Micro Four Thirds Focal Reducer Shootout

Since Metabones released the first general purpose m4/3 focal reducer in June 2013 several more products have appeared on the market. To my knowledge, there haven't been any systematic high resolution comparison tests of these focal reducers, so I decided to shoot a test chart and put together a matrix of 100% crops.

The matrix below shows 100% crops from the center, left side, right side, and all four corners of test chart images taken with a 50mm f/1.4 Nikon G lens (wide open) attached to: 1) the Light Cannon by Vizelex; 2) the Lens Turbo by Zhongyi; 3) the R.J. Focal Reducer; and 4) the Speed Booster by Metabones. The test chart used is a standard Resolving Power Chart available at low cost from Edmund Optics: http://www.edmundoptics.com/testing-targets/test-targets/resolution-test-targets/resolving-power-chart/1665 .  Note that the chart illumination wasn't perfectly uniform, and was a little brighter on the right side.  In each case, focus was adjusted for maximum contrast in the center of the field.  The camera used was a Panasonic G6.

DISCLAIMER:  I'm the designer of the Speed Booster optics.

In the second part of the focal reducer shootout I shot images of a back-illuminated soft box with a square patch of black gaffers tape taped to the front.  The lens used was the Sigma 18-35mm f/1.8 set to f/16.  The reason I used this lens is that its exit pupil varies substantially during zoom, ranging from 70.3mm at the 35mm focal length to 149.4mm at the 18mm focal length.  This is useful because hotspotting behavior is strongly dependent on the exit pupil distance.  Note that 149.4mm is an unusually long exit pupil distance, and most SLR lenses have an exit pupil distance less than 100mm.

Combined with the Sigma lens are 1) Vizelex Light Cannon; 2) Zhongyi Lens Turbo; 3) R.J. Focal Reducer; 4) Metabones Speed Booster; and 5) Sigma lens by itself with a plain adapter (no focal reducer).

In the case of the Lens Turbo and R.J. Focal Reducer the intermediate focal length position was chosen to show the worst instance of hotspotting.  This occurs when there is a sharp ghost image of the aperture stop.  For the m4/3 Speed Booster the worst hotspottiing occurs at the 18mm focal length position (exit pupil distance = 149.4mm).  Hotspots like this always occur in the middle of the frame because they are ghost images of the iris opening, which lies on the optical axis.

The Sigma 18-35mm lens by itself has very low ghosting, as shown by the two images in the bottom row.  Among the focal reducers, the Vizelex Light Cannon performs best because it has only a single optical element.  Unfortunately, this simple design also means that image sharpness is extremely poor, as can be seen in Part 1.

DISCLAIMER:  I'm the designer of the Speed Booster optics.

This is very much appreciated Brian. It is about what i have read, that Speed booster is the best by far, then lens turbo in 2nd, followed by junk. It is however nice to see it with my own eyes. I have debated picking up a SB and this eases my mind. For example the classic 50mm F1.8 offering from Nikon is very good and dirt cheap. With a SB it works out to be a 71mm F1.26 which is just fine for portraits.

Thank you for the tests and design.  The design has revolutionized mirrorless systems to be able to obtain radical lenses.

zuikowesty wrote:

Interesting, as I am just looking into this. But why no OM to MFT version?

I'm guessing: when there is a large number of users, there will be an adapter (as it needs to sell).

Although I'm originally an OM film cameras users, most of the legacy lenses I'm now using with my Micro 4/3 are Nikon. I don't  know if this is the case with other users.

OM lenses are a dying breed because there are no current SLRs to support them. I'm very sorry to say.

Surely not. The SB increases FOV. Applied to the 50mm f1.8, it results in a 35mm f1.26 equivalent.

Fraser

I think Jack Hass is correct in relation to the M4/3 system, which this thread is about. "35mm x 2" is not far off 71mm.

fraserhadden wrote:

Surely not. The SB increases FOV. Applied to the 50mm f1.8, it results in a 35mm f1.26 equivalent.

Fraser

Im talking about on my GX7, a MFT camera. (50mm x2) x.71=71mm. for APSC yes, it would be what the lens actually says, 35mm.

woops, make that 50mm.

ApertureAcolyte wrote:

So why did they make an FD speedbooster then? Contax/Yashica?

The OM mount makes more sense than any other mount. Metabones would have made more sales on the OM speedbooster than the C/Y.

I think even though both FD and OM are no more, there are far more FD's out there due to Canon being a much larger user base, so it makes sense to sell to the larger audience. I agree they should have made both, but id imagine it has to be convincing before the R&D is started.

Brian

Thank you for making these tests  it is appreciated.

I am always puzzled by the condition that creates these hot spots.  I have no proper technical knowledge of what causes the but I have tried to explain them on forums to the best of my understanding.  (For want of an alternative more knowledgeable source).  It seems that the "spots" are shown and commented upon but their technical reason for occurring remains a mystery surronded in optical jargon.

If I repeat my understanding here then you might be able to comment and confirm or advise differently.

My understanding is that back in the film days the film stock was not as reflective as modern sensors and therefore the reflection of a strong light source from the film and further re-relection back off some part of the lens glass back on to the film was less of a problem.  In fact some lens manufacturers did not particularly worry about it as inconsequential.  Different legacy slr lenses are therefore more prone to sensor re-reflection spots/smudges than others.

Sensor reflection can be bounced back via some flatter glass surface in the form of a aperture delineated circle if from an element located in front of the aperture or as a smudge if after the aperture.  According to some Canon literature even a flat filter lens surface can cause this reflection.  Canon seem quite proud that they have a meniscus protective front lens to prevent this re-reflection "unlike some of their opposition".  Canon's advice for those getting this hot spot on their esteemed telephoto range in extreme conditions was to try removing the lens filter.

I believe/think that good lens design can disperse this re-reflection before it is completely reflected back to the sensor surface.  You are to be congratulated as the Metabones product seems to have been designed to reduce or eliminate this re-reflection that happens in more extreme circumstances very well.

I also have to make a disclaimer so that readers might understand: I share your surname, but we are not related.  I have no commercial connection with Metabones nor any other Chinese focal reducer manufacturer.  I buy my own focal reducers, including a Metabones Speed Booster and several other Chinese made focal reducers for the Sony E-mount.  I have one RJ made M42 to M4/3  focal reducer adapter.  I am not suitably qualified to make exacting tests, nor do I have any optical knowledge other than what I have read.

Jack Hass wrote:

ApertureAcolyte wrote:

So why did they make an FD speedbooster then? Contax/Yashica?

The OM mount makes more sense than any other mount. Metabones would have made more sales on the OM speedbooster than the C/Y.

I think even though both FD and OM are no more, there are far more FD's out there due to Canon being a much larger user base, so it makes sense to sell to the larger audience. I agree they should have made both, but id imagine it has to be convincing before the R&D is started.

Ask Brian?  The oracle started this thread, otherwsie we are just flapping our keyboards.

I don't know but I think that there are some bits that stick out of the back of OM lenses that complicate matters.

I'm not sure if it's the same thing as the so-called hot spot phenomenon, but I've definitely seen problems with filters reflecting back the sensor with MFT.  That's the main reason I stopped using UV filters, which I always employed when I shot Canon DSLRs.

Tom Caldwell wrote:

Brian

Thank you for making these tests it is appreciated.

I am always puzzled by the condition that creates these hot spots. I have no proper technical knowledge of what causes the but I have tried to explain them on forums to the best of my understanding. (For want of an alternative more knowledgeable source). It seems that the "spots" are shown and commented upon but their technical reason for occurring remains a mystery surronded in optical jargon.

If I repeat my understanding here then you might be able to comment and confirm or advise differently.

My understanding is that back in the film days the film stock was not as reflective as modern sensors and therefore the reflection of a strong light source from the film and further re-relection back off some part of the lens glass back on to the film was less of a problem. In fact some lens manufacturers did not particularly worry about it as inconsequential. Different legacy slr lenses are therefore more prone to sensor re-reflection spots/smudges than others.

Sensor reflection can be bounced back via some flatter glass surface in the form of a aperture delineated circle if from an element located in front of the aperture or as a smudge if after the aperture. According to some Canon literature even a flat filter lens surface can cause this reflection. Canon seem quite proud that they have a meniscus protective front lens to prevent this re-reflection "unlike some of their opposition". Canon's advice for those getting this hot spot on their esteemed telephoto range in extreme conditions was to try removing the lens filter.

I believe/think that good lens design can disperse this re-reflection before it is completely reflected back to the sensor surface. You are to be congratulated as the Metabones product seems to have been designed to reduce or eliminate this re-reflection that happens in more extreme circumstances very well.

I also have to make a disclaimer so that readers might understand: I share your surname, but we are not related. I have no commercial connection with Metabones nor any other Chinese focal reducer manufacturer. I buy my own focal reducers, including a Metabones Speed Booster and several other Chinese made focal reducers for the Sony E-mount. I have one RJ made M42 to M4/3 focal reducer adapter. I am not suitably qualified to make exacting tests, nor do I have any optical knowledge other than what I have read.

Hi Tom:

Central hotspots are a ghost image of the aperture stop.  The visibility of the ghost image depends on the magnification of the ghost image (smaller means brighter), the amount of aberration in the ghost image (less aberration means a more pronounced ghost), and the efficiency of the lens coatings.

Tom Caldwell wrote:

Jack Hass wrote:

ApertureAcolyte wrote:

So why did they make an FD speedbooster then? Contax/Yashica?

The OM mount makes more sense than any other mount. Metabones would have made more sales on the OM speedbooster than the C/Y.

I think even though both FD and OM are no more, there are far more FD's out there due to Canon being a much larger user base, so it makes sense to sell to the larger audience. I agree they should have made both, but id imagine it has to be convincing before the R&D is started.

Ask Brian? The oracle started this thread, otherwsie we are just flapping our keyboards.

I don't know but I think that there are some bits that stick out of the back of OM lenses that complicate matters.

I would like to see Metabones come out with an OM version.  However, I'm not directly affiliated with the company so I don't know what their plans are regarding this.  I do recall that some of the OM lenses presented more of an interference issue than other brands such as Nikon.