09-26-2015, 07:30 AM
At the end of the day, the excellent/very good/... approach is (almost) nothing else but the percentage figures ...
The charts are divided into 5 segments. The top segment is excellent, the 2nd one from the top is very good, etc. so this can be directly translated to a percentage range. Thus it's merely a matter of granularity and a coarser granularity is simply more appropriate.
I have strong doubts that you can visually distinguish a difference of -say- roughly 200 LW/PH or 'half a mark' in real life images. Without a sturdy tripod (& RAW processing) you'll also not be able to recreate lab quality anyway. This doesn't mean that the findings are irrelevant - all lenses are affected by real life impacts to a very similar degree so a better lens in the lab will remains a better lens in the field. However, fact is that the 4 digit LW/PH or 3 digit percentage values aren't helpful regarding "the point".
That being said I can fully understand that it somehow "feels" easier to imagine that a sensor has say 4000px (vertically) and the lens can deliver 3325px in the center ... although I have doubts that you can imagine how say 1500px would really look like in your images. But now we got the risk of invalid cross comparisons.
So at the end of the day there's simply no win-win in sight. :lol:
The charts are divided into 5 segments. The top segment is excellent, the 2nd one from the top is very good, etc. so this can be directly translated to a percentage range. Thus it's merely a matter of granularity and a coarser granularity is simply more appropriate.
I have strong doubts that you can visually distinguish a difference of -say- roughly 200 LW/PH or 'half a mark' in real life images. Without a sturdy tripod (& RAW processing) you'll also not be able to recreate lab quality anyway. This doesn't mean that the findings are irrelevant - all lenses are affected by real life impacts to a very similar degree so a better lens in the lab will remains a better lens in the field. However, fact is that the 4 digit LW/PH or 3 digit percentage values aren't helpful regarding "the point".
That being said I can fully understand that it somehow "feels" easier to imagine that a sensor has say 4000px (vertically) and the lens can deliver 3325px in the center ... although I have doubts that you can imagine how say 1500px would really look like in your images. But now we got the risk of invalid cross comparisons.
So at the end of the day there's simply no win-win in sight. :lol: