Canon EF lenses on MFT cameras?

Am not commenting on the specific case but adding glass elements through converters always deteriorate quality and I have tested this on land. Of course the speedboster is prime quality so the effect will be next to nothing but it is proven there is an effect at identical equivalent focal length.

What you say about the tokina does not make much sense the speedbooster is actually wider than the normal DG crop factor (1.42x vs 1.6x) so the image if at all will be wider  compared to a normal canon EF-S mount. Likewise the smartbooster keeps the 2x crop so the lens is even longer and narrower. The Smart Converter and tokina should be compared with speedbooser teleconverter and tokina at identical focal length. Am pretty sure the smart converter will do better there to but am unsure if the tokina is actually acceptable or becomes total trash

Interceptor, it is going to become very technical, so the terrain is difficult. In addition, I believe that the issue is not really important in practice...

There is, however, a technical white paper (by Brian Caldwell from Metabones, so there IS a conflict of interest) that states that sharpness and IQ are improved by the 0.71x speedbooster compared to glassless 1x adaption, and not detoriated. It is hard to believe, intuitively I would say (like you) that additional third party glass can only detoriate and not improve. But see here: https://www.dpreview.com/articles/8961802602/metabones-and-caldwell-photographic-annouce-speed-booster-lens-adapter-for-mirrirless-cameras

I did not read this technical paper (and probably never will), but think the improvement works via the reduction of the image circle in its diameter. That means the sensor is using a greater proportion of the circle and lens errors are reduced proportionally, while sharpness is inversely proportionally increased (in the case the additional glass has no or only minor effect on the overall optical quality, what is probably the case with the high quality Metabones booster). As with the Tokina, that is a low quality lens, I can subjectively confirm this improvement in IQ when I compare 0.71x to 1x (both sharpness and CA improve at compareable angles of view), although of course, I did never systematical testing. With the Canon 8-15mm I cannot notice any improvement, probably because the basal IQ of this lens is very good and so there is not much there that could be  improved...

Wolfgang

1 hour ago, Architeuthis said:

Interceptor, it is going to become very technical, so the terrain is difficult. In addition, I believe that the issue is not really important in practice...

There is, however, a technical white paper (by Brian Caldwell from Metabones, so there IS a conflict of interest) that states that sharpness and IQ are improved by the 0.71x speedbooster compared to glassless 1x adaption, and not detoriated. It is hard to believe, intuitively I would say (like you) that additional third party glass can only detoriate and not improve. But see here: https://www.dpreview.com/articles/8961802602/metabones-and-caldwell-photographic-annouce-speed-booster-lens-adapter-for-mirrirless-cameras

I did not read this technical paper (and probably never will), but think the improvement works via the reduction of the image circle in its diameter. That means the sensor is using a greater proportion of the circle and lens errors are reduced proportionally, while sharpness is inversely proportionally increased (in the case the additional glass has no or only minor effect on the overall optical quality, what is probably the case with the high quality Metabones booster). As with the Tokina, that is a low quality lens, I can subjectively confirm this improvement in IQ when I compare 0.71x to 1x (both sharpness and CA improve at compareable angles of view), although of course, I did never systematical testing. With the Canon 8-15mm I cannot notice any improvement, probably because the basal IQ of this lens is very good and so there is not much there that could be  improved...

Wolfgang

I have read the technical white paper it compares a bare lens vs the speed boster but this is not what the smart adapter does.

This guys has done a proper test equalising the apertures and distances and as expected the adapter without lens is the best in terms of aberration and distortion

http://www.verybiglobo.com/metabones-speed-booster-ultra-review-part-ii-canon-ef-501-2-l/

The only area that speedbooster seems to do better (but I can't see it from the image he posts) is centre sharpness while the edges are always better with the smart adapter

I think I am good to continue with smart adapter and Canon 8-15mm besides the fact that the smart adapter is effectively an extension tube means the lens quality is the same of the bare lens and no optical effects are introduced. The resolution of the lens should be totally unchanged and of course if the megapixel count of the MFT is the same of the SLR effectively get the same resolution at two stops wider aperture

Here are the files for 3D printing - I attach the files here: they were made with Tinkercad and are for the original Nauticam Zoom gear for Canon 8-15mm (and Tokina 10-17mm). One for 1x Smart adapter and one for 1x Smart adapter plus Kenko 1.4x teleconverter (as far as I remember this one need to be made a little bit shorter by abraser (or correct the 3D file)). For optimum fixing I treaded three tread for worm-screws into each adapter. These screws are used to fix the extension collar to the zoomgear.

I also insert the files for the Tokina, just in case someone is interested. As far as I remember also these parts need a little bit of abrasing until they fit (and, of course also the worm-screws)...

Canon8-15_1x_v4.stlCanon8-15_1x_K1.4x_v1.stlTokina10-17_1x_K1.4x_v1.stlTokina10-17_1x_v1.stl

All of them work nicely in my hands, I let them print via the site below using the cheap standard material, but no guarantee...

https://i.materialise.com/de/

Wolfgang

great stuff wolfgang I will do a write up when I have all the parts at hand!

Question re canon 8-15 with smart adapter. The lens is now 5mm longer so the glass is closer to the dome
Are there consequences on image quality? There is also a 35 extension ring in case this is not perfect?

Sent from my SM-A505FN using Tapatalk

4 hours ago, Interceptor121 said:

Question re canon 8-15 with smart adapter. The lens is now 5mm longer so the glass is closer to the dome
Are there consequences on image quality? There is also a 35 extension ring in case this is not perfect?

Sent from my SM-A505FN using Tapatalk

Hi Interceptor,

I did not test the 1x Smart Adapter with the 30mm extension (I use it for the 0.71x speedbooster). I always use 35mm (1x Smart) and 55mm (1x Smart +1.4x Kenko; 35+20mm).

Here a listing of the extensions how I use them (they are probably appropriate - but i did not do any test series nor measurement of entrance pupil; I choose them after corresponding with and as suggested by Nauticam and Zen):

Wolfgang

Ok so you are effectively adding the 5mm required

Sent from my SM-A505FN using Tapatalk

Am looking at your FOV table and am not sure how you tested that. 

The Tokina 10-17 with the speedbooster is 7.1 - 12.07 on the MFT body the Canon is 8 - 15 

The theoretical field of view calculations on an 8mm show less than 180 degrees however when you are looking at lenses you need to consider 22.5 mm as the diagonal which means at 8 mm you will get  fisheye anyway and the Tokina should vignette at wide end a little and at the same time because the crop with the speedbooster is 0.71 giving a total of 1.42 the lens will be wider than the nominal 100 degrees more 110 degrees.

At the same time the 8-15 should not go below 88 degrees with the smart adapter

I will run some comparisons with the panasonic 8mm fe

26 minutes ago, Interceptor121 said:

Am looking at your FOV table and am not sure how you tested that.

=> I did not test, but calculate. There are several construction types for fisheye lenses available, and Canon and/or Tokina may not be strictly 100% one type. Formula for equisolid fisheye was taken (FOV (equisolid fisheye) = 4 * arcsin (frame size/(focal length * 4)); "framesize" is diagonal im mm, in case you want diagonal FOV). The differences between types are not hughe, so problem is a little bit academic. Results are very good to estimate what you will get...

 The Tokina 10-17 with the speedbooster is 7.1 - 12.07 on the MFT body the Canon is 8 - 15

=> Exactly. Tokina with booster gives vignetting @10mm (=7.1mm) and one need to zoom out a little to avoid. Canon @8mm with Smart 1x adapter is just perfect (with booster there is, of course, even more vignetting, so 0.71x is even less desireable).

The theoretical field of view calculations on an 8mm show less than 180 degrees however when you are looking at lenses you need to consider 22.5 mm as the diagonal which means at 8 mm you will get  fisheye anyway and the Tokina should vignette at wide end a little and at the same time because the crop with the speedbooster is 0.71 giving a total of 1.42 the lens will be wider than the nominal 100 degrees more 110 degrees.

=> The calculations for equisolid FE gives 170 degree FOV. Since there are different construction types and real lens may be a mixture, 8mm cam give 180 degree for another (as Zuiko or Pana 8mm)...

=> The difference between the (calculated) 170.2 of the Canon 1x @8mm and 180 degree is not so much in reality: note that the spacing of degrees on the diagonal is not uniform with fisheye and the increments become smaller and smaller the closer you come to the corner (this makes the "fisheye" look). I am even not so shure that the specifications of manufacturer are so accurate, when they say it is 180 degree: it could very well mean some degrees less (maybe also more), when precisely measured...

At the same time the 8-15 should not go below 88 degrees with the smart adapter

=> see here the table with exact values (also FOVs for rectilinear WA lens are given - they can be calculated very accurate, of course):

are

I will run some comparisons with the panasonic 8mm fe

I am really eager to hear what your practical experience will be (I am very fond of the Canon 8-15mm with 1x - it became my standard WA lens). AF performance will depend on camera body: with EM1II C-AF+tracking work very well, while with EM5II S-AF is o.k. ...

Wolfgang

I am very interested in the 8-15 with smart adapter as it fills a gap. Right now I have nothing between the 130 degrees of the WWL-1 and 180 degrees of the 8mm fisheye

I also want to see how it performs around 105 degrees and what is the level of distortion as I am hoping to take the 8-18mm out of the mix 

For autofocus I only use single AF with wide angle shooting f/8 mostly for macro AFF but otherwise even just centre on single AF

2 hours ago, Interceptor121 said:

I also want to see how it performs around 105 degrees and what is the level of distortion as I am hoping to take the 8-18mm out of the mix

The level of fisheye distortion becomes reduced as the FOV is reduced upon zooming in. My guess is that the WWL-1 (and also the WACP) have similar amount of fisheye distortion when compared to "zoomed out" fisheye lens at comparable FOVs

Sharpness of Canon 8-15mm @15mm on MFT is superb as @ 8mm and with every fisheye (is'nt the Zuiko (or Pana) 8mm providing superb optical quality ?). The trick how WWL1 and WACP manage to get such IQ might well be that they produce zoomed-in fisheye optics together with the standard zoom lenses.

I did, however, never use WWL-1 nor WACP, so I cannot say for shure. Really interested to hear from someone who uses both WWl1 and Canon 8-15mm on MFT and can compare in practice...

Wolfgang

The level of fisheye distortion becomes reduced as the FOV is reduced upon zooming in. My guess is that the WWL-1 (and also the WACP) have similar amount of fisheye distortion when compared to "zoomed out" fisheye lens at comparable FOVs (the trick how WWL1 and WACP manage to get such superb IQ might well be that they produce zoomed-in fisheye optics together with the standard zoom lens). I did, however, never use WWL-1 or WACP, so I cannot say for shure. Really interested to hear from someone who uses both WWl1 and Canon 8-15mm on MFT and can say from practice...
Wolfgang

@Architeuthis one other consideration you want to make on the fisheye lenses in general is that although the specifications for field of view are given on the diagonal this is not really what matters to us. We are interested in the horizontal and vertical field of view.

When I compared the WWL-1 with the 7-14 I found out that while on the diagonal there is a difference on the horizontal the lenses behave similarly and on the vertical the 7-14 is actually wider. This is an outcome of the fisheye distortion that pulls the diagonal and then compacts it back in the frame in a barrel shape. In my tests the fisheye distortion on the wwl-1 is present until 2.5x zoom on the 14-42 and is less apparent in video just because the 16:9 crop is not affected by the vertical compression so the image is more straight

So in in reality 15 on the canon zoom on mft body with 1x converter is not the same as 12 mm rectilinear but is like a 13 on the horizontal and 14 on the vertical if it makes sense.

The teleconverter on the 1x for me makes very little sense as I am under what I would call true wide angle (90 degrees on horizontal corresponding to 8 linear)

When I look at the tokina on MFT body with speedbooster ultra this does not make too much sense as due to the crop factor being 1.42x and not 1.5 or 1.6 the lens remain too wide when fully zoomed

So the way I see it is that for what I need (once I decide to spend a fortune in lens ports adapters and glass)

Canon 8-15 (fundamentally from fisheye to 13.5 mm) 1x or corresponding to kenko + metabones on the nauticam chart

Leica 12-60 rectilinear (this is mostly for portrait work of mid size fish)

Olympus 60 mm macro

I am hoping to cover every wide angle situation with the 8-15 including split shots and wrecks, split will need a large acrylic dome to float

I will keep you posted as I go along...

@Architeuthis

First tests with tripod confirm what I thought on the horizontal a 12mm rectilinear is the same as the 8-15 at 14mm zoom 

Image quality is pretty much the same at the same aperture, the Canon 8-15 exhibits fringing while the panasonic does not but this is because the camera corrects it for all panasonic lenses. The purple fringing is only visible in corner at 1:1 pixel ratio and once you click remove chromatic aberrations in Lightroom it goes away without any loss of sharpness it is really tiny

Field of view I have not been able to do scientific test the canon is much longer lens and the additional lenght means the Panasonic is closer to the target of around 3 inches therefore the frame looks bigger however I am extrapolating that once the front are at the same distance it will be identical more or less

One consideration is obviously as this is DSLR size the dome is much bigger and so is the port compared to the Nauticam acrylic 4.33 or even smaller 3.5". If you are into wide angle macro this lens is not really the best combination and better rely on the panasonic for that.

Will report on the port system once it arrives...

I deposited some sample images for the Canon 8-15mm, 1x adapter and EM1II that I took during last week's exciting workshop at the Red Sea:

View archived post in: Canon 8-15mm, Nauticam 140mm minidome and EM1II

Wolfgang

Interceptor,  have you compared the 8-15 + 1x adapter vs the WWL-1 in the overlapping range?

How does the barrel distortion and image quality compare?

The overlapping range is 10 onwards I have not done scientific testing

However in general terms:

1. The native 14-42 will have corrections applied in camera for distortion vignetting etc that apply to both video and stills. For video is baked in. The lens has around 5% distortion so there is an element of softening of corners

2. The 8.15 mm is extremely sharp across the whole field of view and it won't be corrected in any form by the body there is an element of chromatic aberration uncorrected

I have taken some test shots with the 8-15 mm and my conclusion is that for stills clearly there is no comparison the 8-15 mm is better and you can correct ca and vignetting in raw processor

For video you need to check the amount of CA in the frame as this is uncorrected with the 8-15. The 8-15 will be sharper but it is not just about sharpness

Personally I would not use the 8-15 mm for video because of lack of IS and because the range 8-10 is not possible to cover with lights

Very interesting post for a newby in MFT.

I have a GH5 with Nauticam housing and I'm wondering if the Nikon lenses from my old equipment (Tokina 10-17, Nikon 105VR) will work properly with the Viltrox NF-M1 adapter. In the pictures it seems to be too big to fit in the housing.

https://cdn.shopify.com/s/files/1/0104/0380/7298/files/NF-M1-__13.jpg?v=1591360988

Any experience with this combination?

If it looks too big it probably is, the Nauticam housing has an n85 port and the ID is a maximum of about 77mm at least on the ports I have - I have seen references to need to remove the tripod foot from metabones adapters for other combinations.  Looking at the adapter, the foot appears non removable and I'd estimate it is about 80-85mm high so too big to fit through a N85 port - which it would need to do as it needs to be right at the port opening.