Saturday, December 16, 2017

“There are two kinds of people in the world, those who believe thereare two kinds of people in the world and those who don’t.”

Ok it’s an old joke but it does sum up those that read my blog, eg those that use my Focus Bar and those that don’t ;-)

But it also is a serious observation on how we think/work differently to each other. For example, some like thinking/learning through images, others through text.

So the latest version of FB (downloadable on the right) adds some additional information to help you decide when to stop focusing as you approach ‘infinity’, irrespective of what type of person you are. 

It provides infinity focus info in two ways. You still see the infinity (vertical and visual) focus breakdown, but now you also, when in ‘full info mode’, get numerical feedback via the ‘Focus Quality Factor’, in addition to the near depths of field data.

The FQF is simply the % increase of the defocus (sic) infinity blur relative the hyperfocal defocus blur (an FQF of 100%), which, as we know, is the minimum acceptable (diffraction corrected) blur that we set in the ML menu as the CoC; as focusing beyond the hyperfocal means that infinity blur gets less and, of course, becomes zero when we focus at infinity.

But we also know that we can not resolve line-pairs less than two sensor pixels. Thus the Focus Bar uses the 2xsensor limit blur to define the maximum focus quality.

These next two screen shots show the FQF in action:

As we can see in the top screen capture, we are focused at 1.27m and the infinity focus breakdown is shown on the right, ie the white total CoC criterion, as set in ML, the blue total blur at infinity, the yellow shows the diffraction component and the green bar shows the defocus component. 

Because we have the Full info selected in the FB menu, we also see the textual (data) on the left, ie the near DoF based on the CoC criterion is 55cm, the near DoF based on the infinity defocus blur is 64cm and that based on twice the infinity defocus blur is 43cm. We now also see the FQF, which is 137%, relative to the hyperfocal point.

The second screen capture shows that we have refocused towards infinity and are now focused at 1.94m. The visual feedback on the right tells us that we are focusing beyond the sensor limit, as the green bar has turned red, ie we shouldn't focus towards infinity anymore and could afford to back off. The focus bar has also turned cyan over the entire DoF, meaning that, once again, we are less than the sensor limit.

We also see, on the left, that the FQF has maxed out and is telling us that at the camera settings we currently have, we can't get a (defocus) FQF greater than 170%.

In the end it is all down to you to decide what information works best for you, however, there is no doubt about it, the Focus Bar provides the best info available, in any camera!

Wednesday, December 13, 2017

Extreme Focus Stacking

Having previously posted about the technical aspects of my Focus Bar script, I thought I would give a demonstration of its power.

My test subject was a small (50mm high) candle holder and my living room. The composition was that I wanted to place the candle holder as close to my (EOSM) camera as I could: about 15cm away.

I wanted everything from the candle holder to infinity to be tack sharp and that my infinity blur was to be optimized to the best possible, ie better than the hyperfocal and finish with an infinity defocus blur of just over the sensor limit of my EOSM.

I first used the ML auto ETTR  at ISO100 and at a focal length of 14mm, to give an exposure of 4 sec at my chosen aperture of F/7.1.

Having placed my focus at the macro end of the lens, ie at a reported 15cm, the Focus Bar script told me the minimum number of focus brackets would be eight, ie at an overlap blur of the defocus blur. However, as I wanted a high quality focus stack, and was aiming at an overlap blur of around 10 microns, I knew I would be taking more that eight brackets.

Once I had taken the first bracket and started focusing towards infinity, the Focus Bar provided me all information I needed when deciding the next focus distance. Once I saw around a 10 micron overlap, I took anther image, and repeated this until I entered the infinity focusing zone and finished my focus bracketing sequence at around an infinity blur of about 10 microns. In all I needed 12 focus brackets and all, thanks the Focus Bar, were optimised for their image to image overlap.

Next the 12 images were ingested into Lightroom and the first image minimally adjusted and the other 11 synced to this base image. Here are the first and twelfth images:

From within LR I then exported the 12 images to Helicon Focus and let it do its magic, before seamlessly returning to LR with the merged image. After a little bit of perspective tweaking, and minimal exposure tweaking, I ended up with this image:

I hope this simple demonstration has helped Canon Magic Lantern users, interested in focus stacking, take a look at my Focus Bar.

A clarification

Someone asked me to clarify how the various blurs compare as you do focus stacking with the Focus Bar.

This chart shows the relationship between two images taken at different focus distances, #1 and #2.

The first image (#1), as well as the second one (#2), shows the meaning of the cyan bar. That is the area of defocus where the defocus blur is less that the sensor limit of twice the sensor's pixel pitch.

The magenta bar that shows on the second image tells you where the depth of field overlaps with the first image's depth of field.

Finally, the overlap blur value, which the Focus Bar reports, is where the depths of field of the two images cross each other. It is this value that you should use to ensure high quality focus overlaps. For example, if this value was about the defocus limit, you would just have an overlap. If this value was zero, you would have a full overlap, ie one image on top of the other. Overlap values less than the sensor limit (twice the pixel pitch) are rather meaningless.

Bottom line: you have all the information you need to make perfect focus stacks.

Sunday, December 10, 2017

Take your post precessing to the next level

As readers of my blog are aware, I'm a great believer in the power of Photoshop, and although stand-alone PS-CC is great, there are some really impressive 'extensions and add ons' out there.

Some of the most impressive are those that help you with your luminosity masks.

Over the years I've used many different luminosity masking extensions, eg TK-Actions and Lumenzia, as well as RapidMask. However, I have always leaned towards Aaron Dowling's ADP Panel:

So I was really excited when Aaron just announced the latest version of his ADP Panel, Version 3.

So, 5 minutes after I received his email announcing the latest version, I had it on my PC.

I haven't the time now to discuss ADP Panel V3, but I will give a hint of the power by sharing and image, from a single capture, that V3 helped me craft.

Saturday, December 9, 2017

Further refinements

A quick post to announce a few updates associated with my Focus Bar and Toggler scripts.

The Focus Bar now displays the sensor-limited focus area on the focus bar. Thus the cyan area is the focus field where the defocus blur is less than or equal to twice the sensor pixel size.

The near and far depths of field are the distances where the defocus blur satisfies the SQRT[MLCoC^2 - Diffraction^2].

The following shows what the FB looks like with the new sensor-limited info.

We see ML is reporting we are focused at 81cm and at this distance (at 12mm and F/6.4) the near and far DoFs are 56cm and 1.52cm respectively. This DoF information is shown twice in the above, because I haven't taken an image yet. Once an image is taken, and focus stacking is on, the DoF info at the bottom will show the DoFs of the last image taken, whilst the focus bar will show the live DoFs.

The (14), in the middle of the two DoFs at the bottom, shows the defocus blur criterion, in microns, based on the ML set Circle of Confusion you have set, ie adjusted for diffraction, assuming you have set it on (which you should have).

The [2] to the right of the near DoF at the top shows that you will need to take at least 2 focus brackets from this point to the hyperfocal. More if you wish to have good focus overlap.

To illustrate the power of this simple feedback, ie image to image focus overlap and the sensor limited zone, let's look at the next two screen captures, where we have moved focus from the first image (at 81cm) towards the macro end.

Note the bottom DoF now shows the DoFs of the last image taken, ie  56cm to 1.52cm. This information will remain as long as you don't take a new image.

The top images shows that although we have focus overlap (the magenta zone) this overlap is above the sensor limited zone. The new information that is provided after the near DoF (12) tells us that the blur at the overlap is 12 microns. Healthy, but not the highest quality you can achieve, which is the sensor limit, ie twice the pixel size.

In the lower screen capture we see that the overlap blur is reported as 9, ie (9) to the right of the near DoF on the focus bar.

Also we see that the number of brackets to the hyperfocal is 3, as we have moved away from the first image towards the macro end.

The Focus Bar script now saves the state it is in at camera shut down. So once you have configured the Focus Bar to way you like it, the camera will remember this.

As for the Toggler, I've created an alternative data arrangement in the script, which I hope makes it easier for others to tweak. Also, I've enhanced the UI so that Toggler shows you the state of the element you are about to toggler, eg on (green) or off (red).

As usual I welcome feedback.


Monday, December 4, 2017

Latest Magic Lantern Lua Fix update

One of the 'issues' with scripting is the risk in getting out of sync with the Magic Lantern updates.

I've therefore revised the Focus Bar and Toggler scripts to be in sync with the latest (as of today) Lua Fix Experimental branch, which is the Build  dated 2017-12-03 22:09 at

I've also withdrawn the links on the right to my other scripts, but will add these back as I bring them in sync with ML.

The Toggler script has been written to work with the Focus Bar. That is you can use Toggler to switch the Focus Bar on and off, as well as switching the FB focus stacking feature on and off.

Friday, December 1, 2017

Thank you Sony ;-)

As readers of my blog will be aware, I have ‘retired’ my IR converted 50D, sold my Sony A6000 and have settled on three-cameras. My trusty 5D3, an EOSM and an IR-converted EOSM.

The EOSMs, running Magic Lantern, my Toggler script and my Focus Bar script, are killer travel cameras: they are small and light. The 11-22mm EOSM lens suits my preference for WA shots.

Although I have augmented the EOSMs with hand and thumbs grips (see here), there is no doubt the small footprint of the EOSMs can, at times, be challenging.

So I’ve been looking for an alternative EOSM arrangement and decided to use some ‘DSLR video’ gear: in particular a cage, which has multiple female screw threads for attaching ‘stuff’. Looking around it was clear that there was not a dedicated cage for an EOSM, but there was something close: one for a Sony A6300: A6300 Cage

The standard A6300 cage, however, can’t be used as it interferes with the battery door and the lens release button, and, if you are really picky, the EOSM rubber grip.

Luckily, converting the A6300 cage for EOSM use is relatively simple, eg hacksaw and file. Plus if you know someone skilled with their hands, thank you John, you can make a really good looking conversion. These two images show the before and after conversion:

Once the cage is adapted to the EOSM, the world opens up for you. So I decided to add a handle on the left and I purchased (under 100 GBP) a HDMI monitor, great for those low angle shots. The monitor is from Viltrox: a model DC-50.

Here are a few shots of my complete set up.

Bottom line: for those, like me, that rely on and love Magic Lantern, the EOSM is a great travel camera. Add a converted cage, and an HDMI monitor, and you have a killer set-up for photography and videography, which, thanks to ML, can be RAW!