Samsung NX1 - FF level quality in an APS-C?? Imaging Resource samples...

[jrista]

So...I was browsing around imaging resource, checking out the NX1 preview. I started comparing the sample images with the 7D II. I was blown away by the quality of the NX1 up through ISO 6400, and even 12800 for that matter. You can clearly tell the 7D II is noisier. So, I decided to compare with the 5D III.


All I can say is...WOW. I'm hooked on the NX1!! I think it may be my new high speed birding camera some time next year, assuming the lenses pan out. I think I'd get this before I got an A7r even...as I am well and truly impressed.


NX1: http://www.imaging-resource.com/PRODS/samsung-nx1/FULLRES/NX1hSLI06400NR2D.HTM


7DII: http://www.imaging-resource.com/PRODS/canon-7d-mark-ii/FULLRES/E7D2hSLI06400NR2D.HTM


5DIII: http://www.imaging-resource.com/PRODS/canon-5d-mkiii/FULLRES/E5D3hSLI006400NR2D.HTM


The NX1 has more megapixels than even the 5D III, and the noise is just about as clean. The NX1 definitely has lower chroma noise. The links above are NON-normalized...I intend to normalize once I get home, and once normalized, I expect the NX1 may even TOP (!!) the 5D III for ISO6400 IQ. Now that would be a thing of wonder...an APS-C with BETTER IQ than an FF?? Noo! Say it ain't so!


The 5D III does seem sharper...not sure why. It almost seems as though the lens is not fully focused with the NX1...but, that kind of seems to be par for the course for IR. I've never felt they get the best example images.


Another source of examples, although not comparisons, is Luminous Landscapes review of the NX1:


http://www.luminous-landscape.com/reviews/cameras/samsung_nx1_first_impressions.shtml


The bookshelf examples near the bottom are just mindblowing. The IQ up through ISO 6400 is just amazing. It starts to fall apart after that, but still...I am extremely impressed by the IQ from this BSI ISOCELL sensor at ISO 6400.


I think Sony has a real competitor. Samsung is another electronics powerhouse...I am very curious to see how they do in the marketplace. I love that they have produced a DSLR-style mirrorless body, instead of some micro-cramped body like the Sony A7 series. I think it is much more along the lines of what I have been looking for, and in comparison to the 7D II...it seems the NX1 tops it in just about every category except lens selection (and, if the adapters work well, that may not even be an issue...and certainly won't be an issue for terribly long, as Samsung is already working on some nice big white supertelephoto lenses.)

 

[AvTvM]

Based on currently available information I also find Samsung NX1 and its sensor very interesting ... but I think,
we should still exercise some caution regarding Hi-ISO images of well-lit studio test scenes [including the LL bookshelf]

in real life, when I turn to Hi-ISO settings it is usually in "light limited" conditions. IQ then is quite different and images (on all cameras) turn out much noisier, even when correctly exposed, and even worse, when underexposed. I'd definitely like to see more real-life use in low light/adverse lighting conditions [red stage lights etc.] before coming to a conclusion whether it really matches or beats 5D III IQ at ISO 3200 and 6400.

 

[zlatko]

Edited: The NX1 looks very good overall. Just the color is duller.

 

[jrista]

Low noise for sure, but doesn't the NX1 image look like it has more artifacts when viewed up close? Also, the color is duller overall.

They all have noise artifacts. The nature of them changes between cameras, but I am not too concerned about that. The A7r had it's unique noise characteristics as well, as does the D800. I also suspect that once normalized, much of those unique characteristics would just fold away as they are averaged out.


I also just noticed that those are all NR2 (default NR) images. I don't think there are any no-NR versions for the NX1 yet, although there are for the 5D III. I would be willing to bet that a lot of the noise characteristics we're seeing are the result of in-camera NR, which I am personally entirely uninterested in (especially with a tool like PixInsight at my disposal...I have multiple multi-scale, wavelet, and contrast-directed noise reduction tools that should make short work of the clean, low color noise kind of noise characteristic of the NX1.)


As for color, color is a matter of processing. Of all the cameras I've used this year, Canon cameras at higher ISO have some of the most difficult color to control, because the color noise is so high. The color noise is very low on the NX1, so I honestly do not believe that improving color would be a problem.

 

[raptor3x]

The 5D III does seem sharper...not sure why. It almost seems as though the lens is not fully focused with the NX1...but, that kind of seems to be par for the course for IR. I've never felt they get the best example images.

Seems like the NX1 is using a more aggresive NR setting for the JPGs.

 

[AvTvM]

Edited: The NX1 looks very good overall. Just the color is duller.

I agree. Especially reds.

 

[jrista]

Based on currently available information I also find Samsung NX1 and its sensor very interesting ... but I think,
we should still exercise some caution regarding Hi-ISO images of well-lit studio test scenes [including the LL bookshelf]

in real life, when I turn to Hi-ISO settings it is usually in "light limited" conditions. IQ then is quite different and images (on all cameras) turn out much noisier, even when correctly exposed, and even worse, when underexposed. I'd definitely like to see more real-life use in low light/adverse lighting conditions before coming to a conclusion whether it really matches or beats 5D III IQ at ISO 3200 and 6400.

This is true...however better data is better data. I keep saying that...I don't know if people understand what I mean. That comes from constantly digging ultra-deep into the levels of shadow that most people never even consider when I do my astrophotography. The NX1 clearly has very, very clean data up through ISO 6400. That isn't surprising, given the technology employed in the sensor. So, if it looks this good in good light levels (I mean, it looks better than the 7D II up through ISO 12800), it should still do better in more challenging light levels. The noise is very, very clean...very low color noise, even tonal gradients without funky color shifts and stuff like that. The noise is lower at a higher frequency level as well, so on a normalized basis, it should be EVEN BETTER. That is so much easier to deal with once you start processing the data and pushing it around. The limits of processability should be higher with the NX1 in more challenging situations than with the 7D II.


I really want to see some raw statistics generated for the NX1. I am very curious to know what it's SNR and DR are at higher ISO. I think this could be a phenomenal camera for astrophotography (with the one major caveat that there really isn't any control software out there that will work with it...most of it is for Canon, some of the newer stuff supports Nikon now, as Nikon is starting to take the lower end astrophotography world by storm.)

 

[jrista]

Edited: The NX1 looks very good overall. Just the color is duller.

I agree. Especially reds.

There is lower contrast in some of the reds, I noticed that on the fabrics. Hopefully IR will have RAW images up soon...I'd like to know if that can be corrected in post. I also wonder if it is due to the NR...

 

[tolusina]

@jrista
Have you done the arithmetic yet to estimate FF and MF mega pixels with NX1 pixel density?

 

[RustyTheGeek]

These observations about the NX1, particularly about the sensor, coming from jrista got my attention. The fact that jrista is impressed, impresses me too.

Just curious jrista... what are your thoughts on the Pentax K-3? If I was going to buy into another system, the K-3 is what I've been interested in due to its superior build, ruggedness, true weather sealing and interesting feature set.

 

[AvTvM]

The 5D III does seem sharper...not sure why. It almost seems as though the lens is not fully focused with the NX1...but, that kind of seems to be par for the course for IR. I've never felt they get the best example images.

Seems like the NX1 is using a more aggresive NR setting for the JPGs.

yes, but only slightly more [@ ISO 6400] ... to my eye, on my monitor.

 

[RLPhoto]

BSI tech will eventually get incorporated into all sensors just like microlenses did. The NX1 is the first step and everyone else will follow.

 

[Aglet]

All I can say is...WOW. I'm hooked on the NX1!! I think it may be my new high speed birding camera some time next year, assuming the lenses pan out. I think I'd get this before I got an A7r even...as I am well and truly impressed.

I also find Samsung's entry into higher end ML to be quite interesting and might consider one myself, especially if there's a nice post-intro price drop next year.

Did you read the link in the post I made a while back?

www.canonrumors.com/forum/index.php?PHPSESSID=8958f61376cc00ed7968518a7299d155&topic=23052.0

it mentioned a bit about the hardware tech inside the camera, an aspect of which I think that you, given your vocation, might find extra appealing.


I returned my 7D2, mine had some serious initial AF lag issues which may well be improved in a firmware update but for now, ALL of my ML bodies and comparably long lenses can AF faster than the 7D2 I tried with its initial firmware. EDIT: that's with the long lens I tested it with, the venerable, original, 100-400mm L.

 

[jrista]

@jrista
Have you done the arithmetic yet to estimate FF and MF mega pixels with NX1 pixel density?

I haven't really done any analysis yet. When I get home, I'm going to poke around with the images in PixInsight. It's got some powerful statistics tools and whatnot, and I can run the math on various things. I really want to see normalized results.


A FF camera with this sensor technology and this pixel size would be...phenomenal, IMO.

 

[Lawliet]

I really want to see some raw statistics generated for the NX1. I am very curious to know what it's SNR and DR are at higher ISO. I think this could be a phenomenal camera for astrophotography (with the one major caveat that there really isn't any control software out there that will work with it...most of it is for Canon, some of the newer stuff supports Nikon now, as Nikon is starting to take the lower end astrophotography world by storm.)

In that context a little detail - the NX1 relies much less then most other cameras on application specific hardware, it's more in line with Samsungs other consumer electronics products, quite programmable and general purpose. Remember the part of the presentation where they talk about what you can do with lots of full res frames, and how the camera being designed to be extendable. Basically the interviewer asked about why only jump/ball detection, with the answer going along the lines of "just load an app, no big deal".
There shouldn't be a fundamental problem in doing ML-like stuff with that camera.

 

[Mt Spokane Photography]

I had heard that, and checked some online comparisons, and it was nowhere near a FF camera, at least at higher ISO levels.

I saw the NX1 at ISO 3200 being about the same as the 5D MK III at ISO 12800, not quite, but close when I view the dark red cloth detail.

I think that different people, me included, see somewhat what they want to see.

Three crops below.

5D3@ISO 12800

NX3@ISO 3200

NX3@ISO 6400

 

[raptor3x]

@jrista
Have you done the arithmetic yet to estimate FF and MF mega pixels with NX1 pixel density?

It would correspond to a 63MP FF sensor.

 

[jrista]

I had heard that, and checked some online comparisons, and it was nowhere near a FF camera, at least at higher ISO levels.

I saw the NX1 at ISO 3200 being about the same as the 5D MK III at ISO 12800, not quite, but close when I view the dark red cloth detail.

I think that different people, me included, see somewhat what they want to see.

Three crops below.

5D3@ISO 12800

NX3@ISO 3200

NX3@ISO 6400

You are comparing non-normalized images here. Aside from the muddying of the red (which I believe is NR), I think the NX1 DOES have about the same noise as the 5D III. On a non-normalized basis, you are comparing a frequency of noise that is smaller than the 5D III (because the NX1 has significantly smaller pixels.) You have to downsample to the same size to compare appropriately. Do that, then tell me if you don't think that ISO 6400 on the NX1 is about the same as ISO 6400 on the 5D III. I am not sure about ISO 12800...that does seem to be where the NX1 starts to fall apart, while the 5D III still largely holds it together. But at ISO 6400, I think the two cameras have very similar IQ. The NX1 certainly does better than the 7D II, even if it doesn't actually match or top the 5D III....it's getting very close.

 

[jrista]

All I can say is...WOW. I'm hooked on the NX1!! I think it may be my new high speed birding camera some time next year, assuming the lenses pan out. I think I'd get this before I got an A7r even...as I am well and truly impressed.

I also find Samsung's entry into higher end ML to be quite interesting and might consider one myself, especially if there's a nice post-intro price drop next year.

Did you read the link in the post I made a while back?

www.canonrumors.com/forum/index.php?PHPSESSID=8958f61376cc00ed7968518a7299d155&topic=23052.0

it mentioned a bit about the hardware tech inside the camera, an aspect of which I think that you, given your vocation, might find extra appealing.


I returned my 7D2, mine had some serious initial AF lag issues which may well be improved in a firmware update but for now, ALL of my ML bodies and comparably long lenses can AF faster than the 7D2 I tried with its initial firmware. EDIT: that's with the long lens I tested it with, the venerable, original, 100-400mm L.

I hadn't read that yet, but yeah. Incredibly impressive technology. Programmable hardware...as hardware? I've never heard of anything like that in a consumer product. That's incredible! Bringing apps to cameras in a way that is actually meaningful to cameras...that kicks ass.


Yeah, I think Samsung has a solid winner here. All they need to do now, now that they have some seriously competitive sensor and ISP technology, is to build out the ecosystem...lenses and customer support. I hope they do that properly.

 

[RustyTheGeek]

With regard to lenses - Samsung could simply build a solid adapter/converter to use other lenses. Make the AF, etc work flawlessly. Even if they had to work out an agreement with the Canon, Nikon, etc it would be worth it to have an instant and excellent lens arsenal that many would already own. Canon, et al. would still sell a ton of lenses and I some folks would probably buy their DSLRs too. Sort of like people buying iPhones and iPads and then later buying Macs because they think they can use an Apple computer to go with the already purchased Apple devices.