Part 1: Comparing the imaging quality of digital cameras and traditional cameras
First, the comparison of imaging quality
I have always wanted to compare the imaging quality of digital cameras with those of traditional cameras. I did not make them because of the limited conditions. I recently borrowed Epson PC850Z\PC300OZ, Kodak DC290\DC48OO, Nikon Coolpix95O\99O and Nikon D1 digital cameras from major companies. Among them, Epson PC3000Z, Kodak DC480O and Nikon COOLPIX99O are all above. The latest products of several companies, while the Nikon D1 camera is a high-end professional camera worth nearly 70,000 yuan. I compared them with the traditional 135 camera.
In the process of downloading photos, I encountered a problem. Many digital camera drivers not only take up space, but once installed, they cannot be deleted. I would like to delete all kinds of system software that must be re-installed, otherwise the shutdown time is as long as several. Ten minutes (my computer's memory is 128M, and the hard disk is 20G). In order to save the space of the computer and to unify the downloading input effect, I used the Feiya CFU-3020d card reader to read the data on the memory card. It is clever that all the three types of cameras are used. CF card. One of the possible consequences of not using a driver is the problem of color reproduction because the individual color management systems of the individual cameras are different. However, it does not affect the sharpness of the image. I
At the same time, the measured dpi values ​​of their resolutions are published, which is why I have to explain my resolution calculation method repeatedly in the previous part 3.
Second, the evaluation results of imaging quality
Table 1 shows the main technical specifications of each brand of digital camera I measured. Line 7 of Table 1 shows the amplification resolution value of each digital camera I measured, and the eighth row is based on their resolution values. And according to their respective frame ratios to calculate their actual shooting pixels. The reader can compare these images with these values.
Surprisingly, their real-shot resolutions almost all approached or even reached their nominal resolution, and the Nikon D1's real-shot resolution exceeds its nominal resolution. This is greatly beyond my expectation. I originally thought that the difference in the actual resolution of the lens and the size of the CCD
The real-time resolution of the first seven cameras will be much lower than the nominal resolution (just as the resolution of conventional images is much lower than its film resolution), and only the resolution of the Nikon D1 can be close to or The nominal resolution (its CCD area is large) is reached, but the results of the test negate my idea.
Table 1: The zoom resolution and the number of real shots of the digital camera I measured:
Digital Camera PC850Z PC3000Z DC290 DC4800 CP950 CP990 D1
Effective number of pixels 1.92 million (1.9M) 3.15 million (3.2M) 2.15 million (2.2M) 3.11 million (3.1M) 1.92 million (1.9M) 3.15 million (3.2M) 2.62 million (2.6M)
Image Resolution 1600×1200 2048×1536 1792×1200 2160×1440 1600×1200 2048×1536 2000×1312
The number of interpolation pixels is 2.96 million (3.0M) 4.84 million (4.8M) 3.36 million (3.4M) - - - -
Interpolation Resolution 1984×1488 2544×1904 2240×1500 -- -- -- --
CCD size 1/2 inch 1/1.8 inch 1/2 inch 1/1.8 inch 1/2 inch 1/1.8 inch 23.7×15.6 mm
Zoom resolution 158dpi 202dpi 173dpi 205dpi 155dpi 208dpi 212dpi
Real shots of 1.88 million pixels (1.9M) 3.06 million (3.1M) 2 million (2.0M) 2.8 million (2.8M) 1.81 million (1.8M) 3.24 million (3.2M) 3 million (3.0M)
At the same time, these data also made me more convinced of the accuracy of my test method. Because: The result of various test errors can only make the resolution lower, and the high-resolution result shows that the system can definitely reach this resolution value. But I still can not explain why the Nikon D1's real shot resolution actually exceeds its nominal resolution.
With the above test data. Make me speak better. Now look at the clarity of traditional photographs. For accurate comparisons. I used a Nikon camera with a 105mm macro lens and a Hasselblad camera plus an 80mm standard lens. I used a Fll aperture and used 135 and 120 films, respectively shot with a negative film and a 135 film with Kodak and Fuji. In film shooting, a total of 8 rolls were taken. This 8 roll film is also for later scanning. These include two volumes of scenery, architecture and portraits.
I used a microscope to pick up two negatives (135 and 120) and zoomed them up to 10 inches, where I enlarged the resolution in the two negatives. Test results are shown in Table 2
Table 2: The zoom resolution and the number of real shots of the two conventional cameras I have measured:
Traditional film type 135 negative (Kodak) 120 negative (6x4.5, Kodak)
Zoom resolution 343dpi 425dip
The number of real shots is 7.48 million (7.8M) and 13.55 million (13.6M).
Third, the objective evaluation of imaging quality
For several years, I carefully compared various digital images and compared them with traditional images. I have found that although their theoretical resolution is compared to the number of pixels, they differ depending on the structure of the image. It is not possible to use the previous formula to perform a simple conversion, but their actual visual resolution can be compared using the previous formula. For example, the resolution of conventional photography (optimal aperture, still shooting) is about 100 lp/mm. Equivalent to 5,000 dpi. The female Q fruit is calculated with a 135-future of 24 x 36mm, which can reach 34,56O,000 pixels. That is, 34.56 million (34.5 M) pixels; calculated in three sizes of 120 × 6 × 4.5 (55 × 42 mm), 6 × 7 (55 × 65 mm) and 6 × 9 (55 × 85 mm), The total number of pixels was 92.4 million (92.4 M), 14,300 (143 M) million, and 18,700 (187 M) million, respectively. However, the actual hand-held shooting resolution is due to people, due to the camera, due to the lens, due to aperture, due to film, due to washing conditions, due to different shooting techniques, to 50lp/mm should be considered excellent, to 4Olp/mm should be considered good, and A pass of 30 lp/mm should be considered. Their corresponding digital image resolutions are: 2,500dpi, 2,00Odpi, and 1,500dpi, respectively, while the corresponding number of 135 format pixels is 8,640,000 (8.6M), 5,530,000 (5. 5M) and 3,110,000 (3.1M); see Table 3 for the number of pixels corresponding to the 120-frame size of 6x4.5, 6x7, and 6x9.
So we can roughly define the standard of image quality by 135 frames and 120 frames:
135 frame: high-definition 10M or more pixels; mid-range definition 5-8M pixels; low-resolution 3M pixels below.
120 frames: high-definition 20M pixels or more; mid-range resolution 15M pixels; low-resolution 10M pixels or less.
The amplification resolutions corresponding to the above data are:
135 frames: high definition above 40dpi; mid-range definition 280-360dpi; low definition below 200dpi.
120 frames: high definition above 56Odpi; mid-range definition 48Odpi; low-resolution below 400dpi. (To be continued)