Sony NEX3 and NEX5 EVIL cameras with new E-mount

ampguy said:
Several folks here were very pleased with 4/3rds sensors, zoom lenses, crop, and use with M lenses, until they tried RD1s and M8s, now they don't talk about 4/3 much...interesting...
Click to expand...

4/3 is basically the same size as APS-C. The vertical height is within 20%, then it comes down to whether you prefer a 3:2 aspect ratio to 4:3. With respect to IQ there's little doubt: the Olympus lenses are terrific and the current Olympus DSLRs are comparable in performance to APS-C gear in many applications. Sometimes a little better, sometimes a little worse. As we now know, for example, the IQ of an E-PL is ahead of the D-3000.

So what? In big-picture terms we're splitting hairs.

With respect to the M8 and RD-1 -- these products were so successful that they are no longer manufactured, while micro 4:3 equipment appears to have captured as much as 20% of (what was formerly) the DSLR market.

Given these facts, I think it's hilarious that you'd claim that the system that's selling well is comparable to an extinct system, while *two* systems that are *actually* extinct are, according to you, not so comparable.

Here is a suggestion: all of the digital systems we've just discussed are capable of yielding wonderful -- and indeed, technically similar -- results. In most cases, without EXIF data or pixel peeping, you'd be hard-pressed to tell the difference. I know that my E-620 gives me results that are technically superior in several respects to the best results I've ever obtained with my Nikons or my M6 and film (say, Provia), and I know that the E-PL1, being a couple of releases later, would be even better.

I still mainly use the M6 because I love using the damn thing, and I kind of like developing film. That's perverse, but I don't apologize for it.

So. Shoot what you want to shoot, and don't worry so much that other people might reach different conclusions.
 
Last edited:
NickTrop said:
@fdigital... Also, your pic is a bit misleading. You have a prime lens on the Oly, the Nikon has a zoom. Put the zoom on the Oly and have it fully zoomed out... then compare.
Click to expand...

No, the lens in that picture is NOT a prime, it's the 14-42mm kit lens. The prime is under half the size of the 14-42mm kit lens.

In fact here's a picture of an e-p1 with the 17mm prime lens next to a canon 35mm f1.4L:

4565102336_444a8d2e99_o.jpg
 
Last edited:
ampguy said:
225mm squared = 370mm squared???
Click to expand...

You mean square mm, not mm squared.

Four thirds: 13 mm high.
APS-C (Canon): 14.8 mm high (14% taller than 4/3).
APS-C (Nikon): 15.7 mm high (6% taller than Canon; 21% taller than 4/3).

IQ generally scales with the linear dimension of a sensor, not with sensor area. It stands to reason: what we're after is the amount of enlargement (magnification) that you can get from a negative or a digital file. An 8x10 enlargement is an ~8x enlargement of a 24x36 mm negative, and a ~4x enlargement of a 645 negative. All else being equal, a Nikon APS-C file might yield a 4:3 aspect ratio print 20% larger (linear) than a four thirds file, at equivalent IQ.

Don't believe me? Then when was the last time you saw an MTF expressed in cycles per mm^2, rather than cycles per mm ? (Answer: you probably never have, because that's not how it's done!)

Thus, 4/3 and APS-C are basically the same except that APS-C has a wider aspect ratio. If you don't care for the wider aspect ratio (and this is an aesthetic question, not a technical one), then there is effectively no technical advantage to an APS-C sensor. It's the difference between an 8x10 print, and a 9.5x12 print -- at most. To do meaningfully better than 4/3, you have to go full frame or MF. Seriously: you need to be shooting a 5D or an M9 or a D3 or a Pentax D645 or a Phase One.

Other factors -- tripod use, image stabilization, RAW or JPEG processing, and of course lens quality -- will almost always swamp that 15-20% difference. This is why the 4/3 Olympus E-620 was smack dab in the middle of the APS-C cameras, and the m4/3 E-PL1 destroyed the Nikon D3000 in the DP Review comparisons. The sensors are close enough that the difference is, in fact, hair-splitting. It is a near certainty that, with two files stripped of EXIF data and identifying features like aspect ratio, you could not reliably tell the difference -- except, of course, that the Olympus lenses are often (though of course not always) better than the APS-C equivalents.
 
Last edited:
@fdigital. Okay, it's a zoom. I stand corrected. When you power up the camera - how far out does the zoom extend?

DxO ratings:

D5000 - Sensor: 72; Color Depth: 22.7; Dynamic Range: 12.5; Low Light ISO: 868
Oly EP1 - Sensor: 54.8, Color Depth: 21.4; Dynamic Range: 10.4; Low light ISO: 536
Oly EP2 - Sensor 55.6; Color Depth: 21.5; Dynamic Range: 10.4; Low Light ISO: 505

In every key metric, the Olys either somewhat lower, to substantially lower in areas that matter like low-light shooting.

I'm not disputing that the 4/3 are smaller. Obviously, they are. Extend the zoom lens out (like I did in the store...) and add a viewfinder and they're closer. But even without, what does the smaller size "buy" you over a smaller, less expensive compact DSLR? That's - again, what I'm saying, and that's the exact same conclusion Stella Dante's article reached independently. Nothing - as far as I can tell, other than compromises and higher cost. Though, admittedly, they are sexy looking. Also, you're comparing it to a Canon F1.4 lens that's quite a bit larger than the 35mm Nikon DX prime. This lens is the same size as your garden variety 50mm f1.8 film SLR lens we're all familiar with, not the massive Canon prime in your picture. (Why are you cheating? 😉
2183_af-s-dx-nikkor-35mm-f-1-8g_front.jpg
 
Last edited:
huh?

huh?

pick any aspect ratio you like from square to wide and APS-C (Nikon) is significantly larger.

Compare the 2 camers (EPL1 and D40x) at dxo mark, and you'll see that the d40x numbers are all higher than the epl1:

Higher for Sensor.

Higher for Color depth

Higher for Dynamic range (over 1 point!)

and

Higher for Low Light ISO performance.


semilog said:
You mean square mm, not mm squared.

Four thirds: 13 mm high.
APS-C (Canon): 14.8 mm high (14% taller than 4/3).
APS-C (Nikon): 15.7 mm high (6% taller than Canon; 21% taller than 4/3).

IQ generally scales with the linear dimension of a sensor, not with sensor area. It stands to reason: what we're after is the amount of enlargement (magnification) that you can get from a negative or a digital file. An 8x10 enlargement is an ~8x enlargement of a 24x36 mm negative, and a ~4x enlargement of a 645 negative. All else being equal, a Nikon APS-C file might yield a 4:3 aspect ratio print 20% larger (linear) than a four thirds file, at equivalent IQ.

Don't believe me? Then when was the last time you saw an MTF expressed in cycles per mm^2, rather than cycles per mm ? (Answer: you probably never have, because that's not how it's done!)

Thus, 4/3 and APS-C are basically the same except that APS-C has a wider aspect ratio. If you don't care for the wider aspect ratio (and this is an aesthetic question, not a technical one), then there is effectively no technical advantage to an APS-C sensor. It's the difference between an 8x10 print, and a 9.5x12 print -- at most. To do meaningfully better than 4/3, you have to go full frame or MF.

Other factors -- tripod use, lens quality, RAW or JPEG processing, and of course lens quality -- will almost always swamp that 15-20% difference. This is why the m4/3 Olympus E-620 was smack dab in the middle of the APS-C cameras, and the E-PL1 destroyed the Nikon D3000 in the DP Review comparisons. The sensors are close enough that the difference is, in fact, hair-splitting. It is a near certainty that, with two files stripped of EXIF data and identifying features like aspect ratio, you could not reliably tell the difference -- except, of course, that the Olympus lenses are often (though of course not always) better than the APS-C equivalents.
Click to expand...
 
NickTrop said:
DxO ratings
Click to expand...

Apples. Oranges.

These differences seen by DxO are due to Sony's best-of-breed sensor tech, as much as sensor size. Same reason that the APS-C Pentax K-x is a lot better than the APS-C Pentax K7 in DR and in sensitivity. The D-5000 has the same sensor as the K-x. It's probably the best APS-c sensor ever manufactured, and not by a little. The thing is a small miracle. What does this mean? Only that everyone else -- Samsung, Canon, Panasonic et al., are a generation behind, and in a year they'll be equivalent.

I note that you are not talking about the DxO marks for the M8 or the RD-1.

If I were buying into a system today, which I'm not, it would probably be a K-x with the 21 and 70 Limiteds: Sony sensor and terrific, compact lenses that are substantially better than the Nikon equivalents. But I have no illusions: I know that the Olympus 12-60 on an E-620 is a higher-resolution setup than the Pentax setup, especially wide open, and especially at my favored FL of 35mm-e. But not by much.

As I said before, the real differences are between full frame and everything else.
 
Last edited:
Here ya go...
zrtn_002p3d5fa16f_tn.jpg
Now add a viewfinder at the top for a more fair comparison (cheater 🙂 )

Again - does the "smaller" - albeit still unpocketable size, justify the higher cost, loss of an optical viewfinder, lower IQ (see testing), worse battery life, over a less expensive compact DSLR? I'm still not getting what this buys you...
 
Last edited:
Oh - also, you're saying this is a "digital PEN"? Well, I'd dispute that. Digital PENs were 1/2 frames, so you could shoot twice as many frames per roll of film - a useful purpose to justify the lower (but still good) image quality. Digitals don't use film, obvs. So what purpose would a "digital PEN" serve? Do you get more shots with it? 😛

EP1-vs-D5000.jpg

Now, extend the zoom lens all the way out of the Oly, put a viewfinder on top, take the zoom off of the Nikon, and put a regular-sized "nifty fifty" 50mm 1.8 prime on it.

I now declare myself - as any fool can plainly see, the winner of this debate.
 
Last edited:
I think as long as dynamic range is over 11, and low light iso is over 600, all is good.

Do any 4/3rds cameras meet this spec?
 
Ah. Here we go.

DxO Mark for:

Leica M8 (1.33 crop): 59.4.
Olympus E-PL1 (2.0 crop): 54.
Leica M9 (1.0 crop): 68.6.
Pentax K-x (1.5 crop): 71.9.

Absurd conclusion: the $490 Pentax K-x (1.5X crop) is a "better" camera than the $9000 M9 (full frame). (The K-x will be under $1500 if you put a Pentax 31mm Limited on it, and -- going by the numbers -- that lens will work as well as or better than the 50mm Summicron-M on the M9, with a wider aperture and arguably better bokeh.)

They don't score the RD-1. It would probably come out on the bottom (but remember: it takes *great* pictures).

Now, do one more thing. Take a look at the DP review comparison showing REAL photographs taken with the D3000 (DxO 62.4) and the E-PL1 (DxO 54). Oh, and take a look at the ISO ladder, too.

As I said, these just are not huge differences in image quality. The numbers might be different, but if you can't see big differences at 100% in test shots, the practical quality difference is negligible.
 
Last edited:
ampguy said:
I think as long as dynamic range is over 11, and low light iso is over 600, all is good.

Do any 4/3rds cameras meet this spec?
Click to expand...

That's a pretty arbitrary cutoff. Velvia, Provia, Kodachrome certainly never offered 11 stops of DR, and the difference between ISO 400 and ISO 640 is (woohoo!) half a stop.

Seriously. If those numbers really matter to you, stop torturing yourself and get a D700 or a used 5D.
 
Last edited:
Why is it an "absurd conclusion"? They ran their test - which other camera test sites use, along with optics manufacturers... that be the numbers.
 
NickTrop said:
Why is it an "absurd conclusion"? They ran their test - which other camera test sites use, along with optics manufacturers... that be the numbers.
Click to expand...

The conclusion -- that the K-x is "better" -- is absurd because the cameras are designed to do different things. A camera is more than a sensor. It is also a user interface and a platform for lenses.

The user interface is the main reason that I still shoot an M6 with film (the one in the picture to the left) most of the time.

Another example: the E-PL1 with Panasonic 20/1.7 has about the same low-light capability as your D5000 and 35/1.8. Maybe 2/3 of stop worse. On the other hand, it has IS that will give you another 2-3 stops at low shutter speeds, where the D5000 has none. Of course, the Pentax K-x, like the E-PL1, has IS.

On the other hand, I've got an anthropologist acquaintance who does photo-ethnography in big developing-world cities, much of it at night, in neighborhoods mainly lacking streetlights. His camera? A 5DII with 35/1.4. He does not screw around with APS-C.
 
Last edited:
semilog said:
The conclusion -- that the K-x is "better" -- is absurd because the cameras are designed to do different things. A camera is more than a sensor. It is also a user interface and a platform for lenses.

The user interface is the main reason that I still shoot an M6 with film most of the time.
Click to expand...

And now the rationalizations start...
 
NickTrop said:
And now the rationalizations start...
Click to expand...

I use digital imaging as a quantitative, technical tool in my job as a research scientist. My decisions about equipment in that setting are driven primarily by technical considerations.

I use still photography to take pictures for personal reasons when I'm not at work. My decisions about equipment in that setting are driven primarily by aesthetic considerations.
 
Last edited:
antiquark said:
How about this one.
509320258_a0911a9c50.jpg
Click to expand...

Wicked! Now put a 5DII or D3 in for good measure.

Of course when I want a really small camera, I have a Ricoh GR-1 that produces IQ as good as I anything ever saw with my Nikon film cameras, and close to what I can obtain with the M6, albeit with a clumsier UI.
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom