I got some story here to tell, but before I go on talking about it I got to announce that I got a new tripod now (headless) and I think I've should have had this long time ago. It fits my VR-head perfectly giving more stability and on the other hand I can fit a previously-purchased tripod head (or panning head) separately. This pan head has one handle to rotate around making it easier (and faster) to handle rather than using two controls for panning sideways and vertically. I couldn't find this brand on the net or on B&H, but it's called Fancier, followed by some model number.
Just as a little review, the tripod is extremely useful for my work specially when it comes to panoramas, and it extends to some lengths longer than my previous tripods (only matched maybe by my monopod). The camera can fixed on the central column of the tripod either way (up and down). The weight is not an issue, but I have to say it is a bit heavier than my previous aluminum tripod. The legs can be stretched out to low levels as well (giving me some ideas now).
Now to the bad points: It doesn't have a handle to carry around. To carry it around I'd have to put it in its case. It does not have a lock around the central column to hold the 3 legs firmly, making sliding legs a possibility.
I've measured the full length of the tripod extended, with central column completely stretched, and having my VR-head above that, it gives out a length of around 204cm (~ 6.69 ft), and that inspired me for a little thing to do, as I will explain later. Now let's start the little story...
1. Chapter One: Idiocy:
Instead of using the slide numbers that I usually get after converting to HDR images for the panorama stitching, I'm using simply numbers for angles. The outer circle is for the horizon or 0 degrees tilt, and the inner circle is for 45 degrees upward tilt. There is no need to use a downward 45 degrees tilt when we talk about vertical panoramas, but maybe in certain occasions it might add some depth by putting more of the background. The middle point "Z" here means the Zenith of the panorama. I will be using the notion "^" for angles on the inner circle.
Now, an example to work with. The first target to work with was the Ardeaglais, as I explained in the previous post last week. This vertical panorama followed with the same plan, but as a comparison I will put it here again...
|Suas agus Síos (up and down)|
To make out this panorama, I've started with frontal angles first [12,1,2] and start from there going upward [12^,1^,2^]. At the zenith, I realized one slide from the zenith won't be enough, so I must take others on the side, i.e. [10^,4^]. Now, the hard part, which had to be done outside in Photoshop then plugging in the results into PTGui. The coming part includes the slides at angles [8,7,6] at 0 and 45 tilts (notice that I keep the order from left to right). I had to rotate these slides upside down to mimic the view of upside down (duh!) because simply, I couldn't change the rotation of these slides within PTGui without affecting the whole stitching process. The stitch was successful, to some degree, and the image was there online.
Later on, I thought that this is not enough. The vertical panorama was too thin to my eyes. So I decided to include more width, and that is by taking angles 11 and 3, and also 9 and 5. This inclusion meant I had to include more from angles 10 and 4, and other work related to flip the images of the lower side and... wait a minute! huh! How come I didn't see this coming...
|The Hanging Dominican|
more width added.
2. Chapter Two: Spark of Light:
In case you wonder why I named the last chapter "Idiocy" and didn't know why yet, well, here it is! After working with the above panorama, The Hanging Dominican, I realized that I'm practically... using the full set of the panorama! Should I flip some and keep some? Is it really easier than I thought? It's really amazing how we lose our conscious when we concentrate about details of procedures and forget about the whole picture or concept! The whole procedure of flipping images before stitch, picking the proper slides, and all the fuss in the memory because of that - all of that could have been avoided by a simple touch. Changing the pitch parameter for the spherical panorama and point it upward to the roof! As simple as that! The width of the panorama can be controlled then later on by cropping in Photoshop after all.
full spherical panorama with changing pitch orientation
3. Chapter Three: Complicated Simplicity:
Now, this is something related to panoramas but it's not related to the previous points before. I explained above some aspects of my new tripod and I had plans to try it out on the roof. I thought such good height that this tripod can provide for me would be a good start to see how it works with a simple panorama from above. I had to climb on the highest point on the house and from there spread out my tools. My aim was a simple panorama; a 360, i.e. one horizontal line.
Despite its simplicity, the stitching took me hours to fix, and I'm still not satisfied with the results. Technically, the process of taking the panorama took me less than 5 minutes, literally. To take this panorama, I used a method that I've read about before but never used it, which uses metering for highlights and shadows of the scene with a fixed aperture (f-number) and then take the shutter speeds that you get from metering, and work in Manual mode, fixing the shutter speed to something in between the two shutter speeds (of the highlights and of the shadows). This way would make a coherent exposure across the whole set, but it didn't work quite well probably because I'm in the open sun (and I don't want to talk about the heat).
For some reasons unknown to me, there was always break in lines even in this simple panorama, and optimization would fix this but with some weird stretching lines (caused by distorting some images to fit some areas down there). The biggest problem was one of the slides didn't have control points to stitch with others, and fixing my own control point did make the situation worse. I tried to optimize the exposure in hope that would fix the control points problem (exposure might be a factor for not recognizing some points or pixels on adjacent images, sometimes), but that didn't work out. There was some crop in some of the slides, but it was not much of a big deal (one to three pixels I believe) so this reason was rolled out.
Finally and just to get the process going on, I gave up stitching in HDR. I simply tone-mapped the whole set, and stitched from there. The set was stitched nicely without any loss in control points! Since I'm satisfied with the colors already, I thought it's not needed to save the control points and stitch back in HDR to tone-map again. No need for this. There were some broken lines and optimization saved the work, but with some stretching in the lower portion (can be apparent even after cropping it in the image above). Left now with problems of noise and grain pixels which were hard to remove without losing some details. I still don't understand why such problems occurred in such a simple panorama as this one...
My books arrived now and I can "waste" some of my time reading. Three books they are:
Of the three, the Night Photography book is the one that captures my interest the most. Did I mention that I have an artificial language to look after?...