The real reason why some m43 lenses are very expensive. But...

It's rather obvious if you understand anything about digital camera design:
Great. Where can we find such a person to get your questions answered?
  • m43 specifies an enormously thick filter stack - just over 4mm compared to 1-2mm on most DSLRS
  • The thicker the sensor stack, the more you need telecentricity from lenses (google this if you don't know why this is so or what it means. Or read
They don't have much to do with one another.

The absolute size of the filter stack has nothing to do with performance. What is important is that a lens designed for one filter stack will perform differently than on one with a different filter stack, and telecentricity does not solve this.

"Telecentric lenses, _provide the same magnification at all distances. An object that is too close or too far from the lens may still be_ out of focus, but the resulting blurry image will have the same size as the correctly focused image would."

All that Roger is saying is that lenses designed for one filter stack will behave differently on another filter stack--pretty much common sense.

Telecentricity was/is a big deal in M43rd for a different reason.

http://theonlinephotographer.typepad.com/the_online_photographer/2008/08/micro-four-thir.html

(NOTE: this is from 2008, but all I could find. Probably telecentricity isn't an issue today.)

Of great importance, if not well understood, is the relaxation of Four-Thirds' rigid telecentric lens standard. CCD imaging chips used in early digicams need light to hit the chip surface at a perpendicular angle, as their photosites sit in depressions that off-angle light can't reach evenly. This creates havoc that the original Four-Thirds standard addressed by demanding system lenses be perpendicular (telecentric). However, newer NMOS chips new used by Olympus and Panasonic don't suffer fatally from angled light, and advanced in-camera processing can address intensity differences that still occur across the frame.

The real reason M43d lenses are expensive because 1.) they don't make that many yet 2.) it is more difficult to make a small fast lens than a large one, because the curvature is the lens is more extreme, making a thicker lens that needs more corrections. An exotic F1.0 M43rds lens is equivalent in curvature to a common F/2.0 FF lens.
 
Since m43 is a development of 43 it continued to use the 43 sensor with its thick stack. I don't think Olympus would have been popular with 43 users if they had introduced m43 with a thin stack sensor. It would have made their beloved HG and SHG glass useless on m43 cameras and m43 would have been dead on arrival without the initial support of 43 users.
There must have been good reasons why they kept the stack other than pleasing the FT users. Unless I misunderstood, had they chosen a thinner stack then all they needed to do for legacy FT users was to put the extra glass in the adapter.
They certainly could've added a glass cover at the rear of the adapter, but they might've feared accidents due to improper use.

Regarding this topic in general, the OP seems to be forgetting just how wide the throat mount diameter is in relation to the sensor size. Among the most popular lens mounts today, MFT has the largest throat diameter to sensor diagonal ratio at about 1.76. Compare this to Canon EF (1.25), Sony E (1.63 APS-C / 1.07 35mm) or Nikon F (1.02). What this means is that lens designers have a lot of freedom to place very large rear elements close to the sensor, and avoid telecentricity/chief ray angle issues simply by designing for a larger image circle (basically what happens when one mounts an FX lens on a DX body, for example, where the bad corners are cropped out). Telecentricity is a non-issue for MFT regarding the sensor stack because the format encourages designs with larger image circles that take advantage of the central 'sweet spot'. It's also the reason why some smaller lenses, like the Olympus 14-42mm EZ and 17mm f/2.8 have disappointing performance - they just can't make use of all the freedom afforded by the MFT mount.

If the 12mm f/1.4 is large, retrofocal, and expensive, that's because it needs to meet extremely stringent design goals, not because of some sensor stack issue.
 
MFT lenses are not expensive. My Nikon 14-24, 24-70, 70-200mm sell for a total of $7,091 from B&H. The comparable f/2.8 lenses for the Olympus MFT cameras, the 7-14mm, 12-40mm, and 40-150mm, sell for a total of $3,347.

The high end MFT lenses are selling for less than half as much as it costs to buy equivalent DSLR lenses. Someone's perspective is in need of adjustment.
 
MFT lenses are not expensive. My Nikon 14-24, 24-70, 70-200mm sell for a total of $7,091 from B&H. The comparable f/2.8 lenses for the Olympus MFT cameras, the 7-14mm, 12-40mm, and 40-150mm, sell for a total of $3,347.

The high end MFT lenses are selling for less than half as much as it costs to buy equivalent DSLR lenses. Someone's perspective is in need of adjustment.
 
Design of a lens is costly but when many thousands of copies are sold not something that makes a difference between an expensive lens and a hugely expensive lens.

The expensive lenses are expensive because of the economic laws of supply and demand. If these lenses were quite a bit less expensive the currently less expensive lenses would no longer sell at their price so everything would need to become less expensive. That makes no economic sense for a manufacturer.

The only "law" that I know here is the law of diminishing returns. Cheap stuff is produced at a price close to actual cost. As stuff becomes more expensive it does not become equally better. It becomes much more expensive than needed to keep the market for each level of product sufficiently divided. The more expensive stuff then attracts a smaller market, which means that development cost needs to be amortized over a smaller number of copies and that then creates the need for the product to be more expensive...
Key factors that drive manufacturing cost are the need or use for high precision components, tight assembly tolerances, and difficult manual alignment proceedures. The same effect occurs for both mechanical tolerances and tolerance of electronic components.

I've seen this in the radio products we've designed, and the report I saw a year or so ago about a tour through the Sigma manufacturing facility indicates that it is a key factor for them as well.

If a design requires a lot of tight tolerance components, besides the fact those components are more expensive and typically hard to get, it often indicates that the design will not tolerate normal variations in component performance and assembly - this means a higher percentage of units failing to perform after assembly, more rework and more units in the discard pile.

If a unit has a lengthy and complex adjustment proceedure, like certain radio filters, but also like some lenses, that also really drives up production costs. Not only is there a lot of labor that does into each unit, but the yield goes down as well.

The final cost of the product has to pay not only for the design and development cost, but also the labor, materials used in manufacture, capital depreciation on manufacturing equipment, and all the units that didn't make it out the factory door. First run yield can be pretty low.

I think that has a lot to do with the difference between a $2000 lens and a $200 lens. True, the $200 lens will be produced and sold at a much higher volume. To make that possible however, it's got to be a lot more tolerant of manufacturing and component variations.
 
You are assuming that filter stack doesn't affect light pathways but only does move dust further.

Meaning, that assumption requires that light angle ain't altered when it hits the stack surface and correct it parallel to sensor pixels like optical (any) glass does.

The same effect has been used in many military applications to move the light pathways from a surface to a another.


If you look that armored windshield, what you would estimate its thickness is? Like 1cm or 2cm? Right?

Here is the exact same armored windshield

That is example of the thick armored glass in very low angle and you still see through it like it would be very thin glass.

Meaning, it doesn't matter is the dust on pixel or on filter stack, it is on the way of the light direct pathway and affects same manner the sensor "regardless the thickness".

The filter stack ain't a lens that spreads light more, it just transmit it straight down and light continues its angle after passing it in same angle as it hit the glass.

In otherwords, it "extends" the flange distance and it becomes "nothing".

You can do the same thing with the normal window glass. Cut it to pieces and stack them to even a 10-15cm thick stack and look through it and you will see that you don't see the edges of the glass but just "through it". It is like a periscope.

74451c51137f44ed877700ad65fe87f5.jpg.png

Now consider what happens when you test a lens without filter stack? You get awul results as the lens has focused to wrong position on the sensor, 4mm in case of m4/3.

So you need to "extend" the light travel distance when it exist the last lens and is heading to the sensor.

Sony has very thin filter stack, but they as well have very short flange distance. Olympus chose 4mm for 4/3 standard and that same 4mm can have benefits for telecentric design, as it doesn't need to be in lens, it is as well in the filter stack, correcting light to hit sensor microlenses in different spread and distance.

If you have dust on that sensor stack, it will have same effect regardless the sensor stack thickness or if it exist at all. As it would block the light entering to the stack and coming out.
 
Last edited:
I would imagine Olympus engineers never imagining IBIS in their cameras at the time of creating 4/3 standard with a 4mm sensor stack.
I imagine that they did have clearly that system in mind when they were designing the 4/3 format basing it to 110 format with assistance of Kodak for final image quality requirement.

Researching ways to do a image stabilization requires that you know do you do it with the lens or not. And if you do not do it on lens, how you do it then?

Olympus 300mm f/4 is their first lens with the optical image stabilization. So why not have them in their 4/3 lenses like 14-54mm f/2.8-3.5 or 150mm f/2 or 300mm f/2.8?

Olympus clearly back then opted for IBIS instead OIS. They just didn't have it either ready for E-1 so E-3 got it first in top line. Wasn't the E-520 the first three digit model to get IBIS too?

So thinking that Olympus just suddenly found and developed a IBIS for E-3 at 2008, four years after E-1 sounds little odd.

Of course it is always possible, but that would mean Olympus planned in first place NOT to use any kind image stabilization in their 4/3 cameras (lens nor body).

Fairly big assumption, don't you think?

And consider that Olympus had E-7 ready to come out way before E-M5 was out, and they tested it side by side with E-M1 to decide that do they push DSLR professional line further, or do they simply jump on both feets to Mirrorless line by presenting their professional model there?

When E-M5 got released 2012, they already had started to research and develop E-M1 Mk2 at late 2011. That is 5 years development for Mk2.

And that is the thing, what you see today in announcements like Nikon D850, those has been in development for years. All kind small pieces in technology that are waiting and trying to catch up. And then now and then there come big "out-of-nowhere" technologies for marketing people like E-M5 Mk2 HR-Mode that likely was just a one or two engineer "side-hobby" before presenting it, easier to do once you have all the other technologies ready inside. But as well shows that even top marketing people doesn't know anything for sure before it is presented to them by the engineering team. That is just "Need to Know" information that ain't leaked anywhere.

Why it is just stupid and worthless to go to camera shows and camera stores to talk the camera manufacturer representatives and expect them to know anything about incoming products if it ain't specifically already presented to them so they can be trained for the features.
 
You begin with reversing the cause and effect, then proceed with several wrong ass-u-m-ption.

The original 4/3 is telecentric from system design concept, not as the after thought. You could google and wiki for tons of informations before start the misconception.

Telecentric lens design not only enables Olympus to create the new system, it allows using of thick glass in front of the sensor.

What's the use of this thicker glass?
Many very likely when it come to digital.

Like look everyone else.

Canon and Nikon are literally stuck to film. Their heritance is decades old lenses. All designed for a film, without any filter stack. So, when they knew they need to go digital, they needed to re-design the whole system from new mount to new lens designs and prepare it for digital sensor in the future, with compatibility for film.

Nikon had the same problem. They just couldn't throw away the legacy.

Sony with E-mount. They got change to go fresh, but Minolta engineers were there likely to hold for some compatibility for film lenses, but likely they opted fully to digital one with thin body.

Fuji with their X-system, got a new change so they opted for very short distance between lens and the sensor.

Olympus is the only one with a thick filter stack. They got totally fresh start after long absent from markets. Designing the 4/3 system from ground up not by basing it to what they had, but what is the image quality expected by the viewers and what the digital computer industry was heading. So far only Sony has managed to do the same predictions but they have their feets deep in the computer industry.

Olympus was pioneering the digital camera designs from digital corrections to all kind digital advancements over optical and electronical. They just never got big.
Apart from additional better IR reduction, it can withstand the rigorous of the ultrasonic vibration of the dust-off mechanism.
You can make a good IR filtering with thinner filter stack, but there are other properties than just that.
Imagine shaking the 2mm thin glass at 2-30,000 cycles/sec!
Not "shaking" but "vibrating". Two different things, but you know that ;)

But the point is that so many is seeing a thick filter glass as some kind "flaw" or "idiotic design choice" comparing Olympus was the only one to implement such. It can always be true that Olympus should have gone for a 1mm or even just 2mm instead 4mm. But there can be many many benefits that we just don't know, like having more mass on sensor assembly means it needs more energy to be moved and less energy to keep it steady when it is floating (like try to get a heavy car moving, comparing to bicycle, but once you get it moving, it is far more difficult to stop that heavy car than the bicycle).

Or maybe better analogy is a boat. A rowing boat vs cruiser. Which one gets more affected by external forces like high winds, high waves and people moving around? But which one requires more force to get moving?

But you can always add weight to sensor assembly, but not really reduce it. So using thick filter glass as a weight, is not logical.... Don't you think? So it must be the optical properties that what you can do with the sensor and lens.
 
Last edited:
from someone without the necessary technical understanding.

Why would thickness of UV and IR filtering, protective etc. glass (ie. "stack") influence need for telecentricity? AFAIU (doubtful) a light ray hitting the glass is bent toward perpendicular the amount depending on the refractive index of the glass. When the ray exits the glass it's bent back to the initial angle. At least that's what it looks like on the graphs I've seen. Net result: light ray displaced slightly toward centre. This effect increases with distance from centre. Net result: the whole image is slightly downscaled. In a thicker stack the ray travels longer at steeper angle thus being deflected more toward the centre. Net result: a thicker stack downscales the image slightly more than a thinner one theoretically resulting in very slightly increased FOV. This wouldn't seem to influence lens design in any significant way as the exit angle stays the same independent of thickness of glass.
 
Here's my summary of suggested reasons for the choice of a thick stack.

Please add any I have missed.

Advantages
  1. Dust less visible
Not so. The dust ain't any further from the angle of the light.
  1. Stronger and less fragile sensor
Sensors are already very strong and doesn't need a protection against physical impacts, unless you are sticking your nail there all the time while mounting lens?
  1. Stiffer and less flexible sensor improves image stabilisation
You can get that with way other means than just with optical glass that is affecting the optical pathways. So no.... Simply using even sheet of titanium back of the sensor would do the same, without affecting optical pathway.
  1. Degrades performance of adapted lenses thus promoting sales of new 43/m43 lenses (advantage to Olympus not to the consumer)
Not really... Olympus had stopped selling the OM system way before 4/3 system. They didn't care anything about legacy like Canon or Nikon. They cared the future and how to make a small and light camera with image quality that meets the requirements (notice, Olympus is only one to define the system based the requirements of the final output quality, not to chase some fantasy of "more is better". They defined the system based what is the most needed image sizes, cropping requirements (based lack of focal lengths and sudden situations) based to empirical research in industry and KODAK.

But that is the disadvantage as consumers are not rational people, they are move by feelings and fear. If you say "You need 47Mpix to get good photos from your first born child" then D850 gets huge sales. Regardless that they images end up to be a 1-2Mpix.
Disadvantages
  1. Lens design made more complex and expensive
And how is that so? That is still the assumption.
  1. Degrades performance of adapted lenses thus promoting sales of new 43/m43 lenses (disadvantage to the consumer, not Olympus)
Why? The average consumer doesn't care about that! Even majority of the Sony A7 users are using system native lenses and not Canon or some legacy ones!

Designing a new system based a old lenses that are not in production and not optically anyways high enough for digital use is bad assumption!
  1. Heavier sensor degrades image stabilisation
You can always make sensor assembly lighter or heavier by limiting what is being moved up to some point. And if you have something heavier, it requires more energy to be moved somewhere (tries to stay dynamically stationary by its own mass) and it means you don't need to be so accurate with the energy to do small changes. While ultra light requires more precision for energy control.

So it would be the opposite that heavier is better. But up to the point. But if you want heavier, you wouldn't make a compromise to make things more expensive and complex in lens design, when all you need to do is to add a couple lead weights to the sensor assembly to make it heavier!
How have the above affected me?
  1. I've never had to clean a m43 sensor
Yes, sum of many factors but mainly the Olympus designed sensor cleaning system.
  1. I'm impressed by sensor stabilisation
Yes, Olympus has well researched and developed their version of the IBIS following Pentax lead.
  1. I've used adapted lenses and not worried about or even noticed any slight image degredation caused by the thick stack
There are, but in normal final output requirements those ain't visible. That is why majority of the lens sharpness tests, discussions, digital or optical correction etc are waste of time as you will not see the effect in final output! Only pixel peepers and gear heads get to see the results when they take the images out of the context.

That is why we can take a 42Mpix Sony A7rMk2 and E-M1 Mk1 and use in both cameras ISO 1600 and do a 20" wide print and you will not see a image quality difference in the conditions. If you take the magnifying glass and you focus to minute detail (1/60 of degree of view, meaning like a 3x3mm area in 20" wide print) you will find difference, but only when you have direct comparison side by side (identical scenes etc, meaning you carry two cameras to take photos from same scene) and when the scene is such that it can reveal the minute difference. And even then if you would step further than 10" viewing distance and place the prints on opposite walls, different rooms or even just little apart from themselves than side by side on table, you wouldn't see the different, nor care.
  1. I've never considered the sensor stack size when deciding to purchase a lens
Who really would? How many even knows what a sensor stack is or what its thickness does?
  1. The price I paid for my m43 lenses was determined by market forces far more important than sensor stack size considerations.
The price you pay is determined by the corporation based their target of profits or their target of quality (if they are having "hobby" business, more like Olympus does).
So, for me, if the thicker sensor stack is one of reasons I've never had to clean my sensor I'm all for it.
If it does allow you make photos as good by quality in final output as best of the FF cameras, then it is fine.

If it does help to maintain the camera performance, go for it!

If it does have more benefits than cons for photography, go!

Of course there is always a change that 4mm filter stack was a huge mistake from Olympus and they can't anymore fix it by any ways as m4/3 transition is gone. And why no one else want to make m4/3 lenses or bodies than Olympus as that thick filter stack is so terrible decision!
 
MFT lenses are not expensive. My Nikon 14-24, 24-70, 70-200mm sell for a total of $7,091 from B&H. The comparable f/2.8 lenses for the Olympus MFT cameras, the 7-14mm, 12-40mm, and 40-150mm, sell for a total of $3,347.

The high end MFT lenses are selling for less than half as much as it costs to buy equivalent DSLR lenses. Someone's perspective is in need of adjustment.

--
"The people can always be brought to the bidding of the leaders…tell them they are being attacked, and denounce the peacemakers for lack of patriotism and exposing the country to danger."

- Hermann Goring
Were you replying to me? Doesn't seem like it.

Anyway, it only makes sense that the lenses cost half as much, since they sport physical aperture diameters only 25% as large as the Nikkors. In fact, 35mm gear is less expensive for comparable performance, simply because MFT can't compete for total light in many instances.

If anything, lenses like the 100-400mm Panasonic and the Olympus 25mm f/1.2 are obscenely expensive for what they bring to the table.

--
"Chase the light around the world
I want to look at life
In the available light" - Rush, 'Available Light'
How much does a 25mm f1.2 for 35mm format cost? See, a 25/1.2 is a 25/1.2, regardless of format. The only thing that changes is the size of the image circle.

All this equivalence talk is ultimately pointless.

I didn't switch from 35mm format to MFT for equal light gathering. I switched for sufficient light gathering, better edge-to-edge sharpness, smaller size, less weight, and lower cost. If what you want is 35mm format, what you should buy is...y'know...35mm format. Duh.

--
If you think digital is hard, try slide film.
http://jacquescornell.photography
http://happening.photos
 
Last edited:
Thanks, there's nothing like a nice graph but don't you think stack thickness would be compensated by simply adjusting focus accordingly? Isn't the focus plane shift uniform across the frame?
 
Thanks, there's nothing like a nice graph but don't you think stack thickness would be compensated by simply adjusting focus accordingly?
That is the question, what exactly does the filter stack do optically? It ain't just directing light like typical glass. You don't need thick one to get great IR filtering or anti-aliasing.
Isn't the focus plane shift uniform across the frame?
It could be made so, or not. It is a one lens more top of the sensor that can be used by many ways to refocus or shift the image plane and light.

It ain't just a "clear filter in sake of protection" as anyone could use 0.1mm thick hardened glass for that that can even withstand vibrations.
 
How much does a 25mm f1.2 for 35mm format cost? See, a 25/1.2 is a 25/1.2, regardless of format. The only thing that changes is the size of the image circle.
Nope, it isn't. F-numbers are only directly comparable between lenses for the same given format. Cross-format comparisons need to be done on the basis of the effective aperture diameter.
All this equivalence talk is ultimately pointless.
It's not, because it shapes the limits of what is possible with the equipment. I'm personally a big detractor of "full frame" for most uses, as it's just too expensive and unwieldy. But for some photographers, it's still the tool that produces the best results, and the physical advantages of larger sensors with larger lenses are undeniable.
I didn't switch from 35mm format to MFT for equal light gathering. I switched for sufficient light gathering, better edge-to-edge sharpness, smaller size, less weight, and lower cost. If what you want is 35mm format, what you should buy is...y'know...35mm format. Duh.
I actually use 1" sensor cameras. My RX10M3 has a lens that is simply unmatched in any other format or camera type, and that probably will remain as the most advanced travel zoom ever made for many years to come. But that doesn't mean I get to be a moron and believe that the Sony produces the same results as a Canon 6D with a 600mm f/4 L IS tele prime (a folly of some Olympus 300mm f/4 users). That's just denying the physics behind optics.

Would I pay or much less carry the Canon kit around? No way. But I'll be the first one to say that my lens is only equivalent to f/11 in 35mm terms, and so it isn't a real replacement for the huge combo either. People don't invest so much time and energy 'just because' - that'd be ridiculous.
 
M43 lenses from Panasonic and Olympus are cheap. Why else would Sigma and Tamron be incapable of producing cheaper and better version of the same focal lengths?
They are capable. Sigma sure is.

Olympus 25mm f1.2 = $1,199.00

Sigma 30mm f1.4 = $339.00

I know, I know. Not the same focal length, not the same speed.

BUT QUITE CLOSE ENOUGH.
I suspect if Panasonic could sell their 25/1.4 in APS-C mount to match the sales volume of the Sigma, they could also match the price. Most if not all Sigma designs for MFT are adapted APS-C designs. The Tamron 14-150 is one of the few 3rd party, AF lenses that is MFT native. And it's cheaper than Oly or Pany, but I suspect they have not made money on it...
 
M43 lenses from Panasonic and Olympus are cheap. Why else would Sigma and Tamron be incapable of producing cheaper and better version of the same focal lengths?
They are capable. Sigma sure is.

Olympus 25mm f1.2 = $1,199.00

Sigma 30mm f1.4 = $339.00

I know, I know. Not the same focal length, not the same speed.

BUT QUITE CLOSE ENOUGH.
Why on earth would you compare a weather-sealed 25mm/1.2 to a 30mm/1.4 for your example, when you can compare the PL25/1.4 with the same aperture and relatively similar build to the Sigma.

The price of that lens fluctuates from $450-600. Still higher than the Sigma, but not anywhere close to your, frankly kind of absurd, comparison.
 
M43 lenses from Panasonic and Olympus are cheap. Why else would Sigma and Tamron be incapable of producing cheaper and better version of the same focal lengths?
They are capable. Sigma sure is.

Olympus 25mm f1.2 = $1,199.00

Sigma 30mm f1.4 = $339.00

I know, I know. Not the same focal length, not the same speed.

BUT QUITE CLOSE ENOUGH.
I suspect if Panasonic could sell their 25/1.4 in APS-C mount to match the sales volume of the Sigma, they could also match the price. Most if not all Sigma designs for MFT are adapted APS-C designs. The Tamron 14-150 is one of the few 3rd party, AF lenses that is MFT native. And it's cheaper than Oly or Pany, but I suspect they have not made money on it...

--
Shawn Wright
http://www.flickr.com/photos/zuikowesty
http://blog.corsequere.com
Yep, APS-C designs with just the lens mount swapped are probably the least expensive examples you can find (C-mount CCTV lens are the only ones even cheaper). This gives the manufacturer high volume with minimal effort. You can find a lot of off-brand manual focus chinese lenses that are such a case.
 
Last edited:
M43 lenses from Panasonic and Olympus are cheap. Why else would Sigma and Tamron be incapable of producing cheaper and better version of the same focal lengths?
Sigma is making some of the Olympus and Panasonic lenses
And if this is true (no comment from me on this...), then it may explain why they don't sell any native MFT designs. There may be an agreement in place preventing them from doing so, in exchange for the (alleged) business of producing those lenses for Oly and Pany. Sure they could perhaps try to fill in the gaps in the MFT product line, but maybe it makes better business sense for them to just fill orders (allegedly again) for others, with minimal upfront design and tooling costs. Oly and Pany are left holding the liability if the lenses don't sell, not Sigma.
 

Keyboard shortcuts

Back
Top