NEX-7 Rolling Review

[Maurício Costa]

Sorry, I don't mean to be like those pricks that pick on you for Madonna's cycle...

But isn't there a better way of doing that comparison? Because as I understood it, you are comparing a 12 Mp crop from a 18 Mp camera to a 18 Mp resize from a 24 Mp...  

[Maurício Costa]

As the test stands we're looking at lines/mm on the sensor and disregarding sensor size, taking it to a logical conclusion the camera in an iPhone will have higher resolution than the NEX-7 if we used the same test methodology.

Maybe you went too far, or as we say here in Brazil - you kicked the bucket...

[bobtowery]

Thank you Michael. I see the debate has already started about the methodology. I see why you dislike posting your findings, so many are unhappy.

Certainly what can be deduced is that Sony has created quite an interesting piece of photographic machinery. For the money, even someone that is a dedicated M9 shooter would be wise to consider one as a backup body. Not to mention those things it can do that the M9 cannot, like AF (given one of their lenses of course), higher ISO and so on.

I don't know how long you plan this review to keep on rolling, but another "comparison" comes to my mind: Bokeh. One of the reasons people invest in an M9 and lenses is for the incredible wide open bokeh.  Since it is viewed that it is the lens that makes the bokeh, would it be the same on the Sony vs M9? Since the subject matter is paramount for bokeh, seems like a comparison where you can move with your feet and get the same approximate field of view is desirable.

I wonder if the greater number of smaller photosites in the Sony's sensor will have an effect on the bokeh? Perhaps not quite as creamy/dreamy as on the M9?  

OTOH if you loathe further abuse I completely understand!

[billh]

I don’t worry about Michael’s methodology - it’s his opinion/conclusion I’m after. In every case I can remember, when I own the equipment he is testing, his opinion has been the same as mine. The only issue I have with this report is it will make the NEX7 even more difficult to obtain! It was also very interesting to discover the reason behind the unavailability of Leica Lenses. Given their cost these days, I could not understand the reason for this situation. Leica must be delighted.

I’ve been using the 5n exclusively with Leica M lenses, and with the instant image magnification, focusing is a snap. I find I have to turn off peaking because it obscures the the magnified image to the point it makes focusing difficult for me. I’m from the same era as Michel, and used Nikons and an M4 as a newspaper photographer in the 60s and 70s. It’s a real treat to again use these wonderful M lenses and focus accurately anywhere I please in the frame.

[dseelig]

Hi Michael
I wish you would do equivalents such as the 35 on the m9 and the 24 lux on the sony any chance? thanks though for the hard work.

[pindman]

Hi Michael
I wish you would do equivalents such as the 35 on the m9 and the 24 lux on the sony any chance? thanks though for the hard work.

Michael,

I appreciate your tests, which compare the quality of the sensors using equal dimensions of each sensor. The real life question is the quality of the image when using each camera at a given distance to photograph a given scene using the entire sensor. From a practical standpoint, in real life I would choose lenses to match the field of view and use as much of the sensor as possible.  As implied above, if I were to use a 50mm on the M9 I would use a 35mm on the NEX-7, or longer lens on medium format. Hopefully you aren't too worn out to do this, as I think it would also be valuable.

Thanks.

Paul

[Faintandfuzzy]

Sorry Michael, but just because that's the way you did it, doesn't mean it makes sense.  In fact, the reasoning is rather bizarre.  Simply use different lenses, or change your camera position.  Cropping out the FF M9 is not the way I've ever seen any review done by anyone else. 

[hiepphotog]

It's not nonsense and neither is the way I conducted my test. It's just the way that I chose to do it.

If I used a different lens then there would be those that complained about that. If I used the same lens then I would have to reposition the camera.

The point of the test was to see how the same lens shooting from the same spot at the same time would compare.

Remember, though I cropped the M9 frame I also reduced the resolution on the Sony frame.

Michael

I find the method you used is the more scientific way to do it. Using 2 different lenses would defeat the purpose of comparing two different sensors. Moving the camera to achieve the same framing would ultimately change the perspective, which is not the right way to do, IMO. However, I would suggest a different method of resizing. After you cropped the M9 image to get the same framing as the NEX-7, whatever the resolution of the crop (e.g. 12MP or 9MP) is the final size you should downsize the NEX-7 picture. That should eliminate any discrepancy.

[Jeff Kott]

Using 2 different lenses would defeat the purpose of comparing two different sensors.

First of all, let's acknowledge that there is no way to test two sensors of different sizes without there being some valid criticism of the methodology. Knowing this, I applaud Michael for his efforts. I remember when Bjorn Rorslett did extensive testing of the Nikon D2X against some full frame Canon and he was crucified on the web.

Having said that, I do think that Michael's 35 and 50 Leica's would both out resolve the M9 and NEX 7 sensors.

[stever]

i don't understand the logic of the "test" and the numbers don't add up, and i'm a bit disappointed - and no, i'm not posting to any other forums

i'm not a Leica fan, but have enormous respect for their lenses.  my personal experience with full-frame vs crop frame cameras is that for an equal linear pixel count, the full frame camera has about a 15% resolution advantage over APSC -- if the APSC isn't lens-limited.  the linear pixel difference between the M9 and NEX 7 is almost exactly 15% which would suggest equal performance - pretty much Michael's conclusion, but i'm not sure how conclusive this is.  It is certainly evidence that the NEX 7 is not a bad camera.

[dreed]

To my eyes at least, the M9 has the advantage on the NEX-7 in the corners in the wider apertures. At f/5.6, they appear close to equal. The part of the corners that I compared was the blue wall, where its shadow is cast onto the red building and the light brown/white building that abut each other above that (there's a fine small shadow between the two and this is not rendered the same.) At the wider apertures, the M9 appears (to me), to be slightly sharper?

If I look at the telegraph/power powers further down that image, the M9 also appears to suffer less from CA?
Does that make sense?

Is it just my eyes/pc?

I can't find any difference in the center shots worth noting.

That's my contribution to the pixel peeping!

[ErikKaffehr]

Hi,

I think you have mixed up the pixel pitches. Sony has 3.9 micron pitch and Leica 6.87. That means that the Sony sensor is much more demanding on the lens than the Leica sensor.

The way Michael tests he down samples to the Sony to 6.87 microns.

In my view the reasonable solution would be to resize both images to a equal size. I normally use 70x100 cm at 200 PPI for my comparisons.

Best regards
Erik

i don't understand the logic of the "test" and the numbers don't add up, and i'm a bit disappointed - and no, i'm not posting to any other forums

i'm not a Leica fan, but have enormous respect for their lenses.  my personal experience with full-frame vs crop frame cameras is that for an equal linear pixel count, the full frame camera has about a 15% resolution advantage over APSC -- if the APSC isn't lens-limited.  the linear pixel difference between the M9 and NEX 7 is almost exactly 15% which would suggest equal performance - pretty much Michael's conclusion, but i'm not sure how conclusive this is.  It is certainly evidence that the NEX 7 is not a bad camera.

[viztyger]

I've really enjoyed reading the NEX-7 rolling review; thank you for all the work. I'm not sure though that the test comparing resolution between the M9 and NEX-7 is valid in the "real world." If I was shooting from the same roof top, I would decide how much of the scene I wanted to capture, then pick the focal length suitable for the job. I feel it would have been more representative of a "real world" situation if you had tested the M9 with a 50mm lens and the NEX-7 with a 35mm lens. Of course you're now introducing another variable - different lenses - but both cameras would now be capturing a similar view. And the test would probably be more representative of how many of us would use the respective cameras.

[Wayne Fox]

After reading the article, Michaels goal of trying to see some "real world" perspective instead of simple pixel peeping seems logical. His method appears to offer something pretty close on screen to what we might see viewing identical size prints side by side.

[paratom]

I dont understand the relevance for the comparison M9-Nex7.

1) the Nex images look sharper than the ones of the M9 - so why does the Nex7 sensor outresolve the lens?? It would outresolve the lens if the lens on the Nex7 would not deliver better images even though the sensor has more resolution.

2) I think we all have seen before that different sensors-even when having the same pixel density can lead to different results, for example cause by micro lens assignment etc. (remember how some m-lenses dont shine on thre Nex5 but look better on the Nex5N)

3) Comparing a downsampled image to a cropped 100% image doesnt tell much IMO.

[JohnBrew]

Hey Michael, that was fun (the NEX-7 vs. M9). And an interesting read. I think both cameras did just fine and for me the bottom line is the print. And who knows which one would win using the paper with which I most like to print? I think a photographer as opposed to a pixel-peeper would be happy with the output of either one.

[bjanes]

It's not nonsense and neither is the way I conducted my test. It's just the way that I chose to do it.

If I used a different lens then there would be those that complained about that. If I used the same lens then I would have to reposition the camera.

The point of the test was to see how the same lens shooting from the same spot at the same time would compare.

Remember, though I cropped the M9 frame I also reduced the resolution on the Sony frame.

Your test may have achieved its intended purpose, but it does not reflect how the cameras would be used in practice. Cropping of the M9 reduces its 18 MP to 7.7 MP. I don't know why one would pay US $6K for an 18MP camera and then use it as an 8MP low end model. In practice one would move the camera or switch to a wide angle lens.

You downsized the Sony to 18MP, but why not to 7.7MP? The Leica has a Nyquist of 72 cy/mm and the Sony 128 cy/mm. Naturally, the Sony will win out under your conditions, even without deconvolution sharpening to offset the effect of the low pass filter.
Regards,

Bill

[bjanes]

After reading the article, Michaels goal of trying to see some "real world" perspective instead of simple pixel peeping seems logical. His method appears to offer something pretty close on screen to what we might see viewing identical size prints side by side.

Since the images are being observed at 100% on screen, in effect pixel peeping is being done. No prints were made. In the real world, one would not crop the M8 18MP image to 7.7MP.

Regards,

Bill

[samdman]

I think it would be just if any of us compare NEX-7 vs Leica M9 using our own best methodology(s) rather than judging what has been done by Michael.

I myself would do it, if I can get my hands on both system at the moment. (And the lens of course). Since it's simply unavailable nor unreached (financially), there is some points that we could take away with. Since there are samples available everywhere on the net. We might just get on with our own comparison method, and draws personal justification out of it. Personally, any method available being used, it would greatly sucking one's energy, time, and (possibly) other resources.

My two cents.

[xxl_and]

Is it possible maybe to do the same comparison versus NEX-5n maybe. I would be very interested in this so that we (owners of NEX-5n) can decide if it makes sense that at some point in the future we switch to NEX-7.

A reply would be greatly appreciated.

Thanks :-)