GH2 Worth buying a Tamron 17-50 2.8?

The lens could be parfocal but the adapter not necessarily. I use Minolta MD 35-70 f3.5 Constant Aperture for most of my filming and the zoom should be parfocal but isn't because of the wider tolerances in the low cost Pixco adapter.
 
Talking about adapters; is it true that when you attach an adapter for a different lens mount you lose a full stop or more light since the adapter eats light depending on how much it extends?
 
Admir said:
Talking about adapters; is it true that when you attach an adapter for a different lens mount you lose a full stop or more light since the adapter eats light depending on how much it extends?
Click to expand...

In the context of GH2 mount adapters, no it's not true. Those adapters are just empty metal rings with no lenses, designed to hold the lens the same distance from the film plane as they would be on a full size DSLR. In terms of light gathering and exposure, an F2.8 lens stays F2.8.

What you are likely referring to are teleconverters, containing optical elements, that increase focal length and therefore reduce f-stop.
 
But if you mount a lens made for a larger sensor size in front of a sensor with surface area of only 1/4 (2x crop), 3/4 of the light of the lens will go past the sensor and does not help illuminate the image. The f2.8 is for a light circle with certain diameter. If the sensor does not cover full light circle of the lens, one actually loses f-stops within the camera system, in this case 2 f-stops for 2x crop sensor.

Sensitivity of the sensor then starts to have a role, and because cameras with different sized sensors have been matched reasonably well, this reduction of light gathering power is no longer seen easily. 2x crop sensor can have the same amount of light gathering pixels as a full size sensor. To avoid complicating things too much it can be said that a f2.8 lens for full size camera remains the same for crop size camera.

But in purely optical devices 2x crop or magnification would indeed mean reduction of 2 f-stops (2x optical extender, for example). The optical extender does the exact same thing as a smaller sensor, crops and magnifies image. For smaller film size this reduction in light gathering power by smaller surface area would be seen when using the same projector and candle power to project same size images. For smaller film size the projector would have to have greater distance from the screen, hence two stops dimmer projected image on the screen (all else being equal).
 
Last edited:
My Rokkor 50/1.4 is a half stop faster than my Olly 45mm (measured at 2.8) which cost ten times as much. I suspect that this is because of the size of the front element of the glass (gathering more light), but I have no idea, really. Anyway, you will not lose any speed. Some lenses are slower or faster than their rating. Maybe this Rokkor is really 1.2, have to test that some day if they ever make a m4/3 1.2 that I can afford.
You MAY have to adjust the focus. Most primes work very well, but a few zooms and the odd prime need to have the infinty focus tweaked slightly. If you don't shoot infinity, it is not a problem. Of course, it could be that the lens need to have the infinity focus reset anyway. Zooms are more susceptible to focus errors than primes.
 
DrDave said:
My Rokkor 50/1.4 is a half stop faster than my Olly 45mm (measured at 2.8) which cost ten times as much. I suspect that this is because of the size of the front element of the glass (gathering more light), but I have no idea, really.
Click to expand...

M4/3 lenses can focus more light on smaller circle, which in turn becomes brighter. They should be brighter and the comparison directly from the lens rating should no longer be valid, because the rating is for different sized light circles.

This is the reason a f0.95 25 mm exists for m4/3, but the fastest equivalent for full frame is f1.4 (2 f-stops difference, as sensor area indicates). One could easily test by taking a lamp in hand and keeping lens back to a wall. Direct light through the lens and take the lens farther from the wall. The light circle grows, but becomes dimmer.
 
Last edited:
Aki_Hartikainen said:
But if you mount a lens made for a larger sensor size in front of a sensor with surface area of only 1/4 (2x crop), 3/4 of the light of the lens will go past the sensor and does not help illuminate the image. The f2.8 is for a light circle with certain diameter. If the sensor does not cover full light circle of the lens, one actually loses f-stops within the camera system, in this case 2 f-stops for 2x crop sensor.
Click to expand...
No, the 75% of light that's lost is in the outer perimeter of the image circle that is cropped out of the Four Thirds frame. Since there is no light lost within the image sensor's frame, there is no reduction in exposure. The only reduction is in the lens' Field of View, compared to a full-frame sensor.

As an analogy, consider what happens when you crop out the center quarter of an image in Photoshop, reducing it to half the width and height. Is the remaining image reduced in brightness or altered in any way?
 
lpowell said:
No, the 75% of light that's lost is in the outer perimeter of the image circle that is cropped out of the Four Thirds frame. Since there is no light lost within the image sensor's frame, there is no reduction in exposure. The only reduction is in the lens' Field of View, compared to a full-frame sensor.

As an analogy, consider what happens when you crop out the center quarter of an image in Photoshop, reducing it to half the width and height. Is the remaining image reduced in brightness or altered in any way?
Click to expand...

That's a good analogy and explains what happens in electronic image systems and why the apparent brightness can be maintained throughout the processing and viewing chain. However, the cropped image has been exposed with fewer photons, hence it must be dimmer. Unless there was more sensitive surface to receive fewer photons.

In Photoshop consider the post card sized window is illuminated by 10 watts. When you enlarge the image to full screen and illuminate it with only the same wattage, the exposure would be lower and full screen image dimmer. For a LCD display the full screen image is illuminated not with 10 watts but with 50 watts, for example. An optical system would only have the original 10 watts worth of photons from the subject. Using amplification and increasing wattage must be used to adjust brightness to same level as larger image that received photons worth 50 watts for exposure.

What comes in is the same that goes out all else being equal. A telephoto lens receives fewer photons from the subject than normal lens, hence telephoto lens is dimmer unless relative aperture size is increased. Same principle applies for different film and sensor sizes.
 
Last edited:
Camera engineers have made sure different cameras output very similar brightness. Jpeg encoding requires great precision for brightness for best results. However, few clues still remain that tell us that a 50 mm f2 normal lens does not become a 100 mm f2 by halving the sensor width. The most obvious clue is the depth of field. Had the 50 mm f2 lens turned into 100 mm f2, depth of field should have become shorter. When this does not happen, the depth of field remains the same, this tells us that the 100 mm equivalent lens is not f2, but actually the equivalent of 100 mm f4.

When we don't see reduction in apparent image brightness is because of the fine camera engineers adjusting the brightness to same level in camera using onboard electronics. What this also means is that potential for m4/3 lenses to be brighter exists, and this can be seen in the Nokton 25 mm f0.95, as well as a few other lenses. Optical image stabilization often prevents using large apertures, hence O.I.S. lenses being rather slow.
 
Last edited:
If I put my camera in full manual and swap out lenses, they all read slightly different, but the camera isn't doing anything that I can see, and my 50mm is reading slightly faster, not slower. My 55mm MF reads the same as my 45mm Olly. That means the lens is the difference.
 
Lenses are different and ratings are different, but my point was that if you put in 40 photons and get 40 photons out (full frame), and then put 10 photons in (2x crop camera, 1/4 surface area) you can only expect to get 10 photons out. When you get 40 photons out, the camera amplifies the photons by factor of 4, which is the equivalent of 2 f-stops. It does not matter how the amplification is done, if you squeeze 16 mp photosites in 1/4 area, that is one form of amplification.

Optically 50 mm f2 lens for full size camera becomes the equivalent of 100 mm f4 on a 2x crop camera, but electronically difference in brightness is done away. Perhaps better think of it only if you are real interested in optics and electronics. Cameras have been matched as closely as possible and are within spesifications. Comparing lenses is always a good idea.

If there is something to consider in this is that the F-rating for the lens does not determine it's brightness. It only determines the maximum aperture / focal length ratio. Brightness is determined by the size of the light circle that the lens outputs. The smaller the light circle, the more focused is the light, thus increased intensity and brighter lens.
 
Last edited:
Aki_Hartikainen said:
Optically 50 mm f2 lens for full size camera becomes the equivalent of 100 mm f4 on a 2x crop camera, but electronically difference in brightness is done away
Click to expand...

I'm sorry but you are 100% wrong about crop affecting stop. The crop and image circle size have nothing to do with the f-stop, t-stop or resulting image brightness, "equivalent" (I'm so sick of this crop equivalent-speak, this is a perfect example of the confusion it causes) or not.
 
Last edited:
Crop doesn't affect f-stop of the lens, but it should if what many people claim was correct. Does 50 mm f2 and 100 mm f2 have the same depth of field from the same distance? If not, then 50 mm f2 can not be the equivalent of 100 mm f2 on a crop camera. It would be the equivalent of 100 mm f4. Optically speaking.

The problem with the whole issue is the lack of basic understanding of crop and magnification factor involved. When there is crop, the image has to be magnified. Magnification is what eats brightness in optical devices, but you can adjust it electronically to the same brightness and that's what is done on the crop cameras.

40 photons of brightness in = 40 photons of brightness out (full size)
10 photons of brightness in = 10 photons of brightness out -> adjusted to 40 photons of brightness out (crop camera)

It is really this simple, if you allowed it to be.
 
Let's not forget this is a camera system and the lens is only one part of it. One can not expect everything to remain the same when you take one part of a system and install it on a different system. The area of the film or sensor is part of the system's light gathering power. If you reduce the area, light gathering power can not remain the same. Optical properties of the system have changed. Smaller area does not gather as much light as larger area.

We can adjust so that often the end result looks similar enough, and that's what matters. We can get similar results from different systems, but that does not mean the systems are the same.
 
Camera lenses are measured in F stops, which is a theoretical number. This explains the difference between photographic lenses, it depends how accurate the calculations were and how well manufactured the glass is.

Cinema glass is however measured in T stops, which is the real number.

Sensor size does nothing to lenses. A 50mm F1.4 is a 50mm F1.4 on ANY size sensor. Smaller sensors crop the image, EXACTLY THE SAME as if you would crop an image in photoshop. The aperture doesn't change and neither does the focal length.
 
Lens remains the same but camera does not. Camera captures the image, not the lens. One has to consider the whole camera system, not just one part of it. Differences in brightness between lenses come partly from different sized light circles, which determine the brightness of the lens.

If you could condense the light circle into smaller area, it would be brighter. Accepted?
 
I made the image for your consideration. Same lens, same light, same f-stop. Allow the business card to represent a m4/3 sensor. You can now see that a genuine m4/3 lens made for that sensor size would produce brighter image with the exact same light and same f-stop.

lightCircle.jpg
 
The conclusion is that we need to start to consider the diameter of the light circle that the lens outputs to determine brightness of the lens. F-stop is not enough. Clearly there are brighter lenses for the same f-stop, as was noted above by others too. This is caused by lens outputting light circle with different diameter.

A genuine m4/3 lens could in theory be 2 f-stops brighter than the same lens for full frame, if the light circle was made to match the m4/3 sensor area. This is seen in the Nokton f0.95 25 mm lens. To make the lens work for full frame would require enlarging the light circle, which would make the lens dimmer. Now these things go hand in hand, we can not just change the last lens element for different magnification and brightness. The whole lens has to be re-designed for different sensor area. Something that not many lens makers are willing to do.

They just use the same lenses and conveniently forget about to report brightness difference for images captured by different sensor sizes. It doesn't help that people keep insisting that "the lens is the same lens". Sure, but the camera is not.
 
The lens to sensor distance does not change, as you had it do in your example. Read up on "flange distance".

The image circle that a lens makes does not change, different cameras just utillize different size chunks of the middle of that image circle.
 
Last edited:
You must log in or register to reply here.
Back
Top