and crisp music and sound effects could rival, and in several cases (such as Chase H.Q.) surpass, those of the C64. Nonetheless, the general perception of the CPC, several decades on, is one of a machine whose commercial success could have been greater.

That said, it is important to remark that numerous software companies from that era, such as Ocean Software, Codemasters, Elite Systems, Palace Software, Incentive, Hewson Consultants, Loriciels and Dinamic Software, among others, released quality titles on a regular basis which kept CPC users more than happy. It is also significant that the CPC had much greater support for serious, non-game software than the ZX Spectrum or (in Europe) the C64, not least due to its 80-column text mode. A large userbase persisted well into the 1990s, even from a smaller start than the Spectrum or C64, largely because of those people still using CPC word-processors such as Protext.

Comparison again[edit]

The comparison section is largely biased towards the CPC and seems to be written by people without deeper knowledge of the other platforms. I rewrote the comparison with the C64 somewhat, but the whole section is in need of a rehaul, really. 08:22, 13 June 2007 (UTC)

I'm not sure that it's fair to say (as the article does) that the CPC had a particularly primitive Basic. I'm not convinced that Locomotive Basic was primitive by the standards of the day (citations?). I can think of a couple of ways it was which inferior to GW-Basic but also a couple in which it was superior. It was not as sophisticated as BBC Basic, but BBC Basic was better than the norm (its case-sensitivity notwithstanding). -20:00, 14 September 2007 (UTC) —Preceding unsigned comment added by (talk)

I always considered Locomotive BASIC to be one of the CPC's strong points, certainly compared to the chronic C64 BASIC. I've certainly never heard it described as "primitive" before, and a statement like that definitly needs to be referenced. But I'm just going to remove that bit rather than waste my time looking for a reference I won't find. :) Miremare 23:40, 14 September 2007 (UTC)
Well, it turns out that the article isn't referring to the CPC's BASIC after all, it's talking about the C64 in that paragraph! I've clarified it though. Miremare 00:08, 15 September 2007 (UTC)

The C64 version of Microsoft BASIC is often criticized because it lacked direct support for the new video and audio hardware. There was no problem with the core of the interpreter, but rather, it lacked the bells and whistles of other systems. This was corrected later with the C128, but C64 owners were forced to purchase third party BASIC interpreters such as Simon's BASIC.

Also, as per the faults of the slow disk drive access, that is not due to an OS issue. Commodore once used the IEEE488 (parallel) bus for floppy drives. Unfortinately, a problem arose with getting the uncommon cables, so just before the release of the VIC20, Commodore came up with a custom serial version of the 488 bus that could use generic DIN cables. The VIC20's 6522 VIA would buffer up to 8 bits via a shift register without CPU assistance, and all would be well. Bzzt, wrong. The 6522 was broken, so the CPU now had to fetch on every bit instead of every byte. This problem was fixed with the 6526 CIA in the C64, but Commodore wanted the new 1541 floppy drive to remain compatibility with the VIC20. So, the C64 had to bit-bang the 6526 as well due to the 1541's xfer method. Problem is, the main data bus on the C64 was a very busy place since the VIC-II was a bus hog. So, the 488 bus had to be slowed down even more so that the 6510 wouldn't miss interrupts from the 6526 while the VIC-II was doing DMA on the bus. —Preceding unsigned comment added by Dinjiin (talkcontribs) 17:57, 8 November 2007 (UTC)

Recent removals[edit]

I am not sure that the removal of the sections on the relationship with its main competitors [5] and on the programming of the audio and video chips [6] has improved the article. While I agree that they need more sources, I can't see how the term "fancruft" applies here. Regards, High on a tree (talk) 08:46, 6 May 2008 (UTC)

I disagree. There is no need for any "my machine's better than yours" statements. The sections were unreferenced, non-neutral, and detailed programming tricks of little interest to non-fans - fancruft. I have re-organised the article and added some, much needed, references with the view of adding more in the near future. Nreive (talk) 10:46, 6 May 2008 (UTC)

Sound chip[edit]

Nitpicking I know, but I think the sound chip used in the later Amstrad CPCs made in Taiwan used a Yamaha chip of the family used in the Spectrum 128k and the Atari ST.

[ For any reading this who may wonder, the CPC had neither hardware scrolling of any type nor hardware sprites. It relied on pure 'cubes' to move the screen memory around. (talk) 06:59, 16 June 2009 (UTC) ] This assertion is false, and due to lack of knowledge of cpc software. To see a scrolling hardware on cpc, just type " list " in BASIC with a listing longer than 25 lines, because all the scrolling are managed with hardware scrolling by the basic itself. Certain games, such as genocide ( published in 1987) used vertical scrolls hardware with one pixel. (Logon System)(and why a discussion about scrolling or hardware sprites in the sound section ??)


Weren't the German versions of this machine named "Schneider"? Article should make some mention of this. How did the German machine differ from the British one? 2fort5r (talk) 21:34, 6 August 2009 (UTC)

I introduced a new section for the Schneider brand of CPC models, "Schneider CPC" already redirects to the article. TBH, I found it quite difficult to place the section because, compared to other home computer line articles (e.g., for the ZX Spectrum or the Amiga line), the Amstrad CPC article is rather poorly structured. I hope the new section is in the right place, but ultimately I feel that the article needs a structural overhaul. --Takimata (talk) 14:03, 20 September 2009 (UTC)


I would like to know more about the devlopment history of the CPC. Did it have any antecedants or did the design team start from scratch? Who were the design team - were they Amstrad people or part of some other company? 2fort5r (talk) 21:34, 6 August 2009 (UTC)[edit] is very slow and often inaccessible altogether. Is this situation permanent? I would like to link game articles to it but that's not practical in these circumstances. 2fort5r (talk) 13:47, 23 August 2009 (UTC)

According to this the site has been taken over by hackers and is downloading malware onto visiting browsers. 2fort5r (talk) 23:13, 28 August 2009 (UTC)

Weasel words, source requests[edit]

With the Nov 1 revisions, the weasel wording has been removed and source requests have been fulfilled. I still feel that the article could need some additional work (more detailed description of the model history, better structure for hardware and extension, perhaps a section on development), and I'll get to that as soon as I find some free time. But can we get at least rid of the ugly weasel word and source request tags now? What do you think? --Takimata (talk) 03:25, 9 November 2009 (UTC)

Seeing as the requests have been fulfilled, I removed the respective tags. --Takimata (talk) 23:53, 21 March 2010 (UTC)

Page rework ... GX4000?[edit]

I started reworking the page, starting with expanding and, in some places, correct the plus range section. I plan to significantly expand the model history (separate sections for CPC464, CPC664, CPC6128) and add a section on development. I'm not sure what to do with the GX4000 section, tho, seeing as there is a dedicated article on the machine I am inclined to leave it with a smaller paragraph here. —Preceding unsigned comment added by Takimata (talkcontribs) 23:58, 21 March 2010 (UTC)


I haven't got time at the moment, but it would be nice to add some information regarding the other peripherals available for the CPC. Off the top of my head:

  • Amstrad light pen
  • Amstrad TV tuner for the colour monitor
  • Romantic Robot Multiface II
  • Dk'Tronics 64k expansion
  • Dk'Tronics mouse expansion (came with Advanced OCP Art Studio)
  • Dk'Tronics silicon disc

Does anyone fancy having a go? Cpc464 (talk) 12:11, 13 April 2010 (UTC)

I've already compiled a pretty extensive list, but I'm not sure how to best go about it. Faced with WP:NOTDIRECTORY, I'm currently researching ways to integrate an overview of peripherals without having a lengthy list of extensions, peripherals, and accessories that is only marginally relevant to the topic. Any suggestions? How are other (top-rated) articles dealing with it? --Takimata (talk) 06:13, 14 April 2010 (UTC)

are you serious? you don't have an article for the 464 with the tape?[edit]

shiiiiiiiiieeeeeeeeeeeeeeeeeeeeeeet. [with the wire accent]. —Preceding unsigned comment added by (talk) 14:46, 19 May 2010 (UTC)

Other than "you" in the Wikipedia context means "we", and you are always free to make improvements: Please elaborate. This is the Amstrad CPC page, and the 464 is one of the CPC models. The article is lacking some detail for the original models yet, but what exactly are you missing? --Takimata (talk) 22:55, 23 May 2010 (UTC)

Inconsistent sales figures[edit]

I just removed the following paragraph from the CPC464 subsection:

It was launched in September 1984 in France and sold 2.5 million units[1].

Besides the launch date already being mentioned in the preceding paragraph, the given number is confusing. Given that Amstrad itself states total CPC sales at about 3 million with the CPC6128 being a major selling model, it is unlikely that the 2.5 million number only refers to CPC464 sales, even less so in the French market context. I believe it would be helpful to know the source's original text to put that number in perspective. --Takimata (talk) 19:14, 11 October 2010 (UTC)

No, I disagree, there is nothing necessarily contradictory about those two sets of figures, the 3 million given by Amstrad is clearly a rounded number, it could very easily have been rounded down from a much higher figure, shaving off, lets say, a possible 500,000 units odd from the total, also, we do not know if the 3 million is a sales total for only Amstrad made CPC's, it may not be counting systems made by Schneider for the German market.
Its pretty much common knowledge that the CPC was a huge success in France, much more popular than anywhere else, after that should be Britain, then Spain (CPC was very popular in Spain, but their computer market in the 80s was very small). So, given the possibility of rounding down, after the 2.5 million for France is removed from the equation, its possible we may still have over 1 million units left over for Britain and Spain, which is well within the realms of possibility. The figure also included a source Les Chroniques de Player One, p.38.
Though I do agree with you that some of the other info you've pointed out seems a little suspect, strange enough to remove the figure for the time being at least, maybe the editor wrote in the wrong section?, miss quoted the source? (total French sales written as CPC464 sales?) Jesus.arnold (talk) 19:32, 10 February 2011 (UTC)
What you regard as "common knowledge" is not backed up by the numbers. In 1987-88, when overseas (in this context meaning "outside Britain") sales were highest and all of the original CPC models have hit shelves, Britain still accounted for 42.5% of Amstrad's sales, France for 19.4%, Spain for 17.3%, and Italy, Germany, and Americas for roughly 5% each (David Thomas, "Alan Sugar: The Amstrad Story", p. 273). Even after factoring in that these numbers include PCW sales, there is no possible way the 2.5 million figure, an 80%+ share, can be attributed either to the CPC464 or the French market alone. The editor probably misinterpreted the source that was probably giving an equally rounded figure of total CPC sales, or he wasn't versed enough to put it into the right context of the article. Hard to say given the source is not open to the public. --Takimata (talk) 18:05, 13 February 2011 (UTC)
You were probably right to remove the figure for France, what I've seen points towards the fact that the CPC was most popular in France, in that it had majority market share there, but the market itself was so much smaller in France during the 80s that a 3rd place (or thereabouts) in Britain constituted higher sales than a 1st place in France.
This source backs up your source in that it indicates majority sales were Britain Amstrad Action 1989
This puts French sales at 650,000 by early 1990, and this put it at approximately 1 million in 1992.
So, I'd imagine French figures were more like between 1 million and 1.3 million odd for all models put together (the 1992 article does state that CPC sales were still strong at time of writing) Jesus.arnold (talk) 23:40, 16 February 2011 (UTC)
Yes, that sounds like a reasonable estimation. The David Thomas book I quoted has spectacularly detailled insight into the Amstrad company in general, and the CPC history in particular, at least up until early 1990 (it ends just before the CPC plus came out, unfortunately; It would have been really nice to get some more details on that range). But the sales figures it states are, with few exceptions, for the whole Amstrad line-up, which leaves numbers for the CPC range up to estimations.
Also, I support your theory about the CPC having a bigger relative share in France than in Britain. The CPC was at the forefront of an emerging French home computer market. And to my knowledge it was the first major computer system localized to and distributed in the French market, so it was right up the alley of the French who were (and, to some extent, still are) notoriously hostile towards foreign products. After all, distribution of Amstrad products in France made Marion Vannier.
Can you please name your other two image sources? There is no reason not to work well-sourced numbers into the article, even if we have to refrain from making assumptions and estimations (not sure about that 1.3 mil figure, but, eh, "no original research" anyway). The "original range" section could use a more generic overview, with the current content put into a "development" section, and that's proper content for that section. --Takimata (talk) 01:22, 17 February 2011 (UTC)
P.S.: Speaking of whom, there's a interview with Marion Vennier in a 1990s "Micro News" in which she states 800,000 sold CPCs in France by 1990, with "100,000 sold each year", most of them being colour 6128s. However, it is reasonable to assume that CPC sales dropped rapidly after 1990 with the original range's end of production. Even if the French were the largest market for the ill-fated plus range, it is very unlikely that sales reached '80s levels again, and went far beyond 1992 at all. (Which is consistant with today's rarity of plus models even amongst enthusiasts.) This puts the CPC figure for France on or slightly below the 1.0 million mark. Which is quite an achievement nonetheless. --Takimata (talk) 01:50, 17 February 2011 (UTC)
they're all from Amstrad Action, the first source is May 1989 pg15 the 2nd source is May 1990 pg16, the 3rd source is February 1992 pg 20
I'm not sure how reasonable it is to assume sales dropped off in 1990 in France, the French articles in Amstrad action indicate that in France the Amstrad is still strong in 1992, much stronger than it is in Britain at that stage (they say that the shops are still stocked with a wide selection of Amstrad titles, even though there's apparently no budget market in France to extend the systems lifespan like we had in Britain) Loriciels are quoted as saying that Amstrad software sales make up over 40% of their total sales in 1992 in France, with 20% Atari ST, 20% PC, and 10% Amiga, of course thats only one companies experience though, and doesn't take into account consoles (which I don't think Loriciels supported), but I think its enough to cast some doubt on the idea that they were definitely following a similar time frame to the UK Jesus.arnold (talk) 02:45, 2 March 2011 (UTC)
I certainly don't doubt that CPC software sales were still going strong by 1992, and possibly even for a little while longer than that, especially in France and with new titles still produced. However, I am arguing that CPC hardware sales pretty much had to drop in 1990 or very shortly thereafter, with the original CPCs no longer in production and the plus machines never having been anywhere as near as big a success. But we are not to assume anything anyway. And, if noone else beats me to it, I shall gladly put the numbers your sources provided into the article very soon, thank you for the references. (Now if I could only find some source for SRPs of the plus models other than the french ones ...) --Takimata (talk) 01:38, 7 March 2011 (UTC)

3" drive manufacturer actually Panasonic[edit]

I believe the 3" drives were in fact manufacturer by [Panasonic (evidence)] and not Hitachi after all.

Does anyone have evidence of Hitachi? (I contacted them in the 80's when my drive failed after reading Hitachi were the manufacturer in an 8-bit ZX Spectrum magazine and they claimed not to know anything about 3" drives)

DamienG (talk) 17:12, 13 January 2011 (UTC)

According to the floppy disk article, the 3" floppy was a joint venture between Maxell, Hitachi, and Matsushita, so there may have been various manufacturers. Miremare 18:27, 13 January 2011 (UTC)

My edit of June 27th, 2012[edit]

I feel I should not let this edit uncommented, especially to the editor, Manbemel. I get that you are proud of, I assume, the recent acquisition of an MP-1 adapter, I do. But I'm sorry, your "Amstrad with MP-1 plugged to TV set.jpg" is a really, really, really bad and haphazardly edited image that offers little to no additional information to the CPC464 image we already have on top of the article--especially when left uncommented like in your edit. Also, I removed the MP-x paragraph in "Extension" not out of malice but because it is redundant. There is already a description of this first-party accessory in "Video", and your paragraph just describes its function again--albeit in far more words. However, I took your picture, sized it down, and put it next to the respective paragraph in "Video". Takimata (talk) 13:08, 28 July 2012 (UTC)

Firstly, I own that module since 1985, it's not a recent acquisition at all, therefore I'm not and I do not need to be proud of anything at all. Secondly, the image serves its purpose, which is to show the module in action, something unavailable in Commons prior to the upload of that image. If you don't like the aesthetic composition of the image (you're entitled to do so, even though it is something secondary, a bad image is always better than the absence of an image at all), remember: it's a Commons image. You can download it, fix it yourself they way you see appropriate and upload it again. Feel free to do so if you think you can improve the image. Finally, those rude manners were completely unnecessary and quite offensive. I uploaded that picture to make it available in Commons without expecting anything in return, I was not expecting a reward or honor, but the least someone would expect was some respect for the work of doing it, regardless of their subjective opinion about it. Next time, please, try to use less bitter words when qualifying somebody's work, since you can unintentionally hurt someone's feelings.--Manbemel (talk) 16:19, 6 August 2012 (UTC)
I realize I came across as mean-spirited, which certainly was not my intention. I apologize. I hope you can still see the reasoning beneath my poor choice of words. --Takimata (talk) 22:44, 12 August 2012 (UTC)
    • ^ Les Chroniques de Player One, p.38 (Pika, 2010) (in French)