The CPU is dead. Long live the GPU.

Adobe just announced that they've developed a new version of their Flash Player that reduces computer CPU use when playing video to as low as 10% of previous versions. It's a notable development because high CPU use has long been a key metric demonstrating Flash's considerable inefficiency with regard to power consumption and heat production. The need to improve these performance characteristics is critical for advancement of the Flash platform.

The improvement was accomplished by shifting more of the image processing from the computer's main processor (CPU) to the graphic processor (GPU.) This shift has been an ongoing trend with many graphics-intensive applications and was bound to happen with Flash eventually. But it's important to remember that the real goal is to reduce power consumption and heat overall -- not simply to transfer those issues to another component. Enhanced CPU performance matters little if the total power/heat profile remains high, with only a new metric of poor performance.

It also raises the question of how much these improvements will scale to mobile, where improvement is arguably the most important. Mobile devices don't generally have the kind of GPU horsepower available on the desktop. Failure to significantly improve power performance for mobile will continue to leave Flash effectively impaired for that critical market.

The new version is currently in beta but should be finalized with the coming 10.2 update. Ultimately we'll need to see benchmark data for the final release to know the full extent of power improvements, but for now we can be optimistic. Hopefully Adobe will also use this as an opportunity to make Flash Player more stable and secure. For many users, crashes and security patches both need to become far less frequent.

Non-Destructive Geometric Corrections: Yes, Please

One of the major advancements in the current generation of photographic image editing software has been the advent of non-destructive editing, whereby user modifications to an image do not actually alter the image source data (pixels) in the master file, but are recorded instead as separate actions that are applied to the data when viewing and/or outputting the image. There are many advantages to preserving the integrity of the source data for future use, while still achieving the necessary modifications for an immediate application. These non-destructive edits have been available to alter a wide range of image parameters such as exposure, color balance, contrast, sharpening, noise, and many other adjustments. But there are other parameters, such as geometric distortion manipulation, that are not yet offered in a non-destructive form. These corrections generally require the creation of a separate new version of a file in order to preserve the modified data. For a variety of reasons this is not an ideal solution, with issues related to storage capacity, application interoperability, and versioning. None the less, these types of image adjustments are very important for subjects such as architectural photography where control of visual perspective is highly desirable. In these cases photographers have generally learned to live with these multi-file issues.

On another photography blog that I regularly read the moderator speculated that geometric distortion manipulation of photographic images would always be a destructive editing process, presumably due to the complexity involved to maintain modified data in a state accessible for further editing. However, I'm not so quick to discount the possibility of these as non-destructive corrections in the future. Any image correction is just 1.) a user-selectable value and 2.) an application-generated view of that transformation. There is no reason why that view MUST actually transform the original data, as long as the software can always regenerate the edited view on-demand from the user settings. Even distortion adjustments are correlated to user settings, and these can be recorded, stored independently, then recalled and applied to an image, no different from any other non-destructive value. It probably takes a lot of processing power to manage those kinds of "live" transformations in real time, and that may explain their current lack of availability, but the ever increasing level of computing horsepower is assured and ultimately this should not be a limiting factor. I fully EXPECT for software developers to offer non-destructive distortion corrections in the future. Perhaps not in their next updates, but certainly sooner, rather than later.

For architectural photographers the advantage of being able to perform all our mostly commonly used image corrections in one application, and without needing to create alternate versions of files, will be a significant improvement to our workflow and image management strategies. I have seen examples (on the Web) of prototype image editing programming like this from one of the big graphics laboratories, though I can't remember where now. I do remember that it was impressive and, of course, I wanted it immediately. I have no doubt we'll see it -- and be paying to get it -- before too long.

What Adobe Should Do About HTML5: Lead

The introduction of Apple's new iPad has triggered much discussion in the blogosphere about the value of Flash as a standard for Web development in the future. Apple's decision not to allow Flash on its mobile hardware (the iPhone and iPad) in favor of promoting new open standards such as HTML5 and CSS3 to deliver dynamic graphical media (video, audio, animation, etc.) has been hotly debated by supporters and detractors of this decision. There is much to like about about HTML5 and its advantages over Flash, particularly for mobile applications, but that's not the point of this post. I'm confident the market will ultimately decide the outcome based on the real world merits of their different strengths. I'm more interested in the direction Adobe is heading with their tools based on these new developments.

Thus far Adobe seems to have reacted mostly in defense of their proprietary interests, without much indication of their intent to actually lead the advancement of new standards for the development community. Ironically, although Flash is a closed proprietary format, their responses cite "open access" as the prime rationale for Flash as a viable media option -- but with little mention of its inherent the liabilities. Another similar response offers some indication that Adobe does understand what open standards proponents want, but with only vague hints of whether Adobe is truly committed to leading the way in developing tools for these, or merely going through the motions while continuing to push Flash. This is important because new, better performing, and open standards such as HTML5/CSS3 are important to the evolution of Web development and Adobe SHOULD be in the game with best-of-class tools to support them.

Adobe says that it's in the "helping people communicate" business, wherever that may lead, but I find it curious that we haven't yet seen any evidence of their work to produce new open standards-based development tools independently of Flash. These tools are beginning to emerge, but not from Adobe. (Such as this, this, this, and this.) The HTML5 specification itself has been around since pre-2007 and includes capabilities for both video and vector graphics. As a graphics software leader, Adobe is naturally positioned to develop software for it. It's been obvious for quite some time to anyone who follows Web technology that HTML5 was coming and would be an important new standard for the Web. But all we've seen from Adobe so far is a vague commitment to develop new tools for open standards in the future, while simultaneously chiding Apple for trying to get there sooner without Flash. This is not a very forward-looking public position.

Something else bothers me: regardless of the reasons for poor Flash performance issues, particularly on Macs, this has been a problem for far too long. Flash-derived browser crashes, high processor cycles, and high power consumption have been the status quo on Macs literally for years. If Flash technology is that THAT hard to optimize, then it was ALWAYS going to be a problem for battery-powered mobile applications. And mobile applications were ALWAYS going to be the future. If Adobe has done their best to mitigate these problems until now -- and I have no reason to doubt they have -- then perhaps it's not the best solution for next generation mobile devices that will need even greater power efficiency. You can only flog a technology for so long before its limitations become too much of an obstacle.

Let's also not forget that Adobe didn't invent Flash, they bought it in 2005 from Macromedia for $3.4 BILLION. They were obviously optimistic about the future of Flash technology and wanted to capitalize on it. No harm in that. But like any investment, the market makes no long term guarantees. For financial reasons I'm sure Adobe would very much like to see Flash continue to dominate the market, but clearly the market is evolving. It's not personal, just the way it goes. It's only a problem if instead of continuing to innovate to meet future demand for better products, you get stuck trying to maximize an investment the market no longer believes in.

I'm a great fan of Adobe and have purchased and used their products for many years. The company has produced some truly innovative breakthrough technologies. Products such as Postscript, Illustrator, and Photoshop, among others, drastically improved the world of graphical computing. I absolutely want to see them continue to be successful. But the relevance of ANY technology product in the marketplace is a moving target. With the development of new Web standards that offer better performance and truly open access, the viability of Flash has peaked. Adobe should embrace the change and lead the way by producing great new tools for the next generation of Web development's best practices. But in any case, they need to lead, follow, or get out of the way. The advance of HTML5 as the dominant open standard is a certainty, and very many Web developers are already looking beyond Flash.

Personally, I'm excited about the possibilities that HTML5 open standards mean for Web communications. I anxiously await the availability of new tools that Adobe can provide to move us in that direction. I have no doubt that they will be rewarded financially for their innovation.

Motion Stills in Final Cut Pro

Recently I’ve been experimenting with different software used to animate still photos with pan and zoom motion for video, otherwise known as the “Ken Burns effect.” These types of effects can add a lot of visual interest to a presentation when producing animated slideshows or video projects. Since video is increasingly becoming popular for many Web marketing uses, adapting high-quality architectural photography to this medium is a natural evolution. With regard to the real estate market in particular, research has shown that animated still images are just as effective as captured video. In fact, the video quality can be even better, through the use of higher-quality stills as source images.

There are several software options available, but I was looking for an approach that would integrate well with my video editing workflow in Final Cut Pro. I’m fairly new to FCP and the learning curve seems fairly steep so far. Although it’s not hard to learn the basics, advancing to more complex projects quickly reveals a daunting set of issues, options and controls that require more than a casual knowledge of video theory. My impression is that full mastery of FCP is probably achieved only by video professionals. Most users only scratch the surface of its very deep toolset, somewhat like Photoshop.

Final Cut Pro natively has all the necessary controls to animate stills, but the process is somewhat laborious, with no provision for automating multi-image projects and requiring manual intervention with each image and transition with modifications in the timeline. I’ve discovered that a better solution is to use a third-party software plugin called PhotoMotion from SlickFX. Using this plugin within FCP greatly simplifies the process of animating still effects, either for still images combined with other captured video, or for entire slideshows built from a series of still images. The software manages both the mechanics of each image’s motion, as well as the transitions between images. This latter is particularly helpful when the user inevitably begins to edit a project timeline, and the software automatically updates the required motion and transition parameters. This makes it easier build projects from scratch and then to experiment with different creative choices in the timeline.

I’m quite pleased to have discovered this software and expect that it will save me a significant amount of time when producing animated slideshow videos. I highly recommend it for those doing similar projects in Final Cut Pro.

Video: Nightfall in New Braunfels, authored in Final Cut Pro 6 with PhotoMotion

What's Lies Ahead In 2010

Offering views of the future seem to be a standard practice on many photography blogs at this time of year, so I'll add my thoughts to the mix.

First, I must profess that I don't find future technology developments to be particularly interesting or important. It's a given that tech will keep moving forward relentlessly, providing more and better gear, with more and better features, and more and better pixels. There will always be new "must-have" models of the tools we use, delivering increased productivity, quality and/or capabilities, in smaller, easier-to-use and more affordable packages. It's less a matter of if and more matter of when any desired tech development will arrive. So I just take it for granted that the state-of-the-art in tools will improve, without being much concerned about the specifics. Truth be told, right now I have all the tools I really need to produce the images I need. I will no doubt embrace new tech as it arrives, but pondering the details in anticipation seems pointless.

There is now much talk among photographers about the importance of video capabilities in our future -- and I have no reason to doubt it. Like most of them I'm also working to expand my video capabilities. Judging from its prevalence on the Internet it's clearly growing in demand for certain markets. But I'm unconvinced that video will soon be a requisite service for others. With regard to my own specialty in architecture and real estate, my subjects are essentially static and motion capture adds little to the presentation experience, at least for common uses. When motion is desired an animated still photo (think Ken Burns) is virtually indistinguishable from a true video pan and is satisfactory for most client needs. Still photography continues to be a very serviceable product in this market, without much pressing demand for true captured video. For my part, in the near term I plan to spend a little time working with true video and somewhat more time with still animation techniques, but the bulk of my development work will remain solidly in the realm of traditional still photography.

One questionable perception about the impact of new DSLR video tools on traditional markets is that still photographers will be at the forefront of all that new work. This may not be altogether accurate. It may be that, as Kirk Tuck speculates, services could move the other way as established video professionals begin offer high-quality stills as part of their packages, thus eroding traditional photography markets even more. The balance will likely settle somewhere in the middle, but it does remind one of the old adage to be careful what you wish for.

For me the more interesting question about the future concerns the overall focus of our work. The interactive nature of the Web is changing our expectations of how visual information is presented, and to what depth. The traditional role of photography for much of the work I do has been mostly about documenting moments or places in time. But I see photography now evolving to be more about story-telling, where a singular great image is less important than a coherent set of images that tell a story or provide a deeper experience of a moment or place. Producing great images will always be an objective, but more as part of a larger body of material that expresses a subject more fully. These may include video, audio, and/or text, but still images can still be at the core of the story, used in creative and interactive ways to flesh-out the overall visual experience, or to present more perceptual detail.

None of this is new, of course. Many photographers, especially in editorial, have been working this way for quite some time. But what I see now is that this is becoming the broader norm for all visual producers. The expectation rather than the exception. The Web provides a means for almost everything to be a richer story and as content consumers we are learning to value that. I've got some interesting ideas about how this might work for architecture and real estate photography and I plan to develop some of these in the coming year.

With the continued decline in traditional editorial work for newspapers and magazines, compounded by the impact of a weakened economy, recent history has not been particularly good for photographers, among many others. But viewed from a different perspective, it may be that we're just seeing a natural transformation the craft. History offers many examples of this kind of evolution. The trick is to embrace the changes and evolve as well. Markets and economies will almost certainly revive again, and when they do we need to be poised with new ways of thinking about what we produce. If we do that, then our perceived value will remain high and our work will continue to be in demand. Now more than ever is a time to adapt and grow.

Best wishes for a prosperous new year.

Gruene Hall Sunset

D700: Enfused Multiple Exposure
Those who are familiar with Gruene Hall know that in reality there are multiple utility lines spanning this view. I've digitally removed the lines in this version to enhance the image.

HDR and Visual Credibility

D700: Enfused Multiple Exposure

I've been working for a while with methods to extend the dynamic range of digital photographic images through various HDR and exposure blending techniques. These techniques can be particularly useful in architectural photography where very-wide contrast ranges are often found when shooting both exteriors and interiors. There are many common situations where these methods can help, such as revealing detail deeper into the shadows of a dark building entrance on a bright sunny day or, conversely, helping to hold bright exterior details in window views when photographing darker interiors. Because the dynamic range that digital sensors can capture is so narrow compared to human eyesight, these techniques can help bring an image closer to how we naturally perceive the scene. This is highly desirable for high-quality architectural images.

For most people the term HDR (short for High Dynamic Range) has become synonymous with glowing, wildly exaggerated color images, often well into the bizarre range. Many people apparently appreciate the strong colors and surreal qualities for their artistic merit. These types of images have also become popular with many photo enthusiasts because the software for producing them is quite easy to use and does not require much precision -- requiring neither color fidelity nor adherence to visual realism to produce a striking image. Indeed, dismissing those qualities is often the artistic intent.

However pleasing those images may be artistically, in professional architectural photography the objective is considerably different. Visual credibility matters very much. The primary goal is to enhance the perception of a building or interior in a natural and realistic presentation. An architect may appreciate having a garish, artistic vision of their work displayed in the proper setting, but it's hardly what they require for their portfolio, promotions or magazine spreads. Similarly, a potential home buyer may appreciate the same qualities in an image hanging on their wall, but it's not helpful when they're considering the amenities of a potential new home purchase. Consequently, the demand for high-quality architectural images requires a more more subtle touch.

I've tried most of the software that's available for HDR -- at least on a Mac, which is quite a lot -- and have found that most of them are not ideal for the type of images I need to produce. I generally avoid the packages that require tone mapping to develop an image. I have no doubt this technique can produce fine results, but for me it doesn't seem suited to producing well-adjusted results quickly and consistently. Every effort seems to require going down a rabbit hole of possibilities to find the optimum settings, and ultimately still doesn't "look" natural enough. More often I get better results using an exposure blending technique based on Enfuse software. (Purists will point out that due to certain data attributes this is not technically "true" HDR, but that's a nominal distinction. Both require the digital merging of bracketed multiple exposures for the same intent.) More specifically, I use a software plug-in to create Enfused images from within my Adobe Lightroom workflow. This excellent plug-in, called LR/Enfuse, was developed by photographer Timothy Armes. With it I can much more quickly develop a pleasing starting point for a blended image, and will often need to perform only minimal adjustments (in Lightroom or Photoshop) for the final. But the main advantage for me is achieving an expanded dynamic range in a more natural looking image.

New software for creating better, easier HDR images continues to emerge and I will no doubt continue to try them in my quest for the "perfect" solution. But for now, this approach seems to be the optimum technique and workflow for producing the images I want.

Photo: C. J. Ludewig Building, Built 1929, New Braunfels, TX

Canon G11 and Lightroom 2.62 RC

One important feature of the Canon G11 is that it supports both JPEG and RAW formats. I've been getting excellent JPEG files directly from the camera, but with RAW files I've had quality concerns with Adobe Lightroom conversions. Canon's own DPP software does a very good job with these files -- as you would expect from the same manufacturer -- but I'm not a great fan of DPP and generally prefer to use Lightroom. Although the most recent version did technically work with G11 RAW files, its profile for this model did not produce very good default results, at least for me. I also tried the Lightroom 3 beta, which supposedly had better support for the G11, but I found those results to be not much better. Fortunately, this problem now seems to be solved with the latest LR 2.6 RC upgrade. This version now appears to have an improved standard profile for the G11 and the default conversions are much more to my liking.

This delay was inevitable since no software can be pre-programmed to anticipate the characteristics of all future camera sensors. New camera models generally require significant testing to create optimal profiles for converting their RAW data. These profiles are essential for producing good RAW conversions and I'm very glad Adobe is committed to regularly updating LR to provide them.

Up until now I've been capturing both JPEG and RAW files with the G11 to have more development options. But I don't like having to manage dual copies of images. My early experience with this update is that the new conversions are good enough that I may resume shooting RAW only. At the very least, I'm much more comfortable with the default conversions that LR now produces when working with G11 RAW files.

LR3 Beta

On the subject of the Lightroom 3 beta, I must profess that so far I find it to be a bit underwhelming. It does have some modest improvements, but mostly for things it already did fairly well. Performance enhancements are always nice, but with major increment upgrades I've come to expect more. For Lightroom what I want most is the addition of features that keep me from having to LEAVE the application in order to do important edits that many of us regularly perform. For instance, things like lens distortion and perspective correction, HDR, image blending, and masking -- all done non-destructively -- are features I could really get excited about.

Yes, I know I can currently export, edit in other applications, and then return to LR, but the more I can do without round-tripping, the better. Not only would this improve my productivity, but it would eliminate the need to create and add alternative versions of files to the library.

To be fair, this is only a beta version of LR3 and Adobe has indicated that there is much more to come. I do anticipate other important enhancements in future releases. I just hope they aren't so protective of Photoshop market share that they withhold some of its most important features for photographers from Lightroom. If Adobe is serious about Lightroom becoming THE essential workflow software for photographers, then they need to expand its capabilities seriously too.