2001: Digital Imaging Leadership

Carly's "growth at all costs" style of leadership alienated HP employees and many veteran HP executives. In February, 2001, Carolyn Ticknor retired from HP, and VJ became president of the Imaging and Printing Systems business.

VJ's was passionate about inkjet technology and about digital imaging. As Lee's new boss, it made sense that we would be asked to put some focus on digital imaging now that the digital internet workflow and commercial printing initiatives had become mainstream IPS programs.

George Lynch had been on VJ's technology staff, and he joined Lee's staff in 2001. George and I always respected each other, and we worked well together and divided the overall Strategic Technology objectives between us. A few years later, George would end up reporting to me before his retirement. Here is how Lee described things:

George Lynch has joined Evan Smouse in recently expanded roles as IPS Strategic Technology Directors. They will drive the development of consistent technology strategies and roadmaps across IPS to reinforce HP’s leadership in core businesses and enable rapid acceleration of growth in new markets. They are leading teams of senior technologists who represent IPS on cross-HP teams working to create transformational solutions. The IPS Strategic Technology function will provide leadership to drive key industry standards and will help to shape strategic relationships with external business partners and with universities. Their work will be tightly integrated with and will reinforce the goals of the IPS Business Development team.

 For quite awhile, my team had been leading the HP color strategy and had been working to define and evangelize "ultimate image quality." So I took the lead in this area for VJ and Lee. It fit well with my earlier work to define the internet imaging model. Applied to the field of digital imaging, this model meant that the HP plan would create an open architecture that to link together personal, online and professional services so that digital imaging becomes:  

• Pervasive – web of connectivity and sharing – open standards that allow images to be created, distributed, accessed, and shared independent of specific brands and models of devices and services - computers, television, printers, internet, mobile devices, cameras, retail photo services, ...
• Predictable – consistent results based on output-ready interchange spaces for straightforward implementation and easy verification by developers 
• Simple – based on implicit color management that just works by virtue of consistent, well-defined color information interchange spaces instead of complex profiles and settings
• High quality – “anywhere, exactly as intended” - user-preferred color reproduction for consumers, more accuracy for professionals, "device-best" color rendering superior to analog systems

HP was steadfast in the belief and insistence that imaging workflows should be based on an open, unified, Web-centric imaging architecture for scalability from low end imaging to professional photofinishing.

One of the main digital imaging competitors, Epson Seiko, decided to take a different approach in February, 2001, by introducing Epson's Print Image Matching system (PIM). This proprietary approach was the antithesis of our model, and it provided a specific target for my own digital imaging leadership campaign for the next year or so.

In a nutshell, Epson worked with some camera manufacturers (Casio, Konica, Olympus, Ricoh, Sony, and Toshiba) to agree on a set of proprietary information to link digital camera images to Epson printer settings. Those camera tags would place proprietary information in the EXIF header of JPG images. Epson then could optimize sharpness, color balance and brightness to match camera-selected preferences. But the PIM-tagged images had no advantage when printed on non-PIM printers (including ones from Epson) or when shared and viewed on other devices. It was an attempt to gain market attention. PIM also required proprietary software for image editing. It didn't work on the imaging software like Photoshop that many people used.

It would have been dangerous if Epson had been able to have their proprietary approach adopted by the JPG standards group. HP and Kodak were colleagues in some open standards organizations and agreed that we should work together to drive the industry in a direction that supported our mutual interests. I modified the Internet Imaging Model pitch that I had made for Microsoft, and VJ presented it to Kodak executives. They bought into the idea immediately. The problem was that we needed resources to influence companies and the JCIA standards organizations where EXIF was controlled in Japan. VJ provided the funds, we enlisted help from HP Japan. We also hired Ramon Garrido as full-time program manager for Digital Imaging Leadership. Ramon and I headed to Tokyo in December after the groundwork had been laid there.

We would eventually gain worldwide support for our proposal to revise EXIF 2.1 to version EXIF 3.0 with (geek alert!) fully rendered sRGB/sYCC color encoding, camera mode metadata (portrait, night, landscape, other camera settings), and capture/enhancement flags for white balance, flash mode, exposure mode, and certain effects. By standardizing these essential elements, digital imaging would flourish as device makers, software companies, and imaging services providers adopted common technology for data interchange. There would be much more work in standards ahead to get critical mass implementation of our Internet Imaging Model concepts.

On September 3, 2001, HP announced the merger with Compaq. That event would affect me soon.

Next: 2002: CTO Multitasking