What is 4 times better?

[danijel973]

ultimitsu wrote:
135 is the best balance between best IQ and cost. With anything bigger the cost goes up exponentially. M43 is 1/4 the sensor, but it is not 1/4 the size and its price is almost the same as 135 system.

Well, I agree that the price is almost the same as 135 if you buy the most expensive cameras and lenses, but if you have a more moderate approach, such as using kit lenses and high quality old manual focus lenses via adapter, you can build a very good system for significantly less money. I paid around $600 for all my m43 gear. But then again my style of photography works very well with a Minolta X-300 and a roll of Sensia:

 

[Lost in Time]

tt321 wrote:

But could you achieve 2x per image height (or width) resolution on FF with the best FF lens compared with the cheapest zoom on M43? This is what the larger sensor area potentially should give you, but not realized by even the best FF optics. I don't see anyone arguing that M43 has achieved image parity with FF in the absolute, but it has exceeded parity by a lot in the relative sense.

That would be possibly only if the pixel density is the same for the two sensors, which is not currently the case (largely due to market/cost issues). If the pixel *count* is the same, then you would expect the limiting resolution in the case of perfect optics to be the identical on both formats (neglecting diffraction, which disadvantages u4/3 in this example).

However even with a pixel density handicap, the standard kit 24-105mm lens on Canon delivers approximately 3x the resolution of the standard kit 12-50mm on Olympus. I am less familiar with Nikon, but there are similar kit lenses that deliver 4x on a D800 (which is still has a huge pixel density handicap compared to a 16MP u4/3 sensor).

There is no inherent optical disadvantage for FF. There are, however, disadvantages for cost, size and weight.

 

[tt321]

Lost in Time wrote:

tt321 wrote:

But could you achieve 2x per image height (or width) resolution on FF with the best FF lens compared with the cheapest zoom on M43? This is what the larger sensor area potentially should give you, but not realized by even the best FF optics. I don't see anyone arguing that M43 has achieved image parity with FF in the absolute, but it has exceeded parity by a lot in the relative sense.

That would be possibly only if the pixel density is the same for the two sensors, which is not currently the case (largely due to market/cost issues). If the pixel *count* is the same, then you would expect the limiting resolution in the case of perfect optics to be the identical on both formats (neglecting diffraction, which disadvantages u4/3 in this example).

We don't have enough data to say one way or another what might happen with a 48 or 64 MP FF sensor. However, it's not difficult to imagine that the M43 optic, being a lot more than 1/4 the weight, size and price of the FF counterpart, will produce a better image than 1/4 of the FF optic can, assuming neither design is inadequate.

However even with a pixel density handicap, the standard kit 24-105mm lens on Canon delivers approximately 3x the resolution of the standard kit 12-50mm on Olympus. I am less familiar with Nikon, but there are similar kit lenses that deliver 4x on a D800 (which is still has a huge pixel density handicap compared to a 16MP u4/3 sensor).

This is simply wrong. First off, we are now comparing an L series lens (£689) with a basic throw-in kit zoom (£262) which is noted to be the worst kit zoom in all of M43 in terms of resolution - its advantages are elsewhere in the weather sealing, macro, and clever zooming facilities as world's only power zoom for video and mechanical zoom for still - none of these features are noted on the Canon L zoom. Even with this unfair comparison, the 12-50's best achieved resolution on Lenstip is 59 lpmm (on a 12 MP sensor) and the L 24-105 gets about 44. Translating to per image height/width and ignoring the format aspect ratio difference, to equal the L zoom performance an M43 lens needs to achieve 88. 88 is alot higher but not three times 59 - resolution is purely a linear (1-dimension) property and perceived as so by humans. Even if we are to talk about area, 88x88 is smaller than 59x59x3, by some margin.

Why not compare lenses of the same price?

For high priced zooms, the 24-70/2.8 zooms on FF can get you from 46-49, whilst the Panasonic 12-35 almost touches 80, which approaches the equivalent of 40 on FF, again on a 12 MP sensor. There is a less than 20% difference here.

And why not compare what the best each format can do in practice?

Lensrentals have put some very high quality lenses on the D800, currently the highest resolving FF platform. The best per image height resolution on their MTF50 data shows something like 1200, increasing to 1250 for the E version of the camera. The best number achieved by M43 through the same testing procedure is above 1000. The difference between the best FF and the best M43 is less than 30%. If you go with the DXOmark P-MP comparison this difference increases to 40%. Neither approaches the 100% better justified by the sensor size difference. (This is calculating in the linear, thus the 2x, rather than 4x, expected advantage for the FF.)

I don't think larger pixels on current FF sensors alone could explain away the deficit. The reality is that M43 lenses do not need to be 1/4 the weight, size and esp. price as their FF counterparts and some of that surplus could be spent on improving resolution and other image parameters. Similarly, medium and large format lenses, when producing 135-size crops, could not compete with native 135 lenses. The engineering reality in almost everything is that when you scale up your device, the quality of service delivered by that device rarely scales proportionally.

 

[Lost in Time]

tt321 wrote:

This is simply wrong. First off, we are now comparing an L series lens (£689) with a basic throw-in kit zoom (£262)

Yes, I know that - that's why I originally compared the $99 50mm with the $400 25mm. I was responding to the question "is anything 4x better", and these are two lenses which are functionally equivalent. I should point out that you are also comparing a $2K+ body with a $1k consumer device.

The problem here is that that all such comparisons are apples and oranges. The lenses are optimised for their corresponding bodies. Most existing FF lenses have been optimised (price/performance) for the 10-20MP range. Newer lenses are being optimised for assumed higher resolution future bodies (examples such the new Sigma 35mm f1.4, or the improbably expensive Canon 24-70mm II).

Also, when you look at resolution, you need to look at what aperture you are shooting, which part of the frame you are looking at, what the aberrations are, and so on. As an example, one of the most popular portrait lenses is the Canon 85mm f1.2. This lens is great in the centre, but very soft at the edges. It is very slow to focus and has a lot of interesting optical quirks. But people buy it for that centre sharpness combined with the character of its bokeh. There is no equivalent to this lens on u4/3 (and arguably on any other system).

The engineering reality in almost everything is that when you scale up your device, the quality of service delivered by that device rarely scales proportionally.

The same also applies when you scale down. The reason why a budget FF 50mm lens outperforms the highest performing u4/3 primes is because it is actually easier to design and manufacture than u4/3 lenses with a comparable field of view.

The reality is that there is no inherent reason (ie physics) why FF optics will perform worse than their u4/3 counterparts.

All that you are really looking at with current lenses is what the market will bear in terms of price/performance - and I should point out that current FF bodies and lenses are usually a lot less than 4x the price of similarly specified u4/3 kit - which in some cases, such as the 25mm manage to be both poorer performing *and* 4x more expensive than similar FF counterparts. That's not to denigrate the 25mm - if you want a 50mm field of view, fast AF and wide aperture on u4/3, it is a superb lens.

 

[MatsP]

DarkShift wrote:

DxOmark MP-scores depends on the camera the lens was mounted on. So sensors should have similar MP counts to make comparison reasonable.

I wonder why hasn't DxO published lens test results for the best m4/3 cameras E-M5 or GH3? Obviously they must have performed the tests as there are modules downloadable for a variety of lenses on those bodies in DxO Pro.

 

[tt321]

Lost in Time wrote:

All that you are really looking at with current lenses is what the market will bear in terms of price/performance - and I should point out that current FF bodies and lenses are usually a lot less than 4x the price of similarly specified u4/3 kit - which in some cases, such as the 25mm manage to be both poorer performing *and* 4x more expensive than similar FF counterparts. That's not to denigrate the 25mm - if you want a 50mm field of view, fast AF and wide aperture on u4/3, it is a superb lens.

Depending on what you meant by "poorer".

The following data is from Lenstip.

Canon EF 50/1.8 II above 35lpmm across three aperture values 4, 5.6 and 8. Max = 38 at 5.6.

PL25 above 70lpmm across four aperture values 2, 2.8, 4, 5.6. Max = 74 at 4.

So the max values have the PL25 inferior by 3% or so - it would have needed to be at 76 for absolute parity. but it does have a wider band of high resolution (over 35/70) apertures. For example, if you for some reason were to need the DoF of 5.6 on M43 and 11 on FF the former would deliver a sharper image. If you want the DoF of M43 at 2.8 and FF at 5.6 the latter is better. But these better/sharper would be very small differences indeed.

In addition, the 20/1.7, whilst cheaper than the PL25, is actually as good (if not better) in resolution, so the standard prime range is actually one of the good quality points in M43.

Your point is of course valid in that a throw-in kit prime worth almost zero money in FF is as good as highly priced units in M43, but as good with a 2x longer sensor means inferior by 50% in real sense, thus partially justifying the prices of the smaller format items.

The PL25 is the most expensive standard prime in the entire M43 lens pool. What is the price for the most expensive standard prime in the entire FF lens pool? Likely multiple times what Panasonic asks for the PL25. The PL25 is also the cheapest 50mm FoV prime lens in the entire M43 lens line, and as you said it costs multiple times compared to the cheapest 50mm lens in the entire FF lens line. Comparing a single item to a large population you are bound to have these results.

 

[ultimitsu]

Marty4650 wrote:

image quality, then it might be 20% better
portraits, then it might be 30% better
landscapes, then it might be 40% better

The whole point of this thread is, How did you come up with 20%, 30%, 40%?

sports and action, then it could be 100% better

Why just 100%? high IQ is maintained at iso 4 times higher, so shouldn't it be 4 times better?

portability and being inconspicuous, then it could be 400% worse

how does one define portability difference, just measuring size? i cannot see how it is 400% difference. a 6D + 24-105 is not 4 times bigger than a GH3+ 12-50.

value for money, then it could be 100% worse

How do you define value for money? IQ/price? in that case it is quite arguable that FF is better value for money since, you can get much better IQ with much cheaper lens.

But why stop at FF? If money or weight don't matter, then there are larger formats that will deliver better image quality than a FF camera will.

Except, they do not deliver better IQ than FF, all that they provide is extra resolution that most people no not need. A4 and A3 may be common printing but A2 and A1 are still pretty occasional for even half the pros. colour depth and DR is a draw bwtween MF and FF, when it comes to high iso performance, current FF are several times better.

But a camera is more than just IQ, current FF also have much better AF, FPS and responsiveness generally.

Even if the cost of a FF sensor came down to pennies, the smaller formats would still dominate, because a smaller sensor means smaller lenses that cost less.

But smaller lenses do not cost less if you are after the same IQ as FF. in fact to get the same IQ, it cost significantly more with smaller lenses than it is with FF. compare 25 F1.4 vs 50 F1.8, or 75 F1.8 compared 100 F2 / 135 F3.5 (which is no longer produced), or 7-14 vs 17-40, etc.

 

[Marty4650]

ultimitsu wrote:

Marty4650 wrote:

image quality, then it might be 20% better
portraits, then it might be 30% better
landscapes, then it might be 40% better

The whole point of this thread is, How did you come up with 20%, 30%, 40%?

My percentages are just guesses. Just like anyone elses. There is no real way to measure image quality without considering the specific need. If you are making billboards, then you might want the highest possible resolution. If you are going on a vacation, you might want something smaller and lighter that has reasonably good resolution.

If the only thing that mattered was shallowest DOF, then the bigger sensor wins 100% of the time. But there is a lot more involved, and a lot more things that people want from a camera.

sports and action, then it could be 100% better

Why just 100%? high IQ is maintained at iso 4 times higher, so shouldn't it be 4 times better?

portability and being inconspicuous, then it could be 400% worse

how does one define portability difference, just measuring size? i cannot see how it is 400% difference. a 6D + 24-105 is not 4 times bigger than a GH3+ 12-50.

value for money, then it could be 100% worse

How do you define value for money? IQ/price? in that case it is quite arguable that FF is better value for money since, you can get much better IQ with much cheaper lens.

Here's how:

But why stop at FF? If money or weight don't matter, then there are larger formats that will deliver better image quality than a FF camera will.

Except, they do not deliver better IQ than FF, all that they provide is extra resolution that most people no not need. A4 and A3 may be common printing but A2 and A1 are still pretty occasional for even half the pros. colour depth and DR is a draw bwtween MF and FF, when it comes to high iso performance, current FF are several times better.

Point being... that if you want your universe to consist only of professional photographers, then there are some pros who shoot fashion photography or do studio work who use medium format cameras. If they don't need them, then why do they buy $40,000 cameras when a $6,000 FF camera is "just as good?" or "several times better?"

But a camera is more than just IQ, current FF also have much better AF, FPS and responsiveness generally.

When you say "most people need" then you have just described the camera phone. Most people don't need FF even if a few people think they need it, and a few more just want to have it.

AF speed isn't an image quality issue unless you are shooting sports/action photos. Not everyone is an NFL sideline photographer. How does AF speed make a portrait or a landscape have better image quality?

FPS? Compare the OMD (9fps) or the NEX7 (10fps) to the D800 (4fps) and 5DIII (6fps). It seems like MILC cameras have a much higher continuous shooting rate.

"Responsiveness" is argumentative.

Even if the cost of a FF sensor came down to pennies, the smaller formats would still dominate, because a smaller sensor means smaller lenses that cost less.

But smaller lenses do not cost less if you are after the same IQ as FF. in fact to get the same IQ, it cost significantly more with smaller lenses than it is with FF. compare 25 F1.4 vs 50 F1.8, or 75 F1.8 compared 100 F2 / 135 F3.5 (which is no longer produced), or 7-14 vs 17-40, etc.

You say "same IQ" but you mean "same shallow DOF." They are not strictly speaking the same thing.

Your friend Mike Fewster was right when he said:

"What I took issue with was the implication that because the area of the sensor is 4X larger, the picture quality is 4x better. It's better, but not by an amount that can be quantified in direct proportion to the areas of the sensors."

--
Marty
http://www.fluidr.com/photos/marty4650/sets/72157606210120132
http://www.flickr.com/photos/marty4650/sets/72157606210120132/show/
my blog: http://marty4650.blogspot.com/
Olympus E-30
Olympus E-PL2
Sony SLT-A55
http://farm7.staticflickr.com/6101/6318442842_7b93cb589b.jpg

 

[Lost in Time]

Marty4650 wrote:

Here's how:

[70-200m f2.8L IS vs 35-100 f2.8 OIS].

You could equally well match the Panasonic against the 70-200mm f4 L IS, which is only $1150 vs $1400 for the Panasonic (amazon).

The only thing that you can consistently say about u4/3 is that the lenses are smaller and lighter than FF counterparts - which is the entire point of the system really.

--

http://www.flickr.com/14989580@N00

 

[olliess]

Lost in Time wrote:

Marty4650 wrote:

Here's how:

[70-200m f2.8L IS vs 35-100 f2.8 OIS].

You could equally well match the Panasonic against the 70-200mm f4 L IS, which is only $1150 vs $1400 for the Panasonic (amazon).

The Canon f/4 L lens is still a full stop ahead of the Panasonic f/2.8 in light gathering. Could also look at the $1400 70-300 f/4.5-5.6 DO IS zoom if you care most about smallness.

 

[Aleo Veuliah]

Marty4650 wrote:

There is absolutely no doubt that a camera with a FF sensor will provide better image quality than one with a 4/3 sensor, but it is nothing near "four times better." The FF sensor is indeed four times larger, but this does not translate into four times better image quality.

You do get "somewhat better" but you have to spend a lot more money to get it, and you have to be willing to carry around a much heavier camera and lenses. This is because the diminishing returns curve really sets in.

Of course, it really depends on what you define the word "better" as meaning.

If we are talking about:

• image quality, then it might be 20% better
• portraits, then it might be 30% better
• landscapes, then it might be 40% better
• sports and action, then it could be 100% better
• shallow DOF, then I might agree that it is 400% better
• portability and being inconspicuous, then it could be 400% worse
• value for money, then it could be 100% worse

The bottom line is your choice should depend on how much money you are willing to spend for something slightly better. How critical are your needs?
For most amateur photographers, the need is met at a lower level than a camera with a 4/3 sensor. For many, a camera phone will do. For others, a camera with an APS-C sensor will do the trick.
Despite all the current hoopla over FF cameras, this is still a niche market, dominated by professional photographers, and amateur enthusiasts with lots of disposable cash.
For those who demand the best possible image quality, then it might make sense to spend a lot more to get it. But why stop at FF? If money or weight don't matter, then there are larger formats that will deliver better image quality than a FF camera will.
Even if the cost of a FF sensor came down to pennies, the smaller formats would still dominate, because a smaller sensor means smaller lenses that cost less. The smaller format will always be easier to carry and be less conspicuous. This might be why you don't see too many people walking around with 8"x10" view cameras today. (And yes, they will provide MUCH better IQ than a 36 x 24mm sensor will.)
People don't spend tons of money for cameras they don't need. And most people don't need a FF camera.

 

[ultimitsu]

Marty4650 wrote:

ultimitsu wrote:

Marty4650 wrote:

image quality, then it might be 20% better
portraits, then it might be 30% better
landscapes, then it might be 40% better

The whole point of this thread is, How did you come up with 20%, 30%, 40%?

My percentages are just guesses. Just like anyone elses. There is no real way to measure image quality without considering the specific need. If you are making billboards, then you might want the highest possible resolution. If you are going on a vacation, you might want something smaller and lighter that has reasonably good resolution.

I think this analysis is flawed. We are not talking about whether it meets your needs - it is an entirely different discussion, we are talking about in absolute terms, how much is FF's IQ better than m43

sports and action, then it could be 100% better

Why just 100%? high IQ is maintained at iso 4 times higher, so shouldn't it be 4 times better?

I would still like to hear your thoughts on this.

portability and being inconspicuous, then it could be 400% worse

how does one define portability difference, just measuring size? i cannot see how it is 400% difference. a 6D + 24-105 is not 4 times bigger than a GH3+ 12-50.

And this too.

value for money, then it could be 100% worse

How do you define value for money? IQ/price? in that case it is quite arguable that FF is better value for money since, you can get much better IQ with much cheaper lens.

Here's how:

This comparison shows the fallacy often presented by this type pf comparison. 35-100 F2.8 on m43 is equavelent to 70-200 F5.6 - without giving the resolution of 200mm on FF. To achieve the same IQ as 35-100 F2.8 on m43, an FF only needs something like 70-300, which cost 649, and it provides much more reach and actual resolution on the long end.

It does weigh more at 630g, but weight and overall system convenience is not part of this discussion - we are not trying to determine which is better for X use, we are only focusing on IQ, and to an extent, the cost of obtaining that IQ.

But why stop at FF? If money or weight don't matter, then there are larger formats that will deliver better image quality than a FF camera will.

Except, they do not deliver better IQ than FF, all that they provide is extra resolution that most people no not need. A4 and A3 may be common printing but A2 and A1 are still pretty occasional for even half the pros. colour depth and DR is a draw bwtween MF and FF, when it comes to high iso performance, current FF are several times better.

Point being... that if you want your universe to consist only of professional photographers, then there are some pros who shoot fashion photography or do studio work who use medium format cameras. If they don't need them, then why do they buy $40,000 cameras when a $6,000 FF camera is "just as good?" or "several times better?"

But my universe does not consist only pros, some pros have a need for MF and their cost is justified for them. But for average joe like me who shoot a bit of everything, MF simply are not up to the task (as well as unjustifiably expensive)

But a camera is more than just IQ, current FF also have much better AF, FPS and responsiveness generally.

When you say "most people need" then you have just described the camera phone.

I am not talking about what most people need, I am talking about the camera performance one get for the money, FF are way ahead of MF.

AF speed isn't an image quality issue unless you are shooting sports/action photos. Not everyone is an NFL sideline photographer. How does AF speed make a portrait or a landscape have better image quality?

The AF point was directed at yoru comment of why stop at FF - because beyond FF, AF start to suck and the camera becomes very limited in its application. Thus there are good reasons why stop at FF. Thought it is not directly related to the main point of this thread.

FPS? Compare the OMD (9fps) or the NEX7 (10fps) to the D800 (4fps) and 5DIII (6fps). It seems like MILC cameras have a much higher continuous shooting rate.

They do not have continous AF and AE durign this burst, if i understand it correctly.

And we are not comparing FF to m43 on this point are we? we are comparing FF to Mf, only because you asked why stop at FF.

"Responsiveness" is argumentative.

Again, this is FF vs MF, not sure why you think it is argumentative.

Even if the cost of a FF sensor came down to pennies, the smaller formats would still dominate, because a smaller sensor means smaller lenses that cost less.

But smaller lenses do not cost less if you are after the same IQ as FF. in fact to get the same IQ, it cost significantly more with smaller lenses than it is with FF. compare 25 F1.4 vs 50 F1.8, or 75 F1.8 compared 100 F2 / 135 F3.5 (which is no longer produced), or 7-14 vs 17-40, etc.

You say "same IQ" but you mean "same shallow DOF." They are not strictly speaking the same thing.

I am not talking about just DOF, I am talking about IQ - which is partially to do with amount light end up hitting the sensor.

Your friend Mike Fewster was right when he said:

"What I took issue with was the implication that because the area of the sensor is 4X larger, the picture quality is 4x better. It's better, but not by an amount that can be quantified in direct proportion to the areas of the sensors."

Why you think he is right, what makes 4 times better, is the core of this discussion.

 

[Great Bustard]

ultimitsu wrote:

Great Bustard wrote:

tt321 wrote:

Great Bustard wrote:

ultimitsu wrote:

Fellow forum poster Mike Fewster and I were enaged ina discussion of merits of FF over M43. Mike Fewster made an interesting statement:

"What I took issue with was the implication that because the area of the sensor is 4X larger, the picture quality is 4x better. It's better, but not by an amount that can be quantified in direct proportion to the areas of the sensors."

I would like to ask Mark and everyone else, What do you consider to be "4 x better"?

Before a meaningful answer can even begin to be given, one would have to define exactly what "4x better" means.

Mathematically, if something is 4x better, it should be 5x as good.

Now you need to define "5x as good". ;-)

So we are starting on a journey already on the wrong foot...

And discussion that attempts to quantify "better" without rigorously defined terms is, as you say, "already on the wrong foot".

That is really what I am trying to show to Mr Mike Fewster - IQ covers a broad range of qualities and is impossible to say X is 4 times better than Y, regardles how much X is actually better than Y. We can break it down to several metrics, such as resolution, DR, colour depth, high iso SNR, etc, there can be a comparison of two cameras on each metric. We cannot however say thus overall X is 4 times better than Y.

Indeed. The way I like to say it is:

http://www.josephjamesphotography.com/equivalence/index.htm#IQ

But what, exactly, is IQ? That is difficult to define -- so much so that the term seems to lose any meaning in an objective sense. However, as will be discussed later in this section, the subjective nature of overall IQ comes from how we value individual objective components of IQ.

To respond to Mike's original contention , I would say in some metrics FF is about 4 times as good, in others, it is not that much better. Repeating what I said in the other thread on FF's advantage over M43, I will copy paste the following.

I would shy away from saying "4 times as good" and say instead "4 times greater".

1, shallower DOF. when standing at the same distance to the subject, using lenses of equivalent FOV and similar aperture ratio, larger sensor gives you shallower DOF, useful in portrait shooting. for example a FF with 85 F1.8 (a moderately priced lens), will give you much shallower DOF than m43 + 45mm F1.8, to get the same look as the FF, m43 need a F0.9.

Sure.

2, better lens shapness. flowing from the previous advantage. portrait images are often sharper with FF because the F ratio does not have to be so large to get the desired isolated look. for example many people use 50mm lens at F2.8 for portrait, all seven 50mm lenses from canikon are bleeding sharp at F2.8, to get the same look one would need to use a 25 F1.4 on m43 at F1.4, it is a lot softer. this is not even taking into account the inherent extra resolution in FF.

Sure, depending on the lens, the DOF, and where in the frame you look.

3, lower cost for shallow DOF. also flowing from advantage number 1 is cost, like I said 85 F1.8 is a moderest cost lens, canikon both offer it at around 500 usd, a 45mm F0.9, if exists, would cost no less than 2000 USD. both canikon also offer a 50 F1.8 at about 100 USD, nikon offers a superior model at 220 USD, the m43 25mm F1.4 cost 499.

Of course, we have to consider the system cost, not merely the cost of the lenses.

4, more resolution in good light. this needs no further explanation. As you have commented, the advantage is not apparent on smaller prints. One thing worth mentioning is that people tend to over-estimate the miximum size where you cannot see the difference. Because all lenses are imperfect, some resolution is eaten up by that imperfection. 7-14, 25 f1.4, 45 F1.8 can be considered top end lenses for M43, right? Even with the 16mp GH2, none was able to achieve more than 10mp . So basically, with m43,, the biggest you can print at 300 DPI is about A4.

DPI (dots per inch) is simply a function of the printer output. I believe you want to say PPI (pixels per inch) which is a function of the pixel count of the image file.

However, even PPI is not the relevant measure, as we can upsample to get any PPI we like. The relevant measure is lw/ph (line widths per picture height, usually measured at 50% contrast).

For sure, more pixels increases the lw/ph, just as does a sharper lens, or a larger sensor. The problem is when we have a sharper lens on a smaller sensor with fewer pixels against a less sharp lens on a larger sensor with more pixels. How do they compare, then? As always, it depends on the differences in lens sharpness, pixel count, and sensor size.

5, Better IQ low light. This is the big one. if you are an average joe who shoots a bit of everything then majority of your shooting will not be done under ideal lighting. most indoor flashless shots, even on a good day would exceed iso100. quality of image is determined by amount of light hitting the sensor and that in turn is determined by sensor size. I will repeat the data I give you before - the maximum iso that one can maintain SNR of 30dB, DR of 9 EVs and a color depth of 18bits is only 618 on G5, 826 on OM-D, on the FF side of the battle we have D800 with 2853 and D3s with 3253.

With the understanding, of course, that the lower noise requires a wider aperture for a given shutter speed, which in turn results in a more shallow DOF that may, or may not, be desirable.

6, Lower cost for lens resololution. Similar to point number 3, it is cheaper to buy lenses that would give you X resolution with FF. As I have told you in point number 4 that those three top quality m43 lenses cannot giev you more than 10 mp, about half of their money, you can buy Canon 17-40, 50 F1.8, 85 F1.8, they will produce more than 10mp of resolution on any current FF body.

Again, we have to consider the system cost, not the cost of just one component, like the lenses.

7, 8, More dynamic range and better colour depth. The jury is still out as to why FF do better in these two areas. But measured data indicate that bigger sensor is indeed better. OMD uses the same sony technology as D800 but has less colur depth and drynamic range. os if these two things are on your priority list, you go FF.

All these things depend on noise, which, in turn, depends on the total amount of light falling on the sensor and the efficiency of the sensor. In many instances, we find the smaller sensor outperforming the larger sensor in these measures (e.g. D7000 vs 5D2). That said, DR, in my opinion, is one of the most misunderstood measures of IQ. For example, the Canon G12 (compact) and Canon 5D2 (FF) have the same DR, but we wouldn't say they have the same IQ, right?

 

[Great Bustard]

John King wrote:

I am also extremely concerned about colour fidelity, and that does not appear to be a characteristic that I would buy any 135 format camera for. My touchstone is to compare the bare print to the object under as close to the original lighting conditions as possible. To this end, I also work in a ProPhotoRGB 16 bit colour space. Many very common colours are out of gamut for aRGB, and sRGB is a bad joke, even when using a 16 bit colour space.

Well, let's see. You said the following about this photo of yours :

There is no way I could have captured this wide DR image with the nuances of tonality, colour and contrast with my E-510. I can guarantee this. With slight tweaks to the WB, exposure, recovery, brightness and contrast in ACR this image goes from merely ordinary to absolutely stunning. I have done it. However, as I said before, the image posted here has had no PP applied to it.

OK. Here are the before and after photos:

http://forums.dpreview.com/forums/post/33935142

"Stunning" isn't the word I'd use, myself, but, well, that's just me. In fact, given that I cannot even get the white balance right with my own pics:

http://forums.dpreview.com/forums/post/38834123

Wonderful happy snap, Joe. Lousy WB ... Thanks for linking the proof.

one might understand why I was unable to appreciate the "nuances of tonality, colour and contrast" in your photo.

OK, so now we have a visual context for your opinions on color, and can proceed from there.

Almost all dSLR cameras can capture a wider gamut than aRGB; No commercially available monitor can display such a gamut and bit depth; Almost all dye and pigment inkjet printers can print a far wider gamut than aRGB-16 ... Even our Canon MG6150 can manage that ... Let alone my Epson R3880.

Colour accuracy is also dependent upon the characteristics of the CFA, a much neglected item in these discussions; particularly for the intermediate colours in the RGB image - cyans, yellows and magentas.

Just a few thoughts on the subject.

So, if I understand you correctly, what we see on our computer monitors is irrelevent, we have to see the print? Do I have that right?

Anyway, what are your thoughts about this?

http://www.imaging-resource.com/PRODS/E5D/E5DIMATEST.HTM

The Canon EOS-5D showed good color accuracy overall, but tended to get pretty carried away with strong reds. This is a fairly common response with this target (the MacBeth ColorChecker), but the 5D falls toward the upper end of the range of oversaturation of reds that we've seen in professional SLRs. Hue accuracy is among the best we've seen though, with only slight shifts in cyans, magentas, and oranges. Average saturation was 108.5% (oversaturated by 8.5%, almost entirely in the reds), average "delta-E" color error was 4.98. (Hue error, after correction for saturation, among the best we've seen.)

or this:

http://www.imaging-resource.com/PRODS/E5D2/E5D2IMATEST.HTM

The Canon 5D Mark II showed excellent color accuracy overall. Hue accuracy was exceptional, with only small shifts relative to mathematically correct colors Average saturation was 105.3% (oversaturated by only 5.3%, mostly in the deep blues and reds, with just a little in some greens). Average "delta-C" color error was only 3.94 after correction for saturation, which is very low, and the best score of the group. (Hard to say without exhaustive checking, but this may very well be the best delta-C score we've ever seen.) All in all, an excellent color response for an SLR. Mouse over the links below the illustration above to compare results with competing models.

How do those tests jive with your comment that:

I am also extremely concerned about colour fidelity, and that does not appear to be a characteristic that I would buy any 135 format camera for.

 

[Martin.au]

Lost in Time wrote:

Marty4650 wrote:

Here's how:

[70-200m f2.8L IS vs 35-100 f2.8 OIS].

You could equally well match the Panasonic against the 70-200mm f4 L IS, which is only $1150 vs $1400 for the Panasonic (amazon).

The only thing that you can consistently say about u4/3 is that the lenses are smaller and lighter than FF counterparts - which is the entire point of the system really.

--

http://www.flickr.com/14989580@N00

Actually both lenses are about the same price on Amazon.

Why would you go FF to capture the same images (or near enough) as m4/3s? If you're going FF, wouldn't the point be to get substantially better IQ?

 

[happypoppeye]

I consider that 4x better...

Hmmm ...4 potato chips is 4x better than 1

...4 weeks vacation is 4x better than 1 ...well, actually probably closer to a hundred times better

 

[Great Bustard]

happypoppeye wrote:

I consider that 4x better...

Hmmm ...4 potato chips is 4x better than 1

...4 weeks vacation is 4x better than 1 ...well, actually probably closer to a hundred times better

Are four wives 4 times better than one, or might it be the other way around? ;-)

 

[StevenAdler]

Only thing that matters is whether you have one with you when you visualize a shot. I've shot some nice photos on an iPhone. Early 20th Century artists like Alfred Steiglitz used glass plate cameras and got superb results.

I chose MFT because they weigh less than FF and with prime lenses get great results. Who cares if there are other cameras that are better. They aren't in my hand and they shouldn't be in your head...

 

[tedolf]

ultimitsu wrote:

Fellow forum poster Mike Fewster and I were enaged ina discussion of merits of FF over M43. Mike Fewster made an interesting statement:

"What I took issue with was the implication that because the area of the sensor is 4X larger, the picture quality is 4x better. It's better, but not by an amount that can be quantified in direct proportion to the areas of the sensors."

I would like to ask Mark and everyone else, What do you consider to be "4 x better"?

What is better about FF compared to u 4/3 is shallow DOF wide angle shots and continuous tracking AF.



That is about it for all practicle purposes and print sizes.



Micro four thirds is better at portrature and telephoto because it has enough DOF to allow you to shoot wide open using a lower ISO than FF.



Tedolph

 

[John King]

As with your previous attempt to pick a fight with me, I am simply not interested for reasons that would be patently obvious to the vast majority of posters here.

--

-