Four-Thirds Systems

It could already be argued that Canon users are the goats ...

I tell you what Paul, in any reasonably sizeable group of underwater photographers I feel very much in the minority. Shooting Canon, with TTL then going through the images on a non-Mac laptop. I get some right funny looks.

Martyn

First of all I would like to say that it is nice to see this amount of participation in the Wetpixel forum in regard to Olympus.

I agree with some that much to much emphasis is being placed on the term 4/3rds and not on the system itself. Olympus never really entered the Auto focus 35mm SLR film market and therefore had no "legacy" lenses to support as Canon, Nikon and others did. As a result Olympus was able to start with a blank page and built an entirely new interchangeable lens digital SLR system without having to support many past products. If you want, Olympus does sell an adapter which will allow the use of Olympus 35mm manual film lenses.

I don't know what advertising James has been reading but Olympus has never said that they are the "only" innovative DSLR company, only that they have built the Olympus E-systems from the "ground up". A statement which I would find hard to argue with. Regarding James statement on innovation I guess he has forgotten about the "supersonic wave filter" which eliminates dust each time the camera is turned on. This item must be of interest to at least some digital photographers because others including Canon are now introducing their own versions of a dust cleaning system. Olympus has in fact succeeded in building the smallest and lightest DSLR, the 10 MP E-400, which by the way has an Olympus housing, the PT-E03 which will be as small as a lot of digicam housings.

Olympus also has the largest selection of made for digital lenses which is the heart of the Olympus system. Olympus E-system lenses are getting rave reviews from testers in many areas of the world. You would be kidding yourself if you believe that you can slap any old film lens onto a digital camera (even a 35 mm chip camera) and get the same results as if you had designed a lens specifically for that sensor. Don't get me wrong, many film lenses (particularly those with a narrow angle of view) work very well with digital cameras. However Canon and Nikon would have had no need to design new digitally specific lenses for the 1.5 and 1.6 sensors if all their film lenses had worked as well with digital as they did with film. Further if just having a 35mm sized sensor was the solution to all problems digital Nikon and everyone else would have had one by now and the price would have come down on these sensors.

Also James, Olympus didn't invent the 4/3 sensor, Kodak did. Unlike Canon, Olympus don't make their own sensors.

Back to the original 4/3 question, it is clear that many don't understand that 4/3 format has been around for years. The medium format 645 cameras have all but replaced the square 2 1/4 X 2 1/4 in the print world. Even the venerable Hasselblad has built a 645 camera and AF lens set for both film and digital, "from the ground up" without having to argue their logic for doing so.

645 and 4/3 are the same thing, four units on one side and three on the other. 645 is 1.5 cm x 4=6 cm and 1.5 cm x 3=4.5 cm. Olympus is 4.5 mm x 4= 18 mm and 4.5 x 3=13.5 mm. This 4/3 or 645 format has been a good thing in the print world because most magazines are closer to the 4/3 or 645 format than to the 3/2 or 35 mm format. In most cases this means less cropping with 4/3 than with 3/2.

It is clear that Canon and Nikon make outstanding products and that users of these products, myself included have been vary happy with the results over the years. The Olympus E-system products will also produce outstanding images as well and the 7 to 14 mm zoom, 11 to 22 mm zoom, 14 to 54 mm zoom, 50 mm macro, 35 mm macro, Sigma 100 macro and 8 mm Fisheye are well suited to U/W photography.

Olympus has a smaller sensor and therefore more noise problems above ISO 400, we all know that. AF speeds are also a bit slower, but have improved greatly going forward. If it was up to me I would be asking Olympus to build a camera with better image quality at ISO 50 and 25 where I would use it most, not at 800 and 1600.

Olympus entered the interchangeable lens DSLR market in November of 2003 and lost millions of YEN in the imaging division through 2005 due to R & D costs. At that time Olympus was selling no interchangeable AF lenses or AF-SLR film cameras or DSLR (interchageabel lens) cameras. As of 2006 Olympus imaging division is showing profits in the millions of Yen and selling millions of E-system cameras and lenses.

Olympus won't be for everyone and they have never said they intend to surpass Cankon only that they intend to produce a quality product at a fare price. I am sure that Companies like Sigma, Panasonic, Leica and others would not have signed on if they didn't think 4/3rds had an outstanding chance for success.

If everyones reluctance to include Olympus as a player is a fear that the system may fold remember that the most successful underwater system of all times, NIKONOS has folded but no one doughts the quality of the images produced by the Nikonos system.

Phil Rudin

I can not understand why the camera manufacturers carried so much baggage across from the film world into the digital world. With the step-change in technology, they had a chance to remove some old confusions.

F-stops could have been replaced with a simpler 0 (closed) to 100 (max open) number.

ISO could have been replaced with a signal amplification percentage.

Focal lengths could have been replaced with a positive/negative zoom index around a central "normal human eye" angle-of-view.

Shutter speeds could be decimal instead of fractions.

I can hear you all getting ready to type: "but the old ways are logical". Maybe they seem logical to someone who has used them for many years. But, if you have ever run a beginner photo course, you'll know the difficulty new photographers have in getting their head around these archaic concepts...

A few of us old photographers would have grumbled and struggled for a while but the new generations would have benefited from more logical systems and easier learning curves.

Oh well.

Regards

Peter

Speaking of a clean slate. That is what I am working with. About a year ago I sold all of my Canon FD gear including several "L" lenses. I never even entered the EOS line up. So I am generation skipping.

Being a Canon user I started looking at their system. I tried the 5D. Very nice, but really big. I schelp a lot of my gear up some really big mountains and down some really narrow dirty canyons. And some of the EOS lenses are pretty big and heavy. I also like the 30D and I know I would be happy with that body. Now for the lenses, my line up would likely include the 10-22mm EF-S, 60mm EF-S, and 17-55mm EF-S. All are digital only lenses which are ground up designs. So are folks buying those lenses also taking a chance that the 1.6 sensors will be around as well?

So if I would be happy with Canon lenses and the 30D why look at the 4/3 systems? I think does have qualities that I might not find in other systems. Yeah, the lens line up is limited but how many lenses does one really need? At this point the lenses I might want are either available or announced.

BTW I work on a Linux box and Mac laptop. I must be really be a goat.

It is very encouraging to read about such enthusiasm to try to be different. My only concern is that the effort to think different doesn't end with the kit. It would seem a pity to go to all the effort to get a different camera - and then just take it to Lembeh and shoot pygmy seahorses like everyone else.

Being a goat should also continue with your photography, and shouldn't just end with equipment choices.

Alex

It is very encouraging to read about such enthusiasm to try to be different. My only concern is that the effort to think different doesn't end with the kit. It would seem a pity to go to all the effort to get a different camera - and then just take it to Lembeh and shoot pygmy seahorses like everyone else.

 

Being a goat should also continue with your photography, and shouldn't just end with equipment choices.

 

Alex

Alex,

You are absolutely correct. I have always maintained that, with computers, photography, and especially underwater photography, the best approach is:

1. get some gear that works reasonably well

2. become extremely familiar with it

3. ignore the gear and concentrate on finding images that communicate

Regards

Peter

Hi James!

It would be a pitty, if Canon wasn't inventive, being the leader of the pack ... :-)

It's interesting though that the 400D got it's now highly appreciated anti-dust system (which for some time nobody seemed to need except forOlympus, who actualy invented it).

Auto-dynamic shutter control (including TTL strobe) was first (and only) implemented in the OM-2, as I remember, so Olympus had a couple of fine ideas pushing photography forward a bit.

Olympus built the first DSLR with LiveView allowing to shoot p&s and through the optical finder as well. A feature I learn to love with my UK-GERMANY housing for the E-330.

I had the chance to see the E-400 side by side with an old OM-10. They are comparable in size. You feel like you're holding an old fashioned SLR body. It's small, lightweight and you won't feel it at all when mounted on the 150mm f2 prime :-)

The 4/3 system may be a niche. The fact that Sigma announced to support 4/3 with their complete lens bag is good news. Also, that Panasonic/Leica now offer bodis. Olympus has announced further 3 bodies early next year so 4/3 is far from dying.

What appeals to me is the concept of starting from scratch. No legacy features, parts etc. that get into the way of a for-digital-optimized design. The smaller sensor is a trade-off for weight and size.

There are limits to the system for sure but there are also limits that apply to Canon and the others. It only depends if your application requires to push the limits. I know pros (people really making their living by taking pictures) working with Canon and Olympus as well (selecting the appropriate tool for a job) or even with Olympus alone.

The 50mm macro is sharp as a knife. The 7-14mm is a great WA lens (89 to 114° angle) and the 8mm FE (180°) is ideal for UW photography. To me the 7-14mm (which I use lots topside) causes "unnatural distortions" by flatting out images (as WA lenses do). The FE seems to picture far more natural to me.

I think that 4/3 is a good choice when you upgrade from a p&s and you haven't already got an expensive bag of SLR lenses.

What I really liked was the way the E-1 fits my hands. It felt like it was built exactly for me. All buttons and knobs were exactly where I thought they should be. The lenses available were what I needed (and far more than I could afford . That were my thoughts in the process of decision.

Helge ;-)=)

What appeals to me is the concept of starting from scratch. No legacy features, parts etc. that get into the way of a for-digital-optimized design. The smaller sensor is a trade-off for weight and size.

There are limits to the system for sure but there are also limits that apply to Canon and the others. It only depends if your application requires to push the limits. I know pros (people really making their living by taking pictures) working with Canon and Olympus as well (selecting the appropriate tool for a job) or even with Olympus alone.

What do you have in mind as legacy features and parts that get into the way? I really can't think of anything that Canon has carried over from the film EOS system that's preventing their DSLR from being better than it currently is. The digital sensor has shown up some problems in some of their wide angle lenses and the smaller acceptance angle of the senor probably has made the vignetting worst in some places. These aren't problems that are cause by features carried over. The system does not prevent Canon from producing better WA lenses than some of their current ones, they just have not done so. There are numerous reports of excellent results from lenses from Leica and others mounted on Canon FF cameras.

Hi!

Olympus says that the diameter of the mount has to be significantly wider for a nearly telecentric design (so the image is projected in a vertical manner onto the image sensor). I don't know if they're right. They have a lens mount as large as the others, despite the fact that they wanted to build smaller camera bodies.

We will see if the other manufacturers will provide WA lenses that get around the problems (significant vignetting, corner shading) in the future, especially on FF bodies.

Helge ;-)=)

"We will see if the other manufacturers will provide WA lenses that get around the problems "

The Leica M8 (1.3x crop) addresses this by utilising angled micro lenses in front of its sensor to ensure better wide-angle performance from existing wide-angle designs. There are always varying solutions to problems such as these. Whether this system is adopted by other manufacturers remains to be seen - there are pros and cons of all existing systems and nothing is perfect.

Whilst I can appreciate Olympus' design decisions and the 'advantages' of the 4/3 system, the history of technology is littered with innovative products which failed for many reasons - which often had little to do with their concept. Unfortunately photography has a very bad past record in this respect and whilst I am am wary of suggesting that we ever learn anything from history, it does have a habit of repeating itself.

We will see if the other manufacturers will provide WA lenses that get around the problems (significant vignetting, corner shading) in the future, especially on FF bodies.

 

Helge ;-)=)

There's really not much evidence that the problem is the lens mount. There are examples of excellent results from lenses made by Leica and others mounted on Canon FF cameras. One thing that seems to missed by many in comparing lens performance between different size formats is that you need to scale the F-stop by the length of the sensor in the comparison. For example, to get the same field of view, the same depth of field, and the amount of light to fall on the sensor, of a 20mm lens at F/8 on a FF, you need a 10mm lens at F/4 for the Olympus 4/3 size sensors.

I agree with Paul that history does at times tend to repeat itself, take the first camera with a full frame 35 mm sensor which was made by Contax. Not only did the camera go away, but so did the entire company. That is no reason to assume that others can't make the concept work. Leica's approach at working around the problem of existing wide-angle lenses is a good one and one Olympus chose not to deal with by starting fresh. Just as Leica and others are moving forward to overcome the problems associated with their systems, so will Olympus advance in the area of smaller sensor noise.

The problem here is that Olympus users seem to have to spend more time defending their investment in the system than talking about better ways to use it.

Phil Rudin

There's really not much evidence that the problem is the lens mount. There are examples of excellent results from lenses made by Leica and others mounted on Canon FF cameras. One thing that seems to missed by many in comparing lens performance between different size formats is that you need to scale the F-stop by the length of the sensor in the comparison. For example, to get the same field of view, the same depth of field, and the amount of light to fall on the sensor, of a 20mm lens at F/8 on a FF, you need a 10mm lens at F/4 for the Olympus 4/3 size sensors.

Well Herb, I am not sure that I agree.

#1 When Contax made the first digital camera with a 36mm X 24mm sensor they abandon there film lens mount and enlarged the mount to sensor size very much the way Olympus has done, thus starting from the ground up to improve on current film lens mounts. This pissed off many owners of existing contax lenses. At the same time they released the N1 a 35 mm film camera with the same mount and a set of Zeiss lenses which are some the the better lenses ever made. Both the film and digital cameras also allowed the contax 645 lenses to be used on the bodies. Giving the 36 X 24 mm sensor a crop factor of about X 1.7 compared to 645. Had this camera become the gold standard perhaps we could have stopped talking about crop factors and understood that if a lens is designed for the sensor size it is being used with there is no crop factor.

#2 And I'll try to make this easy for you. The Olympus 50 mm macro has an angle of view of 24 degrees and the Canon 100 mm macro has an angle of view of 24 degrees when used on a 36 X 24 mm sensor or film.

The Canon 100 mm has a depth of field from 0.99 to 1.02 feet at 1 foot and f-16 for a total depth of field of 0.03 feet ( on the 35 mm sensor).

The Canon 50 mm on a 35 mm sensor and the Olympus 50 mm both have a depth of field of 0.96 to 1.06 feet at 1 foot and f-16 for a DOF of 0.06. Thats is twice the DOF not the same DOF as the 100 mm lens.

This in fact applies to all lenses that are 50 mm regardless of format size.

Your example using the 20 mm is wrong. At 10 mm Olympus 4/3 lens would have the same angle of view as a 20 mm used on a 35 mm sized sensor but twice the DOF not the same amount and a 20 mm on a 6 X 7 body would have twice the AOV but the same DOF as your 20mm.

Last of all if you point an Olympus camera and a 35 mm sensor camera at the same light source, with lenses of the same AOV and both are set for ISO 100 at say F-8 the resulting exposures will be the same.

Having double the depth of field for any given AOV with the Olympus lenses seems to me a good thing for underwater photography especially when you consider the number of low light situation that are presented underwater.

I know that you can crank up the ISO on your camera to compensate but at a loss of image quality.

Phil Rudin

#1 When Contax made the first digital camera with a 36mm X 24mm sensor they abandon there film lens mount and enlarged the mount to sensor size very much the way Olympus has done, thus starting from the ground up to improve on current film lens mounts. This pissed off many owners of existing contax lenses. At the same time they released the N1 a 35 mm film camera with the same mount and a set of Zeiss lenses which are some the the better lenses ever made. Both the film and digital cameras also allowed the contax 645 lenses to be used on the bodies. Giving the 36 X 24 mm sensor a crop factor of about X 1.7 compared to 645. Had this camera become the gold standard perhaps we could have stopped talking about crop factors and understood that if a lens is designed for the sensor size it is being used with there is no crop factor.

The evidence that I've seen are

1) There are existing Leica and other lenses that do work quite well on Canon FF.

2) Nikon have a even smaller mount and there weren't that many complaints about WA lens vignetting from people shooting the Kodak FF.

I have not designed a lens myself and really don't know any more about it than that.

#2 And I'll try to make this easy for you. The Olympus 50 mm macro has an angle of view of 24 degrees and the Canon 100 mm macro has an angle of view of 24 degrees when used on a 36 X 24 mm sensor or film.

 

The Canon 100 mm has a depth of field from 0.99 to 1.02 feet at 1 foot and f-16 for a total depth of field of 0.03 feet ( on the 35 mm sensor).

 

The Canon 50 mm on a 35 mm sensor and the Olympus 50 mm both have a depth of field of 0.96 to 1.06 feet at 1 foot and f-16 for a DOF of 0.06. Thats is twice the DOF not the same DOF as the 100 mm lens.

 

This in fact applies to all lenses that are 50 mm regardless of format size.

Please read what I wrote more carefully. This case does not contradict what I wrote as you seem to think it does. If you are shooting the same subject at the same distance you need half the focal length on the Olympus, so, ignoring the 4/3 vs 3/2 for the moment, 100mm on the FF and 50mm on the Olympus gives the same photo of the same subject at the same distance. If you use the same aperture the Olympus will have greater depth of Field. You need to open up the Olympus 2 stops to get the same DOF. This is essentially what I wrote. Actually for macro the correction is slightly different. That's why I didn't use a macro example.

Your example using the 20 mm is wrong. At 10 mm Olympus 4/3 lens would have the same angle of view as a 20 mm used on a 35 mm sized sensor but twice the DOF not the same amount and a 20 mm on a 6 X 7 body would have twice the AOV but the same DOF as your 20mm.

Again. Please read carefully. You have to open the Olympus 2 stops to get the same DOF.

Last of all if you point an Olympus camera and a 35 mm sensor camera at the same light source, with lenses of the same AOV and both are set for ISO 100 at say F-8 the resulting exposures will be the same.

Once again. Please read carefully. I wrote same amount of light captured not exposure: 1/4 the exposure on 4x the area (again ignoring the 4/3 vs 3/2). This is important because for a given silicon sensor technology the image quality captured ( signal-to-noise at a given resolution ) essentially just depends on the total captured light.

Having double the depth of field for any given AOV with the Olympus lenses seems to me a good thing for underwater photography especially when you consider the number of low light situation that are presented underwater.

The catch is that resolution on the Olympus will be diffraction limited at 2 stops wider aperture. For a given resolution the FF can achieve any DOF that the Olympus can just at 2 stop higher aperture. Actually the larger sensor is better than that at macro shots.

I know that you can crank up the ISO on your camera to compensate but at a loss of image quality.

 

Phil Rudin

A loss in image quality compared to what is achievable for FF not what is achievable compared to the Olympus. Fundamentally, the limit is the light captured: 4x the area 4x the light for a given exposure. Have a look at the noise test charts on dpreview. You have to move the Olympus noise graphs over two stops to the right to match the Canon 5D, and it's not even matching the resolution.

I'll try once again to make it as simple as I can. Assuming I have a Canon lens that is as good as the Olympus counterpart that is 1/2 the focal length and set to 2 stops wider aperture, I can mount that on the 5D and shoot at ISO 400 and get the same field of view and DOF and exposure and capture the same amount of light per photosite if the two sensors has same number of photosites.

I can not understand why the camera manufacturers carried so much baggage across from the film world into the digital world. With the step-change in technology, they had a chance to remove some old confusions.

 

F-stops could have been replaced with a simpler 0 (closed) to 100 (max open) number.

 

ISO could have been replaced with a signal amplification percentage.

 

Focal lengths could have been replaced with a positive/negative zoom index around a central "normal human eye" angle-of-view.

 

Shutter speeds could be decimal instead of fractions.

 

I can hear you all getting ready to type: "but the old ways are logical". Maybe they seem logical to someone who has used them for many years. But, if you have ever run a beginner photo course, you'll know the difficulty new photographers have in getting their head around these archaic concepts...

 

A few of us old photographers would have grumbled and struggled for a while but the new generations would have benefited from more logical systems and easier learning curves.

 

Oh well.

 

Regards

Peter

Peter, I couldn't agree with you more!

When I first became interested in taking images underwatee I had no real experience of film-based photography. So to be confronted with 'this is X (digital) which equates to Y (film)' was (as you rightly point out) confusing to me as a new photographer.

I was trying recently to explain to a group of teenagers the concept of pre-decimilised money - and they just did not get it; moreover, the mental gymnastics required to understand the fractioning required appeared utterly pointless to them - and I had to agree.

So, for how long can we really expect future generations to 'sheepishly' continue to use these totally outmoded units of measurements that are retained in photogrophy?

Hi!

Do you know the story about the Space-Shuttle booster's measurements depending in the long run on the average width of the ass of a Roman war horse?

It's amazing how hard it is to get rid of some "well established" systems or standards.

Another completely overcome (and inapropriately used) thing is the scale reference used to decide if it'S a macro shot or not.

What the hell do they mean by " between 1:10 and 2:1"? OK, I know. They mean that if an object of 24x36mm in size has the same size on a slide, it's called 1:1. But why? According to this, a picture of my hand, taken life size with a 4x5" camera is a macro where the same picture, taken with a 24x36mm SLR is not? If I make a postcard sized print nobody could tell the difference!

As long as everybody was shooting 24x36mm slides the world was intact. Now, with a wide range of sensor sizes and aspect ratios this topic gets more and more confusing.

Another one. Try to get rid of this f**** 1/4" tripod screw. It's absolutely hard to get tools to make them her in Europe. Even hard core metric countries can't help to accept it.

Helge ;-)=)

LEGACY is the word I think!

It has its good points - continuity, compatibility, etc.

And bad points - inflexibility, jargon, etc.

BUT being realistic, we are probably stuck with many aspects of photography which are pretty archaic. On the other hand, photography is evolving as we speak - I hear far more mention of the word 'noise' rather than 'grain' for example. And its interesting that we discuss 4/3 as 4/3 - although the format has been around for a long time in various guises (6x4.5cm & 6x8cm). To take the historical perspective, evolution rarely seems to occur in massive dramatic changes but rather in incremental shifts - which is what we are seeing happening now - and sadly there will be extinctions of perfectly good systems along the way. It remains to be seen as to whether the 4/3 is one of these or not. We live in interesting photographic times.

In an attempt to clarify, the Four Thirds system is so called not because the aspect ration is 4:3, but rather because the sensor surface is the same as the old 4/3" Vidicon Tube that was a long time standard for broadcast video. The 4/3" is only the diameter of the tube, the sensor surface, which was indeed 4:3 aspect, was only 22.5mm - less than an inch.

I have to agree with Herb, although we'll see some great cameras and lenses appear in the FourThirds standard, there is no replacement for capturing as many photons as is practical. For me and many others, that practical sweetspot is 36x24mm. YMMV.

Martyn

Martyn

Thats interesting - I can't think of a parallel offhand whereby a film/video format name has influenced a still format's name (other than of course 35mm film!!!).

Yeah, and then we go and use it the wrong way round. Stupid stills shooters!

Martyn