Hi,
In a ideal wold, sensor resolution would match lens resolution. Image detail that the lens delivers but the lens cannot resolve will be turned into artefacts, that is fake detail.
So for proper rendition of an image there is a need of high resolution. How high? High enough to match the resolution of the lens. A good lens stopped down to f/8 will resolve something like 200 lp/mm, albeit at very low contrast. An excellent lens at f/5.6 will perhaps reach 400 lp/mm. So, in order to produce a correct image, we would need at last 400 lp/mm resolution. That would 1.25 microns. That would be around 550 MP on full frame.
On the other hand, fine detail contrast at 400 lp/mm will be very low, so we may get around with less resolution, say 2.5 microns. That would correspond to around 138 MP.
Now, the images here used to induce hateful comments from a few "besserwissers" on these forums. But, they are a good illustration of the issue at hand. Both images are shot with a 150 mm lens, decent lenses but nothing exceptional. One is the Sonnar 150/4 the other is Sony Alpha 70-400/4-5.6G zoom at 150 mm.
6.8 micron | 3.8 Micron |
18.6 MP (on full frame 135) | 54 MP (on full frame 135) |
| |
Another point may be that we can buy expensive lenses for a low MP camera. We get somewhat sharper images with a lot of artefacts. Or we can use halfway decent lenses with a high resolving sensor, and get better results with less artefacts.
Optimally, we can combine a very good lens with a very good sensor, combined with a well designed optical low pass filter and get pretty optimal results ant optimal cost.
In the future, we may have high resolution sensors that give high MP images, but there may be an option to downsize them to lower resolution in camera firmware.
This image shot with four different lenses shows a lot of artefacts:
Full size:
http://echophoto.dnsalias.net/ekr/Articles/Shoots/BernardSamples/SmallTarget/Center.pngThe pictures below show the top left corner. Note that the top left image shows banding type artefacts. The reason for it is that the lens in question (Hasselblad Zeiss Planar 100/3.5) has severe astigmatism at short ranges. According to it's make, Zeiss, it is a lens intended for long distance photography.
The 120/8 is a macro lens that excels at this distance, causing a lot of artefacts. The 180/4 is optimised to work over a wide range of focusing distances according to Zeiss an yields a lot of artefacts. The Planar 80/2.8 looses some sharpness to the corners, and has less obvious artefacts.
Full image:
http://echophoto.dnsalias.net/ekr/Articles/Shoots/BernardSamples/SmallTarget/UpperLeftCorner.pngYeah, showing those images may cause a lot of hateful comments from those who never made this kind of experiment, I am sorry to hurt their noble feelings.
Why folks don't see these things in everyday work?
- Shooting stopped down to f/16 so diffraction reduces contrast enough to suppress these artefacts
- Using an OLP filtered camera that suppresses the colour artefacts
- Shooting objects where artefacts are less then obvious
- Not achieving optimal sharpness due to technical issues. Non tripod, camera vibrations, non optimal focus small apertures etc... Aliasing effects are only seen when resolution of the optical system exceeds the resolution of the sensor.
Best regards
Erik