View Full Version : SI2K and Metabones BMPCC Speedbooster


Bob Hart
August 8th, 2014, 10:32 AM
Now here is an interesting notion.


The Metabones Speedbooster more or less does optically, some of what a few of us obsessives were tying to achieve with groundglass 35mm adaptors.

A version has been released for the Blackmagic Pocket Cinema Camera. It naturally follows that would should be good with the BMPCC should also be good with the SI2K which has a slightly smaller sensor than the BMPCC's Super16mm cropped sensor.

There remains a crop factor with the speed booster. For technical reasons, the crop factor could not reduced beyond 1.7 to 1.

There is not a lot of advantage to using stills lenses via the Speed Booster on Super16mm digital formats if you already have the superspeed super16mm lenses like Optars, Zeiss, CP Ultra T. However if you want to use longer focal lengths than 85mm which is about as good as Super16mm conveniently gets, then 35mm stills telephoto lenses are the likely direction people may go.

On the SI2K, the best of my Nikons yielded 62 sharpness points on a Siemens-style chart. The longer lenses show softer images on the SI2K which do not match well with those from Super16mm optics.

With the Speed Booster in initial tests, the sharpness numbers of my Nikons have significantly improved into the purposed Super16mm lens zone, though not at the same wider aperture numbers. The low light performance is observable but the last stop of many stills lenses flare as before.

The Metabones Speed Booster will not natively fit the IMS-Mount of the SI2K. However for test purposes it can be offered up to the IMS-Mount with the lock ring and locator pin removed from the IMS-Mount. The IMS-Mount flange to focal plane distance is 19.0mm in air, the BMPCC MFT-Mount is 19.25mm, near enough for rough tests.

For a BMPCC Speed Booster to be modified for the IMS-Mount, a simple IMS-Mount is doable. However, for the wider IMS mount, a 1.5mm overhanging rim on the Speed Booster rear body must be skimmed flat to the rear surface exposed when the MFT-Mount ring has been removed.

The IMS-Mount lock ring is deeper and thicker than the PL-Mount ring it resembles. The Speed Booster body fouls the lugs of the IMS-Mount lock ring. About 0.5mm diameter must be removed from the Speed Booster rear body forward as a shoulder of about 7mm, or a similar amount from the inner diameter of the lugs of the IMS-Mount lock ring for a working clearance to be restored.

I am examining the Speed Booster internal body to see if this material can be removed without compromising the strength and function of the Speed Booster.

Bob Hart
August 10th, 2014, 07:45 AM
Furthur to the above post, the adaptation is relatively straightforward. There are a couple of wrinkles to watch out for.

The machining of the PL-Mount must be P+S Technik standard perfect for the clamp ring lugs to register correctly when tightened. Otherwise it will not be possible to operate the Nikon lock pin release. The rearward movement of the slide button moves into the gap between lugs of the PL-Mount clamp ring.

The slide button also has to be filed down to enable clearance for the IMS clamp ring to rotate. Ideally for an IMS-Mount version of the Speedbooster, a redesign of the slide button to extend the button forward would be desirable. The existing button, once trimmed down for working clearance, is inconveniently deep for a fingertip to easily release it.

When stripping the Speedbooster of the internal parts to enable the case to be machined, one needs to be careful of three springloaded ballbearings inside of the aperture ring.

There is a locking screw which holds the lens group from turning in its threaded internal mount. The fit of the group is fairly loose in the specimen I have. A locking method is needed as the group would shift with vibration from transport and handling.

This screw is threaded into a hole in the MFT mount ring. A much longer grubscrew is needed and its radial hole has to be drilled and threaded into the shoulder of the IMS adaptor.

Bob Hart
August 10th, 2014, 09:46 AM
A couple of low quality quick copy happysnaps via Speedbooster to SI2K. Lenses. 500mm f8 Nikon mirror lens, 1084mm f10.5 MTO "Jupiter" mirror lens. Supermoon through light cloud cover. 0db video gain.

Bob Hart
August 13th, 2014, 01:36 AM
Another frame from the SI2K with speedbooster, Nikon 50mm f1.4 lens iris set at f2 plus IRNDF1.8, Neutral low con look.

Conditions 8/8ths deep cloud cover, 2pm, mid-winter, 32 degrees south, facing east.

No process done of the camera image. No post sharpening has been applied. Routinely I apply about 25 points of sharpening, more or less, depending on the subject.

The subject is locally known as the "bacon and egg plant".

Bob Hart
August 16th, 2014, 11:48 AM
A quick glimpse of the super moon - SI2K/Metabones BMPCC Speedbooster modified for IMS-Mount. Lenses were Nikon 500mm f8 mirror telephoto and MTO "Jupiter" 1084 f10.5 mirror telephoto.

SUPER MOON - YouTube

Bob Hart
August 17th, 2014, 11:36 AM
A bit more on the speedbooster. Ignore the colours. I can't colour grade to save my life. Nearly all the shots were in low light environments.

SI2K BMPCC SPEEDBOOSTER TEST - YouTube

Bob Hart
August 18th, 2014, 09:10 AM
Here is an image of the Speedbooster on the SI2K camera and the IMS-Mount modification..

Bob Hart
August 24th, 2014, 02:15 AM
A few more happy snaps quickly slung together.

One thing very apparent is that ND and IR filtering is very definitely needed. I forgot to take my filters on the day and was forced to high shutter speed and tightest apertures possible with the selection of lenses I took.

This placed most of them in the diffraction zone and the resulting look to me is a bit harsh. The Sigma-for-Nikon 14mm f2.8 and the Nikon 35mm f1.4 AIS seemed to deal with it the worst.

There is a mix of Century anamorphic and cropped image. There seems no advantage to shooting anamorphic versus cropping a 16:9 image. However the newer anamorphics may be a different story.


SI2K SPEEDBOOSTER MIX - YouTube

Bob Hart
August 31st, 2014, 11:49 AM
Here is a bit more experimentation with the SI2K and Speedbooster, this time with an IRND1.8 filter to reduce the light down to a sensible level and get out of the diffraction zone with the Nikon lenses.

ARALUEN RMC 1 - YouTube

Bob Hart
September 13th, 2014, 11:14 AM
In regard the BMPCC for Nikon Speedbooster model, some users on the BMD cameras have reported varying sharpness across the image, with the variation being more pronounced with some lenses more than others. With the SI2K at least, there seems to be conferred by the Speedbooster optic, a skew towards blueness which is very apparent in shadow detail on objects like white shirts.

This may be an artifact caused by or aggravated by interaction with a Truecut infrared filter which I was using.

I recall on the forum for 35mm adaptors, the general theory of a focal reducer in place of a 35mm groundglass adaptor was discussed. Opinion at the time was that it was possible but that an individual solution for each lens was likely needed. It would appear the Speedbooster optic has transcended that limitation somewhat but that it remains.

When using a 35mm adaptor in aerial image mode, with interactive focal adjustments of both the lens on front and the camera lens/dioptre as relay, there was enabled a wide range of tolerance for position of the aerial image plane.

Eventually a corner softening artifact would appear, which signalled the focus was about to fall over if camera autofocus was being used through the 35mm adaptor.

If the Speedbooster optical group could be conveniently trimmed for focus by an external adjustment, such interactivity might work to similar beneficial effect in finding more easily a sweet spot for individual "difficult" lenses. The lens focus witness marks would become inaccurate but the image itself would be improved.

In circumstances of bright backlight and high contrast, the Speedbooster image begins to show a softness and colour separation across sharp edges. In this circumstance, the interior environment was lit by daylight spill and overhead flourescent tubes. Its use will require more careful selection of location and lighting environment.

Within those caveats, the Speedbooster is a handy tool for extending the utility of the SI2K.

I have been filming some pro-bono coverage for the Turkish Honorary Consul in Western Australia of the participation of Turkish Australians at the Araluen Botanic Park 2014 Tulip Festival. I used the Festival as an opportunity to practice with the Speedbooster/Nikon stills lens combination.

Here is an example of how not to use it. When needs must, you are better off with the simpler solution of a Super16mm prime lens. Of course, adding light or changing the location/composition of the shot is the better solution if circumstance permit. I make no claim to being any sort of competent camera operator. This is also an exemplar of why you do not use a partially red/green colourblind person to colour-grade your footage.

ARALUEN RMC 8 - YouTube


With careful collimation, the Nikon f4 12mm - 24mm zoom gains some useful light performance and is telecentric.

ARALUEN RMC 10 - YouTube

Bob Hart
October 30th, 2014, 01:16 PM
For folk with a BMPCC Canon dumb EF-Mount Metabones Speedbooster, this might be worth examining. Not all PL-Mount 35mm cine lenses are going to fit without picking up on the throat of the Speedbooster or clashing the rear element on the Speedbooster optics.

DSLR PL Lens to Canon EF Lens Mount Adapter for Canon 5D2 5D3 BMCC BMPC with EF | eBay (http://www.ebay.com/itm/DSLR-PL-Lens-to-Canon-EF-Lens-Mount-Adapter-For-Canon-5D2-5D3-BMCC-BMPC-With-EF-/361028862956?pt=LH_DefaultDomain_0&hash=item540eff43ec)

With 2.5mm worth of material skimmed from the rear and re-drilling the 2mm Nikon F-Mount screw pattern, this might be a viable option for the Nikon version of the Speedbooster. This will require that sufficient material is available for machining away without ruining the piece.

Bob Hart
January 29th, 2017, 12:47 AM
Some supplemental notes on the speedbooster IMS mod follow.

Furthur investigation has determined that the Canon Speedbooster version cannot be IMS-Mount modifed because of shorter flange to focal plane distance, electrical conductors and wider diameter of the Speedbooster barrel.

The IMS adaptor does not lock the Speedbooster optical group after collimation / backfocus adjustments. The original M4/3 mount has a simple screw on a radial axis which locks the optical group after adjustment. This arrangement has caused problems with the Speedboosters because the optical group is forced out of true alignment with the optical axis and causes one side of the image to be soft. The optical group is not a tight threaded fit in the Speedbooster body

The only way to secure the optical group in the IMS-modified Speedbooster is to use small spots of adhesive on the fine threaded section of the cell where it comes out of the rear of the Speedbooster body. I have used water cleanable non-silicone white bathroom sealer. It is an adequate locker but can be easily removed. Clear silicone sealer must not be used. It is not water cleanable and is impossible to remove spills or smears from coated optical glass without damage to the coatings.

Because of the loose fit of the optical group cell in the Speedbooster body, the only way to be sure the group remains true to the optical axis is to adjust backfocus / collimation to a focus chart with the camera and lens directly facing downward. Gravity keeps the optical group cell squarely in its threads. When the adjustment is correct, the small pinspots of bathroom sealer can be applied to the rear of the threads of the optical cell. The sealer must allowed to set and dry before the assembly is moved from being face down.

Bob Hart
May 27th, 2017, 01:04 AM
Here are a few more happy snaps related to modifying the Metabones Speedbooster for Nikon to fit the IMS-Mount system.

Sorry pix would not upload due to some security token business I do not have the knowing of how to get around. Will keep working on it.

Bob Hart
May 27th, 2017, 01:29 AM
Nother try at uploading pix..

Edit. - Unsuccessful. I must have broken dvinfo. It was a good thought whilst it lasted. I have a near-identical post on reduser with the happy snaps here :-

http://www.reduser.net/forum/showthread.php?158270-Metabones-Speedbooster-BMPCC-to-Nikon-Mount-adaptor

Here follows some supplemental info.

I place four white patches of adhesive around the optical cell at the rear. I had to use this because the original lock screw remains on the M4/3 adaptor which the IMS adaptor replaces. The lock screw on the original M4/3 adaptor is not good design because it forces the optical cell to rest slightly off-axis in the Speedbooster body because the sloppy fit of the threaded section.

I understand that this defect has been fixed in later Speedbooster builds by using a threaded collar on the rear of the optical cell to lock it squarely in place once the backfocus has been correctly set. I have not yet enquired with Metabones if that lock ring is available as a separate part.

This whiter material is water-cleanable white bathroom sealer. I use it because it can be picked away with a needle or sharp knife point if I make an error. Do not put it on the threads. Just use it as a tack across the joint. Do NOT use clear silicon. If it gets on the glass, it is impossible to clean off without damaging the lens coating.

For setting backfocus ( collimating the optical cell in the Speedbooster ) , I position the camera facing directly down towards a test chart with my best known good lens with accurate focus marks pointed at the chart at a measured distance from the focal plane to the chart. I adjust the optical cell until the focus is correct. I do this by removing the lens and turning the optical cell from the front, the remounting the Nikon lens.

The reason why I face it straight down is so that the optical cell rests square to the optical axis due to gravity pulling on the threads along the optical axis. If you try to set the focus of the cell horizontally, it may tilt off-axis because the thread clearance is sloppy.

You will observe the cut-down rear of the Speedbooster in later photos I may manage to get uploaded. The vernier caliper shows the diameter which just allows the lugs of the IMS-Mount clamp ring to clear the Speedbooster body. Your IMS-Mount clamp ring may be of the lighter style used on the old Mini35-400 adaptor. In that instance, you will not have to remove as much as the 6mm forward of the rear of the Speedbooster body as I have. That 6mm you see on the vernier includes the 0.2mm brass shim which cannot be seen in the image.

In this modification, you lose the little lens support but if you make a curved spacer, you can remount the support if you revert the Speedbooster to its M4/3 mount.

The lug on the IMS-Mount clamp ring turns a little past the original position I had arranged because of wear on the metal I have used, aluminium plate from Austal Ships factory scrap bin. I hope you can see the amount of material I filed off the little release slider which pulls the locking pin back iknside the Nikon mount to release the lens. This must be filed down to allow the IMS-Mount lock ring lug to barely pass over without catching .

Be careful not to file too much plastic off the slide handle because it is molded around a thin metal structure and will split away if too much material is taken.

You must dismantle the Speedbooster body and clean out the grease so that metal filings do not stick inside. Be careful not to lose the little ball-bearings inside of the aperture ring. They are an absolute dog to put back in afterwards when re-assembling the Speedbooster.

I hope this is helpful. Good luck with it. I post this with strong recommendation that you do your own independent research to verify what I have suggested here. - You proceed entirely at your own responsibility. As the man says in Rambo 3, "If anything happen to you, I am not responsible".

Joe Gill
August 8th, 2019, 03:30 PM
BOB - THIS IS AWESOME! I will have to try this out and am looking forward to using it on my R1MX with the IMS as well...

Bob Hart
August 10th, 2019, 10:56 AM
Joe.


Unless you are using the R1 in windowed S16 mode, you may be better off with a Speedbooster Ultra optic to the R1's full sensor from full size stills lenses. Because of the front element of the Speedbooster optical cell getting is the way, you cannot use most traditional cine film lenses. However the lenses approved for the Speedbooster Ultra optic will work fine.

I have used one in a custom mount on a Blackmagic "Big" URSA camera which has enough throat space to accommodate the optical cell. You would have to do a custom design for the IMS-Mount but it would not be hard. Unlike the later RED camera families with the sensor set so far back in the camera body, there is more workspace.

I don't know where the IR/anti-aliasing filter sits relative to the sensor on the R1 so that could be an inconvenience if it is too far forward both for the IMS-Mount Speedbooster mod I made for the SI2K and the Ultra Speedbooster mod I made for the "Big" URSA.

The rear element of the Ultra optical cell sits further forward of the sensor than the S16 version for the original BM Pocket camera. The Ultra optical cell extends the usefullness of the "big" URSA in that it can more closely match the BM URSA Mini 4.6K in the available field-of-view for a given lens focal length and also adds nearly one stop to the camera's ability in low light.

I was actually encouraged in the "Big" URSA experiment by similar developments here on the forum with the later generation of RED cameras.

LEEUWIN TESTS 30JUN - YouTube

Bob Hart
August 10th, 2019, 11:00 AM
Further to the preceding reply here is some detail about the "Big" URSA fitment.

https://www.dvinfo.net/forum/blackmagic-cinema-camera/535768-metabones-0-71x-optical-cell-bm-ursa-4k-pl.html

https://forum.blackmagicdesign.com/viewtopic.php?f=2&t=60610&p=362389&hilit=URSA+0.7x+Speedbooster#p362389

If you want to go the Speedbooster Ultra 0.7x route, I will see if I can find my design drawings which would enable you to calculate the offsets for the IMS-Mount.