(SpeedBooster 0.64 + canon EF 50mmf1.8 STM) vs (Zuiko 45mmf1.8)

[Thomas An]

The Zuiko (and later the Panasonic 42.5) is likely the go-to portrait lens for MFT. It is my most used lens to date and gives the right amounts of background blur (good size bokeh balls etc) ... but lately I came across -->this and it got me wondering. Can we do even better ?

1. If that speed booster is paired with a Canon EF 50mmf1.8 STM , the combined width of the contraption would be 2.2in (vs 1.8 of the Zuiko). So it is still a compact lens.
2. The speedBooster has a 1.28 crop factor, turning the 50mmF1.8 lens into an imaginary 64mmF1.8 lens (the Zuiko is the equivalent of a 90mmF3.6; resulting in slightly smaller diameter bokeh-balls eventhough it is 90mm)
3. On top of the above, the imaginary 64mmF1.8 gives us an extra 3/4 stop advantage while shooting hand-held in low light because it has less magnification and we can use slightly lower shutter speeds, making it behave like a 64mmF1.5 versus the 90mmF3.6 equivalent of the Zuiko.
4. On top of that the 64mmF1.8 gives us a little more field of view making it even more versatile

In summary of the above, it seems to me the Speedbooster+Canon50mm combo could be a better portrait lens than the native Panasonic or Zuiko lenses.

Am I imagining things ? Your thoughts ?

 

[biggles266]

The Zuiko (and later the Panasonic 42.5) is likely the go-to portrait lens for MFT. It is my most used lens to date and gives the right amounts of background blur (good size bokeh balls etc) ... but lately I came across -->this and it got me wondering. Can we do even better ?

1. If that speed booster is paired with a Canon EF 50mmf1.8 STM , the combined width of the contraption would be 2.2in (vs 1.8 of the Zuiko). So it is still a compact lens.
2. The speedBooster has a 1.28 crop factor, turning the 50mmF1.8 lens into an imaginary 64mmF1.8 lens (the Zuiko is the equivalent of a 90mmF3.6; resulting in slightly smaller diameter bokeh-balls eventhough it is 90mm)
3. On top of the above, the imaginary 64mmF1.8 gives us an extra 3/4 stop advantage while shooting hand-held in low light because it has less magnification and we can use slightly lower shutter speeds, making it behave like a 64mmF1.5 versus the 90mmF3.6 equivalent of the Zuiko.
4. On top of that the 64mmF1.8 gives us a little more field of view making it even more versatile

In summary of the above, it seems to me the Speedbooster+Canon50mm combo could be a better portrait lens than the native Panasonic or Zuiko lenses.

Am I imagining things ? Your thoughts ?

Do you already have a Speedbooster 0.64x? They are very expensive. Unless you have numerous Canon lenses to make it more worth while, I don't really see the point of going that way (over, say, getting the Olympus 25mm 1.2, possibly on sale or used to save money) but it's up to you if you want to spend the money, and you really want shallower depth of field with autofocus.

Re #2, I think it will result in about a 64mm f1.1 , not a 64mm f1.8. (Which will be like a 64mm with the depth of field of f2.2 in full frame equivalent.) If you don't know how a speedbooster works to change both the effective focal length and aperture, I'm a little surprised you are considering one.

One thing to think of with adapting Canon lenses is that aberrations like purple and green fringing, which are usually quite common on cheaper lenses when wide open, will not get fixed in camera, so will have to get fixed in post. Also, that 50mm 1.8 is not very sharp wide open, like m43 lenses are.

 

[Thomas An]

Re #2, I think it will result in about a 64mm f1.1 , not a 64mm f1.8. (Which will be like a 64mm with the depth of field of f2.2 in full frame equivalent.)

The speedbooster has a crop factor of 1.28 which means the 50mm lens is casting a slightly larger image circle than the MFT sensor and it will look like it was taken from a 64mm lens on a full frame camera (so far we agree).

The 50mmf1.8 has an effective aperture of 27.78 casting bokeh circles of some size on the MFT sensor, but because we are cropping those circles will appear larger by a factor of 1.28 in the final image as if the effective aperture was 27.78*1.28=35.77 and the image was taken by a 64mm at F1.8 (= 64mm/35.77 ). This is in terms of DOF.

How did you come up with F2.2 ?

Now, in terms of light. The light from the 50mmF1.8 (same light intensity as a 64mmF1.8) is now concentrated in an area 0.64 times as big which makes it 1.56 times brighter which is the equivalent of F1.44 = F1.8/sqrt(1.56). So a 64mmF1.4 would produce equivalent light on a full frame as the 50mmF1.8 does with the speed booster.

How did you come up with F1.1 ?

 

[Thomas An]

Do you already have a Speedbooster 0.64x? They are very expensive. Unless you have numerous Canon lenses to make it more worth while, I don't really see the point of going that way (over, say, getting the Olympus 25mm 1.2, possibly on sale or used to save money) but it's up to you if you want to spend the money, and you really want shallower depth of field with autofocus.

The Olympus 25mmF1.2 goes for $1199. The speedbooster+50mm goes for $125+650=$775 (comparing new to new for fairness)

Also, that 50mm 1.8 is not very sharp wide open, like m43 lenses are.

Lets check DXO .

The 50mm has a sharpness score of 29 on DXOmark but the highest MFT lens goes to 16 on sharpness. Also I see the 50mm is tied with the Zuiko for lowest chr. aberration among Panasonic, Olympus, and Canon lenses on that site.

The only place it falls short is in vignetting, but then again there is cropping, some of the vignetting will not make it into the final image.

 

[biggles266]

Re #2, I think it will result in about a 64mm f1.1 , not a 64mm f1.8. (Which will be like a 64mm with the depth of field of f2.2 in full frame equivalent.)

The speedbooster has a crop factor of 1.28 which means the 50mm lens is casting a slightly larger image circle than the MFT sensor and it will look like it was taken from a 64mm lens on a full frame camera (so far we agree).

The 50mmf1.8 has an effective aperture of 27.78 casting bokeh circles of some size on the MFT sensor, but because we are cropping those circles will appear larger by a factor of 1.28 in the final image as if the effective aperture was 27.78*1.28=35.77 and the image was taken by a 64mm at F1.8 (= 64mm/35.77 ). This is in terms of DOF.

How did you come up with F2.2 ?

Now, in terms of light. The light from the 50mmF1.8 (same light intensity as a 64mmF1.8) is now concentrated in an area 0.64 times as big which makes it 1.56 times brighter which is the equivalent of F1.44 = F1.8/sqrt(1.56). So a 64mmF1.4 would produce equivalent light on a full frame as the 50mmF1.8 does with the speed booster.

How did you come up with F1.1 ?

A speed booster increases the effective f-stop of a lens by 1 stop, so an f2 becomes an f1.4. And so an f1.8 becomes an f1.2 approx. (slightly over).

The ultra speed booster, which I know less about, I assume increases by slightly more so increases by about 1 and a third stops. So increasing an f1.8 lens by about 1 and a third stops gives you an f1.1 lens. That's the effective aperture, as it affects ISO. The depth of field however, since mounted on a m43 body, is double that, so you get the equivalent dof of an f2.2 on full frame.

 

[biggles266]

Do you already have a Speedbooster 0.64x? They are very expensive. Unless you have numerous Canon lenses to make it more worth while, I don't really see the point of going that way (over, say, getting the Olympus 25mm 1.2, possibly on sale or used to save money) but it's up to you if you want to spend the money, and you really want shallower depth of field with autofocus.

The Olympus 25mmF1.2 goes for $1199. The speedbooster+50mm goes for $125+650=$775 (comparing new to new for fairness)

Also, that 50mm 1.8 is not very sharp wide open, like m43 lenses are.

Lets check DXO .

The 50mm has a sharpness score of 29 on DXOmark but the highest MFT lens goes to 16 on sharpness. Also I see the 50mm is tied with the Zuiko for lowest chr. aberration among Panasonic, Olympus, and Canon lenses on that site.

The only place it falls short is in vignetting, but then again there is cropping, some of the vignetting will not make it into the final image.

I've never really trusted DXO. I use imagingresource's graphs for my lens comparisons, but I'm going more by personal experience - while I have seen worse full frame lenses, the 50mm 1.8 is only OK wide open, its sharpness and contrast aren't great, and are lower than an m43 lens is wide open, particularly a Pro lens.

Also, the Canon lens is designed for full frame so isn't designed to handle the high pixel resolution of m43, although it will of course work.

If you willing to give up autofocus, by the way, there are a whole lot more possibilities, and cheaper ones, to get ultra fast lenses on m43.

 

[bobn2]

The Zuiko (and later the Panasonic 42.5) is likely the go-to portrait lens for MFT. It is my most used lens to date and gives the right amounts of background blur (good size bokeh balls etc) ... but lately I came across -->this and it got me wondering. Can we do even better ?

1. If that speed booster is paired with a Canon EF 50mmf1.8 STM , the combined width of the contraption would be 2.2in (vs 1.8 of the Zuiko). So it is still a compact lens.
2. The speedBooster has a 1.28 crop factor, turning the 50mmF1.8 lens into an imaginary 64mmF1.8 lens (the Zuiko is the equivalent of a 90mmF3.6; resulting in slightly smaller diameter bokeh-balls eventhough it is 90mm)

Best to work in real dimensions, going 'equivalent' is confusing. Do it later, if you want. The speed booster actually reduces the focal length of the lens, so when you put a speed booster on the back, the focal length becomes 50x0.64 = 32mm. The aperture of the lens stays the same (since you haven't done anything to the front of the lens, which is where the aperture is defined) so the aperture was and still is 27.8mm. Since the focal length has changed, the f-number must change, so the f-number is now 32/27.8=1.15 (call it 1.2). ('Speed booster' is a misleading trade name, the lens doesn't collect any more light, since how much light it collects is determined at the front end, it's just squeezing the light collected down more).

1. On top of the above, the imaginary 64mmF1.8 gives us an extra 3/4 stop advantage while shooting hand-held in low light because it has less magnification and we can use slightly lower shutter speeds, making it behave like a 64mmF1.5 versus the 90mmF3.6 equivalent of the Zuiko.
2. On top of that the 64mmF1.8 gives us a little more field of view making it even more versatile

It never was and isn't a 64mm, f/1.8. What it is now is a 32mm, f/1.2. If now you want to go 'equivalent', what it is 'equivalent' to on FF is a 64mm, f/2.4. 64mm has never been a popular focal length because it's a bit of a nothing.

 

[bobn2]

Re #2, I think it will result in about a 64mm f1.1 , not a 64mm f1.8. (Which will be like a 64mm with the depth of field of f2.2 in full frame equivalent.)

The speedbooster has a crop factor of 1.28 which means the 50mm lens is casting a slightly larger image circle than the MFT sensor and it will look like it was taken from a 64mm lens on a full frame camera (so far we agree).

The 50mmf1.8 has an effective aperture of 27.78 casting bokeh circles of some size on the MFT sensor, but because we are cropping those circles will appear larger by a factor of 1.28 in the final image as if the effective aperture was 27.78*1.28=35.77 and the image was taken by a 64mm at F1.8 (= 64mm/35.77 ). This is in terms of DOF.

How did you come up with F2.2 ?

Now, in terms of light. The light from the 50mmF1.8 (same light intensity as a 64mmF1.8) is now concentrated in an area 0.64 times as big which makes it 1.56 times brighter which is the equivalent of F1.44 = F1.8/sqrt(1.56). So a 64mmF1.4 would produce equivalent light on a full frame as the 50mmF1.8 does with the speed booster.

How did you come up with F1.1 ?

A speed booster increases the effective f-stop of a lens by 1 stop, so an f2 becomes an f1.4. And so an f1.8 becomes an f1.2 approx. (slightly over).

It decreases the actual f-number, since 1.2 is smaller than 1.8. That's because the aperture stays the same but the focal length decreases, so the f-number must also decrease.

 

[Great Bustard]

The Zuiko (and later the Panasonic 42.5) is likely the go-to portrait lens for MFT. It is my most used lens to date and gives the right amounts of background blur (good size bokeh balls etc) ... but lately I came across -->this and it got me wondering. Can we do even better ?

1. If that speed booster is paired with a Canon EF 50mmf1.8 STM , the combined width of the contraption would be 2.2in (vs 1.8 of the Zuiko). So it is still a compact lens.
2. The speedBooster has a 1.28 crop factor, turning the 50mmF1.8 lens into an imaginary 64mmF1.8 lens (the Zuiko is the equivalent of a 90mmF3.6; resulting in slightly smaller diameter bokeh-balls eventhough it is 90mm)
3. On top of the above, the imaginary 64mmF1.8 gives us an extra 3/4 stop advantage while shooting hand-held in low light because it has less magnification and we can use slightly lower shutter speeds, making it behave like a 64mmF1.5 versus the 90mmF3.6 equivalent of the Zuiko.
4. On top of that the 64mmF1.8 gives us a little more field of view making it even more versatile

In summary of the above, it seems to me the Speedbooster+Canon50mm combo could be a better portrait lens than the native Panasonic or Zuiko lenses.

Am I imagining things ? Your thoughts ?

First, a quick correction on your arithmetic. The speedbooster has an equivalence ratio (crop factor) of 0.64x which applies to *both* the focal length and relative aperture. Thus, a 50 / 1.8 + 0.64x speedbooster on mFT literally becomes a 32 / 1.15 which is then equivalent to a 64 / 2.3 on FF, not a 64 / 1.8.

That all said, your question is if a 32 / 1.15 would be a better portrait lens than a 42 / 1.8. I think most would argue that the latter would generally be better, just as most would argue that an 85 / 1.8 on FF would be better for portraits than a 50 / 1.2.

Myself, I would say that it depends on the types of portraits you're doing. If you're doing tightly framed portraits, then go with the longer setup. If you're doing more widely framed environmental portraits, go with the wider faster setup.

 

[Thomas An]

Guys, I see in the responses you are getting your information from the same source, but I am not convinced on some of it.

1. I agree on the F1.15 part in terms of brightness (I discovered missing a power of 2 in one the calculation)
2. I still disagree on the F2.3 part. You would need a 64mm F2.3 in order to produce the exact same background blur as a 50mm F1.8 (since they would both have identical Iris diameters), but this is not what is happening here.
   Once the 50mmF1.8 is attached to the speed booster, the final image will have both larger subject AND larger bokeh-circles as if it was produced by a 64mmF1.8 on a regular full frame camera

Here are my notes with explanations on the matter:

 

[Thomas An]

A speed booster increases the effective f-stop of a lens by 1 stop, so an f2 becomes an f1.4. And so an f1.8 becomes an f1.2 approx. (slightly over).

The ultra speed booster, which I know less about, I assume increases by slightly more so increases by about 1 and a third stops. So increasing an f1.8 lens by about 1 and a third stops gives you an f1.1 lens. That's the effective aperture, as it affects ISO. The depth of field however, since mounted on a m43 body, is double that, so you get the equivalent dof of an f2.2 on full frame.

You are not accounting for the crop.

Your method works from MFT to FF because they are both "full frame" (the lens image circle covers the sensor precisely in both formats). However, the speed booster introduces a crop which changes some of the calculations. (I posted my notes on this --> here )

Also, with the standard method of "doubling" the assumption is that the equivalent FF lens has double the focal length but identical Iris diameter. By doing this the equivalent lens would have a magnified subject, while keeping the background blur the same, this gives an asynchronous magnification (the subject looks magnified but not the blur)

 

[bobn2]

Guys, I see in the responses you are getting your information from the same source, but I am not convinced on some of it.

1. I agree on the F1.15 part in terms of brightness (I discovered missing a power of 2 in one the calculation)
2. I still disagree on the F2.3 part. You would need a 64mm F2.3 in order to produce the exact same background blur as a 50mm F1.8 (since they would both have identical Iris diameters), but this is not what is happening here.
   Once the 50mmF1.8 is attached to the speed booster, the final image will have both larger subject AND larger bokeh-circles as if it was produced by a 64mmF1.8 on a regular full frame camera

Here are my notes with explanations on the matter:

The reasoning seems very convoluted and has come to the wrong conclusion. Let's start off with reality. The 50/1.8 with a 0.64 speed booster behind it is a 32/1.2 - not equivalent to one, it is that. Adding the extra elements behind has made it a different lens. Its aperture still is 28mm. The diagonal angle of view on micro four thirds is 2 arctan(21.5/64) or 37 degrees. The DOF and amount of light on the image projected onto the sensor are determined by the aperture and the angle of view (a simple result of the fact that light travels in straight lines. Every infinitesimal point on the scene radiates the same light to the same size aperture, and if the angle of view is the same, there is the same number of infinitesimal points. So, whatever the format, a lens with the same aperture and giving the same angle of view will result in the same light rays being refracted onto the sensor. To get the same angle of view for FF would need a 64mm lens. If the 64mm lens had an aperture of 28mm it would be an f/2.3 lens.

--
Tinkety tonk old fruit, & down with the Nazis!
Bob

 

[bobn2]

A speed booster increases the effective f-stop of a lens by 1 stop, so an f2 becomes an f1.4. And so an f1.8 becomes an f1.2 approx. (slightly over).

The ultra speed booster, which I know less about, I assume increases by slightly more so increases by about 1 and a third stops. So increasing an f1.8 lens by about 1 and a third stops gives you an f1.1 lens. That's the effective aperture, as it affects ISO. The depth of field however, since mounted on a m43 body, is double that, so you get the equivalent dof of an f2.2 on full frame.

You are not accounting for the crop.

Your method works from MFT to FF because they are both "full frame" (the lens image circle covers the sensor precisely in both formats). However, the speed booster introduces a crop which changes some of the calculations. (I posted my notes on this --> here )

No it doesn't. The image circle of the lens is irrelevant. I just determines whether there is vignetting or not.

 

[Thomas An]

I am sorry, my approach is following the physics (which is reality) step by step. If you can't follow the reasoning, then you can't help either.

All you are doing is reiterating some formulas you memorized. But these formulas you are using are the result of some engineer doing these steps (as I am doing) a while ago.

Show me which part is wrong on those notes and why.

 

[bobn2]

I am sorry, my approach is following the physics (which is reality) step by step.

Sorry, your approach is doing anything but following the physics, which presumably why the answer doesn't accord with reality.

If you can't follow the reasoning, then you can't help either.

I admit, I didn't try too hard, because I already knew you had got to the wrong answer.

All you are doing is reiterating some formulas you memorized.

I'm exactly not doing that.

But these formulas you are using are the result of some engineer doing these steps (as I am doing) a while ago.

Show me which part is wrong on those notes and why.

You're a tough task master. I could say the same. Show me which part of what I said was wrong and why, and maybe you could justify why you're right. My argument was much simpler, and thus had less to go wrong.

Anyway, I already told you where you went wrong, which is making an incorrect correction for image circle. The image circle does not affect either she size of the image projected on to the frame or which part of the image gets into the frame. In fact, you do not know what is the image circle either of the original 50/1.8 or after it's been dealt with by the speed booster (it should be whatever it was scaled by 0.63, but it's possible that the SB restricts it some more).

Back to basics, what we actually have is a 32mm lens with 28mm aperture. A 64mm lens with a 28mm aperture will produce 'blur circles' double the size of those of the 32mm lens, not 1.28 times the size. This is also simple geometry. To view the same size, you enlarge an mFT image twice the amount that you enlarge an FF one, thus with respect to the final image the blur circles of a 32/1.15 on mFT end up the same size as those of a 64/2.3 on FF.

--
Tinkety tonk old fruit, & down with the Nazis!
Bob

 

[Great Bustard]

Guys, I see in the responses you are getting your information from the same source, but I am not convinced on some of it.

1. I agree on the F1.15 part in terms of brightness (I discovered missing a power of 2 in one the calculation)
2. I still disagree on the F2.3 part. You would need a 64mm F2.3 in order to produce the exact same background blur as a 50mm F1.8 (since they would both have identical Iris diameters), but this is not what is happening here.
   Once the 50mmF1.8 is attached to the speed booster, the final image will have both larger subject AND larger bokeh-circles as if it was produced by a 64mmF1.8 on a regular full frame camera

Here are my notes with explanations on the matter:

You are making this much harder than it needs to be (and coming to an incorrect conclusion), but let's use your approach to arrive at the correct answer.

Consider an 50 / 1.8 mounted in front of a FF sensor. The diameter of the aperture for a 50 / 1.8 lens is 50mm / 1.8 = 27.8mm. This makes the aperture area to be 2424 mm². Now, not all of that light passing through the lens falls on the sensor, of course. and this results in some confusion when approaching the matter from this angle, because people incorrectly think the image circle plays a role -- it doesn't (so long as the sensor fits within the image circle).

For example, let's consider two 50 / 1.8 lenses, one with twice the image circle area as the other. One might think that the same amount of light passes through each lens (for a given scene and exposure time) since the apertures are the same size, but since the image circle of one lens is twice the area as the image circle of the other lens, that the light for the lens with the larger image circle would be half as bright since the same amount of light is being distributed over twice the area. However, this is not the case -- the brightness is the same regardless of the image circle diameter.

The reason is because every point on the sensor "sees" every point on the aperture and thus every point on the sensor is illuminated by every point on the aperture. Hence, the brightness of the light falling on each point of the sensor depends only on the area of the aperture (for a given scene and exposure time). One might think that a sensor with 4x the area has 4x as many "points" and thus receives 4x as much light. However, each of those "points" is seeing light from only 1/4 as much of the scene (although that light is coming from every point on the aperture), so that exactly cancels out the 4x greater number of points.

OK, so let's consider a 50 / 1.8 + 0.64x FR on mFT vs a 64 / 2.3 on FF. We see that the have the same [diagonal] angle of view because 50mm x 0.64 x 2 = 64mm. We see that they have the same aperture diameter because 50mm / 1.8 = 64mm / 2.3 = 27.8mm and thus the same aperture area. Thus, for a given scene and exposure time, each sensor "sees" the same amount of light.

 

[khunpapa]

In summary of the above, it seems to me the Speedbooster+Canon50mm combo could be a better portrait lens than the native Panasonic or Zuiko lenses.

Am I imagining things ? Your thoughts ?

Lens for portrait requires something more important than physical aperture number.

If you want bright 32mm f/1.15 lens, this combo is good.

But it you want good portrait lens, and don't want to listen to the opposing thought/idea/reason, go paying your money buying them.

Please keep in mind that, in this world, there's no medicine to cure the 'sorry after purchase'.

 

[millardmt]

The Zuiko (and later the Panasonic 42.5) is likely the go-to portrait lens for MFT. It is my most used lens to date and gives the right amounts of background blur (good size bokeh balls etc) ... but lately I came across -->this and it got me wondering. Can we do even better ?

1. If that speed booster is paired with a Canon EF 50mmf1.8 STM , the combined width of the contraption would be 2.2in (vs 1.8 of the Zuiko). So it is still a compact lens.
2. The speedBooster has a 1.28 crop factor, turning the 50mmF1.8 lens into an imaginary 64mmF1.8 lens (the Zuiko is the equivalent of a 90mmF3.6; resulting in slightly smaller diameter bokeh-balls eventhough it is 90mm)
3. On top of the above, the imaginary 64mmF1.8 gives us an extra 3/4 stop advantage while shooting hand-held in low light because it has less magnification and we can use slightly lower shutter speeds, making it behave like a 64mmF1.5 versus the 90mmF3.6 equivalent of the Zuiko.
4. On top of that the 64mmF1.8 gives us a little more field of view making it even more versatile

In summary of the above, it seems to me the Speedbooster+Canon50mm combo could be a better portrait lens than the native Panasonic or Zuiko lenses.

Am I imagining things ? Your thoughts ?

Thomas (et al.):

It appears that this discussion has so far neglected an important consideration. That is, the optical quality of the Metabones 0.64x m43 Speedbooster XL does not resolve to the same standard as the 0.71x "Ultra" version.

Below is a message I had some while ago from the designer.

Marc

--------------------------------------

Hi Marc:
Thanks for your note!

Our press release for the XL does have MTF, illumination, and distortion data, in case that might be helpful:

http://www.metabones.com/assets/a/s...d_Booster_XL_0.64x_Press_Release_5Jun2015.pdf.pdf

Let me know if you need additional information.

Given that you are interested in still photography and not video, I think the only potential downside of the XL for you is that you may encounter some vignetting with DX format lenses. However, this is lens-dependent. For example, I get very nice results with an inexpensive Nikkor 35mm f/1.8 DX used on the XL in still mode. You should have no problems at all with your 1970's era Nikkors since they all have plenty of coverage.

Performance-wise as measured by MTF, the XL is definitely better than the original m43 Speed Booster due to its more complex 6-element design. Whether or not you would actually notice the difference in sharpness depends mainly on your lens. In many cases the original Speed Booster out-performs the attached lens, so increasing the focal reducer performance will have a less dramatic effect than with really high performance lenses. On the other hand, with crazy-good lenses like the Zeiss Otus you can really see the full impact of improved focal reducer design.

At the risk of letting the cat out of the bag - and please keep this confidential, I will tell you that we will shortly be replacing the original 0.71x m43 Speed Booster with an improved 0.71x version. Compared to the more specialized 0.64x XL, this new 0.71x version will have more uniform MTF across the full still-frame field, and it will also be fully compatible with all m43 cameras, including Olympus.

Thanks again for contacting me, and let me know if you have any more questions.

Best regards,
Brian [Caldwell, Speed booster designer]

Jul 2, 2015 09:40:57

 

[Thomas An]

We indeed need to multiply both Focal length and the fNumber by 2, for proper lens equivalence between MFT and full frame.

That's something I've been taking for granted for years without thinking, but when it came down to the nuts and bolts of the physics and the geometry, the optical behavior of the lens element was not behaving proportionally as I was suspecting throwing out my analysis. Instead it follows the lens equation ( 1/f = 1/f1 + 1/s1)

I had to go all the way down the rabbit hole for this one, pulling out CAD software, spreadsheet, a physics book, wikipedia, and several online resources just to prove to myself mechanically WHY this is the case.

Bottom line a 25mm f1.8 MFT lens will give the exact same blur-circles on the out of focus areas as a 50mm f3.6 FF lens.

Here are my notes exactly why.

 

[khunpapa]

We indeed need to multiply both Focal length and the fNumber by 2, for proper lens equivalence between MFT and full frame.

That's something I've been taking for granted for years without thinking, but when it came down to the nuts and bolts of the physics and the geometry, the optical behavior of the lens element was not behaving proportionally as I was suspecting throwing out my analysis. Instead it follows the lens equation ( 1/f = 1/f1 + 1/s1)

It does not. No camera lens follows such equation.

What is lens's focal length?

How can you measure its focal length?