Two years after the launch of the groundbreaking Dynax/Maxxum 7000i, Minolta launched a totally redesigned and upgraded range of single digit “xi” cameras. The Dynax/Maxxum 7xi was the “prosumer” or “enthusiast” model in the lineup and came second only to the professional grade 9xi. Minolta had seriously shaken up the market with their auto focus SLR’s and, for the time, had packed a serious amount of computing power in to each subsequent model.
Consequently, Minolta had a real head start in the market, had gained a reputation for well built, capable and easy to use auto focus cameras and they were keen to extend this lead in any way possible. These “xi” cameras were meant to do just that. You might expect that Minolta just built on what they had, tweaked a few things here and there, increased the auto focus speed perhaps and just left it at that. After all, if it isn’t broken… But, no, Minolta definitely had other ideas.
Minolta were nothing if not bold. Their initial Dynax/Maxxum cameras had the big breakthrough of an auto focus system that worked without attaching hilariously bulky lenses to the camera, but had been seen as slightly difficult to use with some users finding there were too many buttons to learn the function of. They firmly and resolutely fixed this with their second generation “i” labelled cameras. These were a triumph of user interface design, the ergonomics were dialled in to “almost perfect” levels and the icing on the cake was the introduction of a range of expansion cards.
Cameras like the 7000i sold in droves and you’d be forgiven for thinking that after finding the magic formula they’d do a Canon and pretty much stick to the design language they’d established and refine from there. You’d also be wrong. With the “xi” designs, Minolta once again completely ripped up the design playbook and tried an entirely different approach to ergonomics whilst almost doubling the computing power inside. What do you do when you’ve got more processing power to play with? Well, of course, like it’s 2025 you introduce some Artificial Intelligence.
It was time to live the future like it’s 1991 all over again.
- AI before AI
- Fuzzy what now?
- Auto everything
- Those plastics and ergonomics…
- Taking it out for a spin
- Conclusions and Learning
AI before AI
To understand the 7xi we need to indulge in a miniature computing lesson.
The main component in any computer system is the processor or CPU. This is a chip which has one job – take program instructions and carry them out. They work remarkably quickly, the CPU in your average phone will carry out somewhere between two and four billion instructions every single second. We live in an age today where computing power is so readily available there are pregnancy tests powerful enough to play Doom.
Equally impressive is the fact that these processors are made from billions of tiny switches called transistors. Just like a light switch, a transistor can be on or off and that really is it – all the flashy things your phone, tablet, laptop or PlayStation can do is all the result of turning switches on and off.
Simplifying things a little, each CPU has a native number of “bits” and each bit may be on or off (represented as a one or zero). The more bits a CPU can use internally for an instruction or piece of data, the better. This is because they can handle larger numbers, perform more precise calculations and use a larger amount of memory. The first generation of home computers and games consoles had 8 bit CPU’s. The largest 8 bit number is 255 and therefore to use larger numbers, you have to combine many of these 8 bit chunks together. I’m sure you can see how this quickly becomes complicated to do useful things with.
In 1989, SEGA released their infamous MegaDrive (Genesis if you’re in America) console, which went to great pains to boast about the fact it was a 16 bit machine! What did this mean? It meant more colours, better graphics, faster gameplay and was the driving force behind the fastest hedgehog the world had ever seen. Nintendo followed later with their 16 bit offering, the Super Nintendo Entertainment System.
The MegaDrive CPU worked at just under 8 Megahertz. What does that mean? It means it could carry out 8 million instructions per second. Not bad at all, this alone made it 8 times faster than the original Nintendo Entertainment System which gave the world Super Mario. The SNES from Nintendo ran at almost half the speed of the Megadrive.
Why is any of this relevant to a camera?
Canon were amongst the first companies to start putting computers inside cameras in the late 1970’s. Their A series of cameras, including the now legendary AE-1 Program, were powered by minimalistic 8 bit processors. Although these were cut down versions of their home computer equivalents, they offered enough instructions and consumed little enough power that they could rapidly run the calculations required to work out the correct exposure for a scene based on a few basic inputs. These processors didn’t need to be hugely powerful and were coupled to just enough working memory (a kilobyte or two) to run the small, custom programs that had been written to control the exposure, shutter mechanism and flash.
The development of computing hardware has always moved at a lightning fast pace and it needed to if cameras were going to take on more complex tasks like auto focus. The problem with CPU power is that the more of it you have, the more power it consumes and generally the hotter it gets. This isn’t great for a battery powered device that needs to last 40 rolls of film per battery. The good news is, as each generation of CPU comes out, they become more powerful and less power hungry, which is utterly remarkable when you think about it. This is the equivalent of a car manufacturer releasing a car that is twice as fast as the previous model but uses half the fuel. Every time they release a new model.
By the time Minolta developed the 7000 and 7000i, there had been sufficient advances to enable more complex programs to be run inside a tiny camera at even greater speeds. These cameras, however, still made use of 8 bit CPU’s – but they were clearly enough. No one complained that the 7000i was particularly slow.
The 7xi, though, was an absolute power house. Inside its squat body lay not an 8 bit CPU, but a 16 bit CPU which ran at an utterly mind blowing 20 Megahertz. That’s 20 million operations per second in a battery powered camera. Yes, it was less fully featured than the processors powering the aforementioned MegaDrive and SNES, but think about that for a second, this was a camera with a faster CPU than a games console. Can you imagine today if your Mirrorless camera had the CPU power of a Playstation 5? That’s madness.
Fuzzy what now?
Why did the 7xi need all of this processing power? The answer lies in the sheer weight of automation that Minolta tried to cram into this camera. This camera has one overriding principle and that is to know what you want before you do. If the camera is in the unlocked position, as soon as you place your hand on the grip it comes to life. The 7xi has two metal strips running down the inside of the grip which uses your fingers to complete a circuit that tells the camera you’re holding it. Upon doing so, the “Zoom XI” lens will pop to the focal length it thinks you most likely need and the camera will start to look for things to focus on.
Upon bringing your eye to the camera, two Infra-red sensors are used to tell the 7xi that you’re looking through the viewfinder and it begins to focus on what it thinks is the most sensible thing to focus on based on the scene in front of it. Finally, it keeps evaluating the scene for the correct exposure setting and if something starts to move, the auto focus is triggered to start tracking whatever it is that you might want to keep in focus.
This whole system relies on something called “fuzzy logic” which was the latest and greatest buzz phrase of the early 1990’s. When computers were first invented in the 1940’s, visionaries like John Von Neumann (the person who literally invented computing as you know it today, along with many other things) predicted that computers would eventually become “intelligent” and be able to carry out tasks with a greater efficiency and accuracy than humans. Does this sound familiar? It should do, because today the marketing types are vomiting AI into anything they can think of. Your next toilet? Generative AI which will predict the optimum time to flush itself.
The problem is, teaching something to be “intelligent” when all it understands is yes and no is quite the task. Think about what you’d like for your next meal. Why do you want that? The answer is likely that you just “fancy it” or like that type of food. But why? Why do you fancy it? “just because” perhaps? This kind of answer is complete anathema to a computer, it cannot understand grey areas, only complete and utter certainty.
Fuzzy logic was one of the first attempts to solve this binary “true or false” approach to seeing the world. It allowed for computers to have areas of grey and “maybe” suddenly became an option. Essentially, fuzzy logic is nothing more complex than a series of “if this, then that” type statements that deal with certain inputs. Take the weather today, in a binary world you’d describe it as hot or cold. There’s no in between, yet we know in the real world it matters quite a lot what the temperature is.
Fuzzy logic would take the temperature and split it in to ranges, anything below 3 degrees is “cold” you should probably wrap up warm. Between 3 and 8 you’d get a different outcome, 9-12 degrees and things are warming up, perhaps you don’t need that thick coat. You can see how this works.
If you put this in context, it’s taken us nearly 35 years to get from there to a point where a system will understand a sentence in plain English and spit out what is essentially a neatly formatted Google search for you. Most people understand today that whilst useful and a potential time saver, AI outside of some niche applications is still nothing more than an idea generating machine that struggles with anything beyond the relatively superficial.
Auto everything
It will come as no surprise then that the fuzzy logic equipped Minolta isn’t quite up to its own hype. At first the whole “ready before you are” thing is quite a novelty. It does work, you look through the viewfinder and the camera is already zoomed and focussed on something. 99% of the time it has done quite a sensible job of guessing what to pay attention to in a scene.
There are lots of things that sound great when they work as they should, but become quite annoying when they fail. Minolta went all in on the metering and AF being automated and whilst you can turn off AF, select an autofocus area manually, adjust the metering and so forth it is neither obvious nor straight forward how to do this at first. The camera has two control dials, one for scrolling through program shifts, shutter speeds or aperture depending on the mode and the other for changing the mode the camera is in. But it’s not that simple, if you scroll the mode wheel it puts you in a special “programmed aperture priority” or “programmed shutter priority” which is quite odd, because there’s very little difference between these and the standard aperture and shutter priority modes which are also available.
To switch to these you need to press a function button and then scroll the wheel. To take control of the AF point you need to press this twice and then scroll the wheel around. Arguably, this isn’t the end of the world, and it isn’t to an extent but all of these Minolta’s are desperate to return to their default programmed, fully automated setting. One press of the P button and you’re back to factory defaults. I liked this in the 7000i but the AF works slightly differently in that camera and far less often do you need to select an AF point manually. I’d rather it didn’t then revert back to another setting every time it was turned back off.
The main issue is that the AF area is massive, approximately 1/3 of the frame. In reality, it isn’t one giant area, just that all four points are active at once and that’s the area of the frame that they cover. Stranger still is that there’s a missing AF point – they’re arranged in a plus shape, but the bottom point of the plus is missing. Why? Perhaps no one in the 90’s ever focussed on the floor? I’ve no idea.
Ultimately, the system works until it comes across a scene which confuses it, at which point it hunts all over the shop. Hunting is annoying enough with any auto focus system, but when it’s continual it drives you mad. There’s no way, other than flicking over to manual focus, to stop the system focussing. It is focussing all the time whether you like it or not and when it gets confused, oh my goodness, it goes focus crazy.
This scene drove the system bonkers. I was trying to find an angle that worked with the shiny spanner contrasting against the dark metallic paint of the machine, I thought it’d look quite nice in black and white with the stark difference between the metals. Whilst framing up it just kept shooting between the background and foreground. At times it would focus on the spanner (which is what I wanted) but other times it would select other things in the frame such as the cylinders either side. This is the problem with full auto focus. All of those things are legitimate candidates to focus on in this scenario. Not knowing which is best, you fall foul of the fuzzy logic algorithm. Each part of the scene is given a different “weighting” of importance and when you’ve got two or three objects which are really prominent then any of them can receive pretty much the same score. Which one to go for? Depending on slight changes in light and positioning, one will just win out against the other and trigger a refocus and round we go again.
One thing I must say is that I love the dynamic viewfinder display. Depending on what you’re doing, bits of it turn on and off to communicate what’s going on, but only the bits that you absolutely need to see. It was totally novel at the time of release and it works really well. Today, you’d not think twice about such a feature, but it must’ve been sheer magic to new users of the 7xi in 1991.
There’s one final really cool feature of this camera and that’s a special “wide” viewing area. I had to laugh in our last outing with the Sony Ericsson T610i camera phone which came complete with a dedicated camera button – this is something Apple have just sold millions more iPhones with as an “innovative feature.” It turns out they’ve borrowed another of their “innovations” from the 7xi. On your phone, when shooting slightly zoomed in, it will show you the area outside your frame so you can see things as they enter the frame or what it might look like if you zoomed out. The Minolta will auto zoom the lens to a specific point and then dynamically adjust the viewfinder with a frame guide so you can see what will be included in your picture but also things entering and exiting the frame. Upon pressing the shutter button it automatically zooms to the correct level and takes the picture. Not bad for 25 years BC (Before Cameraphone… Sorry, I’ll get my coat.)
Those plastics and ergonomics…
There’s no getting away from the fact that the 7xi is a tank of a camera. The most striking part of the design is the sleek, low profile prism/flash assembly. Personally, I love the look of it and it’s very reminiscent of the natural, flowing lines design language introduced by Luigi Colani in the T90. Side by side with the 7000i, the change in design becomes very apparent.
Initially, the more angular 7000i looks almost primitive when compared side by side, but look closer and there are some changes that I don’t think were for the better. From the top down you get a real insight into exactly what Minolta were trying to achieve with their redesign.
The 7000i is lighter by some margin and has a very well designed hand grip. On top of this is an LCD display angled towards you to make it easier to read – this was a feature that I had a lot of admiration for in my review of the camera. The camera feels secure in the hand, it’s not too heavy and it is very intuitive to use. The 7xi is a total redesign and the grip is wider due to an extra hump at the back but much, much shallower at the front. This doesn’t make it difficult to hold but it is less natural and certainly less secure in the hand when carrying.
Most annoying is their rearrangement of the shutter button. With the new grip, your index finger does not naturally rest where the shutter button has been placed, instead it nestles on top of the control wheel. You can see they’ve gone to some lengths to correct this by cutting out a finger shaped recess but this seems to me a bit of a bodge. The solution is to put the wheel behind the shutter button and I can’t believe Minolta’s engineers didn’t consider this when they were designing the camera – there must be a reason for their choice but I have no idea what it was.
To put the sheer size of the Minolta in context, the camera on the left in the image above is the professional, top of the line Canon EOS 1n. That’s a camera made for sports photographers and journalists, designed to be bomb proof before all else. The two are, to all intents and purposes, the same size, but note that the grip is a far superior design by all measures and the shutter button is perfectly and ergonomically positioned – your finger naturally wants to be in that exact position.
The 7xi certainly feels solidly built and I didn’t suffer any of the mechanical failures that I had when trying to find a working 7000i. It is reassuringly sturdy and makes all the right satisfying noises when you’re using it. However, it does suffer from the dreaded grip rot that seems to plague all Minoltas from this era. Whatever they did to their rubber/plastic compound, it hasn’t stood the test of time. Initially, I thought this wasn’t the case as my copy came with a perfectly intact grip with none of the usual white discolouring they get when they’ve gone brittle. Sadly, though, on its first outing I picked it up and the bottom promptly crumbled to pieces.
Taking it out for a spin
On one hand, the 7xi is a serious bit of kit. On the other, it’s no more difficult to use than any point and shoot, down to the fact it even automatically zooms the lens in and out for you. How does this translate in to real world use? The answer is quite well, most of the time.
When the camera works, it’s a wonderful machine. You pick it up, it’s ready to go, you reframe an image and it automatically refocusses for you. There’s a lot to be said for the camera doing all of the heavy lifting for you and, at first, it can be odd not to have to half press the shutter each time you want to refocus but it doesn’t take much getting used to.
For me, things start to fall apart when you want to do anything outside of the protective automation bubble. Exposure compensation? Hmm, is it this button or that wheel? Turn the wrong one and you’re suddenly in an odd program shift mode. How do you get back? Press the P button and… oh, you’ve reset all your settings such as having the pop up flash disabled. Start again.
Spot metering? No idea. Selecting the focus point yourself. Err… Can’t remember.
Right now you’re probably raging “just read the bloody manual!” Well, I did, and very well written it was too. I learned a lot about the camera from the manual, which I then promptly left at home and by the time I got round to using the camera, I knew it could do the things I wanted it to, but I couldn’t remember how to do those things.
It’s rather odd, really, because for a camera that is designed to be as intuitive as possible (and it is when you’re comfortably in program mode) it falls flat on its face as soon as you need to do something different. None of the buttons are labelled except for “func” and that’s ok, but it has multiple modes depending on whether you hold it, press it or… double press it.
Here’s the top of an EOS 300:
Firstly, the mode dial is labelled, you can go straight to the mode you want really easily. Self timer? There’s a button with an icon. AF point selection? That’d be the top of the three buttons on the right by the LCD, I know this not because I know the camera off by heart, but because there’s a little picture there that makes it obvious. Drive mode? Press func and roll the dial. How? Well… the button only does one thing and you get immediate feedback on the LCD by the icons at the same. Spin the dial and it takes you through the functions.
Ah! But the 7xi shows you things when you press the function button! Well, yes, it does… inside the viewfinder. That’s great when you’ve memorised the controls, not so great when you press it away from your face and nothing appears to be happening.
I know I’m labouring the point here. If you’d splashed out hundreds of pounds/dollars for one when it was new you’d have read the manual inside out, taken it out with you in your camera bag and memorised the various features to within an inch of their life. I get it. It’s totally different these days when you can buy 15 SLR’s for £25 and run a roll of film through each. The point is more that intuitive design should run throughout a device, not just through a single feature or its intended main mode of operation.
I find these choices or omissions odd, just like the decision to ruin the grip for no apparent reason, but 99% of the time these things don’t really matter. The strangest thing is I really enjoyed using the camera, it’s a really positive experience overall, however some cameras grow on you over time and you get on with them more and more, I found the opposite with the 7xi, I loved it for half a roll and then it started to annoy me when I used it in completely different conditions and the AF started to struggle.
Ultimately, the thing that really matters is the pictures and on that front, the 7xi is an absolute belter.
The metering on this camera is nothing short of remarkable. Considering I forgot how to use exposure compensation, it shot an entire roll without a single under or over exposed frame. Regardless of whether I shot into the light, had huge amounts of backlighting, it didn’t matter, the subjects came out perfectly exposed. Take the picture above as an example, this would’ve been really easy for the meter to under expose based on the background, but not here, the 7xi recognised the subject, metered accordingly and exposed perfectly. Credit where it’s due, this camera pulls shots out of the bag every time.
I shot ISO 100 film at box speed and the camera did its best to keep the shutter speed to sensible levels wherever it could. The kit 28-105 automatic zoom lens is very good for a standard zoom and gives clear, detailed and sharp images with no serious fall off or vignetting at the edges in real world use. I’m normally allergic to cheap kit zooms, but this was worth using even if at 105 it felt a little short at times and at 28 not quite wide enough. I’d sacrifice a little bit of zoom range for a wider aperture any day of the week.
Coupled with the excellent Minolta 50mm F1.7, it produced the kind of crisp, sharp images you’d expect from a decent prime lens. Auto focus with either lens was good, but you lose the predictive/auto zoom when you use the older Minolta AF lenses.
Talking of autofocus, I don’t quite think their predictive, automatic tracking feature was quite up to the outlandish claims made about it in adverts and previews that I’ve seen. Minolta claimed that it could track objects going at up to 180mph, which is fine, but it blurred a shot of my son because he had begun to turn away from the camera. In some situations it couldn’t decide what to focus on so just had a bit of a guess and ended up between objects.
Conclusions and Learning
We have so much to thank Minolta for. Technology thrives when there’s decent competition between manufacturers and the 1980’s was an explosion of new ideas, with all brands racing to include as much new technology as they could. Canon and Nikon had some serious work to do when Minolta turned up to the party with their groundbreaking auto focus technology. They weren’t just a one trick pony either, the 7000, 7000i and arguably the 7xi all introduced novel ergonomics and features.
But somewhere along the line, Minolta fell off the wagon and allowed Canon and Nikon to not just catch up, but absolutely trounce them in the market to the point where they were bought out by Sony (and still live on as the Alpha range today). I can’t help but feel like this is the point in their timeline where that began to happen.
Does anyone need a camera that detects when you pick it up? Not really. Does it change the amount of images you successfully shoot? No. Does it result in times when you have to correct the camera and recompose or zoom in and out? Yes.
Canon, much more than Nikon, were throwing lots of crazy ideas at cameras in the early 1990’s. Eye controlled auto focus, bar code readers on the EOS 10. None of it was really necessary, but more often than not, their innovations stuck. Was it because they were more advanced than Minolta? I don’t think they were at all, but the way in which they did it made more sense.
Canon hit upon a design language that broke the mould with the T90. They carried this into the EOS 600 range and later perfected it with the EOS 1. They recognised this and even in the latest, greatest digital mirrorless cameras, this DNA lives on. It’s evolved, of course, but the evolution has been subtle and (mostly) beneficial.
Minolta, in my opinion, struck gold second time around with the 7000i. If they’d kept pretty much the same layout, replaced the rocker switch on the grip for a control dial and just built on a familiar platform it would perhaps have made much more sense to customers. Did a redesign kill Minolta? Of course not, but it was part of a bigger picture. They went on to make some insanely advanced cameras before finally giving up in the very early 2000’s. Minolta knew how to make a decent machine, but once Canon and Nikon pretty much sewed up the professional market and pumped millions in to producing huge ranges of lenses and providing as much support as their big, important customers could desire, it was inevitable that something was going to give.
The 7xi is a great image making tool with some of the best metering I’ve come across in a film camera. It’s just that it tries too hard to hold your hand when, at the level it was pitched, what you actually want is easy access to all of the features a camera offers. Automate, innovate, add features, but don’t forget to make everything optional and everything easy to access and this, sadly, is where Minolta failed. No amount of fuzzy logic is going to fix that.
Should you try out the Dynax/Maxxum range? Absolutely. They’re really good film cameras at very cheap prices, just try to find one that hasn’t crumbled and if you’re really lucky, try to get a Dynax 7 or 9, then you’re going to have a lot of fun. Perhaps the 7xi should be lower down on your list than those, or even a 7000i too.
Share this post:
