Comparing Full Frame to APS-C Landscapes

In this comparison I’m using my 1DsIII and Xpro2 again.  They are similar resolution (1DsIII = 22MP and Xpro2 = 24MP) so it’ll be a fair comparsion.

However, if you’re a serious landscaper looking for as much resolution as possible, you’ll already have a high MP Full Frame camera and you might as well stop reading here. 😉  However I want to test for myself if just having a Full Frame sensor makes much of a difference.  Sure the 1DsIII is old, but it’s a very good sensor, even today, the Xpro2 is new and probably the best APS-C sensor around at the moment.  So let’s see how they go.

NOTE: This is a VERY rough test.  Both as set to f/11, 1/250 and their lowest native ISO settings.  I just shot hand held out my window.  Take it as you will.  1DsIII is using Sigma 35mm 1.4 Art lens, Xpro2 is using 23mm 1.4.
Note on field of view.

There’s an obvious advantage to a 35mm sensor when it comes to shooting wide angle.  So if shooting as wide as possible is a priority (as it is to many landscapers) then Full Frame can be better.

The shots I’m comparing.

Both Compared (screen grab from Lightroom)

Both Compared (screen grab from Lightroom) 1DsIII on the left and Xpro2 on the right.  Sorry about the different angles.

 

100% crops

Xpro2 100% crop details.

Xpro2 100% crop details.

 

1DsIII 100% crop details.

1DsIII 100% crop details.

 

Probably it’s the slight resolution advantage, but the Xpro2 shows more details and doesn’t have moire (that would be the x-tran sensor’s advantage there).  They’re pretty close though.  Of course, if you get a modern 36+MP sensor you’ll get more details. 🙂

 

NOISE comparison

1DsIII NR turned off

1DsIII NR turned off

Xpro2 NR turned off

Xpro2 NR turned off

There’s not going to be much noise at the base ISO of a camera.  I’d say the Xpro2 is better at high ISO than the 1DsIII but I think the 1DsIII has less noise at these lower ISO settings so long as you don’t try and push it (remember this is a studio camera designed to be shot at low ISOs).  Certainly the Xpro2 can match it with a bit of NR added (25 on the Lightroom slider seems to do fine).  Funny enough Fuji hardly has any colour noise.  The main/only issue it has is a bit of luminance noise.

So noise wise, at base ISO 1DsIII wins.  But the Xpro2 has more flexibility for NR use because it hardly has any colour noise and luminance noise isn’t too hard to deal with.

 

Recovering Whites

-5 Exposure

I pulled them back -5 on the Lightroom exposure slider (I assume this is 5 stops)  let’s see how they recover highlights.

-5 exposure slider Xpro2

-5 exposure slider Xpro2

-5 exposure slider 1DsIII

-5 exposure slider 1DsIII

The Xpro2 added some weird halo detail around the blown area where the 1DsIII just left it blown.  Neither did a better or worse job though I guess I’ll give this to the 1DsIII because it didn’t make that weird halo thing.  HOWEVER, if you do it properly and reduce the whites -100 (how you’d really do it) the Xpro2 fares better.

-100 Whites Slider

-100 on the white slider. Xpro2 = no more weird halo

-100 on the white slider. Xpro2 = no more weird halo

Highlights seem slightly more recoverable on the 1DsIII compared to the Xpro2.  So for highlight recovery it’s a win for the Canon.  I don’t think this is Full Frame winning persay, but rather the fact Canon are so good at highlight recovery.  Unlike what we will see below:

 

Boosting Blacks

+5 Exposure slider

-5 exposure slider 1DsIII

-5 exposure slider 1DsIII

 

+5 exposure slider Xpro2

+5 exposure slider Xpro2

 

This is where the older Canon sensor falls apart.  Canon aren’t known for being very good at black recovery.  You can see banding from looking at the tree and a lot of colour noise is introduced.  The Fuji on the other hand holds up much better to this extreme pushing.  Definate win to the Fuji here.  This is the power of the newer sensor tech, I’m sure a more modern Full Frame would do better.

 

Conclusion

If I were a serious landscaper, I’d probably go for a modern Full Frame like the D810 or 5DsR.  36+ MP is certainly an advantage with landscapes and modern Canon and Nikon sensors don’t suffer when pushing blacks as much (Nikon especially is good at pushing blacks).

However, for the casual landscaper or someone satisfied with around 24MP.  The Fuji system does fine.

Overall what I think I’ve discovered is that there isn’t anything inherently better for landscapes in a Full Frame camera.  Given similar megapixels, there’s nothing a Full Frame sensor gives in of itself that makes it better these days.

 

EXTRA COMPARISONS (for fun)

Fuji's Provia vs Velvia Colour Profiles

Fuji’s Provia vs Velvia Colour Profiles

 

Fuji Provia vs Camera Landscape

Fuji Provia vs Camera Landscape

 

I find the 1DsIII seems to have better highlight falloff for sure.  How true this is of more modern Canon sensors I don’t know.

 

Thanks for reading,

signature

ADDENDUM

How about modern sensors?

As someone pointed out in the comments, the 1DsIII is quite an old camera (2007 tech compared to 2016 tech).  While my purpose here is to show that 35mm sensors don’t automatically mean better (in this case for landscapes) it is fair to say that a modern camera will do a lot better especially when it comes to pushing shadows (as I mentioned above).  However, how about details?  Well, as I said at the beginning, if you’re looking for absolutely the BEST resolution, go full frame and get that resolution.  It comes with a price of course, but if you really need 36-50+ MP then it’s the way to go, still cheaper than Medium Format.  You can keep going up to, why not Large Format film?  I’ve seen people in Hong Kong shoot city shots with them.  Very interesting to see.

Anyway, below is a comparison of some modern Full Frame sensors and some of the best crop sensors out there at the moment.

Without further ado here’s the comparisons (I encourage you to right click and open in another tab or something so you can see them full sized).  If you want to compare yourself go here:

http://www.dpreview.com/reviews/image-comparison/

Scene

CLICK HERE TO OPEN THE PIC BELOW IN NEW TAB

BaseISOcompare

 

If I really needed to print huge or if I were a resolution nut.  I’d choose cameras in this order (based on resolution and clarity):

  1. Canon 5Dsr (There’s no doubt this 50.6 MP beast without an anti-aliasing sensor is the king here)
  2. Sony Alpha 7R II (42.4MP, lots of resolution good clarity)
  3. Fuji X-Pro 2 (24.3MP First Crop camera I’d choose, the lack of AA filter and 24MP sensor make it clear and have quite a bit of resolution)
  4. D810 (has more resolution than the X-Pro 2 but lacks clarity – it seemed soft -, surprising since it’s 36.3MP and doesn’t have an AA filter either)
  5. Sony a6300 (this is below the X-Pro2 because it’s showing moire and therefore lacks some fine details)
  6. Sony Alpha 7 II (24.3MP, similar MP to the X-Pro 2 and the a6300 but  yet it seems to be slightly soft)
  7.  Nikon D5 (similar resolution 20.8MP, less clarity than X-Pro 2, high ISO sports beast)
  8. D500 (20.9MP, so very close to the 7DmarkII)
  9. 1Dx II (20.2MP camera, like the D5 more built for sports than landscapes)
  10. Canon 7D mark II (another 20.2MP camera, lacked the fine details coming out with the X-Pro 2)

I found the 1Dx II, D5, D500 and 7D II to be quite similar honestly.  They all had the same or very similar MP and similar clarity. I was looking at the very fine print to see which one was better.

(Note: I picked jpeg instead of RAW for the reason that adding typical bayer sharpening to X-Trans sensors will make for a worse picture.  So I let the camera manufactures do the work and chose jpeg, that way each camera makers individual styles of processing could be expressed.  If you think it’s unfair and I should have chosen RAW, you can click this link HERE and go check it all out yourself, they have files to download and everything).

ACTUALLY it seems, apart from the standout cameras the 5Dsr and Alpha 7RII they were all pretty similar and none stood out as much better than the other.  The D810 was surprisingly soft, so much so I wonder if it was an error on the testers part.   If I were wanting the best I’d go for the 5Dsr and Alpha 7R II.  They do, however, come at a price.

Listed from most expensive to cheapest (based on current USD price on a popular camera website, body only prices):

  1. Nikon D5 $6,496
  2. Canon 1Dx II $4,599
  3. 5Dsr $3,699
  4. Sony A7R II $3198
  5. Nikon D810 $2,796
  6. Sony A7 II $1,998
  7. Nikon D500 $1,996
  8. Fuji X-Pro 2 $1,699*
  9. Canon 7D mark II $1,599
  10. Sony a6300 $998

*Fuji X-T2 is $1,599 and is probably a better all round camera, not out yet though.

So the choice is yours, I feel the difference between crop and 35mm sensors isn’t huge when it comes to similar megapixel cameras (at base ISO at least).  So for landscapes get the camera with the megapixels you want and forget about sensor size if you’re comparing APS-C and Full Frame.

OF COURSE if you want 36+MP Full Frame is your only choice… that or Medium Format.

 

 

 

  • Pingback: Comparing Full Frame to APS-C Landscapes | Fuji...()

  • jarekzon4

    Fuji X-Pro 2 – release date 01.15.2016
    Canon I Ds III – release date 08.20.2007 !!!!!!!
    It is 9th years older camera. :))) This is a whole era in digital photography.

    • Very true. But at base ISO and similar MP I’m testing if Full Frame automatically makes it better.

      I did say in the article too that a newer sensor would be better in terms of noise when blacks are pushed.

      Also, the 1DsIII sensor is designed to be great at lower ISO levels, so it’s pretty fair.

      My point is a newer APS-C sensor is good enough these days. Full Frame isn’t automatically better.

    • Just for you I checked out and compared some modern FF and Crop cameras (at base ISO, since this is a landscape test) on DPReviews comparison widget. Added my findings at the bottom of the article.

  • Pingback: Final thoughts (Full Frame vs APS-C) – Adrian Evans Photography()

  • Soon Jae Kwon

    Thank you for posting this good tests. Why don’t you use Astia or Nega-std for x-pro2 to test for ‘highlight falloff’. Both film simulation on X-Pro2 have better falloff than provia on it, especially nega-std has far more.

    • Thanks for the tip. I’ll look into that for sure.

  • GaryGarth

    When I owned Canon, I usually exposed pretty evenly, but with my Fuji, I expose for the highlights and push in post. I would love for Fuji to improve on their highlight roll-off, but it seems they’ve done what they can with this sensor.

  • Nobody Knows

    A badly done test can make any camera look good or bad , Comparing JPEG’s is really just a software comparison. Any serious landscape shooter will be using RAW.
    Contrary to your findings the Canon 5DSr has very poor shadow noise compared to A7r { I or II} or any of the D800/D810 series . The Fuji sensor does not compete with these modern FF cameras nor for that matter against the best APS options in several key areas

    • I explained why I chose jpeg instead of RAW. You can’t just slap bayer centric processing on a fuji X-Trans file and expect good results.

  • Paulo Martins

    Hi Adrian
    X-TRANS full dynamic range capability only comes true from 800 ISO onwards. This is true not only for the 800% DR extension with in camera jpegs but also with raw files. With Fuji RAW you pull information form the highlights, with Nikon/Sony, from the shadows, with Canon from the butthole.
    Here’s a comparison I made some time ago with a 1st Gen X-Trans. https://www.dpreview.com/forums/post/52928153
    Cheers! 🙂

    • Really? I’ve heard it doesn’t matter. I’ll have to look into it and do some sort of test. Thanks for giving me a topic for a future post. 😀

    • Hi Paulo, thanks for your comment. It got me researching. I came across this:

      http://blog.thomasfitzgeraldphotography.com/blog/2015/9/simulating-fuji-dynamic-range-settings-in-lightroom

      Seems you can just shoot RAW normally and adjust it in post. No need to shoot ISO 800 I guess because the Fuji ISO-less sensor can be pushed up from ISO200. Better exposing for the whites and pushing the blacks I guess? The guy said by using the built in DR options you can squeeze out a tad more DR, but if you shoot DR100 you maintain the option of shooting at lowe ISO values and you can push it close enough to using the built in DR settings.

      A lot of the time I’ll bracket and HDR in Lightroom anyway if I can. Gets you a lot more DR. 😉 LR is actually really quite good at combining HDR.