Nex-7 outresolves M9 w/ 50 ASPH

[houble]

nikon J1 pixel density

nikon J1 pixel density

Maybe, I misunderstood the test, but wouldn't show the nikon 1 (if someone would build an adapter), with the same lens and some cropping and resizing nearly the same results, because it has the same pixel density than the NEX7. Or even better because it would benefit from the center resolution of the lens even more. Would this also mean that the nikon 1 is better than the NEX7, because its cheaper also.

Sorry, if I am totally wrong...

 

[sol33]

Oh well... If you do the math (24Mpixel/18Mpixel*1.5crop^2) you realize the NEX7 has three times the pixel density of the M9. You can see that in the images. No surprise here. However, in real world situations few M9 owners would crop their images that much, I suppose...

If you could manage to screw the Summilux somehow on a point-and-shoot, you would get an even better resolution. The NEX7 had absolutely no chance in terms of resolution against an ancient point-and-shoot with a tiny sensor. It would be even cheaper too. The only downside would be a huge crop factor. 🙂

The whole test doesn't make a lot of sense to me, but well...

EDIT:

To be fair, it was interesting to see that wide open the NEX has much better resolution in the center but only about equal resolution in the corners. I guess that wide open the microlenses on the M9 sensor do make a difference in the corners. (And even more so in the corners of the full image which were not shown but cropped.)

 

[uhoh7]

Sooner or later there will be no benefit in owning a Full Frame camera

I don't think it will be sooner. e.g. to achieve 28mm fov on aps-c you need a 17mm. In general the lenses get slower--or wildly expensive, and there are all sorts of technical issues--distortion etc.

The n7 has a ton of incredible features, M9 chasing being only one. But if the camera had a ff sensor tuned to take the wide light---even the skeptical might take a breath.

Canikon et al --are they oblivious to the lust for such a machine?

Many barriers are down---the evf is perhaps the first "good" one; the little 5n shoots the CV 15 very well right into the corners, though it's still picky in the wides--not really caring for many of the ziess (though zm18 is very good). But even the M8 still must be prefered because of the sensor size.

Perhaps if nothing else the n7 will prod someone into relenting and producing a nice liitle FF EVIL with a sensor that likes the old glass.

 

[EdwardKaraa]

I don't particularly like MR neither as a photographer, nor as an equipment reviewer. But to be fair, I looked at the test, and it seems to be fine from the technical point of view. The Nex7 sensor may very well be better than than that of the M9, despite the AA filter. But as pointed out by someone above, the Nex7 is neither FF nor RF. So there is no need to be alarmed or feel threatened. If anything, the Nex7 will be yet another excellent body that can use M lenses, not a bad idea imho.

 

[bensyverson]

MR's test is about seeing whether there's more detail to be rendered by Leica lenses that is not captured by the M9 sensor. My take: duh. The results: clear.

His test is 100% valid and right. So, what are we arguing about?

 

[semilog]

His test is 100% valid and right. So, what are we arguing about?

$7000, so far as I can tell.

 

[tstermitz]

Lenses still out-resolve sensors.

Lenses still out-resolve sensors.

Yes, no surprise: the Nex-7 is sampling the image projected by the f/1.4 50mm Leica lens at 24mpix over a crop-frame, while the Leica M9 samples the same image at FF 18mpix. (I'm too lazy to calculate the size of each pixel, and I don't know how to fold in the Bayer pattern and anti-aliasing.)

This review makes me consider several interesting things:

(1) As Reichman said, Leica lenses out-resolve circa 2009 sensors, and also the latest 2011 sensors.

(2) Sensor technology is improving. I would expect the Leica M10 to be an improvement over the M9 in both in ISO and resolution (#pixels).

(3) See the article at Diglloyd.com where he concludes that very few of today's top lenses really match up with new, 36Mpix FF cameras. Lloyd Chambers focuses (so to speak) on his familiar Zeiss glass primes on Nikon or Canon.

(4) Modern Leica lenses are probably a good long-term investment because they are already of a quality to be used on 24mpix or 36mpix sensors.... maybe even edge-to-edge at Full Frame. (WOW!)

Oh well... If you do the math (24Mpixel/18Mpixel*1.5crop^2) you realize the NEX7 has three times the pixel density of the M9.
...

EDIT:

To be fair, it was interesting to see that wide open the NEX has much better resolution in the center but only about equal resolution in the corners. I guess that wide open the microlenses on the M9 sensor do make a difference in the corners. (And even more so in the corners of the full image which were not shown but cropped.)

 

[tstermitz]

Apparently "Pros" need fewer mpixels than "enthusiasts".

Apparently "Pros" need fewer mpixels than "enthusiasts".

Not to hijack the thread, but the more important question should be: what mix of resolution, ISO, handling and performance do we really want.

The latest rumors say that the new Nikons will offer "Pro Nikons" at 18mpix while offering "Enthusiast Nikons" at 36mpix. The sports shooter values autofocus, ISO and frames per second over resolution.

What does a Leica shooter really value?

 

[semilog]

(4) Modern Leica lenses are probably a good long-term investment because they are already of a quality to be used on 24mpix or 36mpix sensors.... maybe even edge-to-edge at Full Frame. (WOW!)

Yawn.

Unless you're shooting on a very stable and well-damped tripod, under field conditions resolution is usually limited by camera movement (vibration) and focus error, not by lens or sensor quality. Has been since the film days.

This was exhaustively discussed in John Williams's book, Image Clarity, two decades ago.

Did I mention, yawn?

 

[gavinlg]

Sooner or later there will be no benefit in owning a Full Frame camera. But just bragging rights.

Not really, I have a Canon 5d (very old) and an x100 (very new) and use d90's, d300s, d3100's at work all day. I would never in a million years trade my 5d for any crop sensor cameras, ever. The 'look' you get from a crop sensor is pretty vastly different to a full frame sensor. Even my fujifilm x100, which I am happy shooting up to about ISO5000 with doesn't compete in overall output to my 5d - simply because the sensor is bigger. Same reason 6x4.5 format trumps 35mm in every case - the larger the format, the nicer the output.

 

[anerjee]

Maybe I am mistaken, but wider and faster lenses would be easier for smaller formats, isn't it? Micro-4/3 has a 12mm f/2, 20mm f/1.7, 25mm f/0.95 etc, all under $1000.

My intuition (and I am very non-technical) is that a smaller format should make for better lens, since you have a smaller picture to paint.

The only reason to stick to larger formats is for resolution and noise performance. And more dof control.

I don't think it will be sooner. e.g. to achieve 28mm fov on aps-c you need a 17mm. In general the lenses get slower--or wildly expensive, and there are all sorts of technical issues--distortion etc.

The n7 has a ton of incredible features, M9 chasing being only one. But if the camera had a ff sensor tuned to take the wide light---even the skeptical might take a breath.

Canikon et al --are they oblivious to the lust for such a machine?

Many barriers are down---the evf is perhaps the first "good" one; the little 5n shoots the CV 15 very well right into the corners, though it's still picky in the wides--not really caring for many of the ziess (though zm18 is very good). But even the M8 still must be prefered because of the sensor size.

Perhaps if nothing else the n7 will prod someone into relenting and producing a nice liitle FF EVIL with a sensor that likes the old glass.

 

[jaapv]

I disagree that a smaller sensor makes a better lens. It diminishes the character of the lens, if you use an existing lens.
To build a high quality lens to fit the smaller sensor can hardly be cheaper. You save on the amount of optical glass, that is significant, but you would certainly lose more on the higher mechanical precision needed.
The reason the 4/3rd lenses are cheaper is because most are consumer grade, and they are SLR type lenses. ( in this context that stands for EVIL and LV as well). If it were needed to build them to rangefinder tolerances the price would be comparable to equivalent 135 rangefinder lenses - or higher.

 

[uhoh7]

Maybe I am mistaken, but wider and faster lenses would be easier for smaller formats, isn't it? Micro-4/3 has a 12mm f/2, 20mm f/1.7, 25mm f/0.95 etc, all under $1000.

My intuition (and I am very non-technical) is that a smaller format should make for better lens, since you have a smaller picture to paint.

The only reason to stick to larger formats is for resolution and noise performance. And more dof control.

Re 12/2

"The field-of-view is equivalent to 24mm in full format terms so it's a moderate ultra-wide angle lens. The max. aperture of f/2 is certainly fast with respect to light gathering but don't expect wonders in terms of depth-of-field - in the MFT scope you are are effectively "loosing" about two f-stops so it "behaves" like a "24mm f/4" here. This is still sufficient for quite shallow depth-of-field images if you move close enough though."

Photozone

24/2s have been common since the 70s/80s

M43 now has caught up, except as you note, in dof. With one lens only. Where is m43 equivalent of the zm15/2.8?

The only advantage of smaller sensors I can see is they are cheaper--- not so?

 

[anerjee]

There is the superb panasonic 7-14mm f/4. And a similar olympus version too.

I am not extolling m4/3; I do not own any m4/3 cameras. I do not understand why the 35mm format is considered the best trade-off between quality and convenience.

Re 12/2

M43 now has caught up, except as you note, in dof. With one lens only. Where is m43 equivalent of the zm15/2.8?

The only advantage of smaller sensors I can see is they are cheaper--- not so?

 

[gavinlg]

There is the superb panasonic 7-14mm f/4. And a similar olympus version too.

I am not extolling m4/3; I do not own any m4/3 cameras. I do not understand why the 35mm format is considered the best trade-off between quality and convenience.

I actually think m4/3 is the best trade off between quality and convenience personally. The lenses are great and small, the bodies are good (just need one x100 style with viewfinder). The older m4/3 sensors are a bit average but the newer panasonic 16mp ones are pretty decent actually. The nex is a better sensor for sure, but the lenses have to be much bigger to even approach the same sort of optical quality - and that's the direct result of having a larger sensor close to the lens mount.

 

[hlockwood]

This is interesting:

http://www.luminous-landscape.com/reviews/cameras/sony_nex_7_rolling_review.shtml#m9

I know, no RF, and it's a crop. Still......

While the tests are very interesting, my evaluation would depend upon a final print. I.E., how different is a 12x18 in. print from the two cameras? Sorry if this has already been covered.

Harry

 

[Hatch]

An M9 is not an NEX 7.

So a bit moot to compare them.

Full frame is not APS-C.

Both have merits.

Everyone makes their own choice.

But it's funny though all the huffin an puffin by the prpoponents/ owners of either.

 

[igi]

An M9 is not an NEX 7.

So a bit moot to compare them.

Full frame is not APS-C.

Both have merits.

Everyone makes their own choice.

But it's funny though all the huffin an puffin by the prpoponents/ owners of either.

IMO, the whole point of this "apples vs. oranges" comparison is to show how the cropped sensor of the Nex approached or surpassed the resolution full frame sensor of the M9... that's like saying this orange tastes better than this apple even if apples are generally tastier.

It's a valid comparison as I see it. They both make pictures anyway... people just have a problem with the test methodology.

 

[jsrockit]

Sensor competitions are so intriguing... barf!

 

[semilog]

The only advantage of smaller sensors I can see is they are cheaper--- not so?

They allow physically smaller lenses, and also smaller cameras.

The reason for the second point may not be obvious. Smaller sensors allow smaller cameras because they dissipate less heat and consume less energy. A major source of image noise is thermal heating. Thus, bigger sensors lead to bigger cameras because they require not only bigger lenses, but bigger heat sinks (to keep noise in check) and bigger batteries. Bigger lenses require bigger mechanical assemblies and, if they are AF or IS, bigger batteries because the motors and actuators must move more physical mass.

In addition, for handheld photography the practical resolution limit is about 2,000 and not more than about 3000 horizontal lines. Beyond that threshold, camera movement and vibration dominate and additional resolution is not terribly useful.

Of course, you can get higher resolution if you use a big well-damped tripod, but the vast majority of people never use a tripod.

Since it's possible to make small sensors with very good sensitivity, noise, DR, and color characteristics, and with >2000 horizontal lines of resolution, for most people there will be few reasons to want a bigger, heavier camera.

The exceptions will be people who use stable camera supports and immaculate technique: studio and landscape photographers. But there are few remaining reasons to use "full" frame for a Barnack-style reportage camera, and "full" frame cameras will increasingly be relegated to specialized applications.