Originally posted by npc Like if I take the same image with k-5 iis and k-3 and up-size the 16 mpix image to 24 - will I be able to really tell the difference?
Originally posted by stevebrot I have my friend's K-50 for a couple of days and might make a couple comparison images between it and the K-3 with the FA 77/1.8 Limited.
Ok, because inquiring minds want to know, here is a head-to-head comparison of the K-50 with the K-3.
Sorry, I don't have access to a K5-IIs.
Methods:- Lens: Pentax FA 77/1.8 Limited
- Aperture: f/5.6
- Shutter: 1/6s, with two second timer, SR OFF
- On tripod (Giottos 9360)
- CDAF in live view at same focus point (lower eyelid), best of three focus attempts
- Calculated DOF is ~1.2" at 36" distance
- Anti-alias filter emulation on K-3 turned OFF
- RAW
- Default import into Lightroom
- Normalized white balance and curves
- +15 clarity, +15 vibrance (my standard practice)
- No additional sharpening
- No len correction
- K-3 RAW exported to full-resolution TIFF
- K-50 RAW exported to 4000x6016 TIFF (upsized to 24 Mpx)
- Full resolution crops sampled from the same area of each 24 Mpx image and exported to JPEG with no resize, no compression
The subject:
The doll is about 20" tall and was photographed from 36" away.
K-3 converted from full resolution TIFF
Comparison at point of focus (lower eyelid):
Note that I need to dust more often.
K-3 K-50
For fairness, here is the non-upsampled section from the original K-50 DNG...
Comparison at collar:
Note hair at center.
K-3 K-50 Links to full-sized files K-3 Full Size JPEG https://farm9.staticflickr.com/8635/16657673815_71f0988a28_o.jpg K-50 Upsized JPEG (4000x6016) https://farm9.staticflickr.com/8571/16631762816_e2459a050b_o.jpg
Should anyone want the original DNG files, let me know and I will provide a drop-box URL for you.
Discussion:
As you can see, both cameras captured an admirable amount of detail. The K-3 did somewhat better overall with the difference being best appreciated when comparing various regions from the TIFFs. I would suggest downloading both and using a tool such as the Lightroom Library to do section-by-section comparison.
Some would argue that I should have used other than Lightroom's built in upsize algorithm. I might do that some time, but not today. Others might notice that the two photos don't look quite the same. I would respond that they were taken with different cameras at different bit-depths with different image processors. Someone else might notice that the K-3 TIFF file is several hundred KB larger than the K-50 TIFF. I may be running low on electrons.
Are the differences traceable to the K-3 not having an AA filter and the emulation being turned off for this test? Yes, that is possible.
All the tech aside, this test still leaves the question open as to whether the K-3 sensor out-resolves my copy of the FA 77 Limited. That would require an image taken with a higher resolution camera like the future FF Pentax or perhaps an adapted to a Sony A7R. What we can deduce is that the FA 77/1.8 Limited is not out-resolved by the sensor in the K-50.
Steve