Part 3 of the PowerShot G1 X Mark II Developer's Interview will feature "Image Sensor and Image Processor". Let's learn about the new 1.5" image sensor and the new DIGIC 6 image processor. The PowerShot G1 X Mark II is the flagship model of the PowerShot G series which is packed with Canon's latest cutting-edge technology. In this series article, the developers of this model shares with us the challenges they faced when bringing this model to completion.
Norio Negishi, ,
Coordination of intelligence to draw out the optic performance.
― How do the new lens, 1.5 type sensor *, and DIGIC 6 work together?
Takahashi Light information collected by the lens is imprinted on the sensor, and after the sensor converts it to digital information, the imaging processor carries out image processing which is the main working of a digital camera. Therefore, no matter how advanced the imaging processor's image processing is, if the "material" received via the optics and sensor is not good, it is not possible to create beautiful image quality. The three parts consisting of the lens, sensor, and imaging processor are extremely important for image creation. What this means is that the lens this time is very high spec, and as the sensor achieves virtually the same cell pitch as that of SLRs, information of extremely high quality reaches the imaging engine (DIGIC). With DIGIC 6, settings were adjusted to bring out lens and sensor performance to the fullest.
Negishi Looking at it from the aspect of coordination during development, there are significant merits from using a Canon brand sensor and DIGIC. Of course there is a tight exchange of information and communication is possible, and as it is within Canon this includes sharing technical information, so it is possible to provide the best sensor designed to match the lens. With the sensor as well, the optimal tuning was carried out for the lens and DIGIC 6.
*Total pixels : Approx. 15.0 Megapixels Effective pixels : Approx. 12.8 Megapixels(3:2)
Takahashi The same can be said for noise reduction as well. For noise reduction, the point is to distinguish between noise and the texture of the subject, and how to remove the noise. We believe noise reduction that eliminates texture reduces image quality, so we try to remove noise while preserving the subject’s texture as much as possible. What is important here is that operation of the lens, sensor, and imaging processor is integrated. The lens accurately imprints the subject’s texture as texture, and noise occurring around the sensor is removed as well as possible by the sensor, making it possible to remove noise with more effective DIGIC noise reduction processing. Even when information enters DIGIC that is of a level that is not possible to distinguish whether the signal is texture or noise, this is because it cannot be handled by the imaging processor alone. The lens and sensor on the G1 X Mark II are both high performance, and the information input into DIGIC is extremely high quality. Because of this, it was possible to make uncompromising pursuit of settings that remove only noise.
Negishi Noise reduction settings on the G1 X Mark II are extremely detailed. Noise reduction consists of tuning by performing minute setting changes according to shooting conditions, and this camera is set with so many conditions that it cannot be compared to other cameras. With just a tripod, this performance can be utilized to easily shoot beautiful starry skies with the Star Mode, and you can shoot extremely clear photos.
― Is there any difference in development compared to SLR or mirrorless cameras?
Negishi As the G1 X Mark II is a compact digital camera with an integrated lens, the lens information is determined. As a result, optimal correction that matches the lens can be performed. As the lens position varies according to the zoom, the sensor will change shading etc. to match, and as the lens is integrated, optimization is easy. This is a merit not found in cameras with interchangeable lenses. At SLR size, if you can imagine a lens of f/2.0 (W) to f/3.9 (T), I think you will see that it would not be possible at this size.
Takahashi DIGIC is also the same. Cameras with an integrated lens can easily grasp lens characteristics and the lens condition when shooting, and they can perform optimum processing according to the various characteristics and situations. Because it is a camera with an integrated lens, this makes detailed image quality optimization settings possible.
The outermost 4:3 area on existing models is recorded as CR2
When the recording aspects are 3:2 and 16:9 on G1 X Mark II, the 3:2 area is recorded as CR2
When the recording aspects are 4:3, 1;1, and 4:5 on the G1 X Mark II, the 4:3 area is recorded as CR2
― RAW recording at 3:2 is now possible, was this with SLR users in mind?
Negishi Yes. It was developed with the needs of people who usually use SLR cameras in mind. With multi-aspect up until now, 4:3 images were trimmed to create 3:2 images. This time however, the 1.5 type sensor* is used effectively, making it possible to shoot photos that take advantage of a 35mm film equivalent 24mm wide angle lens for both 4:3 and 3:2 aspects.
*Total pixels : Approx. 15.0 Megapixels Effective pixels : Approx. 12.8 Megapixels(3:2)
Takahashi This specification was quite difficult technically, and in terms of work load. This is because it was like designing for two cameras with 4:3 and 3:2. Compared to a normal product, as the design work load would be significantly greater, some of the developers had their doubts at the beginning, but as they believed that there would be significant merit for users, they put more effort into the development.
Existing model: 4:3(Left), Existing model 3:2(Right)
A: The image of 3:2 for existing model is the same data as the one which is cut both top and bottom of the image of 4:3.
G1 X MarkII 4:3 (Left), G1 X MarkII 3:2 (Right)
B: Record the areas to the left and right that couldn't be seen in 4:3 images where the top and bottom are slightly cut off.
― Have the AF speed and continuous shooting performance improved in conjunction with the faster processing speed?
Negishi The frame rate of the sensor that imports the data for AF was increased, so the reading speed is faster. The AF speed has been an issue since the G1 X generation, so making it faster was a significant goal. In combination with the improvements accomplished with the optics, almost twice the speed of the G1 X was achieved.
Takahashi: Continuous shooting performance is also improved. On the previous G1 X model, high-speed burst HQ was approximately 4.5 fps, but as a result of revising the sequence related to continuous shooting this time, this spec now has a speed faster than 4.5 fps (TBD), and continuous shooting until the media reaches full capacity is possible. By improving the processing speed and revising the continuous shooting sequence, photo opportunities are better with the G1 X Mark II.
― The DIGIC 6 imaging processor is good with movies, but is movie performance improved on the G1 X Mark II?
Takahashi As I mentioned previously, noise reduction has improved dramatically for movies as well, increasing the range of shooting in low light scenes. Processing is carried out based on approximately 9x as much information as DIGIC 5, so texture and noise isolation precision, and noise reduction performance are both improved. Also, as the f/number is even brighter, utilizing the background blurring expression achieved with the G1 X generation, and shooting made possible in even darker areas, etc., the width of expression and the possible range of shooting has increased significantly.
The focus process was rationalized and streamlined, achieving a great reduction in the shooting time lag.
― New expressive functions such as Creative Shot and handheld HDR were added.
Takahashi Creative Shot that was first included on the PowerShot N shoots three photos with one press of the shutter button, then after detecting the shooting scene it automatically adds suitable effects such as framing, trimming, and color filters, making six images. HDR shoots multiple photos and creates one image by superimposing them over a wide dynamic range, so this function requires a significantly heavy amount of image processing. Similar processing was possible with DIGIC 5, but with DIGIC 6 the shooting time is reduced, and processing time is faster, so it is a more user friendly feature. The fact that five types of effects equivalent to the EOS 5D Mark II can be added means it should offer value that will please SLR users.
In particular, by drawing new awareness to Creative Shot, this function is very popular with advanced users using the PowerShot N. I would like to see people enjoy using it to encounter a new world of photos that they have not shot up until now.
― Do you have a message for the users?
Negishi At Canon we do not carry out development based on relative standards such as, “Other companies have this feature, so let’s aim this far”. Canon goes after absolute standards, and following through to achieving them is our characteristic. As the flagship model compact digital camera offered by Canon, the G1 X Mark II was developed to the point where you could say there is no more room to improve the sensor and imaging processor. As a result, there is virtually no weak points with the current technology, and I believe this camera has achieved a good Canon-like balance. I would really like you to enjoy using it.
Combines two images processed with color filters
Lens max. aperture: f/2.0 (W) - f/3.9 (T)
Lens focal length: 24mm (W) - 120mm (T)*1
Min. focusing distance: 5cm
Image sensor type: 1.5" COMS sensor *2
Image processor: DIGIC 6
*1 35 mm film equivalent
*2 Total pixels : Approx. 15.0 Megapixels, effective pixels : Approx. 12.8 Megapixels(3:2)