B4 adaptor or zoom lens recommendation

CineStill

CineStill

New member
Hi I was wondering if there has been a B4 to M43 adaptor released that may have internal optics to make the native 2/3 lens work well on a micro 4/3 sensor? i have tried the Cielio adaptor (i think that was the name) and doubled the B4 lens but the lens I tried was not a HD lens so looked pretty soft.
I am covering a dance concert with 2x AF100's and want one camera to do the closer shots and prefer a manual focus lens. I have used the Panasonic 14 to 140 lens, but I am not a fan of the servo focus. Does anyone have a lens recommendation for a lens that you can smoothly zoom with approximately 20mm to 120mm zoom range and manual focus?
I have Nikon 24 - 70 f2.8 but I find as I zoom the exposure tends to change?
I am also interested in the new Blackmagic studio camera that has a M43 mount but would only buy the camera if I could get a reasonable zoom lens.
Thanks Richard
 
There are two options:

(i) Yes there is an adapter with internal optics thats let you get full use out of a 2/3 lens but is very expensive if I can remember its a few thousand dollars.
(ii) There is a dumb adapter, no optics but it lets you mate a 2/3 lens to a AF100 but you must switch in the 2/3 lens doubler (if it has one) otherwise you get a big vignette type effect. If your 2/3 lens has a double it works great its approx $300.
 
B4 2/3" CANON FUJINON ENG lens to GH2 G2 AF100 AF101 E-P3micro 4/3 adapter at ebay.com
$(KGrHqV,!iUE9mKv+!CtBPokQpK1)!~~60_1.JPG


Fujinon A18x9BE 12U with Double Extender and a 9 volt battery for zoom


https://vimeo.com/41312959
 

Attachments

  • bild07.jpg
    bild07.jpg
    86.6 KB · Views: 0
Yip, thats my setup, you'll find the adapters on eBay made a Polish company, a great company to deal with and their very well made.




https://groups.google.com/forum/#!msg/soc.men/j5ln4yy7SAA/TAkaAru3VpsJ
 
Last edited:
Thanks Vultch and phd, The video using the above lens looked good! No example of zoom which to me is the point of using the lens but it is still good to see an example.
 
The internal doubler in SD lenses can be really bad on the edges, and with HD lenses, the doubler is better but still not ideal.

Abakus makes a $2500 converter which will work perfectly on the AF100 - it's called the Abakus 200 (B4 to micro 4/3rds). You can get it from ZGC.com in the US, or Lemac in Australia - http://www.lemac.com.au/

Otherwise, what you can do is get Ciecio7's (or another manufacturers) B4 to Canon EF adapter, then put on the 2.0x Kenko DGX MC7 or Pro 300 Canon teleconverter (I'm not sure if both are a true 2x magnification, but they should be, or which is better in the center, which is all you'll be using) then use a dumb (as in not powered or with it's own iris) Canon EF to M4/3 adapter - the nicest of which is probably the Kipon, although this one is the only one I could find with a tripod foot: http://www.amazon.com/DSLRKIT-Canon-Micro-Adapter-Tripod/dp/B007LHDS9U/ .

I haven't used this yet, but I'm trying it out next week (just waiting for parts to arrive.) I'm actually going to be using the Kenko MC4 2x teleconverter, since that one is closer to a 1.83x or 1.84x magnification which should cover the 1080p sensor size of the GH3 and GH4 with less resolution loss (19.83 sensor size covered by 11 x 1.84 = 20.24, which is only a 2% overshoot at most.)
 
There is an adapter made with it's own built in doubler. It is purpose-built for this application and delivers far superior performance compared to other options. It is known as the HD-2 (now the HDx2 Mk. II), manufactured by IB/E Optics in Germany and sold exclusively in the US by AbelCine. The adapter is made with a PL mount, but there is a special interchangeable mount to change it to MFT without fear of changing depth calibration.

I know it quite well as I used to be the Product Manager for it at AbelCine.
 
IBE Optics sells a solution that basically has an optical doubler in it so you can use a B4 mount lens that does not have a doubler. Thing is expensive and I can't seem to find the product on Abel Cine's site anymore. Not sure if it was actually put into production or if the video I saw using it was of a prototype. Here is a link to their shop. You'd also have to pickup their adapter for it to work on the AF100.

The video has been posted on DVX before, can't remember where. Here is the vimeo link

I have an older Fujinon 2/3" lens with a built-in doubler that I use with the Ciecio7's B4 to MFT adapter. It works well on my AF-100. I also bought a cheap cable and battery so that I can power the servo-zoom and have a fully-functioning s/s button on the lens. Got that off ebay as well but from some guy in Florida that sells them cheap. Quality is what you'd expect from a guy who imports them from China.

I haven't mounted the lens to my GH3 yet. I plan on testing that this summer since I will have a little more free-time around the office. Want to see how the GH3's built-in crop mode works with the lens without the physical doubler enabled.
 
Mitch Gross said:
There is an adapter made with it's own built in doubler. It is purpose-built for this application and delivers far superior performance compared to other options.
Click to expand...

Mitch, have you ever tested any of the IBE adapters against Abakus's versions? I haven't seen any tests out there by anyone. I personally wouldn't say that any of the IBE adapters are far superior to Abakus's, based on what I have seen (which is basically the video linked above plus personal experience with one of their adapters) but I've never tested them head to head.

In this case, though, I would say that the Abakus 200 might be the better bet, based solely on the numbers, as it was purposely designed to cover the sensor size of the AF100 and GH2, while the HDx2, having an actual magnification of 2.161 (max output of 25.5 / max input of 11.8), will expand the lens diameter past what is necessary. The result is that the sensor is not seeing about 17% of the area of the lens (23.77^2 / 22^2 = 1.167) decreasing resolution, as well as losing about an additional 1/4 stop of light (6.64*log(2.161) - 6.64*log(2) = 0.223 stops.)

Also, all of Abakus's adapters will take up to f1.4 (the biggest B4 aperture is around f1.45 because of the prism block) while IBE's adapters will only take up to f1.8. Because of this, the Abakus might have less vignetting. If you look at one of the documents for the first HDx2 (seen here: http://www.ibe-optics.de/media/raw/HDx2_Adapter_MFT.pdf) you can see that the adapter itself starts to vignette at about the 18mm mark. While I haven't used the Abakus 200, my Abakus 132 did not seem to vignette over any part of the Super 16 image circle. Also, on my testing of the Abakus 132, it didn't hard vignette until about the 22mm mark or 150% past what it needed to cover. Based on the document above, the HDx2 hard vignettes at 33.2mm or 140% past what it needs to cover (33.2/23.77).

Again, a head to head test would show which is best, but I haven't seen anything like that, so I can only go off of the numbers.

Also, not to knock you, Mitch, you've been an invaluable resource for years and years, I'm just putting out some additional considerations.
 
I don't take it as a knock, I'm glad to see someone coming up with hard data and representing clearly what he knows and does not know, and then asking the right questions.

I have not seen it against the latest Abekus adapter, but we did extensive testing v. the previous models and lots of other optics, including some rather esoteric machine vision devices. The MTF of what became the IB/E design outperformed all the others, particularly as you moved away from the center of the image. You'd be amazed at the falloff on some of this glass, even with double a spherics or triplets involved.

We also tested a boatload of 2/3" optics, and had to laugh at the coverage on some of that glass. Some wide angle zooms costing $30K+ new would actually vignette themselves in regular operation. We found that by the numbers where we should have no clipping of the frame we were in fact suffering obvious vignetting when using the adapter. This is why we had to bump up to the increased magnification.

The HDx2 was originally conceived for the Phantom Flex and its sports counterpart. That sensor is a "true" 4/3" image in 1920x1080 crop. It is exactly 22mm diagonal in that mode while 2/3" is 11mm diagonal -- or at least the industry spec is that. Since the dawn of video, every sensor has secretly been a little bit different in size. It's just how technology works. Some of the lenses just didn't cover correctly so that's why the magnification had to be bumped up slightly. Yes it is for a bigger area than the MFT sensor, but another dirty little secret is that there is no hard & fast absolute on the size of that sensor format either. A little over-projection is a good thing.

Anyway, the MTF, color, contrast, flare & veiling protection, color fidelity and a host of other optical parameters are all top notch on the IB/E adapters. I'm not going to knock Abekus, I'll just state that I have never been as impressed as I have with the IB/E product.
 
Such awesome info. Thanks Mitch! I always wondered if you were involved with the design of the IBE adapters, but I was never sure.

Mitch Gross said:
We also tested a boatload of 2/3" optics, and had to laugh at the coverage on some of that glass. Some wide angle zooms costing $30K+ new would actually vignette themselves in regular operation. We found that by the numbers where we should have no clipping of the frame we were in fact suffering obvious vignetting when using the adapter. This is why we had to bump up to the increased magnification.
Click to expand...

Makes a lot of sense - I was actually thinking about that consideration myself, especially since I have an SD lens that definitely resolves at least 4K worth of resolution (as measured on a DSC Labs 4K Megatrumpet) so, especially in 1080, you could throw some of the resolution away and still be fine.

I always wondered why lens manufacturers let wide lenses vignette a bit. Even the Zeiss Ultra Prime 8R shows a bit of strong vignetting in the corner (aka portholing). I'm guessing that the lensmakers can't fight optics, and while they would like the lens to be perfectly telecentric (for digital sensors - I don't think the UPs or MPs necessarily had digital sensors in mind) for wide angle lenses, it's probably not ideally possible. However, some of the newer sensor designs do have better walls which can direct light coming in at oblique angles to the pixel wells: http://www.youtube.com/watch?v=XPec2EaBSSM .

One thing that I always found interesting with B4 lenses was the offset for each color to accommodate the spacing of the three chips (which, at least initially, supposedly made for easier designed cameras and lenses.) If you could, could you talk about how exactly the IBE adapters correct that? In some videos, it seems that the bare lens doesn't look that bad, but obviously the colors aren't focusing at quite the same distance. While in HD it might not be as big an issue, in 4K its likely going to be more visible.
 
Last edited:
Joshua Cadmium said:
Such awesome info. Thanks Mitch! I always wondered if you were involved with the design of the IBE adapters, but I was never sure.



Makes a lot of sense - I was actually thinking about that consideration myself, especially since I have an SD lens that definitely resolves at least 4K worth of resolution (as measured on a DSC Labs 4K Megatrumpet) so, especially in 1080, you could throw some of the resolution away and still be fine.

I always wondered why lens manufacturers let wide lenses vignette a bit. Even the Zeiss Ultra Prime 8R shows a bit of strong vignetting in the corner (aka portholing). I'm guessing that the lensmakers can't fight optics, and while they would like the lens to be perfectly telecentric (for digital sensors - I don't think the UPs or MPs necessarily had digital sensors in mind) for wide angle lenses, it's probably not ideally possible. However, some of the newer sensor designs do have better walls which can direct light coming in at oblique angles to the pixel wells: http://www.youtube.com/watch?v=XPec2EaBSSM .

One thing that I always found interesting with B4 lenses was the offset for each color to accommodate the spacing of the three chips (which, at least initially, supposedly made for easier designed cameras and lenses.) If you could, could you talk about how exactly the IBE adapters correct that? In some videos, it seems that the bare lens doesn't look that bad, but obviously the colors aren't focusing at quite the same distance. While in HD it might not be as big an issue, in 4K its likely going to be more visible.
Click to expand...
In 2/3" video, the three chips are indeed set to slightly different depths. There is a bunch of mumbo-jumbo out there giving technical reasons for why this is, but let me just cut through it to state that it was entirely for cost. It is a LOT cheaper to produce a prism block with the three primary colors focusing at different depths because that is what the different wavelengths want to do anyway. Cheaper to make the cameras that way and really not much different either way for the cost to make the lenses. But of course this is an issue when using these lenses on a single sensor camera that expects all the colors to focus on the same plane.

The IB/E adapters do in fact do a few optical tricks to bring the split rainbow back together to a single depth. I can't really go into detail, one because that would be wrong for me to divulge and two because I really only have a cursory knowledge of what is actually happening to do this anyway. But there is some real optical work going on here and it does make a big difference.
 
You must log in or register to reply here.
Back
Top