Immersive Flash Panoramas

I have just discovered my first dislike for this WordPress.com blog – Flash embeds are not permitted in posts, pages, or text widgets. For security reasons the tags needed for these to work have been removed.

This is frustrating because I’ve spent quite some time on learning how to create immersive spherical panoramas. After the Supanova cosplay event I discovered another Melbourne photographer’s website, Neil Creek, who has some really neat examples of these type of projections. 360 Cities also has some extraordinary dynamic scenes which boggle the mind.

As for available resources on the web explaining how to compose, execute and stitch the required photographs and then convert the projected composite to view as a QuickTime VR or Flash object, here are some that I found:

• Peter Gawthrop’s Creating Spherical Panoramas with the Canon 5D and 15mm Fisheye Lens
• Outside the Lines’ QTVR for the Rest of Us
• panoramas.dk STITCHING AND AUTHORING PANORAMAS
• Panomundo’s How to Make QTVR Panoramas

Alfred Nicholas Gardens Lake - Equirectangular Projection

After taking a set of images from a standard tripod at the Alfred Nicholas Gardens on Mt. Dandenong I set about trying to develop my own budget method for creating and displaying immersive spherical panoramas. The summary is as follows:

• Panoramic stitcher: Hugin 0.7.0
• Flash and QTVR panorama converter: Pano2VR
• Free Flash object hosting: Google Sites
• Permitted blog embedding: Blogger

The immersive Flash panorama of the above projection can be viewed on a Blogger account I created. Noting the Flash object’s width, height and hosting URL: http://sites.google.com/site/ccdoh1flash/Home/Boathouse_out.swf, the HTML code I used to embed it within Blogger is:

Obviously I have a long way to go in refining the method (particularly in exposure control) but I believe this to be an excellent beginning.

Lunar Eclipse Composite

This composite photograph details the advance of the Moon from the umbra to the penumbra during the August 28, 2007 total lunar eclipse as seen from Australia.

I shot this from my backyard using a tripod mounted Canon 350D camera @ ISO200 and a 17-85mm lens zoomed to 85mm @ f/5.6. I composed the shot such that the Moon would pass across the diagonal of the frame and triggered shots roughly five minutes apart. The individual shots for the composite were bracketed to account for the variation in the Moon’s brightness during the transition. The exposure time was 4 seconds in the umbra, 1.5 seconds in the umbra-penumbra and 1/160 seconds in the penumbra.

Using Photoshop, I added each individual photograph to a new layer with the layer blend mode set to screen to achieve the composite image.

The following schematic indicates what is actually happening to the Moon during a lunar eclipse. Within the central umbra shadow, the Moon is totally shielded from direct illumination by the Sun but is lit by refracted red light passing through the Earth’s atmosphere . However, within the penumbra shadow only a portion of the sunlight is blocked and we see a partially lit Moon.

Geometry of a Lunar Eclipse

Skyscraper Perspective Correction

Last night I met up with some mates and flickr contacts (mr.analog, BluePyjamaSyndrome, hann33 & tim_best_1972) to shoot some photos from the top of the Rialto building in Melbourne. It was the same night the Irish defeated Australia in a Gaelic Football match played at the MCG, hence the fireball of lighting. Unfortunately it was a bit cloudy, squally and drizzly on the night, but the direction I shot this photo in was protected from most of it.

The bad thing about shooting skyscrapers from even taller skyscrapers is the tilt they develop with a wide angle lens. After reading this online tutorial, I used hugin to place vertical lines along building edges to fully straighten them out.

Shown below is firstly an uncorrected image followed by the same image that has been stitched to another and corrected for perspective.

I wonder if I’ll ever get to use a tilt/shift lens to combat this perspective phenomenon?

Layer Masks and Transparency Gradients for “The Heart”

This tutorial will cover digital blending in Photoshop CS3 using layer masks and transparency gradients. It is aimed at those who are interested in seeking alternatives to HDR (High Dynamic Range) processing.

I would consider the usage of these techniques as digital graduated neutral density filters and although it can take some time to implement, I find the benefits to be:

• Preservation of intended colour and tone
• Complete control over contrast gradients
• Elimination of halo effects
• Elimination of fringing

I will use one of my photographs, The Heart, to illustrate the method.

1. Download this zipped folder containing the two images to be used for blending and open IMG_1166 in Photoshop.

2. Press Ctrl+Shift+N and click OK to add a new layer. To place IMG_1164 on this layer go to File > Place and locate it. Click the tick to commit the transform.

3. To create a mask, click the New Layer Mask (grey square with white circle inside) button in the Layers palette. To create a transparency gradient on this mask, first hit G to bring up the gradient toolbar and select a linear gradient. Then click and drag as shown in the following picture.

4. The next layer will be a photo filter so click Layer > New Adjustment Layer > Photo Filter and hit OK. Push the density slider to 100% and select a custom colour with properties as shown in the following picture.

5. Now apply another layer mask and gradient, as done previously, to restrict the photo filter to the sky. Change the blend mode to Linear Burn and opacity to 70% for the following picture. The Linear Burn blend mode is similar to Color Burn, but instead of increasing contrast, it decreases brightness to darken the base color and reflect the blend color.

Now add curves layers to the foreground and then the sky.

6. For the foreground, click Layer > New Adjustment Layer > Curves and hit OK. Then modify each of the RGB channels to reflect something like the following picture.

7. Change the blend mode to soft light and add a layer mask and gradient (this time in reverse) to restrict the curves layer to the foreground. This will produce the following result.

The curves layer that targets the sky looks like this.

8. Change the blend mode again to soft light, opacity to 75% and create a layer mask and gradient to achieve the following result.

9. The final step involves adding a levels layer and changing its blend mode to screen to boost the lighter areas. An opacity of 90% will achieve the picture shown below.

Straightening the horizon and correcting the barrel distortion yields The Heart as displayed at the top of this page. Serena Ho attempted the tutorial and her result is posted here. It would be great to see some other results as well.

Photographing the Milky Way with a standard DSLR camera

This tutorial will cover the development of my image Milky Way from Apollo Bay using a Canon 350D, a wide angle lens, Adobe Photoshop and an image alignment program called hugin.

For this to be repeatable you must be shooting with clear dark skies, free from light pollution.  I used settings of Tv: 20 sec, Av: f/4, focal length: 17mm and ISO speed: 1600. A shutter release cable is also a great tool and can keep your camera clicking as it sits on the tripod. The more images taken, the better for the final result, because this will improve the signal to noise ratio that plagues digital sensors during long exposures. A tracking mount is not necessary with a 20 second shutter speed because the rotation of the night sky is undetectable at such a wide angle.

Please read the following steps for more information. All images are hyperlinked to larger sizes.

1. Download this zipped folder containing four of my unaligned shots of the Milky Way.

2. Load them into hugin_0.7_beta_4 .

3. Manually align images with control points – don’t automatically align. I aligned three images to a common one. Enlarge the screenshot for details.

4. Click Edit – fine tune all points.

5. Click View – preview window.

6. Click Center and Fit buttons to achieve this view:

7. Click Edit – Optimise.

8. Click Stitcher – image format – multiple tiff. Final screen before clicking Stitch Now.

9. If you don’t want to worry about learning how to align the images with hugin, then you can download this zipped folder containing the four prealigned images of the Milky Way.

10. Load each image onto a new layer in photoshop adjusting the blend mode to screen which is good at lightening images without lightening the darkest areas.

11. Add a medium contrast curves layer.

12. Add a colour balance layer: shadows (-90,-25,-10), midtones (-15,-5,-20), highlights (0,-5,30).

13. The final result.

For comparison’s sake, shown below is a typical accompanying jpeg to a raw file I began with for one of the individual images. I used Adobe Camera Raw to extract the jpegs provided in step 1 for processing.

In closing, I’d like to point out this is not the only way an image like this can be captured; there is myriads of possibilities. I have developed this simple and inexpensive method by just experimenting with the tools at my disposal.