RiceHigh's Pentax Blog: NEX-3 Vs K-x Full ISO Shootout (Controlled Test)

Saturday, August 14, 2010

NEX-3 Vs K-x Full ISO Shootout (Controlled Test)

Test Methodology and Conditions:-

1. The two cameras were fine tuned and set as close as possible for everything including the colour/tone response, focal length and frame composition, aperture value, exposure values, noise reduction level and so on..;

2. The test shots were taken at the same place, with the same lighting (diffused daylight indoor), at almost the same time, at the same position, on a sturdy tripod;

3. Detailed settings of the NEX-3 are as follows: "Landscape" colour/tone mode (-1, 0, -1), Spot metering, Contrast Detection Auto-Focus with single central AF point, 2 sec. Self-timer, Av mode at f/8, focal length at 45mm, Noise Reduction=Weak, EXIF data preserved;

4. Detailed settings of the K-x as follows: "Portrait" colour/tone mode (-1, 0, 0, 0, 0), Live View mode, Spot metering, Contrast Detection Auto-Focus with single central AF point, 2 sec. Self-timer, Av mode at f/8, focal length at 45mm, Noise Reduction=Weak, EXIF data preserved;

5. All lens correction and DR range/exposure "enhancement" features were turned off for both cameras;

6. Click on any of the thumbnails below will open an enlarged view in about 4,000 pixels in width in a new tab/window, where you can measurbate different pictures in the same viewing size endlessly! ;-)

Okay, let's go!

ISO	NEX-3	K-x
200
400
800
1600
3200
6400
12800

Enjoy! And, would the new Sony 14M sensor blow the old Sony 12M sensor out of the water for the image quality and noise control? Well, you are the one to find out yourself! I blog, you decide! :-)

Last but not least, I hope you folks just don't and won't spend/waste too much time for measurbation instead of going out to take some photographs! ;-D

Read Also:-

My NEX Sample Photos Gallery

Sony NEX-5 VS Pentax K-x ISO/IQ Shootout (Dual Scenes)

Summary Matrix for Equipment Tests and Reviews

Comments (12)

Logging you in...

Dashboard | Edit profile | Logout

Logged in as

0 Vote up Vote down

DDD · 764 weeks ago

I think it's better to leave NR on auto for Nex. It usually doesn't make a difference, but when there is red with black spots, NR auto preserves more detail. See this on IR
http://www.imaging-resource.com/PRODS/NEX5/NEX5HI...

0 Vote up Vote down

Madcatz · 764 weeks ago

Both sensors ar very good IMO, very hard to choose. Nex has 2 extra MP so considering that it's definitely not worse than pentax :) few years ago couldn't have such great ISO on APS-C. Also would be interesting to compare to canon 18mpx sensor (550D).

1 reply · active 764 weeks ago

-1 Vote up Vote down

RiceHigh 110p · 764 weeks ago

Tested previously, see: http://ricehigh.blogspot.com/2010/06/550d-17-40l4...

0 Vote up Vote down

osiris97 · 764 weeks ago

I think the Nex is a revolution. So I buy that today...

But I want to share the information about new model of pentax from you.

Please show me your dream picture too....

+3 Vote up Vote down

YoukY63 · 764 weeks ago

Good idea. But a few remarks:
-how do you want to compare noise level of 2 different sensors between 2 cameras from 2 different brands not using NR setting "off" and with jpeg images. Did you ever think that maybe Pentax and Sony's NR engine was acting differently? And jpeg engine was working differently? If you want to do it right, shoot RAW, then open your RAW file in the same software using same settings, and here, maybe, you can start to compare.
-why comparing very high iso setting in a daylight condition? Noise is mainly appearing in dark areas, which are absent during day. Sure here the pictures looks great. Now take the same scene illuminated with a few candles. You will probably be disappointed.

Then I would like to admit that I also appreciate the work from Sony for their new generation of EXMOR sensors. They deliver great results. No doubt. Now, how much can they improve them during the next 3 to 5 years?

3 replies · active 764 weeks ago

-2 Vote up Vote down

RiceHigh 110p · 764 weeks ago

I can't see the ground of your argument. I compare two cameras, not the raw files of the two cameras. There must be differences, those are what the test is meant to reveal!

Also, high ISO is not only required under low light, but also when high Tv speed is desired.

+2 Vote up Vote down

YoukY63 · 764 weeks ago

"Enjoy! And, would the new Sony 14M sensor blow the old Sony 12M sensor out of the water for the image quality and noise control? Well, you are the one to find out yourself! I blog, you decide! :-)"

So why do you conclude by comparing the sensors, and not the cameras? You blogged, and still nothing help me to decide here...

0 Vote up Vote down

David · 764 weeks ago

I partially agree, partially disagree with Youk. I have to say that shooting in jpeg vs. RAW makes a lot of sense - after all these are enthusiast/amateur cameras, not pro or even prosumer level models. These cameras will take 95-98 percent jpegs, rarely shooting RAW, even Tiff, so that comparison works fine.

As far as noise reduction, I am always under the impression that it should be turned off, and you make any changes in Photoshop. The old *ist models suffered from excessive noise reduction, and if you can utrn it off, it should be.

I also have a problem with Rice's photos, as they are skewed towards the bigger sensor. We all know digital cameras crave light, and if its darker, the limitations of the sensor show, therefore noise and artifacting are going to happen in darker settings. But if you have a lens that goes lower than F8, you use it.

I see that background, and feel it needs to be lighter, because none of us would manually set our cameras to shoot so dark. And at higher ISO, it would be exacerbated.

I guess one more thing I would like to see is the Auto settings in action on these cameras. Just as most people will be shooting in jpeg with these cameras, most will stay in Auto, or use a priority setting. I would like to see how much different settings would be if each camera tackles the same setting, and how their light meters would work.