Panorama view from the Bergerie

Now that I’ve been back home for a few days, I did spend quite some time going over all the various collections of photos. I took around 6 sets in total, changing things slightly each time to then have a play around and determine what works best.

 

Setting it up

Having brought my tripod for the purpose of taking the panos, I set it all up, ensuring it was level (there’s a handy spirit level built into the Gitzo). Then, put the Kirk ball head on top, then fixed the camera to the head with the Really Right Stuff L-plate I bought a month or two ago. These items aren’t cheap by any means, but oh boy do they make setting up and adjustments an absolute breeze! I certainly have no regrets at all in forking out for these as it’s certainly paying off. In fact, I can’t remember last time I never cursed whilst trying to set something up!

Choosing the lens

I opted for using the 105mm for a couple of reasons. Firstly, there’s no zoom so it reduces the chance of my clumsy fingers changing the focal length half way through a shooting process. Also, despite being designed for Macro, I reckon it’s a very sharp lens when focused at infinity. Also, being of a fixed focal length, the hood is designed to reduce flare and vignetting as accurately as possible (or at least that’s what I thought). I didn’t do any tests beforehand to check it out. That focal length also gave me the most amount of photos and therefore detail in the final stitched pano without going absolutely insane. I’m not competing for those gigapixel titles, at least not in my first attempt!

With the 105mm lens fitted, I had a quick pan around taking light measurements. There was about three stops worth of variation between the brightest point in the sky and the darkest point in the grass I wanted to capture. I then put the camera in full manual mode, ensuring Auto ISO was turned off. That way, the camera didn’t try to compensate for exposure ensuring I had a consistent exposure from each photo minimising the risk of seeing the “joins” in the stitching which can be visible to dark to light transitions (or vice versa).

 

Go!

Then, away I started snapping! The 105mm field of view allowed me to have around 20 shots covering the left to right edges of the pano I wanted to make. There’s an overlap of about a quarter of the frame each time. Having more helps in the stitching as it allows the software to recognise more overlapping points and make fine adjustments during the warp and stitching process.

As the camera was in portrait mode, the longer side of the frame was covering the vertical axis and three rows gave me enough range from the sky to the grass in the foreground. I would have liked more sky, but you do need to have the mountains visible at the bottom of each frame to have the recognisable stitching points! I suppose I could have added some “fake sky” above, but I’m not that good with Photoshop or other editors and wanted to have a 100% authentic pano, at least for my first attempt. 🙂

The final result was about three rows of twenty photos, so 60 photos in total to process. Not having the time whilst I was out in France, I took the multiple sets with a view of doing some proper editing and stitching when I got home.

 

Post Production – generating the pano

And so I loaded up all the photos in Lightroom and went through making some fine tuning adjustments on exposure, contrast, saturation. I do a bit of pixel peeping to ensure all’s OK, and shock horror, there are lots of dust spots and even strands of what look like hair! Argh! Not only that, but those are replicated across all 60 photos so it’s not just one shot I have to tidy up.

That’s when I learnt (too late of course!) that it’s vital to ensure you have a clean lens, sensor, etc before you start. As the photos were taken with quite a closed aperture, all the dust spots really show up in the full resolution photos, and that was a quick check I completely forgot to do whilst out there. And so here I was stuck with those dirty photos. 😮

 

Oh no, Dirt!

Fortunately, Lightroom has a handy dust spot clone/heal tool which allows you to quickly take a sample of an adjacent area and replace your dirty bit with a duplicate of the sampled area. As the dust sports (which are dark) were mostly visible in the sky, it’s an easy thing to fix because just about any surrounding area is of pretty much the same uniform blue colour.

Second saving grace was that you can synchronise your edits between photos, and that includes the spot corrections. So I was able to fix on one of my photos from the top row and then replicate those corrections across all the other top row photos. I did then spend a minute or so checking that all was OK. There were a few photos where the sloping hills had been included and so I ended up with a corrective sample of “trees” in the sky. Whoops, floating blobs of tree in the sky! But all was quickly fixed.

Then, it’s exporting all the corrected photos to high resolution high quality JPEGs to send to the stitching application. For the utmost quality, I could have gone for TIFF, but I already knew my PC would struggly with the JPEGs and it would just collapse if I were to send 60 TIFFs to be stitched together!

 

Stitching it all together

It was then a matter of loading up all the photos in the stitcher (I’m playing around with PTGui at the moment) which is a pretty powerful stitching program allowing you to really control how the photos are merged, what perspective they are, etc.

Once it had aligned all the photos, the final output is a massive 43267 pixels wide by 11653 pixels high! This will then need cropping slightly as it still contains the jagged edges of the photos. I started the stitching process just before I started writing this post, and it’s now only about a third of the way through and that’s on my Quad core 3GHz desktop with 6GB RAM! It looks like I’ll have to upgrade my PC next if I want to speed this up a bit…

 

20 minutes later…

All done! However, due to limitations in Photoshop and PTGui, the biggest JPEG version is about 30000 pixels in any dimension. That means that the biggest JPEG I can make is scaled down. I’m not sure why that is the limitation, the JPEG standard allows up to 65535 x 65535 pixels resolution.

A scaled down version can be downloaded from here. It’s sized at 3840 pixels wide so feel free to use it as a wallpaper on your desktop

If you’d like a copy of the very large JPEG (around 35MByte file size, 30000 x 6830), please drop me an email and I’ll send you a link to it.

It’s certainly been a very informative experience doing this for the first time and I’ve learnt a lot to make things run smoother next time. There’s no question that having a good support with a camera on a head you can just click, rotate, click, rotate, click, etc not only makes the actual photo taking process a pretty quick one, but also maximises the quality as you won’t have any rotation between your photos that need adjusting.

In fact, this is probably the widest ranging photo I’ve ever taken in terms of distances. It covers approximately 180 degrees of view, and has the grass just a few metres away to the nearest hills a few kilometres away, to the further mountains around 20km away all the way out to the moon (top centre of photo) around 360,000km away!

Thanks for reading this rather long post. I’ve certainly enjoyed the whole process and the time taken to compute it all gave me the opportunity to write this up whilst I was waiting 🙂

Till next time, happy snapping 🙂

no comments

Your email is never published or shared. Required fields are marked *

*

*

There was an error submitting your comment. Please try again.