NCTech iSTAR Fusion -Conclusion / Reflection
(Based upon the 2 case studies and familiarization period / tests)
Overall, I would consider the iSTAR FUSION as (as tested):
- A budget conscious solution
- Convenient and fast to setup
- Intuitive setup and control via the build in touch screen
- Generating variable image results – from very good to quite poor (native (default) stitching software seems unreliable).
- Faithfulness of color reproduction seems erratic
- Images suffer from significant chromatic aberration
- Images lack the crispness / fidelity one would expect a 50MP capture would offer. Panoramic (zoomed out) views look great, close up (zoomed in) look soft, fuzzy and disappointing.
- Compromised by very unstable software (camera) – experienced software crashes and erratic battery status.
- Can achieve very good ‘post processing’ results of digital negative via native NCTech Immersive Studio software. The iSTAR uses a process called Exposure Fusion to balance out the exposure data it captures during HDR modes. This is different to the ‘tone-mapping’ process employed by the Spheron system.
- Offering limited, default output options, as third party solutions are required to facilitate higher quality image output (standard panorama image format is an 8 bit JPEG).
- Demonstrated dubious battery life (compounded by unreliable battery status indication during tests). With a full charge, battery dropped to approx 50% after 3 captures (wifi NOT enabled in any tests)
- Less susceptible to luminance changes in scene / capture area – shutter capture is fast enough to avoid atmospheric changes generally (eg cloud movement).
- Susceptible to lens flare / light corruption of image.
- Offers the flexibility of a compact camera, whilst lacking the integrity and output of a DSLR (as an analogy) i.e. offers the versatility of a ‘compact camera’ (small, compact and quick setup), but also the fallibility of such – i.e. does not always deliver a perfect capture.
The system I tested was (one can consider) the flagship product for the iSTAR range; the iSTAR FUSION offering the best in HDR capture capabilities of the range – essential if its to be compared to the 32-bit Spheron SceneCam.
Setup is very fast, with one merely attaching system to a tripod / resting on a flat surface.
Camera control is generally very good via the on board touchscreen, with pretty intuitive GUI. The leveling screen is very useful / thought out, providing rapid ‘leveling’ of the system. After that, there aren’t too many options to choose:
- Delay (tested)
- Walkaround (tested)
- Time Lapse (tested)
- Live View
- Lighting Type (prevalent in scene)
- ISO Speed
- HDR Mode
- Review Modes
Results were very variable overall, with an almost random probability of a totally faithful capture being achieved. The biggest issue was its failure to accurately stitch the images together, either creating a real mess at the zenith or boundary stitching issues between lenses. Experienced ‘shearing’ of key geometric features in a number of shots.
Exposure control of processed image was very good on the whole, producing a very good balanced exposure for complex scenes like the aircraft interior examples.
Speed of capture was good, although nothing like the claimed ‘5 to 20 seconds‘ quoted on NCTech website – several minutes was the usual norm* (achieved in the field). Additionally, the camera is essentially disabled whilst it processes the data immediately after capture, so reckon on 4-10 minutes between captures / pressing the ‘go’ button.
*The majority of my 70+ scans were taken using the 9 exposure / HDR Pro setting.
Further investigation on NCTech website confirms:
HDR 3 takes around 10-20 secs to capture but HDR Auto can take a lot longer, up to 3-5 mins in some situations.
My experience (using HDR Pro mode) suggested similar times as for the HDR Auto statement above – that is several minutes.
Battery life seemed an issue with the unit I tested – after starting with 100% charge, the system showed approx 50% battery remaining after only 3 HDR Pro captures. The system reverted to full charge status on a later capture, then dropped to 50% again. Software glitch? Unclear. NCTech quote Approx. 6 to 8 hours battery life in use…
In very low light conditions, the iSTAR system does not have the option of attaching an additional light system, which would be an issue in some situations. Clearly, lighting could be brought in, but these would be either sources of further contaminant (at say a crime scene) or provide obstacles / obstructions to viewing pleasure.
The footprint is also a significant issue – with a huge area being omitted below the camera during capture. If one considers this for real estate documentation, crime scene audit or point cloud colorization, this will be a problem for some. Some users add the mirror ball effect to hide this, others lock the camera orientation to prevent users looking down (towards nadir area).
Positioning the camera closer to the ground will resolve this, but that may not create the perspective / line of sight that is desirable or truly representative of the environment. My preference is ‘first-person’ perspective on environments, so this I found an issue.