Home · Register · Join Upload & Sell

Moderated by: Fred Miranda
Username  

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Leica & Alternative Gear | Join Upload & Sell

1       2       3              18      
19
       end
  

Archive 2014 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test

  
 
FlyPenFly
Offline
• • • • •
Upload & Sell: Off
p.19 #1 · p.19 #1 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


Someone else did a Fuji vs Noctiron test and conclusions seem to match up with AdmiringLight.

http://translate.google.com/translate?sl=auto&tl=en&js=n&prev=_t&hl=en&ie=UTF-8&u=http%3A%2F%2Fwww.dslrmagazine.com%2Fpruebas%2Fpruebas-de-campo%2Fnocticron-vs-m.zuiko-digital-vs-fujinon-xf.html

or if you're using Chrome or another browser that does live translations and want to see the images

http://www.dslrmagazine.com/pruebas/pruebas-de-campo/nocticron-vs-m.zuiko-digital-vs-fujinon-xf.html



Mar 25, 2014 at 08:24 PM
alundeb
Offline
• • • • •
Upload & Sell: Off
p.19 #2 · p.19 #2 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


FlyPenFly wrote:
I agree with Jordan's article on it, if you disagree that's fine.

I'm not sure how you can get "the same picture" when the Aspect ratios themselves are significantly different. It's really not the same picture at all.

Personally, after my first year shooting, the DOF argument seems pretty bunk to me.


The aspect ratio is a good point. Cropping the FF sensor to the 4:3 aspect ratio would probably reflect how people frame images, and not the other way around. Personally I use 5:4 a lot.

The image height crop factor is 1.85 and can be rounded to 1 2/3 stops. Using 1 2/3 stops as a rule of thumb for equivalence is probably better than using 2 stops, both for DOF and noise.

Note that the rule of thumb wil be inaccurate for noise if we compare cameras with different sensor efficiency.

The explanation why so many people think equivalence is bunk, is largely due to these three factors:

- The crop factor is 2 only for the diameter of the image circle (image diagonal). For the image area (which is used by dxomark.com) the crop factor is 1.94 and for image height it is 1.84

- Many people have upgraded from an old sensor tech FF camera to a new sensor tech u43 camera.

- Olympus and Fuji use more inflated ISO values, something many users were not aware of at first and many still not are.




Mar 26, 2014 at 07:17 AM
RobCD
Offline
• • •
Upload & Sell: Off
p.19 #3 · p.19 #3 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


FlyPenFly wrote:
MFT sensor technology has certainly exceeded that of the most commonly sold APS-C cameras.

I haven't read all of this thread so could be missing some background information but I'm not sure what this means. Are you saying that MFT sensor tech is better than the D7100, D5200, D5300, Sony NEX 5n, 6, 7 or are you saying that these cameras aren't some of the most commonly sold.... or is this a play on words in an attempt to limit the comparison to Canon sensors when everyone knows Canon's sensor tech lags?



Mar 29, 2014 at 08:13 AM
FlyPenFly
Offline
• • • • •
Upload & Sell: Off
p.19 #4 · p.19 #4 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


RobCD wrote:
I haven't read all of this thread so could be missing some background information but I'm not sure what this means. Are you saying that MFT sensor tech is better than the D7100, D5200, D5300, Sony NEX 5n, 6, 7 or are you saying that these cameras aren't some of the most commonly sold.... or is this a play on words in an attempt to limit the comparison to Canon sensors when everyone knows Canon's sensor tech lags?


Yes, if you look at sales, Canon APS-C outsells everyone by a lot. The specific cameras you mention except maybe the 5N outperform MFT sensors. The MFT sensor since the OM-D E-M5 are essentially a slightly noisier 5N but tonality and dynamic range are par.



Mar 29, 2014 at 08:43 AM
alundeb
Offline
• • • • •
Upload & Sell: Off
p.19 #5 · p.19 #5 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


FlyPenFly wrote:
If you can provide a better quantified tested source than DXO, I'd be happy to look at it.


I am happy with DXO data. The problem is application out of context as you did now.

You presented data for dynamic range. Can you explain the driteria for dynamic range at dxo and how it affects image quality, and why you chose to present dynamic range instead of signal-to-noise ratio?

Dynamic range at dxo is engineering dynamic range, where the noise floor is where SNR is 1:1. For acceptable photographic quality we often require a much higher SNR, and then then engineering dynamic range does not tell the whole story. When people say that they see much greater dynamic range in FF sensors, that is because the apply a higher criterion for visually acceptable noise. Numbers for this you can olny see in the full SNR curves. They are presented for each individual sensor and not when you compare sensors at the dxomar.com website.

Another thing, it is well known for some time now that the A7 sensor has lower efficiency than the rest of current FF sensors. This is probably a sacrifice to get the good phase detect AF. The A7r is better and actually1 2/3 stops better than u43 for SNR. If you want to go full frame, the A7r is a better choice in my book, as it also provides better resolution, something you cannot get with smaller formats. If you don't require more than 24 MP, there are many reasons not to get the A7 at all.



Mar 31, 2014 at 03:40 AM
alundeb
Offline
• • • • •
Upload & Sell: Off
p.19 #6 · p.19 #6 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


If we look at the dxomark numbers for dynamic range at ISO 400, it looks like the A7 has only 0.5 stops advantage over the E-M5.But that is for a SNR of one which is unacceptably high noise level to most of us.

The Full SNR graphs are per pixel and need to be normalized for resolution. If we apply a criterion for 12.4 dB SNR in the minumum acceptable shadows, which is quite noisy (similar to midtone at ISO 25600 or something) , and correct for inflated ISO, we get a usable photographic DR at true ISO 400:

E-M5 9.0 Ev
A7 10.3 Ev
A7r 10.5 Ev



Mar 31, 2014 at 05:27 AM
FlyPenFly
Offline
• • • • •
Upload & Sell: Off
p.19 #7 · p.19 #7 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


I use DR because I found it to be more reflective of real world performance. For example, the SNR graphs do not show the real world advantages of the 6D sensor above the D600/610 above ISO1600. The DR graphs do for real image quality. Dxo explains in detail what each graph means in their about sections. The actual difference in image quality is very noticeable if you look at samples.

The Print mode graph are all already normalized to 8mpx. If you look at dxo, you do not see anything close to the two stop advantage claimed for FF above last year or two years ago MFT sensors. Olympus has yet to refresh the MfT line as the em5 and ep5 are due for replacements, hopefully we see a new generation of sensors soon.




Mar 31, 2014 at 05:39 AM
FlyPenFly
Offline
• • • • •
Upload & Sell: Off
p.19 #8 · p.19 #8 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


On a personal note, I've been looking at full frame samples quite a lot these days because I need to get really great ISO3200 and ISO6400 shots for a specific type of photography I want to explore (astroscapes). Full frame has an advantage, especially the 6D and D4/df but it's not miles apart if you use the right lenses.

For example, a wide angle (28mm and below) fast lens with minimal coma and vignetting is really what's needed. All the great wide angle lenses perform pretty terribly on FF on the corners except the hugely expensive and literally huge Nikon 14-24 and Zeiss 21 (it does worse here a bit due to high coma). Everyone else needs to be stopped down from F2.8 and that's where you lose the stop performance advantage of FF. On APS-C the very large Sigma 18-35 F1.8 and smaller Fuji 14mm 2.8 are probably the best balance I've found in measurements and tests that can match the performance of something like the tiny Olympus 12mm F2 that show almost no coma and relatively lower vignetting. The Fuji X-Trans is an intriguing possibility for astroscapes because of the very low chroma noise. (However, in everything else you have to deal with a modified workflow for post processing and still other lingering issues, something I'm not sure I want to do.)

I've also looked at the Sony A7R and it's really poor for this kind of photography and in general is of questionable value. Sure, nice small body but bigger lenses are required, it has the worst AF by far of any FF camera, has light leaks, and Sony has yet to release a great wide angle lens.

So it's really a question of how much do you want to spend in lenses, how much size do you want to carry, and a couple of other factors. Unfortunately the answer is never easy.

Ideally, I want a 6D sensor in a GM-1 body with lenses the size of the Olympus 12mm at the cost of a E-PM2.

Edited on Mar 31, 2014 at 06:07 AM · View previous versions



Mar 31, 2014 at 06:01 AM
alundeb
Offline
• • • • •
Upload & Sell: Off
p.19 #9 · p.19 #9 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


FlyPenFly wrote:
If you look at dxo, you do not see anything close to the two stop advantage claimed for FF above last year or two years ago MFT sensors.


Your ignorance is unbelievable.

The Canon 5DIII, 6D, Nikon D610 and Sony A7r all have a 1.5 stops advantage in medium to high ISO DR over the E-M5, the 6D about two stops.

The FF Exmor sensors have about 1.5 stops advantage in engineering DR (dxo DR graphs) at the base ISO of the E-M5.

All of the mentioned sensors have about 1.5 stops advantage in usable photographic DR at low / medium ISO.

All the mentioned FF cameras have about 1 2/3 stops advantage in midtone SNR.

I said that the "Full SNR" graps are per pixel. Those you cannot see in comparison mode, you have to open the database for a single sensor. Of course I know that the graphs you see in the comparsions are normalized for resolution.

You also Ignore that I adjusted the crop factor from 2 full stops to about 1 2/3 stops for more realistic equivalence both in terms of DOF and noise.

DXO data fully supports equivalence with an error margin of about 1/6 stop.




Mar 31, 2014 at 06:06 AM
FlyPenFly
Offline
• • • • •
Upload & Sell: Off
p.19 #10 · p.19 #10 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


Those are all newer generation, look at when those cameras were released relative to when the E-M5 came out. No need for adhom attacks.

1.5 is 75% of 2. In my world, 75% isn't that close to 100%. Especially when you consider sensor generation. Even the latest E-M1 is not truly a current generation sensor as it come from the always behind Panasonic. The E-M10 is a low cost budget model using last generation technology.

The Canon 6D is an interesting case, at low ISO, last generation MFT sensors actually outperform it but at higher ISOs, it beats the current Nikon FF cameras.



Mar 31, 2014 at 06:08 AM
alundeb
Offline
• • • • •
Upload & Sell: Off
p.19 #11 · p.19 #11 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


FlyPenFly wrote:
Those are all newer generation, look at when those cameras were released relative to when the E-M5 came out. No need for adhom attacks.

1.5 is 75% of 2. In my world, 75% isn't that close to 100%.


Sorry, I didn't understand the meaning of "above" in your claim.

There is no point in comparing old FF sensor tech to current u43 sensor tech in a generalized discussion where there are plenty of affordable relevant newer cameras.

1.5 is close enough to 1 2/3 stops, which I now claim for equivalence between FF and u43.



Mar 31, 2014 at 06:16 AM
FlyPenFly
Offline
• • • • •
Upload & Sell: Off
p.19 #12 · p.19 #12 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


I would say when you look at the best FF sensors such as the D600 or D800, the advantage for FF is apparent and obvious and I would agree it's 1.5 stops at HIGHER ISOs. When you look at the worst sensors, it's not as obvious and there's no clear cut winners. I don't know about you but 95% of my keeper shots are at base ISO or close to base. I don't do sports photography and have yet to actually take astroscapes.

However, it's never really that simple especially on the wide angle (wider than 35) front. The large sensor that wide angle lenses have to cover force you to stop down to match MFT or APS-C performance unless you're using the most expensive and largest of FF lenses. This really does negate a lot of the advantages in sensor performance when your optics have to be so expensive and so big. Sure if you don't mind that it's okay. You pay 200-300% in weight and dollars for that last 10% of performance.

When the newest MFT lenses come online such as the Olympus 7-14mm F2.8 Pro lens, it'll be interesting to compare again.



Mar 31, 2014 at 06:23 AM
alundeb
Offline
• • • • •
Upload & Sell: Off
p.19 #13 · p.19 #13 · Fuji 56/1.2 vs PL 42,5/1.2 Controlled Test


FlyPenFly wrote:
The large sensor that wide angle lenses have to cover force you to stop down to match MFT or APS-C performance unless you're using the most expensive and largest of FF lenses. This really does negate a lot of the advantages in sensor performance when your optics have to be so expensive and so big. Sure if you don't mind that it's okay.



The point of equivalence is not that FF is supposed to be superior to smaller sensor. The point of equivalence is to compare performance between images with similar FOV, DOF and shutter speed. Of course smaller sensor systems then have a lot of advantages. They are marginally more efficient and the lenses are smaller.

What I object on, is if you also say that the lenses for smaller formats are better when you compare at wildly different DOF.



Mar 31, 2014 at 06:32 AM
1       2       3              18      
19
       end




FM Forums | Leica & Alternative Gear | Join Upload & Sell

1       2       3              18      
19
       end
    
 

You are not logged in. Login or Register

Username       Or Reset password



This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.