fotofanatic
Member
sounds so amazing. and i so want it to be amazing. it's really good, but the panasonic 20mm f1.7 pancake that ends up being a 40mm kicks ass over my 28mm elmarit (which is a 56mm), my 35mm cron (which is a 70mm) and my 50mm cron (which is a 100mm). so sad. one really great thing though is my 90 elmarit on the lumix. which acts like a 180 and super fast. that is a dream.
JoeV
Thin Air, Bright Sun
This is somewhat similar to my experience with adapting manual Minolta MD glass. The real value is in fast, compact lenses that end up being great for indoor, available light shooting at moderate telephoto lengths. Like the 58/1.2 that ends up being an equivalent 116/1.2. Great for venues like nightclub bands where you need fast glass and a bit of telephoto length to get closeups of the performers.
For the wide stuff, the Lumix 20-1.7 is the bee's knees.
~Joe
For the wide stuff, the Lumix 20-1.7 is the bee's knees.
~Joe
biomed
Veteran
Wide angle lenses designed for film cameras seem to share traits of poor performance when used on digital cameras. A lot of digital cameras require telecentric (correct term?) lenses. The image projected by wide angle lenses onto the sensor strike the sensor at such an angle that it softens the image at the corners of the frame. This is not as critical with film cameras as the lenses were designed to work with that media. Perhaps another member could give a more technical explanation. In my experience lenses 50mm or longer work best on my GF1. The CV 50/1.5 is my favorite adapted lens sor far. That said, I have seen some very nice photos takien with WA lenses in MFD bodies. I would love to see a 10.5mm Lens from Olympus or Panasonic. Maybe even Sigma.
Mike
Mike
Finder
Veteran
I really like my 12mm Voigtlander Heliar on my E-P1. I am really not seeing the smearing people keep talking about.
I wonder if the 20mm is so good because of in camera processing? The Leica lenses would have none.
I wonder if the 20mm is so good because of in camera processing? The Leica lenses would have none.
JoeV
Thin Air, Bright Sun
Perhaps the Heliar has a more telecentric design? I'm just surmising here...
~Joe
~Joe
Finder
Veteran
Joe, the lens is certainly not telecentric. I am wondering if the folks that notice "smearing" are not actually seeing the unfocused area on the image--DoF is not unlimited and the lens depth of field scales on the lens will not be accurate for an m4/3 camera (they would be about a stop off).
kbg32
neo-romanticist
There is no problem using "film" lenses on digital cameras. I have no problems using CV, Zeiss, and Leitz lenses on my M8. These lenses translate well to the chip of the M8 or the Rd1. The MFT chip is of a different nature. I find that all of my M mount lenses perform a lot less then stellar on my G1 and GF1. The worst being my 15 Heliar. My example is just crap. Even on my M8. Film, no problem. My CV 12 is much better, but not as sharp as it would be on one of my M mount film cameras.
The lenses that were designed for the MFT system perform flawlessly and sharper then adapted lenses.
The lenses that were designed for the MFT system perform flawlessly and sharper then adapted lenses.
Vickko
Veteran
I thought the f-number would be doubled on the GF1.
e.g. 50mm f1 Noctilux ends up being a 100mm f2 lens on the GF1. Still a bokeh killer.
Is that right?
Vick
e.g. 50mm f1 Noctilux ends up being a 100mm f2 lens on the GF1. Still a bokeh killer.
Is that right?
Vick
Not quite, it's still a 50mm f/1 lens with f/1 light gathering, however the field of view is effective 100mm, and the depth of field is effectively equivalent to f/2.
Vickko
Veteran
Yes, it is the DoF that I was referring to.
I have a telephoto f2 lenses (say 90mm Leica, 100mm Canon), but in superfast lenses for the GF1, my 50mm normal lenses become "equivalent 100mm f2 (Nocti), f2.8 (Summilux) or f4 (Summicrons)".
...Vick
I have a telephoto f2 lenses (say 90mm Leica, 100mm Canon), but in superfast lenses for the GF1, my 50mm normal lenses become "equivalent 100mm f2 (Nocti), f2.8 (Summilux) or f4 (Summicrons)".
...Vick
Finder
Veteran
Wish the crop factor was never invented as I cannot tell what anyone is talking about.
Basically, the depth of field scales on the lens shift when you put a 35mm lens on a m4/3 camera. If the lens is set to f/8, use the f/5.6 DoF scales.
Basically, the depth of field scales on the lens shift when you put a 35mm lens on a m4/3 camera. If the lens is set to f/8, use the f/5.6 DoF scales.
PCB_RF
Established
I'm sure there are many exceptions to prove the rule, but I'd say the sensor and lens combination puzzle basically looks like this:
1. Most rf wideangle lenses are less telecentric than slr wides; slr rear elements need to clear the mirror, rf wide rear elements generally sit closer to the sensor/film, requiring a more acute light angle to reach the corners
2. Most rf lenses 50mm and longer are as relatively telecentric as slr lenses
3. The M-4/3 sensor requires a highly telecentric light path all the way to the corners
4. The Leica M8/M9 sensor is constructed to require a less-telecentric light path in the corners, in order to accommodate current Leica wideangle rf lenses
So, m-4/3 corner smearing is much more common with wideangle rf lenses than with longer rf lenses, and the Leica M8/M9 has less corner smearing than the m-4/3 sensor. Native m-4/3 wides usually outperform legacy wides in the corners, especially legacy rf wides.
I haven't done any testing, but I'd guess that slr wides should show less corner smearing than rf wides. They don't intrude as far into the body, so their light path should be more telecentric.
Lens light path, sensor design and onboard image-processing algorithms are now connected, and it shouldn't surprise that the $400 Panny lens outperforms the $4,000 Leica lens on the sensor it was designed for.
I don't use legacy rf wides on m-4/3 bodies because I think they'll outperform native m-4/3 lenses. I use them because I can easily manipulate the focus distance and aperture to shoot quickly using hyperfocal focusing. And I use them because I have a couple that are wider than anything offered in native m-4/3 primes. My cameras fire faster with the C/V 12mm & 15mm than with the 20/1.7 or 17/2.8, and both C/V lenses are wider.
I've seen mention of Panasonic announcing new m-4/3 sensor technology that would, IIRC, install the pixels on an aspherical base, allowing tweaking of the corner pixels to minimize corner issues without having to use the complex angled microlenses on Leica's sensor. That would be pretty cool. They could maybe just dial-in the onboard image processing to optimize performance with current native m-4/3 lenses?
1. Most rf wideangle lenses are less telecentric than slr wides; slr rear elements need to clear the mirror, rf wide rear elements generally sit closer to the sensor/film, requiring a more acute light angle to reach the corners
2. Most rf lenses 50mm and longer are as relatively telecentric as slr lenses
3. The M-4/3 sensor requires a highly telecentric light path all the way to the corners
4. The Leica M8/M9 sensor is constructed to require a less-telecentric light path in the corners, in order to accommodate current Leica wideangle rf lenses
So, m-4/3 corner smearing is much more common with wideangle rf lenses than with longer rf lenses, and the Leica M8/M9 has less corner smearing than the m-4/3 sensor. Native m-4/3 wides usually outperform legacy wides in the corners, especially legacy rf wides.
I haven't done any testing, but I'd guess that slr wides should show less corner smearing than rf wides. They don't intrude as far into the body, so their light path should be more telecentric.
Lens light path, sensor design and onboard image-processing algorithms are now connected, and it shouldn't surprise that the $400 Panny lens outperforms the $4,000 Leica lens on the sensor it was designed for.
I don't use legacy rf wides on m-4/3 bodies because I think they'll outperform native m-4/3 lenses. I use them because I can easily manipulate the focus distance and aperture to shoot quickly using hyperfocal focusing. And I use them because I have a couple that are wider than anything offered in native m-4/3 primes. My cameras fire faster with the C/V 12mm & 15mm than with the 20/1.7 or 17/2.8, and both C/V lenses are wider.
I've seen mention of Panasonic announcing new m-4/3 sensor technology that would, IIRC, install the pixels on an aspherical base, allowing tweaking of the corner pixels to minimize corner issues without having to use the complex angled microlenses on Leica's sensor. That would be pretty cool. They could maybe just dial-in the onboard image processing to optimize performance with current native m-4/3 lenses?
