Chp. 9: Content Conversion

Chapter 9 – Content Conversion

 

This chapter contains advice and tips for converting older content and preparing it for use with your new system. As always, make sure that you adhere to all copyright laws and user licenses. For older purchased shows, check with the original show producer to see if digital versions of the images are available. Even if it costs a few bucks, it may well be worth it to you, especially after reading the next two sections.

There are primarily two methods of reusing older content, real-time programming and pre-rendered fulldome video. These will be discussed later in the chapter, but first we need to get our images and video clips ready for use!

 

Single Slide Images

 

Where possible, you will likely want to replace your old slide images with natively digital ones. Natively digital images are those provided originally in digital form, rather than scanned from slides or printed photographs. Easy enough with images from the Hubble Space Telescope or the recent Mars missions. Not so easy with images from older space missions or celestial events from earlier days, although more imagery from the earlier missions is becoming available, such as those from www.apolloarchive.com.

 

If a digital version is not available to work with, check to see if you have a high-quality print version. Those lovely NASA lithographs come to mind.

 

Scanning from a high-quality print is something many of us are accustomed to. Make sure to give the resulting file a meaningful name, especially if you have many of them to scan.

 

If there are no other alternatives, you will need to scan the slides themselves. You will want to use the cleanest, clearest, most vibrant copy you have. If the slide is in a glass mount with the glass fogged from age, try to clean it off or remount the slide in a newer mount. Unmounted slides can be scanned, but require extra care to not touch the film chip.  In most cases, you will also want to remove any masking film chips before scanning.

 

Most desktop flatbed scanners these days have good quality attachments for scanning slides. If possible, use the software that comes with your scanner. As with image scanning, you will want to give a meaningful name, although it doesn’t hurt to use the original slide number from the show script if you still have the original slide list. Keeping the images in a separate folder for each show is a good idea. Just ask the authors how we know this.

 

If you have a very large collection of slides, a specialty slide scanner can save time to speed up this process.

 

After scanning the slides and saving the resulting images, you will want to examine the images closely. Do they need to be brightened or have enhanced contrast? Are they too red? Too blue? Are there scratches or other defects that need to be corrected?  For this you will want to use some sort of image manipulation software. Examples include Photoshop, Gimp, Paint.net and PaintShop Pro.

 

After correcting any defects, you will want to crop and/or resize the image to get it to the desired pixel dimensions. If you want part of the image to be transparent, you can do that in the software too. Make sure to save your images into a format that preserves transparency, such as the PNG or DDS formats. Such formats support alpha channels. Most digital images for video display use 3 or 4 channels. One each for red, green, and blue. The 4th channel is the alpha channel - it defines what parts of an image are to appear transparent. 

 

For more reading on scanning and processing slides for digital use, see the following:

[Put references to Chris’ paper and any one of Dale Smith’s slide scanning papers.]

 

Panoramas and Allsky Images

If at all humanly possible, you will want to scan in or photograph the source art from panoramas and allsky images. The resulting long rectangle (panorama) or circle (allsky) can then be used with little or no additional modification. To recreate them from slidesets is tedious and can require a great deal of trial and error. 

For re-created panorama and allsky images, the process of scanning is the same as for regular slides. But then the images need to be stitched together. Whether to perform corrections to the scanned images before or after stitching is up to you. Always save a backup copy of the originals in case you need to try a different way.

 

If the images have no overlap, they will need to be aligned by hand in your image manipulation software (Photoshop, Gimp or similar). A separate layer for each image segment is advisable, so that brightness and contrast for each slide can be visually estimated and manually tweaked. Thankfully, the stitching process is a bit easier when the images overlap. Not much easier, but a bit. You can use the same method to align those with no no overlap, or you can use an image stitcher. Image stitchers are not the easiest types of software to use, but they can yield better results for some panoramas and allsky sets.

 

There are several image stitching software packages on the market these days, plus some free alternatives. An excellent free software stiching package is called Hugin. Image stitchers are great for making new panorama and allsky images from a series of overlapping images taken with modern digital cameras. Cameras these days record everything about the camera’s settings, and the stitching software reads that data and does most of the work for us. But for scanned images or slides, we will need to do much of the work ourselves. 

 

Stitching software takes two or more images and creates what are called control points. A control point is a point in one image that corresponds to a different point in another image. If two images overlap, they would share those control points in the shared overlap region.

 

Regardless of which software package you use, you will need to load up your scanned images. You may need to guess at what lens and exposure settings were used for the original slide images. As long as the image settings are consistent, you should be able to get close enough results to go back and tweak settings if necessary.

 

Most stitching software packages have a feature that will let the software try to automatically determine the control points. You may need to select the images in pairs and run the control point detection for each side-by-side pair.

 

Next inspect the control points for each pair of images. If the software could not detect any, it should permit you to place them manually. Even if the software did detect control points, you will need to check for mistakes. This happens a lot when there are clouds (or nebulas) in the final image. Delete any control points that do not actually match. In the worst cases, you may need to place each control point by hand.

 

Finally, select the desired output settings and format for your final image. There may be plenty of options or settings. As long as you save your project before rendering your output, you can return to try other output settings until you find what you like.

 

You will still need to do additional cropping and editing with image manipulation software, but the hard part is done. Whew!

 

Scanning slides and stitching them together to recreate panoramas and allskies is time-consuming and tedious, but well worth it if the panorama or allsky is a beloved old favorite.

 

Video Clips

 

If your older video clips were already in a digital file format, you might be able to use them as they are. If they were burned to DVDs, you can open the DVD with a file browser on your computer. DVD video files usually have the file extension VOB. Copy the VOB files to your hard drive. Then rename the .VOB to .MPG (or .MPEG, as they are the same format). The files should play on your computer. If your new dome system will play the MPEG format (most do), try playing it through your system.

 

If the file doesn’t play on your dome system and doesn’t play on your computer, you will need to find another way to get that content. You may need to contact the original publisher or check around to see if a good substitution is available. Don’t be afraid to ask your fellow planetarium professionals. (We don’t bite, at least not much!)

 

If the file plays on your computer but not on your dome system, you can use video editing software to convert it to another format if you plan to use it in a real-time show (see below). Be aware that you will be recompressing and already compressed file. This will degrade the quality of the video.

 

Your vendor will be able to help you know the preferred video format for your software, and may be able to provide playback and conversion tools. Some systems may include a third-party video player installed that the planetarium software uses for playback.

 

Real-time Show Production

 

Many systems now have the capability to display images and 3D models, play video clips, perform preprogramed astronomy functions and audio soundtracks in real-time. These actions are performed on the fly according to a script or timeline set up by the user. This is analogous to using an automation system to run slide and video projectors while playing a soundtrack.

 

This is an excellent capability, and affords you the greatest flexibility for recreating any of your older shows. And just as with slides, if you wish to replace an image with a newer one, you swap out the image. And just as nice is the ability to place images anywhere on the dome. Standing still, spinning, zooming , slewing –your old images can do some amazing new tricks!

 

For many real-time systems, the image size in pixels is best kept at powers of 2. Pixel dimensions of 256x256, 512x512, 1024x1024 and so on work well because that is how video cards “think” when allotting memory space for storing the image. If an image dimensions are not a power of 2, the system will round up to the next higher power of two. This means that an image with pixel dimensions of 600x600 will take up the same amount of memory as an image with pixel dimensions of 1024x1024.

 

If your image isn’t a power-of-two, don’t despair. You can use your photo-editing software to put it on a transparent (alpha channel) background that is. Putting non-square images on a square alpha channel background can also make them easier to scale and manipulate in the software later. When using alpha channels make sure to save as a file format that uses them, such as .png or .dds. Your system vendor will be able to help you choose which file formats will work best with your system. JPEG files do not save alpha channels and will give you the ever hated white box around your image.

 

Do you miss your old box of fuzzy slide masks? Alpha channels allow you to get the same effect with your digital image files. Learning to use them will let you replicate, and improve on your old shows in your fulldome system.

 

As for file types to use in real-time production, acceptable file types should be listed in your system documentation. Even if your preferred file type works on your system, you may find that some file formats take up less memory space than others.

 

Most system vendors have users groups that share content. A great way to start understanding how to add real-time content to your system is to check out the files available from your users group. You may also want to talk to and ask questions of some of the contributors, who should be able to share tips and tricks specific to your system.

 

3D Models

 

Part of the amazing power of fulldome video is the ability to display 3D models as a part of the system’s simulation of the Universe. Rather than just look at a picture of Hubble, you can visit the actual telescope!

 

Though they add some incredible features and are a lot of fun, 3D models can add a lot of complexity to work with. There are a variety of formats, which have different features and limitations, and a given system may only support a few of them, and may only support a few features. Your vendor can tell you what model formats they support and how to use and provide them with texturing and illumination in the system, and what software programs can create them.

 

Most systems come with an excellent library of pre-built models, and have users groups where models (as well as images, videos and scripts) generated by other users are shared in easy-to-use packages, but you may find yourself wanting to make your own. If you don’t have experience doing this, there is training available! In addition to working with your vendor, and online or real-life classes, GLPA hosts regular workshops on working with free 3D software that can open up this world to you.

 

Prerendered Fulldome Video Production

 

In our field we have become increasingly accustomed to fulldome video shows. These shows play on the system as a movie file. In the case of (usually) larger domes with multiple video channels to cover the entire dome, there will likely be several separate video files playing simultaneously. These video files have been split from the original set of video images and warped to match correctly on the dome.

 

These video images are called fulldome master files. Each frame of video is a separate image. Some new shows are provided as a sequence of these image files. Your system vendor should provide you with tools for creating the necessary fulldome video files. Many show vendors can now provide shows already formatted for your system.

 

To create your own fulldome master files from original content requires using either a 3D software package or a video editing/compositing software package. The idea is to combine your images, starfields, smaller video clips and models into the properly distorted final format so it will all look good on the dome.

 

Using a 3D software package can be more awkward unless you are already familiar with using the software. Within the software, a virtual scene is created with a virtual camera to create every image frame of your video. In this scene, you would create filled squares and rectangles. Instead of filling the squares and rectangles with colors or patterns, you fill the squares and rectangles with images or video clips. In the virtual 3D scene, you would then position the squares and rectangles to where you would want to see them in the dome. Again, it is an awkward method, but might be worth a try if you are already familiar with 3D graphics software.

 

To use a video software package to create fulldome videos requires a plugin designed for properly distorting images and video clips to look right in the final projection. Each plugin and software package has its own specific ways, but the gist is this: The software will display a virtual dome view and a timeline. Position your timeline marker to the appropriate time, then select your image or video segment and drag it onto the virtual dome view in the desired position. You may need to provide information about the image, such as whether it is a panorama, an allsky or a regular image. There will be settings for when and how long to fade the image on or off, and to change its position, scale and rotation. The video can either be rendered out as a series of images for slicing on larger systems, or as a single video file for systems using only one channel for projection.

 

Sharing Prerendered Fulldome Video

 

In addition to producing your own prerendered fulldome content, it is good to know how your system will handle prerendered content from other producers. If you are exchanging prerendered content within your vendor’s user group, there may be only minimal work involved to install and configure the content in your system, particularly when sharing with others who have the same projector layout.

 

Also, your vendor may have provided you with tools to convert “dome masters” provided by other vendors into the movie files for playback in your system, or they may help you do the conversion by “slicing” the files for you. Make sure you have this conversation with them and you are comfortable with how to get content into your system.