A7s, Depth-of-Field, and the Micro-4/3 Advantage

[Jacek]

If I can find the time I will shoot some tests with my medium format dof adaptor vs full frame vs aps-c later today.  There is a difference.  Neither one nor the other is better or worse.  They're just different.

 

That would help a lot (to explain/understand).

[Michael Tapp]

I'm a bit confused. The sony A7s is a full frame camera, but you don't have to use full-frame lenses to cover the entire sensor? When people are shooting with vintage lenses on the A7, are they using the entire sensor or just shooting in an APC mode? *scratchng head*

[dafreaking]

A lot of vintage lenses are/were designed to cover full frame. 

[Julian]

Pretty much all vintage lenses are. Photography lenses of course, not cinema stuff.

[KrisAK]

Yes it does, in theory. MFT is 2x crop over fullframe. That means you get 2x the dof, but since the sensor is smaller, you can also say the low light performance is 2 stops worse. So ISO 6400 on fullframe gives the same results as ISO 1600 on a MFT sensor.

 

With raw photo files this theory is pretty much reality. You can compare this at dpreview.com. For example:

 

 

When shooting compressed movie files results might be different because of all the image processing going on vs raw files, but fullframe will always have an advantage when it comes to noise. I think the A7S will hold up at ISO 6400 vs ISO 1600 on the GH4.

 

So in a way neither camera/system has a dof advantage or disadvantage.

 

 

Thanks. So managing deep-focus in FF shouldn't be more challenging than m43 (assuming FF noise & DR performance holds up.)

 

Which begs the question: will the A7S at ISO 12800 be cleaner than what I'm getting from a GH4 at ISO 3200? (And will the codec break down, or the sensor overheat, etc.)

[noone]

I'm a bit confused. The sony A7s is a full frame camera, but you don't have to use full-frame lenses to cover the entire sensor? When people are shooting with vintage lenses on the A7, are they using the entire sensor or just shooting in an APC mode? *scratchng head*

Most people are using native FE mount lenses or adapted full frame lenses from most systems (including some that can have autofocus and stabilization if available depending on the adapter).

 

In addition the a7 takes normal E mount lenses (same mount) and you can use them either in crop mode or normal mode with vignetting.     Crop mode means reduced resolution (around 10mp for the A7, about 15 I think for the A7R, wonder what it will be for the A7S???)        Some E mount lenses are useable in normal mode for at least some of the time.     The 10-18 for example for part of its range.

 

I don't have any E mount lenses (only one FE and a lot of legacy lenses) to see the difference in video between crop mode and  normal FF mode but from posts around the internet, videos shot in crop mode do not seem as good as full mode for the most part though.     Not sure what the difference is from the sensor.

[richg101]

I quickly shot a test to show the difference between aps-c (sorry I don't own a m43 camera) and full frame.

 

first is an aps-c shot was taken with an 85mm sonnar f2.8 (at f2.8)

second is the full frame shot was taken with a 135mm sonnar f2.8 (at f4.8)

 

I used a tokina +0.4 diopter on both shots it enhance the effects so they are clearer

 

 

 

 

What I see from these shots is that though the actual dof is very very similar indeed, it is the full frame with longer lens closed down slightly that has the edge due to being stopped down and thus obtaining way greater image quality.  The smaller sensor needs the lens to be opened wider in order to get as shallow dof and thus is closer to its theoretical limit.  

 

I see this difference as rather a minimal factor in videography terms, however as resolutions increase and when a subject taking up less space on the frame is captured in shallow dof this is where the resolution might create more of a 3d pop.

 

On medium format we'd need a 200mm lens set at f8 to match this fov and dof combination and thus the lens performance goes up another notch due to being that much closer to the optimal f11 point.

 

Opinions?  

[Lucian]

thanks for shooting these.

 

If we are examining for "shittier DOF" on smaller sensors mentioned by araucaria, i don't see it. When viewing the entire image, especially for moving images, I'm don't think I could tell the difference.

 

The resolution factor is interesting though , as you said maybe this might be a perceivable factor to consider when shooting at super high resolutions in the future idk.

[GMaximus]

The smaller sensor needs the lens to be opened wider in order to get as shallow dof and thus is closer to its theoretical limit.  

 

Nah, to keep that DOF with smaller sensor you can have a longer lens and make a few steps further away )

 

To me, a larger sensor system just gives you more options in shooting -  you have more control in capturing light, angles, geometry, dof, in exchange of mobility and money.

 

So you think of your shooting and check if you can make it with this or that gear. That's one approach.

 

'>

For example, can you shoot like this with GH4?

[dishe]

 

 

'>

For example, can you shoot like this with GH4?

 

Sure, I've taken a shot very similar to that by throwing a 135mm f/2.8 on a focal reducer. 

[Nikkor]

I made a little test for fun.

Two images matched, 28 1.4 @ 2.8 cropped to m43 vs 60mm f2.8 @ 5.6 fullframe.

 

[Nikkor]

Second one (both were carfully focused at the same spot)

[Andrew Reid]

If you have the lens focussed at the same position and the camera at the same position, crop factor of the sensor does not change depth of field at all, it will be just as shallow on both shots. Only difference is the composition changes and to compensate we use a wider focal length on the smaller sensor which results in a less shallow DOF or we move the camera further back, which changes the focus point and results in different DOF.

[Nikkor]

Oh man, once you've figure it out will you come back and tell us about it?

[Andrew Reid]

Oh man, once you've figure it out will you come back and tell us about it?

 

What's wrong with what I'm saying exactly? Oh I forgot... Nothing.

[Nikkor]

You haven't even looked at the images I posted which show the difference. A difference, btw, who doesn't have to bother you if you don't intent to watch the footage on a big screen.

 

I'll erase the tip to the facts, it is my gift to you.

[Andrew Reid]

Also the smaller the sensor, the shittier the DOF behaviour. Someday I will write a text on the reasons because I couldn't find any online. This means that 24 f0.95 will not look as good as 50mm f2. neither will 24mm f2.8 look as good as 50 f5.6.

But for now, trust me ;)

 

A 24mm has a deeper DOF than a 50mm lens. It is not 'shittier' and it has nothing to do with the sensor size, but everything to do with the lens.

 

And what pictures anyway? Where?

[Nikkor]

A 24mm has a deeper DOF than a 50mm lens. It is not 'shittier' and it has nothing to do with the sensor size, but everything to do with the lens.

 

And what pictures anyway? Where?

I attached 2 pictures above, "I made a little test for fun.

Two images matched, 28 1.4 @ 2.8 cropped to m43 vs 60mm f2.8 @ 5.6 fullframe."

 

Dof on both is the same, they are both taken from the same point and at different apertures, if you take a dof calculator it will tell you that both are the same.

[pietz]

@araucaria

 

the one thing i like more than telling people how wrong they are, is actually learning to know i was wrong. that way, ill wake up the next day knowing something more. that in mind, i would highly appreciate if you ever published the facts, because as of right now i cant see them to be true.

 

based on the fact that full frame is PRECISELY twice as high and wide as m43 and we're also comparing SAME aspect ratios, im certain to say (as of right now) that 30mm f2.8 will give you the exact same look on m43, as 60mm f5.6 will give you on FF. just based on the optical physics i learned. now comes your turn for saying: dude look at my 2 comparison shots and see how the bokeh differs. and i get you. but i also immediately see how you didnt use 30mm but 28mm. now you go: dude, do you seriously wanna talk about a 2mm difference? really?! and i go: absolutely, because the difference in bokeh is just as small as this difference to my eyes-

 

now one of two things could have happened:

 

1. to get the exact same image you cropped away some of the 28mm picture

2. FF is not precisely twice height and width of M43, maybe because of different aspect ratios

 

either way which of the two is the case, your comparison shot is not a valid comparison anymore.

 

if you cropped away some of the image you OBVIOUSLY get a deeper dof in the m43, as its the same as cropping a 50mm picture to 200m equivalent will never get you the same image as an original 200mm shot at the same f-stop.

 

if you didnt crop away anything than the crop isnt PRECISELY 2 and therefor going up PRECISELY 2 stops doesnt give you the same image either.

 

your fellow believer "rchg101" posted two comparison shots of which i dont see the purpose in this discussion, as the dof seems to be precisely the same. the fact that in-focus points are sharper on a ff camera are nothing new either, but that doesnt have anything to do with this topic. however righg101, i wanna thank you for your focal reducer statement from the bottom of page 2. if taking a speedbooster into this discussion it should always be seen as part of the lens. therefor changing its light sensitivity and its focal length.

 

saying that it changes the sensor size or its crop factor seems to be true in many examples, but its not the same if we look at it more closely. and since this discussion is getting pretty anal, we need to be specific.

[Nikkor]

You are wrong. I wouldn't care if I were wrong, fuck the EGO. But you are simply wrong. I always thought this was bullshit until I got into it and understood why. Since nobody pays me for this I'm not going to write a paper for you because it would take me a lot of time. You don't want to believe me, I don't care.

You don't need to be precise to do the comparsion, the difference isn't going to change with 2mm. But you can try it on your own with a zoom lens (I sold my zoomlenses).

And remember, the two images were shot at different apertures, half along your post you seem to have forgotten it.

 

Richs comparsion shots are between APSc and fullframe, so the difference will be even lower, anyway, the lens he uses is so bad wide open that you can't tell much, so I see a difference in them but you can't really point it out.