Olympus M. Zuiko 12 mm f/2 OR Voigtlander Heliar 12mm f/5.6 ?

[Zupfel Gumba]

Hello G'day there!

I'd like to kindly ask if anyone would be so nice as to give my some feedback/thoughts on using either a
Olympus M. Zuiko Digital ED 12 mm f/2.0 (micro-four-thirds lens, 35mm EFL = 24mm)
or a
Voigtlander Ultra-Wide Heliar 12mm f/5.6 ASFH II (35mm full-frame lens)
on a micro-four-thirds camera (Olympus EP-1, crop factor 2) for sports photography. After looking up both lenses' specifications, I find myself undecided which one may suit my needs better. I'd be very happy if someone with a better understanding than mine could help me out with a few comments! Thank you very much in advance!

What do I shoot?
I'm into sports/action photography and am mostly using wide-angle lenses. I do photography as a hobby, and am typically shooting once a week. A typical example shoot would be motorcycles racing close by, with the camera only centimeters away or a mountain bike jumping over some small ramp with me behind/under the ramp shooting from below. I'm not using a flash and do typically use exposures < 1/1000 to freeze motion. I've purchased a micro-four-thirds system as 1) I needed a small/light weight camera to carry around (in my backpack while riding my mountain bike), 2) I wanted a reasonably priced system that could be replaced if damaged by impact or dirt/sand and 3) I was eger to get a live-view camera as I'm typically not taking off my helmet while shooting these pictures (I can't use a viewfinder as my helmet is in the way). I'm currently using an Olympus M.Zuiko Digital 17mm f/2.8 (35mm EFL = 34mm) pancake lens and would like to purachse a wider lens as 1) the photos I'm taking are often clipped too much and 2) I'd like to get closer to the action to exaggerate perspective for more 'action-look'.

Why I'm undecided
Both lenses are rectilinear 12-mm-focal-length primes and are roughly the same price (around $800). The Olympus is a micro-four-thirds lens (image circle 23mm) whereas the Voigtlander is a full-frame lens (image circle 44mm). Mounted on a micro-four-thirds camera with an adapter (which can be purchased for around $100), the Voightlander lens should give the same angle-of-view than the Olympus lens (is this actually true?). I like the idea of purchasing the full-frame Voightlander as I can potentially use it with other mirrorless cameras if I happen to save up some money (Sony NEX with crop factor 1.5 or full-frame Leica M9). The main draw-back is that the Voightlander is a f/5.6 lens whereas the Olympus is a f/2.0 lens -> that makes the Voightlander roughly 8 times less sensitive (5.6^2/2^2) - at least this is what the theory says. I'm unsure as to whether or not this is actually true in practise - it doesn't seem right to me at all, the Voightlander is a much larger lens and a factor of 8 would imply that I would have to shoot in ISO800 instead of ISO100, drastically decreasing image quality.

I don't have enough experience to answer the question of whether or not this theoretical difference in sensitivity will actually turn out as a factor of 8 in practise, which my gut-feeling tells me it won't. I have recently used a Peleng 8mm Fisheye f/3.5 M42 lens on my EP-1. When comparing my ISO and shutter settings between shooting with the Peleng 8mm f/3.5 and shooting with the Olympus 17mm f/2.8 I find that I'm basically using the same settings even though the Olympus 17mm ought to be 1.5 times (3.5^2/2.8^2) more sensitive. I don't see this difference in sensitivity in practise and am unsure whether or not one can 'trust' this difference derived from simply comparing f-numbers.

Can anyone help?

Thank you so much,
Gumba

 

[Hamish Gill]

Hi there
Welcome to the forum!
Sorry for the delay in activating your account ... It was night time here

Right ...
Complicated question ...
At least I feel you may be making it complicated by your distrust of the maths!

An f2 lens set to f2 will have 3 stops more light gathering capability than a 5.6 lens at 5.6 ie wide open the Olympus lens is a faster lens ... That is fact!
Sensitivity is not the word I would use though ... But that is maybe just semantics!

If you are not going to shoot in low light, and you don't mind not having auto focus or proper auto exposure then why not give te voigtlander a go?
I love cosina voigtlander glass, I have a good few lenses (see my signature) and I am never disappointed ...

Of course, buying the m mount lens will allow you to mount it to te NEX range or indeed an m9... But are you likely to switch systems? Buying for cross system compatibility doesn't seem to me to be the logical thing to do when you are loosing so much functionality...

If i were in your shoes, I would probably go for the oly! with the added light gathering power and autofocus the Olympus lens offers a much more versatile package!
The m4/3 cameras are great cameras, focusing manually is not always easy, and noise starts creeping in at fairly low isos ... With those factors in mind I think the oly is the right lens!

It is also my opinion that the adapters for fitting M mount glass to m4/3 bodies are designed for those with a m mount camera and lenses set up wishing I use them on an inexpensive digital body ... Not the other way round ... If that makes sense?

Just my opinion, it's based on specs and logic rather than having used the kit in question ...
That said, I do have a lot of cv lenses and I did have a Panasonic gf1 ... I found using the cv lenses on the gf1 unnesasarily difficult!

 

[Zupfel Gumba]

Thanks heaps for your reply Hamish,

I am shooting with exposures <1/1000 so sensitivity/f-number is of big importance to me. Using f/2.8 on the 17mm Olympus I can typically get away with ISO100 shooting in the forest (sort of dark) during lunch time.

I think I better trust the f-number comparison - but a factor of 8 difference still seems very high to me.

I don't use autofocus - I'm setting the focus before the action as otherwise shutter-lag is far too high. I'm shooting in contineous mode as soon as the action is there.

I've actually just found out about another lens option:
SLR Magic 12mm f/1.6 (T/1.6 basically a f/1.4 by design)

Not as sharp as the Olympus but 'only' $500 including shipping. Also gathers more light. I just got a reply from sales@noktor.com saying that they are delivering this lens already and delivery time would be February for pre-orders placed right now. I'm really excited about this lens, especially considering the price and am now undecided between the SLR Magic and the Olympus.

... maybe I should wait a bit until Voightlaender decides to launch a counter-attack with a 12mm MFT lens?

 

[Zupfel Gumba]

Coming back to this this issue about sensitivity - I have made a sketch to more clearly demonstrate a possible effect for the fact that a theoretical f-number comparison might overestimate the difference in light-gathering power of two lenses.

When the sun throws a shadow behind the earth there's an area of complete darkness as well as an area which is only 'half-dark'. Going from complete darkness into the half-dark area is obviously continuous, not as abrupt as sketched in that attached drawing.

Now the same is true for the image circle generated by a lens - the brightness falls off towards the edge of the image circle, due to the fact that only the area in the center can catch photons coming from all extremes of the object. A pixel at the edge of the image circle can only catch photons coming from a certain direction as the aperature is blocking most other pathways. This is also referred to as vignetting.

This further leads to the conclusion that when comparing two lenses, where both lenses have the same f-number, the larger lens which generates an image circle much larger than the CCD chip (for example a full-frame lens on a MFT camera) will be a more sensitive lens as compared to the lens which creates an image circle which is just large enough to cover the CCD-chip (a MFT lens on a MFT camera) - again, with both lenses having the same f-number and focal length.



The question remains, however, how strong this effect is - i.e. when comparing these two lenses from my first post (the full-frame Voigtlander lens versus the MFT-format Olympus lens), is the difference between the two lenses really 8 or maybe just 5? 6? or 7? or is the effect typically negilible?

Looking at this sketch, this effect may be only small for tele-lenses, where the object is far away from the aperture and the 'half-bright' area is much smaller as compared to the 'fully bright' area. For wide angle lenses, however, where angles-of-view are of the order of 90 degrees or more, this effect may be (must be) quite dramatic! I can't see how a wide angle lens can produce a flat intensity-field.

I wonder if anyone is aware of a proper investigation of this topic?

Thank you!
Gumba!

 

[Hamish Gill]

Well, I for you would be very interested to hear your thoughts on te slr magic lens if you go down that road
I would be wary of overall image quality not being as high as te Olympus...
Cheaper and brighter ... Does make you wonder just how good it is going to be ... And why do they use t stops not f stops? Is it intended for video use???
Your call on that one, just remember the lack of electrical conversation between the camera and lens, prefocusing is all well and good, but of you want to use it wide open, close quarters even on that fairly small sensor you are going to have very narrow DOF and without any proper viewfinder focusing is not easy!
My advice would be to switch the camera to manual focus with you current lenses and use te camera in every possible circumstance that you might use this new lens and see how you get on ... That will answer that concern if there is any concern at all!

As to your second question ... Well...
I take your point!
Yes, it is widely recognised that lenses vignette at the open end ...
Most good lens test do look into this phenomena and report the edge fall of!
My favourite site for lens tests is photozone
Here is a link to the page that looks into the vignette issue on the oly lens
Olympus M.Zuiko 12mm f/2 ED - Review / Test Report - Analysis
as you can see the vignette causes a 1.3 stop (1.3ev) fall off when at f2 although this falls to 0.3ev by the time the lens is set to f2.8 ...
So yes at the open end I would guess that over all the frame less light would hit the sensor than of using a f2 full frame lens ... But I would definitely put the difference into the category of negilible ...

Think about it this way...
All modern cameras meter through the lens ...
Put your camera on a tripod so it is seeing a set scene
This hyperthetical scene using matrix metering (ie metering from most of the scene) gives you a meter reading of 100 ISO, 250th of a second at f2.8
Now adjust the lens to f2,
The camera will now read 100 ISO 500th sec f2
Same EV!

If lenses didn't do this, people would complain!
When was the last time you read anything anywhere on the Internet about a lens that didn't have the expected effect on metering?

If it was only by a factor of 5,6,7 it would read somewhere between 250th and 500th not 500th ...

The EV scale, the interaction between ISO shutter speed and aperture in terms of exposure is basically the fundamental basis for photography, if a lens were to break that rule ... Well, it couldn't ... They wouldn't be able to call it an f2 lens unless it had the effect of an f2 lens on the metering of the camera ...

1.3 EV does seem a lot ... But I can only conclude that that is only right in the very corner and that the effect gets less dramatically enough away from the corner for it to still retain the same EV ... And therefore it would indeed have light gathering power 8x greater than the 5.6 lens ... Or at very least 7.9

And so ... as far as brain can work out .. The difference must only be negligible based on te above ... No??

And anyway, one way or another an f5.6 lens, regardless of its image circle is never going to allow a camera to meter the same settings as an f2 lens wide open...

What do think? Does that all sound sencible?

Edit ... Photozone also has a review of the slr magic lens ... It reads quite well!
It seems it is t stops over f stops as it has a clickless aperture ... Ie better for vid, but not a huge drawback for photos SLR Magic Hyperprime 12mm f/1.6 - Review / Lens Test - Samples & Verdict

 

[Hamish Gill]

You got me racking my brains now ... Trying to think of a circumstance when my camera hasn't metered as expected based on a change of aperture ...
Certainly an interesting subject this id say...
I wonder if anyone else here has any thoughts??

 

[Zupfel Gumba]

Thanks so much for this discussion Hamish!

Maybe that camera meter is only measuring in the center? Or do these meters measure over the entire area?

I've done enough of research on the SLR magic now and found that it basically is a CCTV lens, made out of brass therefore the weight. It's certainly not in the class of the Olympus f/2.0. I speculate that the Olympus 12mm f/2.0 has a larger image circle than the SLR magic, therefore it's much more sharp towards the edges. I'll buy the Olympus although a lot more expensive! I assume if one places a small spacer between the MFT camera and the Olympus the 12mm can effectively become, say, maybe 11mm?

Hm, let's put this vignetting-sensitivity issue to a test. I've just opened MATLAB and read in the following f/1.6 test data: http://www.photozone.de/images/8Reviews/lenses/slrmagic_12_16/vig.png

I've removed the f/1.6 writing in the centre replacing it with the same bright pixels that surround the writing. For the mean intensity (integrating/averaging accross all pixels) I get 183/256 whereas the maximum intensity found in the center is 203/256. So the mean intensity/sensitivity of this lens is only 90% of the value in the centre.

So let's compare the 12mm SLR MAGIC at f/2.0 and the full frame 12mm Voigtlander f/5.6 which we assume doesn't vignet at all (image circle is 44mm, MFT diagonal length is only 23mm so this assumption is considered a good one). The difference in sensitivity is actually 7.0 (0.9*(5.6^2/2^2)) instead of 7.8. That's significant but not as significant as I thought. Note that I used the vignetting data at f/1.6, using the vignetting data at f/2.8 I get 7.3 (7.84*(186/198)) instead of 7.8 for the theoretical f-number comparison (mean intesity at f/2.8 is now 94%).

So we can finally say that instead of around a factor of roughly 8, the actual difference in sensitivity/light gathering power is only about 7.4 (0.92*8)!

I'm happy now. I'll buy the 12mm Olympus f/2.0!

 

[Hamish Gill]

Camera meters work how the user chooses them to work, there are usually three types of metering mode
Spot metering - measures from a small point in the centre of the frame
Centre weighted - measures a larger portion of the centre of the frame and balances readings toward the centre ... There are oftens settings to adjust how this system measures
Matrix metering - a more modern advanced version of average metering in simple terms takes a reading from the entire frame

Faced with a white wall, evenly illuminated I would still say that for an f2 lens to be called an f2 lens it would have to give te same reading regardless of what metering setting it was set to ...

I've never consciously tested this but I'm pretty certain it is the case!


I'm going to have to admit to be a bit baffled by your maths ..
But I want to ask one question
Looking at this picture

Does your maths take into account the way the light falls off ... Or it only effects the very corners and is not a gradual fall of right from the centre ... Like I say, I don't really understand the maths, but I can't see how you could calculate anything without measuring and specifically understanding the exact point at which the light fall of starts ... ??

Also, for my ease ... Can you translate the effect you are fining into the EV scale ...
ie the difference between 2.0 and 5.6 is 3 EV
2.0 - 2.8 - 1 EV
2.8 - 4.0 - 1 EV
4.0 - 5.6 - 1 EV
I understand that each time we move one EV we are doubling or halfing the light
Double 1, 3 times and you get 8 ... I understand that's where you get "a factor of 8" but fr me to translate "7.4" back to the EV scale is impossible for me as I am aweful at maths!

What I am trying to determin, is IF your maths is right then what effect on a camera meter this might have ...
Camera meters at their most precise work in 3rds of a stop
So between 2.0 and 2.8 we have 2.2 and 2.5
2.0 - 2.2 - 0.333333 EV
2.2 - 2.5 - 0.333333 EV
2.5 - 2.8 - 0.333333 EV
I know the maths isn't right, but for the sake of photography ... That's how it's worked
I'd call any one of those differences "a third of a stop" or "a third EV"
If your factor of 7.4 was strong enough fall off of light you might expect the effect to be enough to make the camera read as if the lens had been adjusted to one of these apertures ...

Do you follow my train of though...
I'm probably not the best to have the convo with, I'm not the smartest cookie academically speaking ...
I know how to use a camera and how all the settings interact etc ... And I'm enjoying the conversation, so hopefully we can come to some sort of conclusion that we are both satisfied with

Anyway, Im pleased to have helped you choose a lens.... After all this, I hope to see some of your photography too

 

[Pete Askew]

Sorry to be so late to this one. An interesting set of calculations and observations on the effect of light fall-off.

I think you have made the right decision with the Olympus lens as its designed 'speed' will be of more value. The critical issue, as you noticed, is the effect of vignetting. The Voightänder lens does vignette quite noticably but it is of course designed for the full 35mm format and so, given that this effect is not linear across the format, it will much reduced on µ4/3 and may even not be detectable. The Olympus probably has less vignette, due to the simpler task of making a lens for a smaller format, but it will be evident. Adding the mount converter will change the field of view slightly I suspect and, as Hamish noted, the Voightländer is a manual lens.

Working out transmission of light through a lens as opposed to the theoretical F-number is a complex process and is usually achieved by measurement rather than calculation (as you have done in an indirect way). This is done with motion picture lenses and there is a good summary here:

F-Stops Vs. T-Stops Explained - The Photoletariat

Hadn't come across SLR Magic before and the 12mm f/1.6 does sound interesting (and might be a good choice for my µ4/3 video camera!) but I think for you the flare (and vignette) might be an issue. As I said at the start, I think you've made the right choice for your needs.