It gets off to a bad start by misspelling Alan Miles' name as 'Allan'. 'Alain' might have been a more forgivable typo, since that's his nom de plume as a writer of fiction. However it happened, though, misspelling the name of one of the key players in the SAM's creation in the article's second paragraph sets the tone for its low editorial standards.
Additionally, we have:
- "It loaded Spectrum software from tapes twice as fast as the Spectrum itself": This is physically impossible, as it would require running the tape at a higher rate. It's a misunderstanding of one of the SAM's lesser-used features - it was possible to adjust the rate of data transfer to tape when saving, thereby making it quicker to load back in... But only the SAM's own data, not data intended for the Spectrum, and certainly not data accessed via one of the Spectrum emulators.
From the User's Guide...
Really simple BASIC instruction, though it's unlikely to have been used much because very few SAM owners would have chosen tape storage, particularly for larger files.DEVICE T use tape running at standard speed to SAVE data
DEVICE T35 use tape running at fast speed to SAVE data
35 is the top speed likely to prove feasible for saving data on to tape, 112 is a standard, relatively slow speed. Unfortunately, fast speeds are less reliable than slow ones. There is no need to remember at what speed you saved your data, because the LOAD command automatically copes with the full range of speeds.
The author has cited a piece written by Mel Croucher, for Crash issue 71 (December 1989, see also on Archive.org), which he has simply taken at face value:
Croucher also wrote the SAM Coupé Users Guide, so there's clearly some bias at play, but this article is also riddled with factual errors and misunderstandings, so it's a terrible source for quoting.DAY SEVEN
OK, I admit it. After five days I’m impressed. The SAM Coupe is the best 8-bit machine I have ever experienced. It seems to save and load off cassette twice as fast as the Spectrum, and gameplay is fast.
- "The first release was sold without a floppy drive, contributing to the machine's out-of-date image": This is entirely untrue. In fact, the floppy drive was an optional add-in at first, so one could buy a driveless SAM if preferred - tape software was available, after all - but a one-drive setup was considered optimal at launch, and twin-drive SAMs were available from the get-go. One reason for this was that MGT were behind the DISCiPLE and Plus D floppy drive upgrades for the ZX Spectrum, and SAMDOS was a development of the DOS used by those. As a result, users of these drives could opt to plug in their existing floppy drive via an external interface, thus saving some money.
Later on, the driveless SAM option was dropped, and was not revisited by SAM Computers Ltd. after MGT went into liquidation during the 1990 recession. However, there was a tape-only 'Gamestar' package being pushed by Format Publications, later on, alongside their "SAM élite", in which one of the built-in floppy drive ports was bodged into a Parallel port.
- "By the time the ROM was updated, the lack of third-party support had doomed the new machine.": The ROM was updated quickly and promptly despatched - free of charge - to early adopters of the SAM. It had zero effect on third party support, which had only ever been theoretical. In fact, lack of third party support was effectively the least of the SAM's problems. Additionally, the ROM bug was little more than an inconvenience, with a simple command input after booting allowing the floppy drive to be used (mostly) as normal. It didn't affect the running of any software.
The computer press at the time had long been reporting conjecture and blue sky thinking as confirmation of support, as most software publishers of the day had a "wait and see" attitude, none wanting to be the first to commit to supporting the new machine (wise, if disappointing, considering what had happened to the likes of Imagine by overstretching their resources).
On a more positive note, the lack of third party support only gave smaller publishers (eg. Enigma Variations, FRED, Format) and PD libraries more room to thrive in the SAM's early days, though SAMCo's own software publishing arm, Revelation, no doubt contributed to their downfall through spreading their resources too thinly. MGT hadn't had the budget to properly market the SAM, SAMCo certainly didn't have the budget for that and software promotion. Third party support was doomed by SAMCo's demise, not the other way round... Which leads us neatly to...
- "The few software publishers who had expressed an interest in it cancelled all SAM Coupé projects. US Gold had considered bringing Strider to the Coupé using code from the Spectrum version of the game and the graphics from the Atari ST, a technique Miles and Gordon vainly hoped would catch on. When MGT when bust, the project was abandoned.": This is the most commonly cited bit of SAM apocrypha. As mentioned above, this was only ever conjecture - a representative of US Gold had been interviewed by the Speccy press and theorised that Strider could be converted this way in about two weeks. There were sadly never any plans to do so, and a project not undertaken cannot be 'abandoned'. I don't recall any interviews with Alan Miles or Bruce Gordon where they passed comment on this kind of conjecture, but I'd imagine they would have liked the sound of it... but preferred a little action over all the talk.
Games like Prince of Persia, Lemmings, KLAX, Escape from the Planet of the Robot Monsters, etc. were basically put together by bedroom coders within the SAM scene and presented to publishers as faits accomplis, dramatically reducing the risk of publishing - all any of them had to do was print up some 'SAM Coupé Version" stickers to slap on their existing stock of boxes. Most didn't even bother amending the manuals, preferring to put in a photocopied sheet of SAM-specific addenda.
- "The rubber feet on which the SAM Coupé stood were often lampooned.": I don't think the author of the article knows what 'lampooned' means. The blue 'feet' were certainly a talking point in the press, but they were hardly ridiculed or criticised, they were simply part of its design. Not sure where the author was going with this point. The case may occasionally have ridiculed, due to the fact that the side view looked a bit like car, and its built-in wrist rest wasn't exactly comfortable, but the built-in drive slots were generally praised for their design efficiency.
- "Text Modes: 32x192, 512x192; Graphics Modes: 256x192, 512x192": This is just confusing and woefully incomplete. The SAM has four screen modes: MODE 1 is Spectrum emulation mode (256x192px broken up into 32x24 2-colour attribute blocks of 8x8 px each) with support for BRIGHT, FLASH, etc. and with 16 colours (technically 8+8, due to the BRIGHT bit). MODE 2 is similar to one of the MSX screen modes, same resolution, but broken up into 32x192 2-colour attribute blocks of 8x1 px, and with 16 colours available. MODE 3 is the 'productivity' mode, with a screen resolution of 512x192 offering 64 column text without affecting the legibility, but with only 4 colours. MODE 4 was the SAM's unrestricted 16-colour mode, still running at 256x192px, but with no attribute blocks to worry about.
The bit where David Ledbury is quoted as saying that 12,000 machines were sold is likely to have been cut-and-pasted from one of his Facebook comments in the SAM users group, as David wasn't consulted on the article (he's actually rather upset that he wasn't asked before being quoted - that's just bad form)... and it's pretty insulting that he's described simply as an "enthusiast" considering he worked at SAMCo and was, for a while, an independent publisher of SAM software himself.
While the conclusion of the article is harsh, I can't call it untrue or inaccurate... However, there's no mention of the SAM's continued existence or the current state of the scene - Quazar would certainly warrant some mention if nothing else. There's no mention of the SAM's six-channel stereo sound, built-in MIDI connectivity, networking support (intended to make the machine appealing in an education environment), or even the earliest hardware add-ons.
The SAM scene may not be thriving in quite the same way as the modern Spectrum scene... but it's certainly not dead and buried, as the article suggests. I think it's generally accepted that the SAM may have been more successful had it been made even just one year earlier but, with 16-bit machines already available, the market was moving on. At the end of the day, though, the pitch of the article is all wrong: The SAM Coupé was never intended to 'save the ZX Spectrum'. Prolong user interest in the 8-bit market, perhaps, but it was an upgrade from the ZX Spectrum that (theoretically) wouldn't require a whole new library of software thanks to its Spectrum compatibility.
But the other big misconception is about the SAM's oft-mentioned 'Spectrum compatibility'. Yes, its screen MODE 1 is designed to emulate the Spectrum's display (and I gather invoking MODE 1 also slows down the processor..?), but you can't just plug a Spectrum cassette into the SAM and expect to be able to LOAD "". The fact that the SAM requires the use of Spectrum emulator software means it's exactly as 'Spectrum compatible' as any computer running a Spectrum emulator.
Aside from all that, here's a misconception of my own:
Something that came up in @Alessandro's excellent Spectrumpedia Volume 1 was an interesting nugget about the SAM's palette, for which there are several conflicting stories. The book states that the SAM palette:
This was taken from WorldOfSAM.org's page about hardware bugs:should have included 256 colours, but insufficient space on the ASIC chip cause it to be halved
but, there's another page about the SAM palette with a slight variation/expansion on that story:The limit of 128 colour palette (64 colours with 64 BRIGHT) as opposed to the desired 256 was allegedly due to not having enough spare physical pins on the ASIC’s die to accommodate it. As it stands there are two pins for each colour (Red, Green and Blue) and a BRIGHT output meaning that there are three bits for each colour, top two bits as per the colour pins output and the least significant bit set by the BRIGHT setting.
I'd always believed it was a bug in the ASIC, rather than a physical hardware issue, that led to the SAM having 128 colours as opposed to 256 - that the BRIGHT bit was supposed to be applicable to each channel independently, rather than all three at once. I had a conversation about this with David Ledbury some time back, and his recollection was that it was originally only going to have 64 colours, but a last-minute change of supplier for the ASIC offered the opportunity to add the BRIGHT bit, doubling the size of the palette.Originally the machine was envisaged to have 256 colours but limited to 64 colours on the development machine due to a shortage of pins on the ASIC that was to be used, but had the extra intensity (BRIGHT) added on Simon Goodwin's suggestion in it's pre-ASIC design stage to give 128.
According to an interview with Bo Jangeborg intervention whilst writing Flash! also upped it to 128:
"I convinced them to do away with the hardware flash in the high colour modes. That allowed for a significantly larger pallet. I think it went from 64 to 128 colours. That also allowed me to transfer some pictures from the Atari without too big a degradation."
Interesting how there are different stories about this, and a couple of different people claiming responsibility for boosting the palette from 64 to 128 colours.
Standby alert 
