conurus

I work for Metabones and let me confirm that this is indeed a firmware bug as well as to apologize to you for the inconvenience. This problem is that IBIS is OFF and you cannot turn it on with the camera's menus. It occurs with only one lens, the Sigma 50-100/1.8 DC HSM Art. A firmware fix will be published in a matter of days. I acknowledged this problem on eoshd at the beginning of the month and promised that a fix was on the way, but I suppose my comment might have been taken out of context. It certainly was NOT a blanket "GX85 IBIS didn't work with Metabones". (Link to my previous comment below.) There is not much else to say, except that the fix will be here soon, and we will have the ability to select between OIS and IBIS (something that even Panasonic's own lenses do not support, for example if you have an old Leica 14-50/2.8-3.5 in SLR Four Thirds mount.)

Sorry about that but I will have to discuss with our tech support about this. Something like "we are having engineering look into this" would have sufficed but I will ask them to reply even in situations where they don't have a real answer yet. Yes you will be able to choose between OIS and IBIS with any IS lens. This is not final but the caveat is you can't change your election without powering down the camera and powering it on again.

I am in charge of firmware at Metabones and I would like to give an official response with regard to GX80/GX85. We do not expect any EF mount to be able to take advantage of Dual I.S. in the foreseeable future. Either OIS or IBIS should be used, but not both. Andrew reported that "They seem to both operate and ignore each other, leading to problems." We could confirm that it is indeed the case when we downgrade to 1.4 or before. This was fixed in 1.5 a long time ago (Feb 2015). It is easy to tell one way or another: if both OIS and IBIS are active simultaneously, vibration will be overcorrected by twice as much and there will be just as much jitter as no IS, which is indeed what we see with adapter firmware 1.4 or earlier. If you are seeing just occasional slow jitter, it is likely you are observing the result of the vibration exceeding OIS' capabilities of that particular lens. We have a 50-100/1.8 on order, but before that arrives, the most likely hypothesis drawn from Andrew's and our own observations is that adapter firmware 1.5 and after breaks the 50-100/1.8 and the old firmware 1.4 and before breaks all IS lenses (no AF and double correction). If this hypothesis is correct, you are okay with the current adapter firmware except for the 50-100/1.8 and except for the fact that you will be always using OIS but never IBIS. A new adapter firmware will be around the corner with which we expect the 50-100/1.8 to be fixed as well. There will be a new option to choose between IBIS and OIS on Panasonic in the next firmware, something that is just natural for Olympus but not yet possible for Panasonic. (On Olympus if you turn off lens IS, IBIS comes on, but on Panasonic the lens IS switch turns off both OIS and IBIS.) The procedure to turn off OIS and turn on IBIS on Panasonic is not going to be pretty, but it will be doable. I enjoy reading Andrew Reid's very informative review and tests, we greatly appreciate his work and we are always listening to customer reports of any issue which might arise. However, we would like to advise our customers that writing to support@metabones.com is the only way to ensure a problem is looked into in a timely manner. I do read eoshd often, but it is not a reasonable expectation that just because something is reported or discussed on eoshd we must be working on it. We appreciate you informing us of any issue via the official support channel in advance. That would greatly expedite the fix.

Rolling shutter is caused by taking a long time from the reading of the beginning of a frame to the end of a frame, and this is exactly what it is. Full readout of 24 megapixels is so much data that it takes a huge amount of time to readout. That the A6300 had worse rolling shutter than everything else should be no surprise. It would get better at 30p a bit, I guess to a similar level as A7RII S35. I think camera manufacturers can introduce low rolling shutter modes by turning off oversampling but use line-skipping instead. (This will of course create other issues, there is no magic bullet in engineering.) For instance A7RII full-frame has lower rolling shutter than S35. Give people a choice and see how it ends up, Sony.

20mp sensor is 5184 x 3888. How can it be 6k readout? 6k is 27mp in 4:3. I hope they will do exactly that and go back to multi aspect sensor like gh2. But 5k oversample to 4k, no crop at 60p is what makes people buy the gh5, imo.

If we do the math, 2.4x oversample at 24p would have meant that 1.92x oversample is possible at 30p, but only 1.6x is used. There is a 20% headroom remaining for heat dissipation. This time, it is 24p overheating that we have to worry about, although without the actual camera this remains pure speculation. If it were Blackmagic, they would simply put a fan in the camera, but Sony couldn't do that. It seems like with firmware alone Sony has resolved most of the issue, because the occurrence seems to have become rare recently. I am very hopeful that Sony has learned and pay extra attention to heat dissipation in its newer body designs. Fingers crossed.

I think the existing ULTRA 0.71x is adequate for the following reasons: The "6" series has a 2-year cycle. 2 years later image processors will be faster and be able to do a full readout at 30p (The counter argument to this is, what if the sensor resolution increase more quickly than image processor performance increase?) There is no crop at 24p. The people doing cinematic work, which is Speed Booster's primary audience, are already happy. The crop at 30p is not as bad as it sounds. My calculations arrived at 1.9x, which after boosted becomes 1.3x. Canon 1D-C also has a 1.3x 4k crop. Do we HAVE to beat 1D-C up? If we do, there is always the A7RII S35 + Speed Booster, and for A6300 we have already done that at 24p (see 2). Dwelling more on point number 1, I conjecture something magical will occur at the next resolution increase. The horizontal resolution will become exactly twice UHD, i.e. 7680, for a 7680 x 5120 resolution (39MP). A simpler oversampling algorithm can be employed, and UHD 60p is not unthinkable, without consuming any more processing power. 39MP is the magic resolution which enables 8k. It is just a matter of time, if the industry is indeed heading towards what it says is heading.

Let me try to rationalize the crop from an engineering standpoint. Look at the A7RII. Can it do a 42MP full sensor readout and downsample/record to UHD? Can't do that. Why? You can only read so many pixels so quickly. There is always a bandwidth ceiling, be it the actual data rate coming out of the sensor, or the speed of the processing engine required to encode it. The APS-C crop of A7RII is about 18MP, and we can do a readout of that center crop (about a 5K readout) and downsample/record to UHD at 30p. The A6300 is subject to a similar readout ceiling (5K max). That, according to my calculations, corresponds to a 1.9x crop factor, akin to the 'two-ish' crop reported by reviewers who have actually tried their hands on the camera. If you lower the frame rate, you can read a larger area of the sensor within the same bandwidth. So, a 6k readout is only possible at 24p. At 30p we can do a 5k readout max. GX8 had worse 4k crop than G7 because the former had higher resolution. So what Sony could have done was to lower the resolution of A6300 until it can do a full sensor readout at 30p. But that would have been hard for them to swallow a feature regression compared to previous models and even the A5100. We should have been happy that Sony did not further increased the resolution beyond 24MP, which would have made the crop even worse. So, Sony has in fact made the best design compromise here - same 24MP means neither the pixel counters nor the high ISO affectionados would be (overly) offended, and as a corollary the cinematic purists for the first time after NX1 and A7SII had their full sensor 4k, a feat that Panasonic had yet to be able to match. This design compromise has made the least number of people unhappy, so I would see this as a glass half full instead of a glass half empty. This approach is exactly what Panasonic should learn and follow, IMO. Oversample over a larger area of the sensor to reduce the crop factor, even when it is only possible to do so at lower frame rate. Imagine a G7 with no 4k crop at 24p. Technically it is entirely doable.

Please see my calculation at the bottom of page 11. A 2.3x crop would mean NO oversampling at 30p which doesn't sound like what the press release is saying. Using the figures from the press release I arrived at 1.9x which can be boosted to 1.3x, which would be better than NX1 because the latter cannot be boosted.

The glass element protruding from the rear didn't clear the rectangular baffle of my NEX-7, and the A6300 had a similarly shaped rectangular baffle.

On 30p crop: the press release says 2.4x oversampling (6k equivalent) at 24p. This 2.4x is area, not linear, because the press release says "2.4x as many pixels". Linear would have been 1.55x. 3840 x 1.55 = 5949 (6k, in fact). The press release says 1.6x at 30p. Not having the camera itself, let's apply the same calculation: 1.26x oversampling (linear), 4857 pixels across (5k readout). We know S35 A7RII can do a 5k readout at 30p and downsample to UHD with excellent quality, so A6300 is just bringing that technology over. The sensor width is 23.5mm. So the cropped width is 4857 / 6000 * 23.5 = 19mm, that is slightly more than half the width of full frame. When people say the crop is 'two-ish' it is in comparison to full frame. Hope my logic is sound! The 30p crop is not as bad as it appears.

All M42 to EF adapters by necessity interfere with the electrical contacts. (Milling a slot to clear the contacts would weaken the structural integrity of the adapter.) Anodized aluminum or black paint if not scratched are good insulators (you can tell if the color is black). If the color is chrome it is not possible to know for sure without asking your adapter vendor. Both untreated aluminum and chrome plating conducts electricity, but clear anodization is also an insulator (the downside is you may not be able to tell if it is scratched or not). Ask your M42 adapter vendor whether their adapter would short, and use a multimeter to verify if the adapter conducts electricity to be sure. If the insulation coating is scratched and starts conducting electricity stop using it and replace it.

Dumb with no electronics whatsoever.

Duty free does not equate to VAT free.

There will be no electronics and no VR support for Nikon G. Pricing has not been finalized but is expected to be around $449.

KineRAW S35: the mount is Short-FFD KineMount (Sub-PL Mount). Does that mean it is physically a PL mount which simply sits closer to the sensor but the PL mount "adapter" is actually a PL mount extension tube? Does anyone know what the flange distance of that is?   M4/3: no good news on that yet, unfortunately. Manual focus Speed Boosters for MFT will come first but active Canon EF will appear later.

Anyway, Metabones just specifically tested this exact setup: ZE 35/2 + Speed Booster + FS700. Works with or without WO button pressed. There has been only one firmware for Speed Booster (version 16) so far, so your SB already has the latest firmware. We can thus conclude you have a defective unit. I will send you their address for product returns.

A Canon lens works regardless of whether the Speed Booster's WO button was pressed or not?   I spoke with Metabones on the phone about your case. They were like... we had a FS700 and a ZE 35/2... so of course it worked! But I asked them, whether they had tested that exact combination, and they were like... errrh... maybe... not 100% certain. Maybe they did test that exact combination but they couldn't remember. Anyhow they said they were going to retest that exact combination. Let's see how it goes.   By the way did you happen to have a multimeter handy? You may contact me directly at conurus at gmail dot com. I would like you to do 1 more test if you wouldn't mind...

The account named "Metabones" on Twitter was not affliated with Metabones. I spoke with Metabones and they had no idea who that was. I didn't think Metabones had ever attempted to take it down either, so its disappearance probably had nothing to do with Metabones. Anyhow, in my personal opinion they could have explicitly stated they were not official, at the very least.

The official position is that Metabones is contemplating that possibility, but could neither confirm nor deny whether there are concrete plans to do a medium format Speed Booster. Basically everything about the roadmap that can be disclosed is already in the FAQ section of the Metabones site. All the rest are "neither confirm nor deny". We are all ears to your ideas but your understanding is appreciated that we have to operate with a certain degree of secrecy but for various reasons not everything could be disclosed publicly.   If you would allow me to put down the official hat but just state my personal opinion, however, a medium format to full frame SB would not have rocked the world like the full frame to APS-C SB had. By that I mean, for APS-C there is no way to get a 17/1.0 other than 24/1.4 + SB. However, name any medium format lens, multiply both its focal length and aperture by 0.7 or thereabouts, and there is always already a full frame lens which could do its job. e.g. Mamiya 80/1.9 => 56/1.3, pretty close to a 50/1.4. There are no medium format lenses that I am aware of which were fast enough or wide enough for some breakthrough to occur in the full frame world.

Within Metabones we had discussions of a "universal" solution where you stack another adapter on top. EF appears to be the favourite candidate for a "universal" mount but we did not want to leave out Canon FD and Minolta MD. Therefore, we haven't reached a concrete proposal on that yet, but we are open to ideas. For now our highest priority is to ship what we promised (C/Y, Rollei, Contarex, Nikon G, m43). The second highest priority is to follow up on customer suggestions on Canon FD/Minolta MD (the group which is shorter than EF in flange distance). We must remain focused on these first, but once they are done we are open to ideas.

I suppose you are thinking of the Digital Bolex? The C-mount has a very small opening diameter. The upcoming m4/3 Speed Booster optics have a large rear element and wouldn't fit in. However, if it is indeed the Digital Bolex you are thinking of, the saving grace is, from their specs sheet...   "C-mount comes standard; Optional PL, EF, B4. M, Micro 4/3, turret in development."   They could use a m43 mount, even if passive. (C to m43 adapters are readily available, after all.)

They have a ZE 35/2 - the exact same as one of the lenses you have. So, I am pretty sure this is not a compatibility issue. Also, other users have been using ZE lenses just fine. This is very mysterious indeed.

That sounds very strange... but let me ask a couple of questions... if you try holding down the WO button on the adapter while you attach the lens, does your camera boot up successfully? was your regular Metabones Mark I or Mark II? What was its firmware version?

The Speed Booster is shorter than a regular adapter. The optics shortens the flange distance so unless there is some way we hacksaw 6mm away from the EF BMCC it is not going to work.