And Nikon launches a Mirrorless....

[Tanegashima]

You really think the nikon crop is better than the sony in this one? What I think your trying to illustrate is the nikon has more detail? However, with all the noise in the J1 image, what do you think is going to happen to the so called "detail" when you apply any sort of noise reduction to the file in post. Bye, bye details. At least with the Sony file, working in RAW, you'd have a chance of pulling the details because there isn't the base luminance noise in the file. I don't really care too much about what a camera JPG engine does to a file, I care about what the RAW files look like. If the noise is that bad in the J1 files AFTER running through the in camera jpg engine, I can't imagine how bad they are in RAW with no NR applied...

And I'm also not 100% sure that part of what is going on in that photo is the result of the expand DOF in the J1 due to the small sized sensor. From the photo's I've seen from the 1 series on the web, the is basically no background blur on any image I've seen from these cameras, even wide open. The reason I say this is the bundles of thread on the NEX image appear soft on that image. These stick farther out into space than any other thing in the still life (At f/10 this surprises me a little bit, and I'm not 100% sure it's the right answer.) Yet the text on the grey board behind that is sharp on the NEX image. However, you look at the head of the Rosie the Riveter, it appears sharper on the NEX image.

NEX

J1

Edit: If you look at Amazon's top 100 DSLR's, the NEX-7 (with the kit lens) is the top selling mirrorless right now trending upwards in the mid teens (14 when I posted this). The NEX-7 body only is trending up in the low 20's, two versions of the NEX-5n are in the mid to high 20's. The first 1 series offering is the V1 two lens kit in the mid to high 30's with the 1 lens kit is 1 spot lower. In just 6 short days, the J1 has fallen out of the top 100. There are 9 different NEX offerings in the top 100, compared to 3 Nikon 1 series offerings. It looks like lots of people had waited to see what Nikon had up it's sleeve and have decided to pass in favor of the NEX system. Nikon 1 Series = too little, too late.

Nikon J1 is 10MP, NEX5 isn't.

Nobody's saying that Sony is better or worse. It's just you comparing the Nikon 1 to the NEX5, and you got what it is.

you'd have a chance of pulling the details because there isn't the base luminance noise in the file.

No, what you'll get will be jagged contours.

Luminance noise isn't a problem at all… :loco:

If the noise is that bad in the J1 files AFTER running through the in camera jpg engine

Who said you that it passed some kind of NR?

Whatever, it has more details than the Sony's, look at the color strings, Sony looks like a p&s camera.

 

[Lonnie Utah]

Nikon J1 is 10MP, NEX5 isn't.

Well, Nikon is the one that decided to bring a knife to a gun fight....

Nobody's saying that Sony is better or worse. It's just you comparing the Nikon 1 to the NEX5, and you got what it is.

I am. I'm saying the J1 is worse than the NEX. I think lots of other people share my conclusions. I would say that early sales figures for these systems validate my conclusions.

Another 800 iso sample. This one from http://www.imaging-resource.com/

NEX5

J1

Again, looking at these images, you can see more noise and lack of detail from the J1. This holds true in ever comparison of sample images I've seen. In this case, down scaling the images to the same size makes no difference in the results, the NEX image is sharper. Here's the same image cropped down to 640 pixels wide....

NEX

J1

I understand brand loyalty, but I also understand reality.

 

[jonoslack]

Again, looking at these images, you can see more noise and lack of detail from the J1. This holds true in ever comparison of sample images I've seen. In this case, down scaling the images to the same size makes no difference in the results, the NEX image is sharper. Here's the same image cropped down to 640 pixels wide....

NEX

J1

I understand brand loyalty, but I also understand reality.

Surely it would be more realistic to upsize the Nikon image to the same size as the Sony image? After all, you're throwing away detail from the Sony crop to compare it with the Nikon?!

I haven't really done the research to check this, but it seems to me a perfect example of what MR at Luminous Landscape was saying with respect to judging noise - everyone does it at 100%, which is okay if you're looking at the same mp file (even if it's a different sized sensor), but if you aren't (as in this case) it's actually rather unfair on the sensor with more pixels, because you're looking at a smaller proportion of the image.

 

[Bob]

Surely it would be more realistic to upsize the Nikon image to the same size as the Sony image? After all, you're throwing away detail from the Sony crop to compare it with the Nikon?!

I haven't really done the research to check this, but it seems to me a perfect example of what MR at Luminous Landscape was saying with respect to judging noise - everyone does it at 100%, which is okay if you're looking at the same mp file (even if it's a different sized sensor), but if you aren't (as in this case) it's actually rather unfair on the sensor with more pixels, because you're looking at a smaller proportion of the image.

Very good point, Jono.
Noise should be judged based on an image of the same subject covering the same dimensions and then processed to produce images of the same size.
-bob

 

[Lonnie Utah]

Surely it would be more realistic to upsize the Nikon image to the same size as the Sony image? After all, you're throwing away detail from the Sony crop to compare it with the Nikon?!

Very good point, Jono.
Noise should be judged based on an image of the same subject covering the same dimensions and then processed to produce images of the same size.
-bob

Well I agree, but I didn't won't folks to scream that I was being unfair to the Nikon by up scaling it's images to match the Sony. However, on the other hand, Nikon chose the sensor size and knew the limitations of that when they made that decision.

Jono, could you link us to that discussion at LL? I'd be interested in reading it.

 

[Bob]

Well I agree, but I didn't won't folks to scream that I was being unfair to the Nikon by up scaling it's images to match the Sony. However, on the other hand, Nikon chose the sensor size and knew the limitations of that when they made that decision.

Hmmm, I didn't mean it quite that way.
Adjust the scaling of the original image via distance to subject so that the same subject is covered in whatever the number of pixels are available.
"Optical" scaling so to speak.

The process both for comparison to produce the same print size. Here one would use digital scaling.
Then feel free to make equivalent area crops, but at realistic print dimensions.
-bob

 

[Lonnie Utah]

Hmmm, I didn't mean it quite that way.
Adjust the scaling of the original image via distance to subject so that the same subject is covered in whatever the number of pixels are available.
"Optical" scaling so to speak.

If you look at the original image resource images, this is what they did. The part that is missing is second part, scaling the images to the same pixel dimensions digitally...

I'll agree with the pixel peeping comments, but doesn't logic dictate that if and image looks better at 100% zoom, it should look better at full sized print resolutions? The better it looks, the larger prints you can make too. This also doesn't address the fact that having better detail and noise handling qualities at 100% gives the user the ability to crop the image and still retain decent quality when printed.

 

[Bob]

If you look at the original image resource images, this is what they did. The part that is missing is second part, scaling the images to the same pixel dimensions digitally...

I'll agree with the pixel peeping comments, but doesn't logic dictate that if and image looks better at 100% zoom, it should look better at full sized print resolutions? The better it looks, the larger prints you can make too. This also doesn't address the fact that having better detail and noise handling qualities at 100% gives the user the ability to crop the image and still retain decent quality when printed.

You might think that logic would dictate that but actually it does not seem to work that way.
Independent viewers often prefer a certain degree of noise in the image and consider the image to be too "digital" unless there is some noise. Film is very noisy compared to most of what we now shoot.
Besides noise amount (rms luminance noise) the nature of the color noise as well as pattern noise or lack thereof is very important from a subjective point of view.
I would say that most or many reviewers make the wrong judgement about noise at 100% since there is just not enough of the image being judged.
So no, some noise is good providing that it is of a certain type and no noise looks worse to many viewers.
So do the whole experiment THEN make a judgement.
Noise and resolution is appropriate and can be judged only based on the end use and size of the image.
-bob

 

[jonoslack]

I'll agree with the pixel peeping comments, but doesn't logic dictate that if and image looks better at 100% zoom, it should look better at full sized print resolutions?

Only if the two files have the same number of MP. Certainly not in this case

The better it looks, the larger prints you can make too. This also doesn't address the fact that having better detail and noise handling qualities at 100% gives the user the ability to crop the image and still retain decent quality when printed.

But still it depends on the actual number of MP in the images you're comparing.

 

[jonoslack]

Jono, could you link us to that discussion at LL? I'd be interested in reading it.

HI Lonnie
it's in the A65 review:

A65 first impressions at LL

Scroll down to where it says 'A word of caution'

all the best

 

[Bob]

Oh the size of print and the resolution needed depends utterly on the viewing distance.
There are a lot of 10 mpx billboards out there.
Generalizations are a trap.
-bob

 

[jonoslack]

More seriously.
I was always in the "let's have bigger pixels and less of them" camp until I had to do two wedding books at the same time - both shot in grisly english autumn light.
One was with the D700 (universally acknowledged as a HIGH ISO winner)
The other was with the Sony A900 (universally accepted to have poor high ISO characteristics).

Both were done as Blurb books - so the output size was the same.

In actual fact, the higher resolution of the Sony more than made up for the extra noise (in effect I was practicing pixel binning by scaling down to 12"x15" or whatever the large blurb books are). But there was also extra 'croppability' in the A900 file.

These days I like the extra resolution - even if I'm not going to use it for huge prints (although it does of course mean that I CAN do huge prints if I want to). Larger sampling rates are usually better than smaller sampling rates! (even if that is a generalisational trap!).

all the best

 

[Bob]

Yup, and one reason I shoot 80mpx files usually.
More flexibility and versatile.
Also makes you look taller, people buy you more drinks, and members of the opposite sex take notice.

Ah, I am kidding of course.
-bob

 

[jonoslack]

Yup, and one reason I shoot 80mpx files usually.
More flexibility and versatile.
Also makes you look taller, people buy you more drinks, and members of the opposite sex take notice.

Ah, I am kidding of course.
-bob

Mmm yes - but I already drink to much, and members of the opposite sex have always found me irresistible . . . perhaps that's why I've never gone medium format?oke:

and so am I kidding of course (at least about the opposite sex).

 

[Terry]

But actually this topic was front and center in a stream of emails I had this weekend. The question was why not get the NEX 5N for better ISO performance than go for the NEX7's resolution..... I want the best of both worlds. I want the resolution so at low ISO I can print or crop however I want and at high ISO by the time I reduce 24mp back to 16mp (a full 1/3 reduction) I'm probably no worse off and get a similar performance.

 

[jonoslack]

But actually this topic was front and center in a stream of emails I had this weekend. The question was why not get the NEX 5N for better ISO performance than go for the NEX7's resolution..... I want the best of both worlds. I want the resolution so at low ISO I can print or crop however I want and at high ISO by the time I reduce 24mp back to 16mp (a full 1/3 reduction) I'm probably no worse off and get a similar performance.

Exactly Terry - I'm almost certain that 'pixel binning' in post processing on the NEX7 will give you equivalent high ISO performance . . . but upsizing on the NEX5 will certainly not give you the same resolution.

 

[Bob]

Exactly Terry - I'm almost certain that 'pixel binning' in post processing on the NEX7 will give you equivalent high ISO performance . . . but upsizing on the NEX5 will certainly not give you the same resolution.

There may be a relationship that one could draw between rms luminance noise and resolution tradeoffs, but like Jono, I usually go for more resolution since no amount of processing can get that back

 

[Lonnie Utah]

There may be a relationship that one could draw between rms luminance noise and resolution tradeoffs, but like Jono, I usually go for more resolution since no amount of processing can get that back

^^ This. Which is why the NEX-5n beats the V1 hands down. Not only does it have more MP, it also has better noise handling characteristics than the V1. Win, win. (It also costs $200 less so win, win, win.)

 

[Terry]

Yes, but I think there are some interesting qualities in the J1 that should not be overlooked. The J1 also makes a lot more sense for someone invested in Nikon glass over the NEX.

On my trip to Ireland in 2010 I took 1 A900 body and 2 NEX 5's as backups. The NEX or Nikons make a perfectly acceptable "digital back" solution to the kit and allows you to have a back up that does double duty. It serves as a real camera backup and as a compact system at the same time.

I don't look at the sales of Sony or the Nikon system on Amazon right now since neither are available and it isn't clear who is taking pre-orders how many people are double or triple ordering, etc. Reliable sales information on either system is jut not reliable at this point in time.

 

[jonoslack]

Well, having made my remarks about resolution I think that the V1 is a thoroughly sensible and well thought through concept. . . . and unlike some, I think it'll be a spectacular success, mainly as a single system for enthusiasts.

I'd just rather have a NEX 7! (or a NEX 5n come to that). But then I don't need to shoot sports at 60fps, and I'm not happy with a 10mp limit (whether I'm right or wrong about it).