Metabones Speed Booster

Home · Register · Search · Software · Hosting · Software · Join Upload & Sell

FM Forums \| Alternative Gear & Lenses \| Join Upload & Sell
1 2 3 … 7 8 9 … 38 39 end

alba63
Online
• •
Upload & Sell: Off

p.8 #1 · Metabones Speed Booster

eosfun wrote:
Still photographers, especially those here at the boards of Fred Miranda and even more so in the alt section, have higher quality standards than average.

Yes, many here pursue the holy grail of the golden lens, and I have had the feeling that most (not all) images shown on this forum end actually here, on monitors, in an online forum, and are never printed. Some here even say they hardly ever print.

In real life photo display the subtle quest for the finest rendering qualities watched in 100% magnification has little to no meaning, as most of it is lost in a print, and even more, when seen from suitable viewing distances.

When let's say a 16x24" print made on an APS-c camera with speed adapter, seen from a distance of let's say 1-2 meters looks visibly worse than one made with an FF camera, then it will have a meaning.

I predict it normally won't.

But let's see and wait. There is of course a possibility that this adapter from Metabones will underperform and produce just bad results.

Jan 16, 2013 at 10:38 AM

cputeq
Offline
• • • •
Upload & Sell: On

p.8 #2 · Metabones Speed Booster

alundeb wrote:
cputeq wrote:
I am unsure if this also causes the m43 to magically gain 2 stops of light, though.

You are unsure, but still write lengthy posts trying to explain how it works?

Nice catch but slightly different topic

My (rather long, sorry) posts were most about light-per-time and whether that aspect has any real meaning in the context of a generic "FF vs crop" argument.

I am unsure of this particular aspect above because the details on the m43 adapter are sparse. For instance, knowing that the adapter only has to cover the m43 surface area, does this mean they will design it to not collect and condense all the light from a FF lens (but instead just sample an APS-C area of the lens, even if the lens is FF?) I could imagine this as actually being a possibility, to reduce the angles of light captured (which, combined with the fact they're from the very edge of the FF lens, might lead to some pretty horrific aberrations).

One would imagine if the m43 version of this adapter were to condense this FF-area light into m43, they would have no problems proclaiming a potential 2-stop gain, but I haven't read about this in any of their material or on the web. Perhaps they're being careful about any claims until they've actually developed a prototype for m43, or perhaps there is only a condensing of an APS-C section of the lens instead of the entire lens as guess above, etc.

Or perhaps they're holding off on this information until people start buying it - knowing that a lot of m43 users have a rather high ISO floor of 200, they don't want people to figure out they'll need to buy a lot more ND filters once they start shooting with lenses boosting their EV by 2

Edited on Jan 16, 2013 at 11:00 AM · View previous versions

Jan 16, 2013 at 10:54 AM

eosfun
Offline
• • •
Upload & Sell: Off

p.8 #3 · Metabones Speed Booster

Bernie, the "bundling process" is an important step that does effect image quality itself. If the downscaling was done hypothetically perfect and in a image neutral way you would be right. In reality though, every lens (and air-lens surface) effects the stream of photons such that the original image quality is touched. The biggest challenge for Metabones engineers without a doubt has been to "downscale" the image while keeping distortion and field curvature of the speedbooster under control and at the same time not introduce all kinds of colour fringing. So it's not as simple like the teleconverter magnifies aberrations while a condensor like the speedbooster makes weaknesses smaller. Both types of additional glass to an existing lens system introduce their own optical challenges.

It's true, we have to admire the creative optical engineers at Metabones. Dusting off this old idea and implement modern technolgy to it is interesting. It would be interesting to see if the same basics of thinking could lead to front mount lens converters to make existing telelenses faster, or have a stacked version of the speedbooster to mount on a Pentax Q for instance

You see, there are much more possibilities than you might have thought. It starts with thinking out of the box, but I am afraid there are still quite some technical (and business) hurdles to be taken before it's all going to be as nice as we want it to be. And then I believe a lot of the progress of optical engineering will first go into innovations in our smartphones, not so much in our alt gear or D-SLR's

Jan 16, 2013 at 10:54 AM

carstenw
Offline
• • • • • •
Upload & Sell: Off

p.8 #4 · Metabones Speed Booster

alba63 wrote:
- It is the opposite - as many have said - of a teleconverter which magnifies everything, also the weaknesses of the lens. This makes the opposite, it makes the weaknesses smaller.

I am not certain on this point, but I don't think this is right. The CA, for example, imaged on the sensor will be smaller at that location, but by the time you look at the full image on a screen or print again, the sensor size is irrelevant and the CA is back to the same size. Plus, of course, any CA that the adapter adds. Similarly for other aberrations.

Jan 16, 2013 at 10:57 AM

cputeq
Offline
• • • •
Upload & Sell: On

p.8 #5 · Metabones Speed Booster

carstenw wrote:
alba63 wrote:
- It is the opposite - as many have said - of a teleconverter which magnifies everything, also the weaknesses of the lens. This makes the opposite, it makes the weaknesses smaller.

I am not certain on this point, but I don't think this is right. The CA, for example, imaged on the sensor will be smaller at that location, but by the time you look at the full image on a screen or print again, the sensor size is irrelevant and the CA is back to the same size. Plus, of course, any CA that the adapter adds. Similarly for other aberrations.

That seems to be my thought process also. It's not as if the adapter will somehow (that I know of) reduce aberrations that the lens has already output, and might potentially introduce even more as it now compresses the exit light of the FF lens into a smaller APS-C area, which is going to require even more bending of light.

I mean, they're touting MTF increases with the lens, which I could see (over a "bare" adapter) but I haven't come across some really good A/B comparison images at full size, especially in situations that would induce bad CA, etc to compare edge image quality.

Jan 16, 2013 at 11:16 AM

theSuede
Offline
• • •
Upload & Sell: Off

p.8 #6 · Metabones Speed Booster

Looking at the optical traces of the adapter, it follows all conventional rules for non-anomalous dispersion glass. That is, it has the same polarity of errors as any normal photographic lens would have.
This means from the image's point of view that any normal aberration in a lens will be worse, not better. The SIZE in µm of the aberration might be smaller, but the size in part of image width will be larger.

So, even if an aberration shrinks by a factor of 1.3 (13µm displacement >>> 10µm displacement) which seems about right when looking at their published MTF over image height, that aberration then needs to be enlarged by 1.5x to get the same image.

Lets say you have two 10MP cameras, one FF and one APS. That's about 4000 pixels image width, same for both. Their pixel sizes will be 9µm and 6µm respectively, to get 10MP.
If you have an aberration that's 13µm on the FF camera, that's the same as 1.44 pixels in the image.
That same aberration will be 13/1.3 = 10µm on the APS camera (with adapter). 10µm/6µm pixels = 1.67pixels.

So even if the aberration is smaller, the impact on the actual image is larger. 1.67 pixel widths versus 1.44 pixel widths in the original.
........................

Giving it some thought since yesterday, I've slightly changed my mind though. About the value of the adapter...
For people accepting manual focus as something a part of normal photography, this is well spent money. With the improvements in ergonomics with the latest NEX cameras (also µFT) this might actually improve a lot of things.

What you get:
More field of view per focal length (good, shorter focal lengths are often worse for adapted lenses)
Higher aperture value on the adapted lens to get "the same image" - this in itself is the most important part!
Slightly (7mm?) shorter overall camera/lens package than using same lens with normal adapter (though you'll need a longer FL lens to get the same image)

One of my main reasons for using larger format systems is that the image quality per DoF (at shallow DoF's) in a real image increases almost exponentially. My 130€ AFS50/1.8 @ F2.8 ISO400 on my D800 literally walks all over the 700€ Zeiss 35mm @F2.0 ISO200 on a NEX camera. If just a little bit of that disadvantage disappears, priorities might change - considering the size difference of the packages.

Jan 16, 2013 at 11:28 AM

alundeb
Online
• • • •
Upload & Sell: Off

p.8 #7 · Metabones Speed Booster

cputeq wrote:
alundeb wrote:
cputeq wrote:
I am unsure if this also causes the m43 to magically gain 2 stops of light, though.

You are unsure, but still write lengthy posts trying to explain how it works?

Nice catch but slightly different topic

Oh, that way

It seems that the m43 version will get the same 0.7x conversion factor. The white paper includes an example of the Nikon AF-S 35 f/1.8 DX lens converted to 24.5 mm f/1.2.

Jan 16, 2013 at 11:31 AM

alundeb
Online
• • • •
Upload & Sell: Off

p.8 #8 · Metabones Speed Booster

theSuede wrote:

What you get:
More field of view per focal length (good, shorter focal lengths are often worse for adapted lenses)
Higher aperture value on the adapted lens to get "the same image" - this in itself is the most important part!
Slightly (7mm?) shorter overall camera/lens package than using same lens with normal adapter (though you'll need a longer FL lens to get the same image)

What is your take on the claimed improvement in telecentricity?

I think the EF-NEX version saves 4mm in the package depth, compared to using the same lens without any adapter on an imagined FF camera..

Jan 16, 2013 at 11:40 AM

eosfun
Offline
• • •
Upload & Sell: Off

p.8 #9 · Metabones Speed Booster

In real life photo display the subtle quest for the finest rendering qualities watched in 100% magnification has little to no meaning, as most of it is lost in a print, and even more, when seen from suitable viewing distances.

That assumes that people are printing, which less and less people are doing these days. However, the amount of people who do pixelpeeping is enormous. And I am not talking about pixelpeepers, photographers like us here at Fred Miranda. I am talking about pixelpeepers like casual shooters, for instance my own mother (76) when she, with the well known movement of her indexfinger and thumb, zooms in on an iPad to just view the face of her grandchild and says "this picture is a failure becuase John's face has purple edges" or "it's a pity you can't clearly see Ann in the corner smiles". She doesn't know anythin about, astigmatism, CA, and smudged corner pixels. Our habits of looking change rapidly and pixelpeeping isn't just for quality seeking photographers anymore, but a common habit for casual shooters and many people we think of as less quality conscious.

Jan 16, 2013 at 11:42 AM

alundeb
Online
• • • •
Upload & Sell: Off

p.8 #10 · Metabones Speed Booster

eosfun wrote:
In real life photo display the subtle quest for the finest rendering qualities watched in 100% magnification has little to no meaning, as most of it is lost in a print, and even more, when seen from suitable viewing distances.

That assumes that people are printing, which less and less people are doing these days. However, the amount of people who do pixelpeeping is enormous. And I am not talking about pixelpeepers, photographers like us here at Fred Miranda. I am talking about pixelpeepers like casual shooters, for instance my own mother (76) when she, with the well known movement of her indexfinger and thumb, zooms in on an iPad to just view the face of her grandchild and says "this picture is a failure becuase John's face has purple edges" or "it's a pity you can't clearly see Ann in the corner smiles". She doesn't know anythin about, astigmatism, CA, and smudged corner pixels. Our habits of looking change rapidly and pixelpeeping isn't just for quality seeking photographers anymore, but a common habit for casual shooters and many people we think of as less quality conscious.

Pixel peeping, not to mention stock agencies.

The idea of sharing high resolution images digitally allowing for dynamic viewing is a scaring thought for many professional photographers. Amateurs and technology together will force this to happen though, and professionals will have to find a way to deal with legal rights.

Jan 16, 2013 at 11:50 AM

theSuede
Offline
• • •
Upload & Sell: Off

p.8 #11 · Metabones Speed Booster

alundeb wrote:
What is your take on the claimed improvement in telecentricity?

I think the EF-NEX version saves 4mm in the package depth, compared to using the same lens without any adapter on an imagined FF camera..

Oh, yup 4mm. Remembered 7mm or something, my error

Telecentricity is better for all use cases, given that the lenses you adapt are of similar construction. But it works even when comparing a "worse" case (symmetrical) against a "good case" (distagon/telecentric). Let's compare the Zeiss Contax 50/1.7 (Semi-symmetrical Gauss/Ultron) with the modern ZF35/2 (Distagon):

Exit pupil distance & size:
50/1.7 >> 38mmBFD+29mm = 67mm, 41mm diam (almost symmetrical).
35/2.0 >> 38.5mmBFD+27mm = 65mm, 35mm diam (short distagon) - from memory, might be wrong!

Now give the 50mm lens between +5-10% in exit pupil distance, this seems to be about right for a near-symmetric lens from the data they've published, and you get:
50/1.7 >> 33mmF1.17 with e.p. at 72mm distance, 38mm diam.
This means that you get a 33mmF1.2 with the same worst-case angle (cross-angle from furthest aperture edge to furthest sensor edge) as you get from the 30F2.0, and this is very good.

This gives a slightly lower maximum edge angle compared to the 50mm lens on a FF-size sensor, but also hard mechanical vignetting on large aperture lenses. Note that they state "no aperture restriction with main lenses > F1.26" - but only for the paraxial chief ray, i.e only the absolute center point of the image.
So worst case possible (at the NEX APS sensor edge) is about the same as a symmetrical 50/1.2 without adapter, in almost all other configurations you get about 2-15% better (less angle) than original.

What the adapter can't do is to be reconfigured for short-flange main lenses, that is: You can't make something like this for an M-mount lens.

Jan 16, 2013 at 04:48 PM

alwang
Offline
• • •
Upload & Sell: Off

p.8 #12 · Metabones Speed Booster

eosfun wrote:
The enthusiasm shown for this metabones speedbooster so far is mostly from videographers. As long as they work in the 720P and 1080 resolution area, and not in the 4K class, this speed booster is a nice additional add on to their setup.

Not sure why: don't most DSLR videographers use 5dmk2/3 cameras anyway? I guess there are some GH2 users, but not sure what advantage you'd have using a GH2 with this adapter over a FF DSLR for video. It's not like you really care about the smaller camera size for video, since you usually need more bulk to hold it steady anyway.

Jan 16, 2013 at 05:13 PM

snapsy
Online
• • • •
Upload & Sell: On

p.8 #13 · Metabones Speed Booster

alwang wrote:
Not sure why: don't most DSLR videographers use 5dmk2/3 cameras anyway? I guess there are some GH2 users, but not sure what advantage you'd have using a GH2 with this adapter over a FF DSLR for video. It's not like you really care about the smaller camera size for video, since you usually need more bulk to hold it steady anyway.

The GH2 has noticeably higher video resolution than the 5DM2/5DM3 (true resolution, so no moiré), and the hacks provide flexibility on data rates.

Jan 16, 2013 at 05:17 PM

ISO1600
Offline
• • • •
Upload & Sell: On

p.8 #14 · Metabones Speed Booster

The only FF HDSLR's that are REALLY well suited for video at this point are:
5DIII
1DX
D800

Everything else has too low of bitrates, too much moire, or other problems.
The 7D, with newest firmware, AFAIK, is a VERY capable HDSLR, really only limited by sensor size. This overcomes that.
The NEX cameras I'd imagine are very capable vid cameras as well, especially with their form factor and tilting LCD screens.
This speed booster/adapter seems largely targeted at NEX and Sony E-Mount large-sensor camcorders like the FS100.

Jan 16, 2013 at 05:42 PM

michael49
Offline
• • • •
Upload & Sell: On

p.8 #15 · Metabones Speed Booster

eosfun wrote:
....

- Something that I look forward to see, is how this speedbooster effects bokeh quality. To my taste, in general tele converters have a destructive effect on bokeh quality. I am afraid this won't be different for the speedbooster and that the design leaves the bokeh quality of the original lens intact. But the fact that lens design becomes decentered for the position of the aperture (an important factor for bokeh quality) is not very promising. I hope my suspicion is wrong, since the smaller DOF which hopefully comes in a nice bokeh quality, should be one of the selling points of the speedbooster.
....

This will be the big one for me.

I don't want to put my Zeiss C/Y 50 1.7 on a NEX with the speedbooster for the ultimate in image quality, I want it for the true 50mm FF perspective and the narrow DOF and associated pop that such a lens provides when paired with a FF sensor - and bokeh is key here.

Edited on Jan 16, 2013 at 06:00 PM · View previous versions

Jan 16, 2013 at 05:56 PM

theSuede
Offline
• • •
Upload & Sell: Off

p.8 #16 · Metabones Speed Booster

The real kicker (and also the main part of their customers) will be the adapter on the BlackMagic box, or the Sony FS- or VG-series APS videocameras with NEX E-mount.

Real resolution
BM: 2432x1366
GH2: ~1500x950
Mk2&mk3: ~1350x800

The video users are typically less sensitive to item price and manual everything.

Jan 16, 2013 at 05:57 PM

AhamB
Offline
• • • •
Upload & Sell: On

p.8 #17 · Metabones Speed Booster

ISO1600 wrote:
The 7D, with newest firmware, AFAIK, is a VERY capable HDSLR, really only limited by sensor size. This overcomes that.

Actually, it doesn't. This type of adapter only works on mirrorless cameras with a short registry distance. It could work on a 7D with medium format lenses.

Jan 16, 2013 at 06:12 PM

alwang
Offline
• • •
Upload & Sell: Off

p.8 #18 · Metabones Speed Booster

ISO1600 wrote:
This speed booster/adapter seems largely targeted at NEX and Sony E-Mount large-sensor camcorders like the FS100.

Yeah, it will be interesting to see comparisons between the FS100 with this adapter against the VG900, both using the same lens.

Jan 16, 2013 at 07:26 PM

ISO1600
Offline
• • • •
Upload & Sell: On

p.8 #19 · Metabones Speed Booster

AhamB wrote:
ISO1600 wrote:
The 7D, with newest firmware, AFAIK, is a VERY capable HDSLR, really only limited by sensor size. This overcomes that.

Actually, it doesn't. This type of adapter only works on mirrorless cameras with a short registry distance. It could work on a 7D with medium format lenses.

haha oh yeah, forgot that little detail

Jan 16, 2013 at 07:34 PM

briantho
Offline
• • •
Upload & Sell: On

p.8 #20 · Metabones Speed Booster

I'm interested in this adapter, especially as it will be available for Contarex (!). However, it looks like the 21 Biogon will be a no go with this adapter, or would it work?

Jan 16, 2013 at 07:38 PM

1 2 3 … 7 8 9 … 38 39 end
FM Forums \| Alternative Gear & Lenses \| Join Upload & Sell 1 2 3 … 7 8 9 … 38 39 end