The miniaturisation war - Sony RX1 vs 5D Mark III

[maxotics]

Andrew, I think it might help if you write your articles with two very distinct sections, photo and video.  I'm not criticizing!  I love your stuff.  Just trying to keep my brain together ;)  For example, when stealth and photos are important, who wouldn't want a 35mm f/2 on their EOS-M?  However, if they're going to shoot RAW video, and they can't tolerate moire, they're going to have to shoot in crop mode, which means they're now at 105mm.  If that's the case, it's not a slam-dunk over i-frame video from a GH3, say, with a comparable lens.  Also, call me crazy, in high dynamic range video, I'd rather shoot RAW on the EOS-M, in crop mode, at 1280x720, then most 1080p in-camera mp4.  As for the Rx1, if it doesn't have RAW, how much of a future does it really have for the,as you say, serious photo/videographer?  I've been shooting RAW intensively for a couple of weeks and yes, most of my footage is horrible, but I see the potential.  You've said it's a game-changer.  You were/are right the first time.  But the technology is truly bleeding edge at this point.  Do you really think, for someone serious about video, they're going to use an Rx1 over a hacked Canon?  Or Black Magic?  Again, for photography, you're right on.  In small form video, the Panasonic G6 and hacked EOS-M are the cameras to beat, IMHO. Sony will probably never do/allow RAW in their consumer cameras for a whole host of reasons.  For photography, we can always use more full-frame cameras.  Again, thanks for such a wonderful site! 

[JohnBarlow]

If you are interested in F2~/5-15mm 3x or F2~/8-120mm 15x full manual pro zooms (L glass) for crop mode EOS-M in either H264 or RAW, I guess you will need to watch my posts in the not too distant future. :)

[Andrew Reid]

Still I think it's kind of misleading. I've read many times over how people actually mean to say that f/2.8 on Micro Four Thirds is equivalent to full frame f/5.6 in light gathering terms, which is absolutely wrong. I say do not feed those "arguments".

I agree full frame has an advantage on how good the noise performance is in higher ISOs but that's an entire different matter that even has to do more with each particular camera or sensor rather that full frame as a whole and that isn't clear in the article.

 

F2.8 is F2.8 in terms of light gathering, and I never said any different.

 

I am making the point that a full frame sensor is less noisy at high ISOs than the average Micro Four Thirds chip.

[maxotics]

If you are interested in F2~/5-15mm 3x or F2~/8-120mm 15x full manual pro zooms (L glass) for crop mode EOS-M in either H264 or RAW, I guess you will need to watch my posts in the not too distant future. :)

 

I await your definition of "not too distant" :)  Put me on your email list, max@maxotics.com

[JohnBarlow]

Sure

 

Forgot to add that both zooms are fitted with anti-diffraction irises, in the case of the 15x its 8 blades plus auto ND.

 

Mouthwatering ?

[Ernesto Mantaras]

F2.8 is F2.8 in terms of light gathering, and I never said any different.

 

I am making the point that a full frame sensor is less noisy at high ISOs than the average Micro Four Thirds chip.

And I understood that. I just think the way it's phrased could lead to misinterpretation, it's may not be very clear to some. But it's alright!

[sgreszcz]

Still I think it's kind of misleading. I've read many times over how people actually mean to say that f/2.8 on Micro Four Thirds is equivalent to full frame f/5.6 in light gathering terms, which is absolutely wrong. I say do not feed those "arguments".

I agree full frame has an advantage on how good the noise performance is in higher ISOs but that's an entire different matter that even has to do more with each particular camera or sensor rather that full frame as a whole and that isn't clear in the article.

 

Here is a good blog post that talks about this: http://admiringlight.com/blog/full-frame-equivalence-and-why-it-doesnt-matter/

[Ernesto Mantaras]

Here is a good blog post that talks about this: http://admiringlight.com/blog/full-frame-equivalence-and-why-it-doesnt-matter/

Good article! There's a lot of misunderstanding on this by some people, and some fanboyism as well just like its writer mentions.

I for one love to be able to shoot at wider apertures and letting in much more light without having to worry about impossible focus (thus shooting with most films regular aperture of f/5.6 or f/4 but with higher sensitivity!). Full frame for video is pointless to me (with the exception of better high ISO noise handling).

There's a very didactic video as well explaining how F stops work and are calculated for those who still have some doubts:

http://www.youtube.com/watch?v=KmNIouLByJQ

[Nikkor]

Enhorabuena ya sabes lo que son los f stops , if you have watched the video you will have learned that f stops refer to "amount of light".

 

Now you have to understand that that amound of light is projected on an image circle. 

Imagine that the amount of light is "number of photons", you use the same fullframe lens on a full frame camera and on 2x crop sensor.

Inside that image circle we have two rectangles, a big one (the full frame sensor) with an area of 40 and a small one (the 2x crop sensor) with an area of 10. Now you know that actually a 2x crop sensor has 1/4 times the area of a full frame.

The amount of light is 40 photons on the full frame, they spread uniformly around the sensor area. How many photons do we have per area unit? 40 photons/40 area units = 1 photon/area unit. How many photons will the 2x crop sensor recieve? 1photon/area unit * 10 area units= 10 photons.

Imagine that both cameras are 10pixel cameras (the same as 10mega pixels, just 10*6 times less pixels).

So how many photons do we have per pixel?

Full frame-> 40photons/10pixels=4 photons/pixel

2x Crop -> 10photons/10pixels=1photons/pixel

 

Which sensor do you think will have less noise?

 

But what has this to do with the number of my f stops, i have written on my lens "25mm f2" so on my 2x crop camera this is a 50mm f2 lens, stop talking bullshit

 

Ok imagine , in the example before, we were talking about a 25mm lens, your famous equivalent 50mm f2 lens. On full frame we will need a real 50mm f2 lens, this will give us the same view angle as the 25mm f2 on your 2x crop. Since it's the same f stop, we will have the same amount of light, 40photons, and therefore we will have the same 4photons per pixel.

 

So now we have the same image, one with 40photons, and one with 10. If f stops refer to amount of light, howcome on one images there are 4 times more photons than on the other? HOW CAN THIS BE? THERE IS WRITTEN F2 ON MY LENS, F2.

That's because f stop, is just a number, focal length / aparent aperture diameter, and focal length written on the lens is always the same because it's the actual distance between the planes.

 

I ALSO WANT 40 photons on my sensor

 

That means you want 4 times more light, I hope you watched the video with attention, every f stop means twice the light, so we will need to look for a bigger aperture. So we will go down in the numbers, from f2 to f1.4, now we have 20 photons, we want more, f 1.4 to f 1, 40photons. So you will need a 25mm f1 to have the same performance as a 50mm f2 on a full frame.

 

Where is my Depth of field, its gone.

 

Welcome to the world of phisics.

 

Now I understand why there are C mount lenses labeled 25mm f0.95 like this one:

http://www.ebay.com/bhp/c-mount-lens-angenieux

http://www.flickr.com/photos/wendine/sets/72157622080643184/

https://vimeo.com/9435241

 

Yep, in the age of film everything was a little bit more clear, if you were on 8mm f0.95 wasn't so special, but the 50mm f0.7 Kubrick http://en.wikipedia.org/wiki/Carl_Zeiss_Planar_50mm_f/0.7 used on Barry Lyndon was a very special one. I mean, if this thing I'm trying to explain were fake, Kubrick just would have taken an 8mm camera with an 25mm f0.95 instead of using that very expensive lens. Don't you think so?

 

If you still don't bellieve me try to compare the same image taken with the same generation full frame camera and a 2x crop camera. On full frame you will use a 50mm @f5.6 (cheapest lens out there will do this) and on 2x crop a 25mm @ f2.8 (common zoom lens range). They will have to have the same megapixles (this will be the hardest part), on the 2x crop use iso 100 and on the full frame use iso 400, you should see that they perfom equally. Same depth of field you will notice.

 

Just to make sure you don't refer to the super biased article, comparing the omd-5 to a 5d classic.

http://www.dxomark.com/index.php/Cameras/Compare-Camera-Sensors/Compare-cameras-side-by-side/(appareil1)/842%7C0/(brand)/Panasonic/(appareil2)/834%7C0/(brand2)/Nikon/(appareil3)/795%7C0/(brand3)/Canon

 

If you go to the SNR comparsion you will see that iso 100 on the gh3 has the same SNR as a mkiii or d600 at iso 400, therefore the same noise.

Now try to go and buy equivalent lenses to a 50mm 1.2, 85mm 1.4, 135 f2 on a 2x crop sensor. They don't exist...