I had a dream, a dream of a Gigapan mounted on a moveable device with a time delay option. Unfortunately one was not available on the market, so after defining my main criteria I decided to build a system myself. My main criteria were as follows.

• Ability to move from along a defined but not necessarily restricted path
• Try not to limit to a linear track, and allow for corners/bends/hills if possible
• Ability to pan the camera on the top of the device between 2 angles
• Ability to define the number of shots to take along the route
• Ability to define a pause time between each shot
• Ability to support 5kg of camera hardware
• Ability to trigger the camera automatically (I have Nikon, but try not to limit it)
• Ability to be modular in design so the device can be extended in the future
• Lightweight and transportable – fitting in the smart car with my other equipment
• Be supported by standard photographic tripods
• Ability to run off internal battery power for at least 4 hours
• Be made from as many off-the-shelf components as possible

Phew, that’s a lot of requirements… now you can probably see why I had to build my own.

the payload

The important bit of the equation was my camera and lens. The device has to be strong enough to support about 5 kilos of camera gear without dropping it so I decided I needed a wide base to give general stability and to allow for support of the rotation disk – now I see why people shy away from building this bit. I also need the buggy to not slip off whatever rail system I ended up with so the attachment needed to be strong and sturdy. I looked at the teflon coated plastic bearing sliders, but as soon as you add a large camera and lens they are a whole load less slidey.

I eventually looked at a roller-coaster style wheel setup as this attached to the rails both top and bottom, meaning the buggy was not going to fall off easily, but also allowed for spacing sideways as well. If I could refine this it would also go round corners. If I could make the clearances adjustable then the track was less of a problem.

the rails

The railing system actually gave me the second biggest headache as this was the bit that needed to be the most modular. The best option I came up with was a ladder style track that broke down into metre long sections. That way you can pack the tracks into the car, but also you can bolt as many as you need together to give flexibility. Making it out of Valueframe slotted T frame was a really flexible way of prototyping the rails.

the drive system

With the buggy designed and the track built, how would they connect to each other. Originally I wanted a gear system as this would give the best power to the track as possible, but that would significantly reduce cornering ability, and the connection to the rail looked quite complex and require a very precise line-up between track sections.

I looked at belt drive. this seemed the best compromise with traction and corners, however the length of track would define the length of belt needed or visa-versa. I could order a long belt and have it coiled on the floor, but then I’d probably still need multiple lengths of belt, and this would not be optimal.

Friction drive seemed the easiest to start with and then I can upgrade to belt drive later. A radio controlled car wheel with air filled tyre was ordered to attach to my motor. This worked quite well to start with, but as I started with a slight incline and a heavy load the grip started to slip and the drive became a lot less accurate.

After a bout of testing, it seemed that corners and bends in the track would add so much complexity that it would be difficult to work properly. So I implemented belt drive which could give me cornering later on.

the motors

I initially looked at servos for both the drive and rotational components of the system. These proved to be quite expensive for a heavy duty option and the control was rather complicated. Both were swapped with stepper motors as these give 360 degree rotation control as well as the precision I needed for the linear drive system. The initial motors have about 2 kilos of holding torque, but as the system approached the 10kg mark, the buggy had to be slowed down too much to get decent pull. I’m in the process of upgrading the drive motor. I could also work out a gear mechanism that would reduce the motor and increase the torque quite significantly, but again, speed would suffer.

the controller

I wanted to use a Raspberry Pi (I kid you not) however these devices are very early in their development lifecycle and the ability to write software was a little complicated, not to mention nearly impossible to get at the time of writing. So, Arduino it was. I needed to control 2 stepper motors, have 4 push buttons for the user input, a missile styled power switch (yep, an absolute necessity but not quite a requirement) and an LCD screen for showing the user what’s going on. Optionally I wanted an SD card to store configuration data and the settings between power cycles as well as Bluetooth connectivity. This all fits into the Arduino Uno so is a lovely small package. I may extend this to fit on the Arduino Mega so that I can get more I/O and more importantly more RAM for a better menu system.

what can it do?

Glad you asked, check out the stuff at the bottom of the page to see what I have done with it during testing. I’ll add some more stuff as more testing goes on.

what can’t it do?

I didn’t really need it to go vertical so I didn’t build that in. I am in the process of upgrading the drive motor as it is very slow when horizontal and doesn’t do inclines, but to get true vertical with the weight of the rig you’d need an even more powerful stepper motor, gear drives and move to Lithium Polymer batteries at a much higher cost.

The menu system is pretty limited at the moment, but this is not a major disadvantage as pretty much everything you need is right there. If I move to the Arduino Mega controller board this could be fixed. This may be an improvement I make in version 2. This will also save some of the button combinations that are needed to get some of the features, but actually I like these as they are pretty quick to set the option you need.

Camera support is pretty basic. I have a Nikon camera with the 10-pin adapter and Nikon do a cable (the MC-22) that you can easily wire in to the board. This cable is compatible with the Nikon D200, D300, D700, D3, D3s, D3x etc. Interestingly it is also used by the Fuij S3 and S5 Pro. Any camera that allows the simple shorting of a wire to trigger the focus and shutter will work with the hardware, it’s just a case of modifying a release cable. This means you can easily get a cable for a Nikon D80, D90 and the majority of Canon DSLR’s. Neat-o. However, since I don’t own any of these cameras, I have not tested any of them and do not plan to do so yet.

thanks and stuff

The following people helped me harm things:

• My buddy Russell was a great help with his urge to change everything (I think he would like to call it encouragement).
• My buddy Malcolm was a great help with his metallurgy and electronics skills.
• The guys at Makerbeam are awesome and have a great product.
• valuebeam is a bigger version of Makerbeam, and makes great rails.
• The guys at proto-pic and I are all on a first name basis.
• PCB Train have the greatest express PCB prototyping service in the universe.
• Amil of Aoo 3D really helped out with his CAD drawing expertise.
• Pat from London Waterjet provides an excellent and friendly service for waterjet cutting.
• Oh, and my mate Andy who said some things sometime about some stuff and just wanted to be on this list.

where can I get one?

If you are interested in getting one then I can build one for you. Bear in mind though that this is still in pre-production which means the edges may be sharp and the bolts will stick out a bit. It will also take a few days for me to order the parts and build it all. contact me for more details. At least if you live on the UK mainland shipping will be free, and if you live locally to me I’ll even hand deliver it and walk you through it all

what do I get?

The Base unit consists of the buggy, 2 lengths of the 1 metre track system with end stops, mounting bolts to attach to your tripod supports, an allen key for the rail attachments and a charger for the internal battery.

You will need to provide your own support tripods for the rails and any tripod head you want to put on the buggy, a phillips screwdriver to manage the belt attachment. At the moment, you will also need to work on your own cable, or I can provide one for you at additional cost, however, as I have mentioned I have not tested anything but the Nikon MC-22. If you want to temporarily donate a recent digital SLR I can look at producing a cable for you, and updating the hardware to support it.

what options are there?

Currently there are a few options that can be incorporated, or are being planned.

• Lithium Polymer battery – doubles the runtime, half the charge time and saves nearly 1.5 kilos.
• Updated controller – better menu system, better controls, more advanced control models and options.
• Colour buttons – for some reason the nice colour buttons seem to cost a lot.
• Carbon fibre power switch – the red cover is cool enough, but carbon fibre is carbon fibre.
• Mount pack – tripods and heads to sort out the mounting.
• Additional lengths of track – lengths can be added by the metre.
• Bluetooth upgrade – this option will allow you to monitor and control the rig from your android phone.
• Cable release system for more cameras, the link here explains all the pin-outs needed.
• E3 connection for the Canon EOS 350D, 400D, 450D, 500D, 550D and the 60D.
• Pentax cameras also use the same E3 style connector.
• N3 connection for the Canon EOS 10D, 20D, 30D, 40D, 50D, 5D, 5Dmk2, 7D and all types of the 1D.
• MC-DC1 connection for the Nikon D70, D70s and D80.
• MC-DC2 connection for the Nikon D90, D3100, D5000 and D7000.
• Olympus have 2 connector types that could be used.
• Sony have their own exotic connector, but it can be used.
• Panasonic/Lumix use a couple of resistors to control things, but it is not so complicated.

how much will it cost?

I am still working on the price, but if you can help in any way I can discount stuff for you. It’s probably best if you contact me for more info. Please bear in mind though that this is not an entry level system.

demo stuff

This was a test of the camera firing mechanism. The images were resized and exported in Lightroom, then affects added in Photoshop via a batch job, and then joined together using ffmpeg. Total processing time was a couple of hours.

Watch this video on YouTube.

I have been asked a couple of times about what it is that makes a good photograph. The simple and honest answer is that it’s both. You can also pass off, or convert a ‘bad’ photograph with some artistic license as well so really, there is no such thing as a bad photograph

There is also another aspect that is overlooked, and that is the element of luck. Being in the right place at the right time is actually the most important part of any photograph. Assuming you are able to control the environment enough to sort this out then what next?

There are some artists out there that are able to produce stunning images on their iPhones. They probably would not be able to win “wildlife photographer of the year” with any of those images, so does that make them a bad photographer? No!

I think that having good equipment takes out some of the pain of having to think about certain things and allows you to concentrate more on composition and timing, but I also think that having been through all that pain with equipment that needed thinking about makes you look at some things differently and possibly even understand some things a little better.

For example I use a sigma 150-500mm lens on occasion. As you move from 150mm to 500mm the f-stop changes from 3.5 to 5.6 meaning the resulting image gets darker. Unless you compensate for this, no matter how good the timing and composition, you are going to have to repair anything you take. This lens also has a bit of a softness at 500mm, so again, compensating to always back off a bit gets me better photos.

Knowing what looks nice and using composition effectively is a key tool you can use across everything, but understanding what your equipment can do for you, and more importantly what it cannot do is much more valuable. The learning curve as you come into photography is quite steep, but I think it does provide you with good instincts and is well worth going though. If you really want to learn how things are working then practice – I have about 600 shots of my watch on a table as I experimented with depth of field and another couple of thousand of a coin as I experimented with macros. No one will ever see them, but I learn’t a lot creating them

There are a lot of people that would crop up an existing image to make it longer down one side and call it a panorama, and they would not be wrong. To be fair there is no real great mystery about panoramas, but sometimes cameras just cannot capture the sheer scale of a place and then, a panorama would be nice. For example the shot above actually covers 180 degrees left to right – not many lens’ allow for that without severe distortion. Also, generated panoramas can scale to… well way bigger than real size in some cases, more on that later. So, the basic steps of how to create a stitched panoramas are:

• Take a series of shots with some overlap
• Stitch the images together
• Blend the joins and post process the final result

The photos

As with nearly all projects, the key is in the preparation. With photography even more so since you will never be able to recreate this particular set of location, weather, view etc. The first question is which way up do you hold the camera. This is not as dumb a question as it seems. If you are doing a left to right panorama then the first thought would be to hold your camera and do landscape shots – you’ll need far less photos. If you use you camera in portrait you’ll need more but you will end up with a bigger image to trim down.

Camera settings. This is the second biggest headache for the panorama stitching later on. Ideally you will want to use full manual mode – including manual focussing. If you only have a point and shoot, you’re very limited in this, but set up as best you can. The reason you have an issue is that in any automatic mode, your camera will meter for each shot. This means that two shots next to each other, containing a lot of the same data (the overlap) could have different exposures or focussing. It’s something to think about.

The next question is how many shots. Well, there is no real answer to this, basically you will start with one end and keep going till you get to the other, leaving a fair bit of overlap in each image. The nearer anything is the more overlap you will want.

To tripod or not to tripod. Over the last few years I have honed my ability to take handheld panoramas that are straight and level but this is tricky. A tripod will help with this, but can introduce it’s own issues. In order to get a good quality stitch later on, you need to rotate the camera about it’s middle. well, technically around the nodal point of the lens, but in lay-speak, you need to move around the camera, not rotate on the spot with the camera held out in front of you.

There are a couple of tripods and heads that help a lot with this, for example the nodal ninja and the manfrotto 303SPH VR head, but I’ll leave that discussion for the advanced master-class

And so to the overlap question. Basically you need overlap in your images so that the stitching software you use later can detect key points in the photos and compare them for matching and stitching. There is a problem however. Things that are closer to you move faster when you rotate your head then the things that are far away. This is called “Parralax” and computer games in the 1980′s were really cool if they had this, but I digress. Parallax caused the single biggest headache for stitching software since it causes massive changes in the bits of the images that are closer to you. A fully calibrated VR tripod will help a lot, but you should still use more overlap if things are closer.

The first of the 2 pictures above, has a lot of detail in the distance, so 6 shots were used to capture the 100 degrees or so. The second image has a lot of change in the sides of my garden, so I needed more shots there. I could have thinned out for the middle bit, but basically I just kept going, producing 28 images for 170 degrees.

As a general rule of thumb a good two thirds of the last image should be in the one you’re taking now. So, looking though your view finder, take a shot, then look for something a third of the way across the shot in the direction your moving and rotate your camera to put that at the edge of the next shot.

The software

My personal favourite peice of stitching software is autopano giga. It wins in nearly all categories of ease of use, functionality, control, quality and visual interaction. It fails on price as it’s not exactly cheap, and, at the time of writing its does have a bug if you’re stitching more than 50 images on a 64 bit operating system, but in the most part it is awesome. One feature puts Autopano Giga at the top of my software choice is it’s ability to handle 360 degree panoramas and click on the bit that will be the centre of the unrolled image. While other software does do this, nothing is as easy to use. The image on the left is a good example of this.

A rather good bit of free Microsoft software – yes, the words “good”, “free” and “Microsoft” were used in one sentence – is a product called Microsoft Image Composite Editor (or Microsoft ICE for short). The reason I rate this is in terms of ease of use for the cost – it’s pretty awesome. It’s also quite fast and can produce some nice output.

When you use images that have severe changes between them, for example pedestrians and moving cars, different software will make different choices about how to deal with this, most will exclude things that do not appear in at least 2 of the images for example some will ignore blurry stuff. The image on the right shows that the pedestrian in autopano giga was replaced by a large truck in Microsoft ICE.

Also worthy of a mention in the software department is the Hugin suite. If you have a PhD in astrophysics then this is a cracking bit of free software. There is a lot of help on the internet so it may be worth having a look at.

Photoshop is not high on my list of recommendations. I don’t like the really squat images it seems to produce. Newer versions do have a really neat feature called “Content Aware Fill” that is pretty good. If you use it, you will categorically have to do some blending around the edges but it can be worth it if you have a chunk missing that would require a severe cropping to get round. See the images on the left as an example – it’s also animated

Photoshop does not seem to like 360 degree panoramas that much either, and you cannot tweak any of the viewer settings. I am not sure why, maybe it knows better than I do, but I like to tweak and refine and Photoshop does not allow me to do this. Your mileage may vary.

Autopano Giga has a really cool rendering view that you can watch as it generated your panorama. The novelty does wear off quickly but the image on the left highlights a few interesting points about the process.

See the dark bands caused by my lens being darker towards the outer edges. Photos are also reshaped to correct the distortion in your lens. Some software knows about specific lenses and knows how to compensate, most generally guess pretty well. The blending stage of nearly all panorama software takes care of all of these issues to some degree or other, but it’s quite nice watching a few times. It’s kind of like slowly unwrapping presents at Christmas

The post processing

Due to the fact I try and use manual settings to take the base images, the resulting image generally needs the levels tweaking if it came from Autopano Giga. Even if I process the base images before exporting them to stitch, they can usually benefit from some standard workflow based processing, for example sharpening. Oh, and dust removal from a dirty lens

Using Microsoft ICE also produces pretty good images however it does not handle the banding caused by the lens vignetting you saw earlier in the rendering, so large areas of clear colour (plain blue sky for example) can have noticeable issues that need fixing.

Where to begin with photoshop. I think the understanding with photoshop is that if you have paid the excessively high cost of owning a copy, then you probably know how to use it pretty well, therefore the stitcher does not need to do the best job of touching up the finished article.

You can see from the image on the left that the additional trees from the stitching look a little funny. There is actually only one tree in the real scene, but I guess it must have been pretty lonely out there at the edge of the shot.

You can also see the line between the photo and the fill-in from the content aware fill, and this will need blending out.

The final word

So, what next? If you remember all the way back to the beginning of the article, you my remember I mentioned about panoramas can become huge. At the time of writing the largest panorama is an 80 gigapixel panorama of London in the UK.

There is also a US based company that provide motorised mounts that you set your camera up on ith the biggest lens you can muster, and then leave it for a couple of hours to do its thing. Check out Gigapan.

I just had an interesting dilemma – which cheapish compact camera to buy so I can go “light” from time to time. After a lot of research and comparing stuff I narrowed it down to the following 3 options

• Panasonic Lumix TZ10 (approximately £200)
• Canon Powershot S95 (approximately £300)
• Canon Powershot G12 (approximately £350)

On the face of it the Lumix had the functionality down to a tee, but didn’t shoot in RAW and some of the reviews hinted that the ease of use left something to be desired. The S95 was the compactest of the lot and shot in RAW but had some extreme lighting issues. The G12 was basically just a bit of a brick, but had really cool old-style dials to twiddle. All the camera’s had full auto modes and full override modes for aperture and shutter priority, as well as scene and program settings.

So I walked into my local camera shop and started fiddling. I found the Lumix first as this was top of my list. I was gutted – the ease of use issue was a serious issue for me. Stepping down from an SLR that talks in aperture and shutter settings I really struggled as I was constantly overthinking stuff. in full auto mode it took nice looking images, but stepping up from a regular compact would probably involve a read of the manual. My next play was with the S95 and I loved the simple way the dial around the lens let you change the setting you currently had set as a priority. It was very easy to use and very compact. I almost bought it on the spot since I didn’t like the thought of a brick in my pocket.

So I picked up the G12 knowing that my heart was set on the S95 and a strange thing happened. I looked at all the dials, and they all made sense. I twiddled the dials and things worked as I expected. I loved the dials. The dials had me. I bought the brick.

In summary

All the cameras are capable of taking great photos if you don’t tax them at the extremes and all of them suffer from flash issues that are prevalent in nearly all compact cameras. In terms of general use and picture quality, all three of the cameras are pretty much on par with each other, so choices came down to the aesthetics and useability.

I loved the lumix on paper. If your’re a digital SLR measurebator you will love it as well, although the fastest the lens will go is not all that great. The Lumix is a great medium level compact but there are cheaper standard compact that don’t have all the SLR control bits that are tricky to get to anyway.

The G12 is a great camera and is so easy to use if you want to start the process of stepping into the SLR world. It is on the price cusp of “Bridge cameras” and there are probably better ones of those out there, but if you’re used to lugging a Nikon D3s and all the gizmos, the brickyness of this camera is not really a reason not to get it.

I loved the S95 because it shoots in RAW has a nice interface and is so compact it’s quite frightening. The reviews on image quality would not impact you since you’re probably not going to be trying to print out 30×20″ posters from any of these cameras. In think that out of the three cameras I originally narrowed it down to the Canon S95 is actually the best all round option if your stepping up from a compact or stepping down from an SLR.

The 3D Revolution is upon us. Since the 50′s with the terrible blue and red glasses the wish to see pictures of things in 3D has been almost as cult like as aliens. In fact some of the images that were produced in this old style are pretty alien, but I suppose that’s the point of a cult, right?

Holograms and 3D TV’s are becoming commonplace now and there have been various alternate methods for photography put forward over the years, but the best is still less than a year old, and is still very expensive and premature. The Fujifilm W3 is a nice gimick, and has been known to produce dome nice images, but it’s a little ways off being a run-away success, mind you, I still don’t like stills cameras that have video in them so what do I know about what is cool.

I stumbled across a website called Start3D a while back that could take 2 images of the same thing taken a few inches apart, and produce a 3D print that you can buy on-line. Why is this cool? well it means you can use your existing camera and get kind of the same results on your computer monitor, and you’ll see why “kind of” below. This is still limiting in that you can’t do action shots for example, and large cameras can’t be placed together at the right distance, but this should be a simple case of mechanics so it won’t be long before someone makes a D3 mount I can buy so I don’t have to make one myself For example, a start for smaller cameras is the twin mount bars here.

So, what’s so cool about Start3D, well here, take a look.

Because of the movement, your eyes magically make the 3D work. To be fair there is probably a lot of complicated maths involved in edge detection frame blending in the animation, and I can’t find a photoshop plugin anywhere. So what options do you have yourself? Well, animated GIFs have been around a long time, but the technology is mildly limiting either in size or speed, but it seems to be all there is. There seems to be a following for the twitching between 2 images, which is quite harsh but can produce ok results.

You can also have a simple go in PhotoShop instead of downloading stereo maker above. Ok, probably an exercise in PhotoShop things rather than anything practical.

So, first, the images. You have to take 2 images focussing on the same point but having moved to the side slightly. For things that are close, don’t move very far and things that are further away, move more. As a rule, you can stand with both feet shoulder width apart and put weight more on one leg… take the photo, then put your weight on the other leg, focus on the same point you did previously and take your second shot. It’s probably best to use manual mode on your camera so that you can produce 2 similar images exposure wise.

Here are my 2 images. I focused on the end of my daughters nose so that this will be the bit that needs aligning later and will form the focus of the shot, by not moving in the animation.

Into PhotoShop. Load the images in. Then copy and paste one into the other as a new layer. On this new layer, set the opacity to 50% so you can see some blur.

Next, using the move tool align the bits you focused on. You’ll see that things that are closer and futher away are blurred and this is 3D in action.

Set the opacity of the top layer back to 100%. Now you need to make an animation, so ensure the “Animation” pane is enabled on the Window menu. Create a new frame on the window at the bottom.

Now, hide the top layer in this frame.

On the delay drop down under each image, select 0.1sec

That’s it, now save for web and devices from the file menu (there may be a size based complaint, but that’s ok), ensure GIF is selected and away you go.

I’d also recommend resizing the image to about 400 pixels or so.

Here is the final animated GIF, and another one I tried. In Hindsight, I probably moved a little bit too far when taking the images, but that’s the fun bit with experimenting.

When you mention macro photography to most people they seem to go off the idea because of the cost of the special lenses you need. Nowadays, this is actually untrue, and while purists may ague that what you’re using isn’t a true macro lens because X, Y or Z, I’m a believer in being judged on your results. All macro photography is really, is the ability to get a really close-up shot of something. There is very little difference now between a prime lens and a macro lens other than the fact the minimum focal distance on a macro lens is less. For example, my 105mm macro lens has a minimum focal distance of 20cm (meaning the subject must be at least 20cm in front of the lens) where as my 70-300mm zoom has a minimum focal distance of 150cm.

So if you already have a kit lens you like then you can still use that with some tweaking of your setup. There are four main ways of getting macro results from a standard lens, or improving an already excellent macro lens:

• Lens reversing – no trickery its what it says. This works better with short wide angle lenses;
• Dioptre filters – a filter size attachment like a UV filter that does a bit of magnification;
• Extension tubes – fits between your camera body and lens reducing the minimum focal distance;
• Teleconverters – fits between your camera body and lens increasing the focal length (100mm becomes 300mm for example).

Each option comes with its own pros and cons and generic uses, so lets have a quick review.

Lens reversing

By far the cheapest option, the clue is in the name. Basically you take your lens off your camera, turn it around and hold it against the body of the camera. You can also buy reversing rings which will fit onto your camera body and attach to your lens’ filter holder. This process works best with wide angle lenses because of the physics of the lens. A wide angle lens (working the right way round) takes a very large area and makes it small enough to fit on your sensor, so in reverse it takes a very small area and makes it very large.

There are some severe downsides to lens reversing, most notable is how tricky the process is to get working. You have no control over aperture so it will default to whatever you lens does by default, you must focus manually whilst holding the lens, using a tripod is completely impractical, and probably the worst one is all the metal sticky-out-bits that connect into your lens now are in close proximity to the glass front of your lens – get a UV protector. The main downside that I can see is the complete lack of control over anything.

The three photos below show how I tested this. The first shot is my 18-105mm zoom at 18mm as close as I could get (~50cm) with the meter showing as in focus. The second shot is my 105mm macro lens in as close as I can get (~20cm), and the third shot is the 18-105mm zoom reversed at 18mm with the aperture selector opened with one finger, as close as I could get (~5cm), with the lens set to focus as close as possible.

Dioptre or ‘close-up’ filters

First, I’ll admit it, I don’t have any of these so any opinion is purely subjective. As I can see it, these seem to be an easy all-round option. They can be quite cheap and are easy to take on and off because they fit on the front of your lens like a polarising filter. However due to fitting on the front of your lens, you’ll either need a set for each lens or choose a favourite lens. They work like eye glasses for long sighted people (reading glasses) and change the focus point of what you’re looking at – not actually magnifying what you’re seeing, although there is a little bit due to physics.

Like lens reversing and extension tubes dioptre filters allow you to get closer to your subject than the lens would allow you naturally. There is not a great deal of interference with the quality or quantity of light so seem to be a nice option, but then you can’t really get in all that close with the big dioptres without the edges of your shot going really soft. If I had the biggest available on my 105mm, I could probably get it to about 12cm, and I’d end up with only the middle two thirds of a shot I could use.

Extension tubes

I think extension tubes are the most variable on price, ranging from £15 to about £200. A decent set at the higher end of the cost bracket will have the CPU/lens contacts passed on so you can still adjust the aperture and focus via the camera. Extension tubes work by allowing you to get closer to the subject and adding to the focal length of your lens. They are full of air so have no impact on the quality or quantity of light and generally come as a set of three that you can use in combination to get between 12mm and 68mm of extension. So, by adding the 68mm extension to my 105mm lens I end up with about a 170mm shot (at about 8cm from the subject)

The photos below show my 105mm straight and with the 68mm extension tubes, both as close in as I can get.

Teleconverters

Teleconverters work in a completely different way to all of the above. All of the options so far have been about getting physically closer to your subject but teleconverters tackle getting closer by upping the power of your existing optics. They come in a range of powers, and generally manual focus is required for regular lenses. I have a kenkoo 3x converter so the effective focal distance of my lens becomes 315mm. The big upside to teleconverters in that they are not just for macro use, and can boost any lens. I originally bought mine so I can use it with my 150-500mm lens for shots of the moon (an effective 1,500mm telescope like shot).

The big downside to teleconverters is they eat light. My 3x converter needs 3 stops of light correction, auto focus is not possible and a tripod is really helpful if not essential depending on how much light you have left because the tiniest vibration is magnified meaning a fast shutter is essential.

In summary

There is no one magic bullet for marco photography, but there are several pewter ones. Out of all of the options I think the teleconverter is the most versatile as you can use it with non-macro photography so this could be a sound investment, but watch the light compensation you will need. I think Extension tubes are a good thing as well as they don’t interfere with the light, and this is always a good thing. My recommendation though, it to always by the best glass you can afford, and that is not close enough you can always crop your photos.

Stacking

As you’ll see from all of the shots above where the aperture is wide open, the depth of field suffers considerably. There is a way of taking many “slices” by just focussing on a separate bit of the object and combining them into one shot, and this process is called stacking. It only really works where you have complete control of the subject – i.e. a mounted insect, or a 20p on a table top. The following 4 shots:

can be stacked to make the following image:

Examples in action