Why crop sensors are better than full frame

[Jonesy Jones]

So this is just a simple little post that has nothing to do with creating division between anyone regarding sensor size. I have no emotional attachment to any particular sensor size, so if I am wrong, I would love for you to tell me why. That's actually why I am creating this post. I would love to know if there is something I am not considering. This may be old news for some, but it is something that has just kinda occurred to me lately and I thought it was worth sharing. Maybe I'm right, and maybe we need to let camera makers know that crop sensors are better.

 

My reasoning behind this has almost everything to do with FOV (field of view), and the combination of our lenses with the advent of speedboosters. Lets say I have 3 lenses, a 20mm, 35mm, and 85mm, AND a speedbooster. And let's say I have a camera with S35 sensor and the speedbooster gets me the equivalent of full frame (or close enough for discussion). By using this set up, I have now doubled the focal length possibilities of my lenses. In essence, I now have 6 lenses. Used WITH the speedbooster I have something close to a 20mm, 35mm, and 85mm, and WITHOUT the speedbooster I basically have a 30mm, 50mm, and 130mm (or thereabouts). What I'm saying is that a speedbooster and crop lens combo give us 2 focal lengths for each lens. Now, that may not seem like a big deal at first, but if you choose the right focal lengths, you may only need to get half the amount of lenses, which is kind of a big deal. This of course works with other crop sensors too, though I feel like the S35 size has the best of all worlds.

 

I guess the first thing full frame lovers will point to is shallow depth of field, and the loss of some of that with crop sensors. However, I believe you get most of that back with a speedbooster, but more importantly, is that really a big deal? Now, just so you know, I've been shooting seriously since before DSLR's, back in the 1/3" (and smaller) days. I know all about the crappy DOF before DSLR's. When DSLR's came out I too got all hyped on the shallow DOF that could be achieved. I'm guilty of overusing it. Way guilty. I've heard people time after time after time talk about shallow DOF and wanting to have the option and how it helps achieve the "film" or "cinematic" look. But then I actually watched movies. Where is this shallow DOF? I don't really see it. I mean, pick a movie that really looks like a movie, Jaws, The Godfather, The Good, the Bad, and the Ugly, Shawshank Redemption, or any other movie for that matter. There is very little shallow DOF, if any. I mean, I'm shooting at like F8 or F11 to get the cinema look.

 

The other argument that I guess full frame folks would make is light, and the xtra stop you get on full frame. But again, don't you get most of that back with a speedbooster. Not too mention, doesn't it seem like these sensors will soon be ultra-sensitive anyway. I think the A7S (yes I know its full frame) is probably a harbinger of things to come with sensors at all sizes. Light will be easy in the very near future.

 

So, am I missing anything? If not, maybe crop sensors are better and we need to vocalize that when possible. These are just some of my thoughts. Would love to hear yours.

[maxotics]

When it comes to consumer video, all DSLR sensors are "cropped" in some fashion for video.  For example, the GH4 sensor size is 4608 x 3456.  But its highest resolution in video is 4096 x 2160.  Therefore it either crops from the sides 4608-4096=512 and 3456-2160=1296, or it must average/in, or throw out, 512x1296 pixels, 663,662. 

 

In video, all sensors are cropped in some fashion (either side, lines of pixels, or individual pixels by pixel blocks).  The question is how much.  In full-frame and APS-C, vs MFT, more pixels have to be averaged ( which creates and draws a lot of power, so is seldom used in consumer equipment) or thrown out.  This thrown out image information leads to aliasing/moire problems.  The reason the GH4 is such a great camera is that it THROWS OUT LESS pixels than other cameras meaning less image data falls between the cracks (quite literally!).  

 

When you crop from the sides on a sensor you are increasing your focal length.  Therefore, most camera makers ether "bin" the pixels (average them), or throw out pixels (often in lines) to maintain the expected focal length of the lens.  Focal reducers work by concentrating the image from a larger lens onto a smaller sensor (or part of the sensor).  They degrade the image in the sense that if you took 2 full-res photographs from a Canon 50 on a full-frame, against a Canon 50 + focal reducer on MFT, the second would not be as optically accurate.  In video, because the lower resolution and, other factors, these compromises are not noticed.  In short, with a focal reducer, a videographer trades edge to edge full-sensor sharpness for more light and increased FOV on a smaller sensor.

 

What Panasonic has done for resolution, with the GH4 4K camera (more pixel data), Sony has done for light gathering power (more accurate color, especially in low light; that is, less noise).  Sony increased the sizes of the individuals pixels on the a7s.  The larger each sensel on the sensor, the fewer of them there are, hence the 16MP, instead of the usual 24 on the a7.  

 

In order for the GH4 to match the a7s in light gathering it would have to increase the size of the sensels which would increase the size of sensor.  In order for the a7s to match the resolution of the GH4 is would have to save 4k worth of pixels (which it can't do internally, probably because of heat/power requirements).

 

In a perfect world you want both cameras.  They have different strengths and weaknesses that cannot be designed away IMHO.  

[Andrew Reid]

"Better" is a kind of catch-all term though isn't it... and "better" should be on a needs-tested basis.

 

Speed Booster is very good at giving you those 'extra lenses' so you're right on that count. However with 4K you can have APS-C simply by cropping and you don't need Speed Booster to get the full frame look obviously, so for me that is a superior solution.

 

"Better" overall? Not when it comes to wides.

 

Say you have a wide-ish lens, aka 35mm. On Micro Four Thirds that isn't wide obviously. On Speed Booster or 1.5x crop Super 35mm it isn't wide either. Only on full frame is 35mm considered as a 'wide' of view.

 

Say you have a 35mm F1.4 and a 35mm F1.2. Barely any difference in brightness to be honest, but the overall look is massively different on the F1.2 - bokeh looks totally different at that aperture. Only full frame will give you a 35mm F1.2 wide angle. I have one here - a Voigtlander 35mm F1.2. It is like the Nokton 17.5mm F0.95 but full frame. The look is just spellbinding.

 

So for wide angle shots it gives you that extra look... of very fast apertures at longer focal lengths yet still wide! 12mm F2 on the GH4 or a 24mm F1.4 on Speed Booster / GH4 at 1.5x crop looks very different.

 

Essentially each camera has to be taken on overall merit and not on sensor size.

 

The GH4 has strengths over the A7S and visa versa.

 

For me it all comes down to the lenses. There are more full frame ones than there are for crop sensors and if you want to use them as intended then you need that sensor size.

 

In some areas they cancel each other out. If for example you need the more manageable focus at F1.4 then a small sensor will give you that. But then it is horses for courses because if low light is a priority and that is your reason for shooting F1.4 then a small sensor might not give you the low light performance of a larger one, like in the A7S even when stopped down to F5.6.

[Bob Wall]

You asked if you were missing something - ya, you are.  You're pointing out that speedboosters are great, basically, and you're absolutely right, they are.  They've only been commonly available (although an old design) in the last year or 2 though, so it's a very new thing.

 

But what you're missing is that to use a speedbooster, you have to have a lens mount that can take the speedbooster adapter on the camera.  So this excludes all EF mounts.  This excludes several other common mounts.  If you've got an E mount s35 camera, then all the points you're making are of course true.  But as far as I know, there are no other s35 cameras that can take a speedbooster (probably there are and I'm forgetting them.

 

You were making a point about fullframe vs crop, so as far as I know, s35 is the only sensor size that can be boosted to full frame.  M4/3 are out.

[Andrew Reid]

Yes good point Bob.

[Bob Wall]

Thanks.  Also Jonesy, remember that with the exception of the FS100 and 700, all of the cameras that can take a speedbooster, whether s35, m4/3, whatever, are stills cameras.  The speedbooster has slow autofocus transmission to the body, so if you're a pro photographer, you're going to opt for the full frame, native lens mount every time so as to be able to actually use your autofocus quickly. 

 

That said, I did think it was quite interesting that it was pointed out that the speedbooster DOES work on the A7s in crop mode, opening up some interesting possibiliites in lens choice. 

[Axel]

I've heard people time after time after time talk about shallow DOF and wanting to have the option and how it helps achieve the "film" or "cinematic" look. But then I actually watched movies. Where is this shallow DOF? I don't really see it. I mean, pick a movie that really looks like a movie, Jaws, The Godfather, The Good, the Bad, and the Ugly, Shawshank Redemption, or any other movie for that matter. There is very little shallow DOF, if any. I mean, I'm shooting at like F8 or F11 to get the cinema look. 

 

I agree.

 

So, am I missing anything?

 

I think so:

 

 

The other argument that I guess full frame folks would make is light, and the xtra stop you get on full frame. But again, don't you get most of that back with a speedbooster. Not too mention, doesn't it seem like these sensors will soon be ultra-sensitive anyway. I think the A7S (yes I know its full frame) is probably a harbinger of things to come with sensors at all sizes. Light will be easy in the very near future.

 

Light is an argument, low light isn't. Actually there is nothing magic about a scene shot with candlelight or practical lights, it's done either for the sake of it, then it's embarrassing, or out of carelessness. Hey, my camera shoots in low light, I can just forget about lighting. Looked okay for a year or two, but no longer.

[noone]

Many full frame cameras also have a crop mode.

 

Indeed the A7R crop mode is about the same pixel count as M4/3 and still has a larger sensor area.

 

That applies for stills  of course.

 

I want  to get (soonish) a better focal reducer than my Light Canon (Nikon-m43) in m43 mount and try it on my GX7 and then A7 in both crop (APSC) mode and FF with various lenses.  The Light Canon is just a little too thick it seems to mount on my A7 on a m43 to E mount adapter.

 

I do think even the Light Canon (woeful as it is as a "speed booster") might be ok for video with a decent lens on some cameras.

[Jonesy Jones]

It seems like what I'm hearing is that it boils down to DOF and bokeh (yes I know those are different). Again, I'll go back to my point that if you are interested in making movies that look like movies (which I am and I realize not everyone is), then those two things aren't really that important. I could be wrong but I seriously doubt that classic cinematographers were at all concerned about dof or bokeh. What I hear coming from their mouth is that it's about composition, which is directly related to fov and camera placement. Yes, dof and bokeh will often give you a nice instant gratification, at least that's been my experience, but I am learning that it's really all about composition (fov and camera placement). 

 

I guess the point that ef and nikon mount cameras can't take a speedbooster is valid. I hadn't thought of that. But E mount aren't the only cameras that can take a speedbooster, m43 can as well. And my point isn't completely about getting back to the full frame equivalent, but getting 2 lenses for the price of 1.

 

As I understand it, a 35mm on a S35 camera with speedbooster gives you roughly a 35mm, and without the speedbooster is more like a 50mm. 2 lenses in 1. I realize there are some bokeh and dof differences between a 35 on a FF and a 35 on a S35 with speedbooster, but stop them down and I don't think the difference is much. 

 

I think everyone has different objectives and so this may not apply for everyone, but for my interest (making movies that look like movies) I believe I'm on to something. But there are a bunch of other things in that equation as well.

 

 

jonesy

[Jacek]

I guess the point that ef and nikon mount cameras can't take a speedbooster is valid. I hadn't thought of that. But E mount aren't the only cameras that can take a speedbooster, m43 can as well. And my point isn't completely about getting back to the full frame equivalent, but getting 2 lenses for the price of 1.

 

So FFs have advantage here actually. Some of them have crop mode option in video (like Sony a7S) so you will get exactly what you want: 2 fov-s from 1 lens. But in opposite to smaller sensor + SB:

- you don't have to unmount lens and mount SB to change FOV. Just press a button.

- you don't have to pay for expensive SB (even worse if you want to use lenses with different mount-s).

- you get 2 fov-s even with native lenses (with their advantages: perfect fit, autofocus. IS, size, software in-camera corrections etc.)

[leeys]

The speedbooster has slow autofocus transmission to the body, so if you're a pro photographer, you're going to opt for the full frame, native lens mount every time so as to be able to actually use your autofocus quickly. 

That depends on the lenses, doesn't it? I switched to m4/3 for photo work and the Panasonic zooms are a treat. I could do with shallower DoF on the wide angles, but that's just about the only thing I'm missing.

 

Light is an argument, low light isn't. Actually there is nothing magic about a scene shot with candlelight or practical lights, it's done either for the sake of it, then it's embarrassing, or out of carelessness. Hey, my camera shoots in low light, I can just forget about lighting. Looked okay for a year or two, but no longer.

I think this is where many people are missing the forest for the trees: Sure you can shoot at ISO 400k in an area lit by two candles, but as a long time photographer, I find that this results in crap images.

 

Good images (be it stills or video) need good light. The current crop of sensors can do at least ISO 1600 competently, with ISO 3200 being workable as well. I find I don't really need more than that most of the time. Certain people might need more, like photojournalists, concert photographers, or documentary shooters, but if you can have a say in the lighting, what're you doing chasing after super high ISO cameras?

[pablogrollan]

I could be wrong but I seriously doubt that classic cinematographers were at all concerned about dof or bokeh. What I hear coming from their mouth is that it's about composition, which is directly related to fov and camera placement. Yes, dof and bokeh will often give you a nice instant gratification, at least that's been my experience, but I am learning that it's really all about composition (fov and camera placement). 

 

I believe your view on a DP's job is oversimplified. Yes, they do care -a lot- about DOF and bokeh, because those are also composition tools. To think they just worry about camera placement and framing is like saying an actor only has to memorize some text.

 

When you shoot a closeup, sometimes you want the character blended in the environment, but many times you want to isolate the face with very shallow DOF to center the attention on his/her expresion. And the bokeh orbs that you see in the background are often lights or elements placed on purpose to serve as defocused background, and have nothing to do with the real background or what you see in the wide shot.

 

For wide shots such as establishing shots (descriptive) deep focus is usually desirable, since you want to show a general impression of the setting and not focus the attention on a specific point. The problem comes when some unseasoned or untrained filmmaker decides to shoot every establishing shot with an 18mm @ T1.5, just to "show off" shallow DOF. The same could be applied to many closeups: you do not always need the background that blurred.

 

DOF is not an aesthetic element that looks "cool", but a narrative element to tell a story. And composition is probably the last thing a DP worries about. Composition is relatively easy. Its rules are common knowledge. Camera placement and lens choice pretty much define it. It has already been agreed in preproduction and any decent camera operator would be able to frame properly. DPs are more concerned with lighting. If you are looking for that cinematic look, you need proper lighting and lenses, and the extra stop of light provided by the Speedbooster becomes irrelevant (though may come in handy in docu).

[Axel]

I think this is where many people are missing the forest for the trees: Sure you can shoot at ISO 400k in an area lit by two candles, but as a long time photographer, I find that this results in crap images.

 

I think one can use these capabilities of one's big sensor camera without producing crap images. This was actually a test shot directed by David Fincher with the Red One:

 

http://www.youtube.com/watch?v=jJpLgdD7r3c

 

Note, how the background is lit ever-so-subtly (and cinematically ...)

How everything works together.

How Fincher found a purpose for the 'lit by a cigar' - situation, to suddenly reveal something interesting, dramatic.

How the colors mix. There is faint, blueish moonlight from the upper floor. There's a yellow lantern outside. Both cause fine gradients to - what? Black? Not completely. Then the flame is so bright, it even covers the blue and yellow.

Note, that the skintone of Leos face is near perfect, under this circumstances only achievable by good CC. The clip was once downloadable as a big ProRes file, flat as hell, and you could try and grade it yourself. That was when I went and bought new monitors ...

 

 

Note, how the girl - in complete darkness - holds the lantern to her face. She mimicks enjoying the stars in the sky, not convincingly. She's not a trained actress of course, but that wouldn't have helped anyway, because everybody knows that a lantern 3 inches from your eyes would automatically fade out all stars.

 

Note, what the light does to the colors. There is little else than sepia, very little definition despite 4k. Looks videoish to me. Cinematic only if you find Blairwitch Project cinematic. That's what I mean with 'embarrassing'. Especially since it's obviously meant to be a jawdropper.

 

Imagine she was looking for a friend who left the party with the awful DJ in anger, despair or drunk. She can't see shit on the beach. She is worried. 

 

This then would have better been made either shortly after sunset, with any camera and sensor size, or at noon (day for night). 

 

Good images (be it stills or video) need good light. 

 

Yes, and good shadows. What's good and what's bad depends on the intention of the whole film/photo. 

[jgharding]

I think the ideal is probably to have a sensor larger than S35, where 4K is the central S35 portion, 5 or 6k is your full frame, and smaller crops from the center are capabale of higher frame rates. Then using an amazing compressed raw codec.

 

But wait, That's a Red Epic!

 

Red do get some things right ;) shame the footage just doesn't look as nice as Arri's

[Jonesy Jones]

I believe your view on a DP's job is oversimplified. Yes, they do care -a lot- about DOF and bokeh, because those are also composition tools. To think they just worry about camera placement and framing is like saying an actor only has to memorize some text.

 

When you shoot a closeup, sometimes you want the character blended in the environment, but many times you want to isolate the face with very shallow DOF to center the attention on his/her expresion. And the bokeh orbs that you see in the background are often lights or elements placed on purpose to serve as defocused background, and have nothing to do with the real background or what you see in the wide shot.

 

Can you give me an example of this? I'm not trying to be argumentative at all, but I just don't see this, really ever. What's 1 example of super shallow dof or creamy bokeh from a classic film or a modern film shot is classic film style (not interested in TV or internet videos either). I'm sure they exist, I just can't think of any.

 

Also, I have not suggested that a dp's only job is to place and frame cameras, but rather it's the most important part of their job.

[Jonesy Jones]

I think the ideal is probably to have a sensor larger than S35, where 4K is the central S35 portion, 5 or 6k is your full frame, and smaller crops from the center are capabale of higher frame rates. Then using an amazing compressed raw codec.

 

But wait, That's a Red Epic!

 

This is an extremely good point. Then you kinda get the best of all the worlds.

 

However, with Red, as I understand it, they use pixel binning. That's why they offer 5K and 6K, so that they can result in super crisp image once debayered to 4K (I may have that wrong but that's how I understand it). So shooting with a Red in cropped 4K mode won't actually give you a terrific 4K image, unless you're delivering in HD.

[Jonesy Jones]

And composition is probably the last thing a DP worries about. Composition is relatively easy. Its rules are common knowledge. Camera placement and lens choice pretty much define it. It has already been agreed in preproduction and any decent camera operator would be able to frame properly. DPs are more concerned with lighting. If you are looking for that cinematic look, you need proper lighting and lenses, and the extra stop of light provided by the Speedbooster becomes irrelevant (though may come in handy in docu).

 

Again, don't want to be argumentative, but this kinda flies in the face of what I have heard dp's (Roger Deakins for instance) say. That's not to say that lighting isn't extremely important, because it is. But composition is priority #1.

[Jonesy Jones]

So FFs have advantage here actually. Some of them have crop mode option in video (like Sony a7S) so you will get exactly what you want: 2 fov-s from 1 lens. But in opposite to smaller sensor + SB:

- you don't have to unmount lens and mount SB to change FOV. Just press a button.

- you don't have to pay for expensive SB (even worse if you want to use lenses with different mount-s).

- you get 2 fov-s even with native lenses (with their advantages: perfect fit, autofocus. IS, size, software in-camera corrections etc.)

 

Question, do FF in crop mode give you the same res? IOW, will FF 4K give you 4K in crop mode? I don't think so, which would kind of defeat the purpose.

[Axel]

Can you give me an example of this? I'm not trying to be argumentative at all, but I just don't see this, really ever. What's 1 example of super shallow dof or creamy bokeh from a classic film or a modern film shot is classic film style (not interested in TV or internet videos either). I'm sure they exist, I just can't think of any.

 

Wong Kar-Wai films, speaking about Deakins.

 

The first two mainstream films where sDoF was used quite often were Die Hard and Alien3 (though only in 'selected shots'). You are right. sDoF is not typical for cinema.

[jgharding]

This is an extremely good point. Then you kinda get the best of all the worlds.

 

However, with Red, as I understand it, they use pixel binning. That's why they offer 5K and 6K, so that they can result in super crisp image once debayered to 4K (I may have that wrong but that's how I understand it). So shooting with a Red in cropped 4K mode won't actually give you a terrific 4K image, unless you're delivering in HD.

 

Ah I didn't know that! I've not used a lot of Epic to be honest, but a few larger shoots.