Metabones announces 'Speed Booster' lens adapter for mirrorless cameras

actually avbee is right. Even with prime lenses there come a time when the extra correction you can introduce by adding a new elements is offset by the incremental loss in sharpness and contrast by adding new surfaces. A good f/4 300mm lens is better than a fast shorter tele with a teleconverter no matter how good the converter is. Having said that I somewhere I still have an old 2x teleconverter - at the time a best of breed model - because I uses long lenses rarely and I was more likely to carry the converter in my camera bag then always load myself down with a large tele.

It is different in this case. The extra lens actually reduce the image circle of the full frame lens meaning that there is more light falling into the smaller sensors hence the faster f-stop.

This is true, but you're also taking 36mm worth of coverage and reducing it to cover 24mm. It's similar to scaling a 36mp image down to 24mp - using a pretty good scaling algorithm you'll the perceived sharpness will increase. If the optics are of reasonable quality you'll end up with a pretty good result.

If you want to fuss over this feel free, but any quality adding optics reduces they scaling will probably negate, and more. You might notice a touch more CA in worst-case scenario compositions, and perhaps a bit more distortion, but you can also run your ISO a stop lower. You lose some here, you get some there, it's probably a wash in the end quality-wise.

Comparing digital resolution reduction to optical image circle reduction is comparing apples and oranges. Did anyone read the white paper? The glass and design of this adapter appear to be of the highest quality; read the white paper and look at their image samples. The results are quite wonderful. BTW, the two gentlemen who designed this adapter are also responsible for designing 12 new prime lenses for Panavision's 70mm motion picture system, so I'm betting they got a mirrorless lens adapter right.

< Not everyone cares if every pixel is equally sharp>

For USD 600.00, and pixels may suffer? For an adapter? Smart investments are different. Just get a contemporary digital camera with lens off the shelf.

ok, calling it "miracle" may be a bit too "emotive", but I haven't seen such an approach in the past, which is why I really like this thing!

How about diffraction? Did you look at the device. The light is not being passed through a tiny aperture. It's passing through a large (relative to the wavelength) optical element. Any diffraction will already present in the light output from the lens, which will depend on the aperture chosen. This is just focussing down the light further from the objective lens at (back of the attached lens) and has been designed to introduce almost no additional aberrations and reduce the aberrations already present. The MTF curve is actually improved at least in the centre for NEX!

Now of course I'd like to see the real world results, but this is exciting news

Yes, also surprised that diffraction is not dicussed, but as an advantage.
My question is, with the interplay of effective speed, reduced focal length, sharper image and different depth of field, will it allow uis to go back to shooting sharp macros at f22 or 32 and getting things in focus?
I'd pay $600 for that.

No, it's not a miracle worker. It can't undo the effects of diffraction already present at using f/22 or smaller. The extra sharpness might highlight the diffraction better.

Do you own any lens worth more than $1000 bucks?

Progress comes from disputes. So I will value your contribution. Nobody else anyway will understand this discussion much further. Just buy a camera today for USD 600 (the price of the adapter) and enjoy your pictures.

Nobody should start out with adapters, but for some of us who already have old glass, it might be worth considering. At this price, though, it has to deliver. I'm waiting for reviews.

Adapters are cool and this is indeed a good news!
just try using some legacy MF lenses and u'll see;-)
... it's the only way to be 'creative-on-the-budget' IMO

--
@bodziupunk

You realize that with this adapter image is not cropped (as much) than with simple mechanical adapter. It projects larger area from a back element into much smaller area. With traditional adapter imaging area of a lens is not changed.

@noirdesir

Well, that explains everything.

I take a picture of a landscape with my crappy SX230HS, you take a picture with the D800 and a top of the line lens, we each make a 40x60 print of our captures, and your take is that the
SX230HS HAS OUT-RESOLVED THE D800!

You can't make it up. Honest to God, you can't.

I'm not sure that ignorance is bliss, though. I've found life to be better knowing my strengths and weaknesses, and trust me, I have a lot of weaknesses. I'm wrong about things, all the time, in a big way. Just not about this. Oh, and I'm not wrong about it being a bad idea to make up stuff, lie, and stick to my guns. The truth is always the best way to go. I think that others have pointed that out in the past. Cheers.

manuel and bob, get a room.
Bob are you claiming you know more than whatever engineer built this? Do you really think everybody else is wrong (including the engineer) and you are right?
When you use a FF lens on a 4/3s or aps c sensor do you also not believe the image will be corner to corner sharper? You seem not to be able to grasp a simple concept. Did you spend too much money on a full frame and this rubs you wrong?

manuel does your equivalency theory makes you right or does my camera being able to use a stop faster of shutter speed make me right? You can say its not true until your blue, and the fact is my shutter speed will be one stop faster. Or do you really believe it will not? is my camera lying to me? brainwashing me?

@bobbarber

1 - Get your favorite m43 camera and mount your favorite legacy lens. Take a picture wide open.

2 - Now mount the very same lens on the very same camera with the Booster in between. Take a few steps FORWARD so that you can take a picture with the very same FRAMING.

Check the MTF for both pictures and then come back to us. I bet you will have learned something.

Manuel: tele extender makes the focal length longer, but does not enlarge the physical aperture. Thus the f-stop value changes to the worse (like from 2.8 to 4 with 1.4X extender).

This wide-angle reducer works exactly the opposite way: it makes the focal length shorter, but does not make the physical aperture smaller. Thus it changes the maximum aperture larger (from 2.8 to 2 with a 0.7X reducer).

It is that simple. Or do you not believe that tele extenders change the maximum aperture either? If you do, I can not help you.

Guys, I will try to respond without using insulting language.

@ManuelVilardeMacedo -- I believe you are incorrect. This does add a stop of light. What it doesn't do is add resolution.

@the reason -- The engineer who designed this does not claim that it adds resolution, not that I've seen. If you have a link, please post it. Don't post a link to the white paper, please. There are no claims in the white paper that this adapter increases resolution. It can't increase the resolution of the original lens. That is physically impossible.

@duartix -- The adapter + lens may increase resolution when compared to the lens without adapter itself on a cropped format, but it will not offer increased resolution over what the lens offers on the lens' native format, which was the the discussion. I really don't understand personalities like yours. You are evidently bright enough to understand this issue, but you try to frame it in a sneaky way to get around the discussion. That's weak.

@bobbarber
Scenario C captures more information than scenario B. That is what is meant when it is said that this adaptor increases resolution, you capture more information compared to just using an empty tube as adaptor.

Your general problem seems to be that you incapable of even considering that resolution is a term that can describe several things. You can have resolution per millimetre or resolution per picture height. If the lens on your SX230 had the same resolution per millimetre than a typical lens used on a D800, it would have only about a quarter of the (linear) resolution per picture height (and thus a 1/16 of resolution measured in MP).

The point is comparing a native FF lens on a 1/2.33" sensor vs. a dedicated lens for that sensor size not comparing it with a FF lens on a FF sensor. By having a much smaller image circle the SX230 lens can be designed to have a higher (central) resolution, smaller lens elements also allow for higher precision.

In the white paper there are comparisons of MTF curves of the original lens and lens with an adaptor. In the centre of the frame, the MTF values are higher with the adaptor. Unless you claim these MTF curves are pure fabrication, this is unambiguous as possible.

it definitely improves resolution on a cropped sensor a lot but not to the level you can get on a 35mm one.

This doesn't change aperture at all. How would it? By pulling the iris blades further into the lens barrel? The f/ number refers to the width of the orifice formed by the iris blades, so it is inaccurate to say this adapter increases aperture. What it does is change two variables taken into account when it comes to depth of field - the distance needed between the subject and the focal plane and (equivalent) focal length. That's why it appears to have gained an extra f-stop.
What I find surprising is that so many people write about this MetaBS as if they had actually used it - when it hasn't even hit the shelves yet...

@ Kinematic Digit

Bingo.

But people were not saying in this thread that this adapter "increases the MTF rating for an APS-C sensor". They said that it increases the resolution of the lens, which is incorrect.

I am very familiar with adapting legacy lenses to digital cameras. I've been doing it since 2005. I currently shoot m43, sometimes with adapted FF lenses.

IF my goal is MAXIMUM RESOLUTION from my current crop of legacy lenses, I DO NOT buy this adapter, I buy a FF camera. Period. There is no way to get better resolution on m43 or APS-C than on FF with FF legacy lenses, which is what so many people are having trouble understanding. I can only get an increase over the cropped resolution, which is SUB-OPTIMAL to begin with.

The appeal of this adapter to me is 1) more light, and 2) native focal length. Increase over cropped resolution? OK, but again, IF my goal is maximum resolution, I don't go this route.

Of course this also depends on whether or not the sensor is capable of resolving the MTF or LP/H of the projected image. This comes down to sensel sizes and density. The pixel sizes of something like a G12 is 2.7µm. This means it will out resolve the very best full frame dSLR lenses. The lenses made for much smaller sensors have to have a higher MTF rating in order to resolve down that small.

a 22MP FF sensor is approximately 6.5 µm while an 18MP crop sensor is around 4.3 µm. This precisely a factor of 1.5. Meaning if the MTF rating was calculated on a 22MP FF sensor, then a focal reduced version will also be equal on the 18MP sensor because it will have the same density of resolving pixels to the MTF of the lens.

Bob, it's semantics. However you are correct, not an increase in the lens.

However, lets take the fact that the NEX-7 actually has a higher MP count of the 22MP sensor (24.3). Since this adapter is designed for the NEX camera at the moment, this is an advantage. For those who are already using L glass for example and does not have a 5D series camera, means they can purchase a much cheaper NEX-7 and have full frame advantages (including resolution) in a much smaller form factor.

@Kinematic Digit

Since you interested in having a genuine discussion, I will tell you that in my experience the best legacy lenses outresolve 4/3 sensors, at least the 10, 12, and 14 mp variety that I've owned, even when cropped, when the lenses are stopped down a little bit.

Some examples, Minolta MD 50mm f1.7, Nikon 105mm f2.5, Nikon 180mm f2.8, etc. At f5.6 or so, these lenses are quite sharp, even cropped. I've compared them in some cases to dedicated lenses, and the difference is not much. For example, I compared the Minolta MD 50mm f3.5 macro to the Oly 50mm f2.0 macro, both at f5.6, and the Oly was better, but not by much. We're talking pixel-peeping here. So there won't be any resolution "jump".

What you get, and what I want, is the old focal length back! That's worth it to me. That and the gain in light.

Regards.

@ManuelVilardeMacedo the problem is people are using aperture to describe the light gathering ability. In fact, the F range doesn't change, the location of the projected image has changed from the optics. This shouldn't affect the depth of field properties because the lens is still a 50mm F/1.8 with or without the adapter. What the optics does is move the projected point to a different location, but also reduces the size to fit a much smaller sensor.

However focal reducers have been around for decades, it just hasn't been used in this way until now. It changes the cone of light and increase the transmission quality of the light. To say an F/1.8 lens turns into an F/1.2 lens is a bit confusing, but general public understand that. It would be better to say the T/1.6 lens turns into a T/1.1 would be more accurate.

@ ManuelVilardeMacedo: f-number refers to the ratio of focal length divided by the aperture = f:2 = focal length divided by 2. F-number is NOT the aperture, it is RELATIVE aperture: focal length divided by the aperture.

With this gadget the focal length is shortened by a factor of 0.7, but the physical size of the aperture remain the same, so the f-stop values are bigger (or smaller number) by a factor of 0.7 also = one f-stop. Exactly the opposite of a teleconverter. That is all there is to it.

With a 1.4x teleconverter the light entering the lens is spread out on twice the area, so f-number is multiplied by 1.4, like form 2.8 to 4. With this wide-angle converter the light is concentrated on half the area, thus the f-number changes from 2.8 to 2.

Interesting results there Bob, but also one thing that hasn't really been discussed or explored is the Diffraction Aperture Limits. I've also done my own tests with my Hasselblad, Contax and other variants of Carl Zeiss lenses adapted all the way down to my E-P3 and found interesting results (my Hasselblad lenses were surprisingly good on both the 135 format and the micro four thirds).

The density of the sensor is still going to depend on the DLA. DLA cannot be changed by optics as I understand it, therefore, in theory, and I may be wrong on this, even with an increase in MTF of a lens, the DLA limits might restrict the very best at those smaller pixel sizes. I've seen DLA issues with D800E shots I've taken, and I don't believe it can be corrected by optics. From my understanding the DLA is based on photon sizes, and optics cannot shrink a particle of light, only increase the amount that is thrown at it.

Resolution is still going to be dependent on the sensors resolving abilities

Of course it is not going to be better than a full frame lens mounted on a full frame camera. Given the same generation of sensor technology - the image off a 24mp full frame sensor is going to be better than the image off a 24mp APS-C sensor - even with the same amount of light concentrated on it.

However it might be a better way of using full frame lenses, especially wide angles, on an APS-C mirrorless camera than a using a straight non converting adapter.

Let's wait and see some real world image samples taken by competent reviewers before passing judgement on the image quality.

I expect it is going to work much better with some lenses than with others.

"That is exactly what people are saying. Please read the thread. People have posted that this adapter will increase the resolution of the lenses that it is attached to." - bob

Really. Got any examples. If so, why not correct them accurately, rather than this mishmash of confusion you've been demonstrating so far. Seems to me that the person who's been most ambiguous with the terms "resolution" and "information" is you

Petka: sorry, but I'd rather believe my photography teacher - from whom I got the answer that there's no such thing as «equivalent aperture» - than you. Nothing personal, really.

The tiny sensor does out resolve the bigger one. Look at reviews of with the same lens on a full frame or APS-C and Nikon CX sensor. Also, keep in mind that the smallest format people generally view the output of that thumbnail sized sensor is an equivalent enlargement as the D800 printed at 16x24!

in principle there is nothing to prevent the same concept being used for a FF lens[es] and e.g. APS-C crop dSLR body - mirror or mirrorless have very little to do with the topic here !!

@jpr2

According to the white paper this is not possible because of SLR design and much longer flange distance. Flange is the same for EF and EF-S. Same for Nikon FX and DX.

@Juck

It depends on the situation. For fast action yes, but otherwise PDAF is not miracle maker. Fe. S-AF on my D800 is not as accurate as CDAF on my mirrorless OM-D. The problem seem to be field curvature of certain lenses and focus point accuracy which means focus is not as consistent as with OM-D.

With CDAF there's no need for frustrating AF calibration and focus point coverage is larger.

I got tired of backfocus/frontfocus years before mirrorless came out. Whats the point of uber fast AF if half the images are gonna be slightly out of focus anyway?

You're right, I don't get it. I don't see how this adapter could possibly add resolution beyond the limit of resolution, what the lens this adapter is attached to resolves.

Or is this one of those things, you know, like with the government, where I'm supposed to play along? I'm not very good at that, but I'll do my best.

You guys are loster-er (is that a word?) than I thought.

You are not talking about resolution, you are talking about equivalent focal length.

Consider: the front of the house fills the frame. Now, with adapter, the front of TWO houses fill the frame. But that doesn't mean that you have more RESOLUTION! It just means that you are taking a different picture. The part of the picture that filled the frame at first, exists at a reduced resolution in the second picture, because it only uses half the pixels.

Honest to God, give it up, please. You are completely wrong, and this is getting ridiculous.

I will repeat, I LIKE the idea of this product, I WANT one, it sounds FREAKIN' AWESOME, it DOES NOT, CAN NOT, INCREASE THE RESOLUTION OF THE ORIGINAL LENS.

I just re-read your posts, and my jaw is practically hanging on the floor in disbelief.

I'll boil down your argument for you, because you are running around in circles trying to prove that something that is false is true.

Your argument is this: that microfilm has MORE RESOLUTION than the source material (e.g., newspapers), because it has more lines of resolution per unit area. Did I get that right? Unbelievable. So according to you guys, if I blow the microfilm back up to the size of the original image, more information will be there than was in the original, correct? Because, "resolution", i.e., information per unit area, is greater in the microfilm.

Does a 1200x1800 pixel image of a scene, printed at 2"x3", have "MORE RESOLUTION" than a 4000x6000 pixel image of the SAME SCENE printed at billboard size, just because the smaller print is denser.

Stop, please, IT'S EMBARRASSING!

This does not increase the resolution in subject realm (whatever small features the lens resolved on the house will be at most resolved with that adaptor). But you know have two houses in the frame, two houses with the same detail per house means twice as much information.

A FF lens projects information on an area of 24x36 mm. Take a photo of only the central part captures only part of that information, obviously. Take a picture of the whole area, and you can in principle capture all the information.

"Your argument is this: that microfilm has MORE RESOLUTION than the source material (e.g., newspapers), because it has more lines of resolution per unit area. Did I get that right?"

You get it wrong. In your example, how could you get the MTF measurement of the newspaper? You cannot compare resolution of a picture to a thing. What you can do, however, is compare picture vs picture. In your example, compare microfilm vs microfilm, not microfilm vs newspaper.

forpetessake, it does not need to be ideal. 58% of the len's resolution is wasted using it as an APS-C lens. The magnifier can lose a great deal of light and still provide a benefit. If the quoted f-stop figures are correct and distortion free (they are cropping their magnification down a bit to smooth off corner issues no doubt), the adapter can improve resolution with significant losses.

Of course, this is assuming an existing FF lens for comparison. If the choice is between a FF lens+ this adapter or an equivalently priced and equivalently large APS-C native lens, I don't think the FF lens would deliver more resolution, no.

1) I am 100% for this product. I want one!

2) It is IMPOSSIBLE to get a resolution advantage from this product. IMPOSSIBLE. In the vernacular of the discussion of this product on this forum, a 4x6 print from a 1200x1800 pixel image would have "increased resolution" when compared to a 13x19 print from the same 1200x1800 pixel image. That's because the same information is stored in a smaller space, which to some people is increased resolution. IT IS NOT INCREASED RESOLUTION! YOU DO NOT CAPTURE ANY INFORMATION ON A SMALLER SENSOR WITH A FF LENS THAT WOULD NOT BE CAPTURED ON A FF SENSOR. THAT'S IMPOSSIBLE. YOU ARE IN LOONEY, HOCUS-POCUS TERRITORY! COME BACK TO EARTH, PLEASE!

And I'm FOR this darn thing...

Look at the buildings not the supports of the rail (which are curved, not distorted). I imagine that designing a lens system to take a focused image and scale it down at a fixed offset without significant distortion or aberration is doable (and it looks like it's been done).

perspective and distortion are not the same thing.