antoniojr wrote:I also find my XL Speedbooster really soft when is paired to my Sigma 35 art. 1.4. This lens is not soft by any mean. I don't know if its a general problem with the BMPCC 4k and this SP or if it's my metabones copy.
If anyone have experience with this combo and can share their experience, it would be really helpful.
This is an interesting topic and you can get a lot of people reporting results based on "feeling". Here are some things to take into account:
- The Sigma 35 art is a lens designed for a "full frame" sensor. When using it in smaller sensors, the outer part of the lens misses the sensor and therefore the result departs from its intended use. Even with the XL's generous focal reducing you are still at 1.2x crop, which while it sounds "almost there", it still means that about 15~20% of the lens is cropped. While that area of the lens is usually less performing, especially on wide apertures the result is more and more compromised, because of that missing part of the glass.
- No matter what work the Speedbooster is doing, everything still ends to a smaller sensor. It's only natural that the result will be compromised compared to the large sensor. Especially if comparing against photos taken with that lens on a full frame camera - they are totally irrelevant to how this lens should perform in this MFT sensor sized cinema camera.
- Testing for things like sharpness, requires very careful design and implementation of such tests and can't be just judged by the eye or feeling. A common and newbie error many do in such tests, is to compare the same lens with and without focal reducer while maintaining the position of the camera. Obviously the focal reducer loses in that case because those testers just zoom at the same area in post and compare. The FOV with the focal reducer is larger and therefore that detail smaller and naturally less sharp. A proper test would require moving the camera to match the exact same FOV. In that case the focal reducer will most certainly win, with the main reason being what I've explained above: glass is designed to resolve a resolution; magnifying/cropping into the glass simply makes that resolution lower. The statement that focal reducers compromise the image and thus you should "not have another piece of glass between your lens and your sensor" is a myth and easily provable with the above test done properly - always talking about the high quality Metabones XL/Ultra Speedboosters of course, cheap focal reducers or older Metabones models could indeed at wide open apertures compromise the sharpness to a greater degree than the benefits gained by using more of the glass.
- In combination to the above error, another newbie mistake is to compare the results of the lens at the same aperture with and without a focal reducer. The correct way would be to compare the lens after adding that 1+1/3 stops that are gained by the focal reducer. For example f2 without a focal reducer then f3.2 with the focal reducer on - which will absolutely and most definitely be sharper on the focal reducer every single time. Except that by doing that you actually get the same DOF for correct comparisons, you should think that 1+1/3 stops you are gaining with the XL as a "gift" - that extra f0.9~f1.4 range (f1.4~f2.2 on your lens) is a gift. Close down your lenses to f1.4+(1+1/3 stops) in order to reach the same FOV as f1.4 without focal reducer, and compare again. Opening the lens at f1.4 with a focal reducer on and comparing the results with the lens at f1.4 without a focal reducer (on any camera) is absurd and inaccurate.
- Yet another mistake is to compare the results vs a full frame camera. Of course the lens won't have the same performance for the reasons I mentioned above; both because of the lens and the sensor. What is the lens expectation on a full frame camera has nothing to do with how that lens performs on crop sensors.
- Finally, another terrible test is comparing a lens between two different cameras that use different process / codecs etc. Sure that can determine how the lens looks on each camera but says nothing about the lens sharpness itself - unless comparison is against another lens.
- If you are interested to see if you can get sharper results with this camera you should try other lenses and compare them with what you are getting with the art 35+XL. For example a ~22mm MFT lens or ~22mm full frame lens without a focal reducer.
- Finally the Speebooster XL was designed mostly for Panasonic GH cameras. Perhaps the performance could be refined for this camera and that's what the new Pocket4K exclusive Speedbooster Metabones announced could be all about. That said, in my experience, the difference is very little and technical and for most use cases won't make any difference. The points I mentioned above, including the logical and practical errors in testing, are what has the biggest effect.
Fri Dec 14, 2018 2:33 am
