Viscount Omega
Carbonite Member
The thing is, I've done these matching DOF equivalence tests and I and other people still see a difference and I'm trying to figure out why.
Biggest problems with the A7S Ii
- Thread starter filmguy123
- Start date
I suppose what would help this topic would be visual examples. Something that could be corroborated.
I imagine a combination of using more diameter of the lens (for FF cameras) leads to higher resolving power. The larger photosites of FF leads to less noise and higher dynamic range, generally. The shallower DOF (with equal f-stops) looks pleasing and actually holds up better to h.264 compression. All these differences and others might lead to an impression that FF looks different than APS-C at equivalent focal lengths. I'm not sure.
deltoidjohn
Veteran
Agreed.
Maybe show us some side by side examples that prove what you're saying.
You are thinking about it back to front though. Do not "draw lines to the sensor". Draw lines going out from the camera representing it's angle of view for a given sensor. Then position the camera appropriately so that, in the FF vs APS-C comparison, the lines match up. If they don't match up, you're not using lenses that have the correct equivalent angle of view. It should be noted that in order to make the "line" match up, you'll probably need to move the smaller lens/sensor backwards - this is due to different optical formulas and physical dimensions of the lenses which effect their point of convergence.
Maybe show us some side by side examples that prove what you're saying.
You are thinking about it back to front though. Do not "draw lines to the sensor". Draw lines going out from the camera representing it's angle of view for a given sensor. Then position the camera appropriately so that, in the FF vs APS-C comparison, the lines match up. If they don't match up, you're not using lenses that have the correct equivalent angle of view. It should be noted that in order to make the "line" match up, you'll probably need to move the smaller lens/sensor backwards - this is due to different optical formulas and physical dimensions of the lenses which effect their point of convergence.
Last edited:
Samuel H
Major Contributor
In my very simple example, the size of the glass plays no role, because in theory it shouldn't. The size of the aperture does, and it's highly correlated with the size of the wall. In my example, that's the width of the pin hole.
And the quality of the glass would be the quality of the wall. Some are super flat and smoooth, clean, white (that would be a modern "clinical" lens with good contrast). Others are not so white (less contrasty glass), or not so flat, or not so smooth. I still think quality of the glass is what you're seeing.
Viscount Omega
Carbonite Member
Here's a little 10 second video I shot last night. Didn't move the camera. I had to crop the 35mm full frame a little to make it like 36mm.
What's happening with the desk? And the frame lines remain about the same but the room perspective seems to shrink and tighten with the APS-C.
Full screen: https://www.youtube.com/watch?v=yRe10tquXU8
Maybe it's because the 35mm full frame is 1mm wider than the 24mm aps-c and I had to crop it? Still, it's quite a difference.
What's happening with the desk? And the frame lines remain about the same but the room perspective seems to shrink and tighten with the APS-C.
Full screen: https://www.youtube.com/watch?v=yRe10tquXU8
Maybe it's because the 35mm full frame is 1mm wider than the 24mm aps-c and I had to crop it? Still, it's quite a difference.
Last edited:
It would help to know which cameras and which lenses you were using. To me, it looks like the culprit is distortion caused by a cheaper lens on the APS-C shot. It's softer in the corners and looks like it suffers from some pin-cushioning. Is one of the lenses a native lens where the camera does some electronic distortion correction? That's my guess. I don't see any difference in how perspective is rendered, however.
P.S. Is the APS-C camera a Sony FS camera? It has the tell-tale ugly green fringe on the bottom of the light on top of the right speaker.
Last edited:
Samuel H
Major Contributor
The APS-C camera seems just slightly taller than the FF one, but I don't see much difference in terms of perspective. There's a difference in focus, though: either DoF is not the same (APS-C being Canon-like 1.62x instead of the 1.41x you corrected the aperture for) or focus is set a bit closer on APS-C. The lenses have different sharpness, distortion, blooming, etc.
Viscount Omega
Carbonite Member
Camera is Sony A7SII. Lenses were Minolta 24mm and 35mm primes. Original versions, probably 30 years old. Used with an LA-EA4 adapter. Distortion correction was set to "auto" but I'm not sure if it does anything. I've never seen a difference.
I thought using APS-C and the center of the lens was the best quality part of it. By cropping to the center it eliminates a lot of the aberrations and vignetting, right? Not sure.
You don't see or feel the room tighten? I always feel things tighten when comparing shots of rooms with APS-C and full frame.
Maybe I'll do another test with different lenses.
p.s. Samuel, I did another test shot and checked the focus and it was "correct"--that's just the way the DOF and focus renders with these old lenses and sensor sizes.
Last edited:
Looking at it again, I can see what you're talking about. On the 35mm Full Frame shot, the foreground is actually a tiny bit closer to camera than the 24mm shot. If you are absolutely certain that the camera was in the exact same spot for both shots, my guess is the following: the actual physical length of the 35mm lens is longer than that of the 24mm, making it protrude just a bit farther out from the camera and thereby a little closer to the chair and the desk. There also seems to be a tad more optical distortion on one lens compared to the other.
Im guessing the 24mm has a maximum aperture of f2.8, and the reason it looks softer is because lenses are generally softer when shot wide open.
Viscount Omega
Carbonite Member
I was careful not to move the camera and, yes, the 35mm is about twice as long. And yes, the max. aperture on the 24mm is 2.8 and it's a little soft wide open.
Regarding the APS-C tightening effect, it feels like the framing is about the same, but everything is slightly squeezed toward the center from the sides inward, from top to bottom, and from foreground to background.
p.s. it's hard for me to quantify scientifically but it just "feels" like it.
Last edited:
deltoidjohn
Veteran
If you had to crop it, you weren't using an exact equivalent lens at the correct distance.
Each lens also has it's own imperfections in design which introduce things like distortion, CA, focus breathing, etc. This is part of what you are seeing in the above video.
Each lens also has it's own imperfections in design which introduce things like distortion, CA, focus breathing, etc. This is part of what you are seeing in the above video.
Viscount Omega
Carbonite Member
Yeah, I had a suspicion about the cropping and what effect it would have.
However, I always sense the tightening effect whenever I look at APS-C stuff. I see it on tv all the time. Especially (for the 100th time) in shots where there's some 3D context in the scene instead of wide shots and close-ups.
p.s. I don't have any lenses that match perfectly so all my tests would have to be cropped, e.g. FF 35mm to APS-C 52.5mm to match a FF 50mm.
Last edited:
Viscount Omega
Carbonite Member
Here's another short video. This time I'm further back and stopped down the lenses to minimize the depth of field and focus issues and remove them from the equation.
Again I cropped a little bit to make the 35mm like a 36mm, but being further back I think it has less effect. I still see the left side desk expanding and contracting.
The green fringe is gone from the lamp so I think stopping down helped with that.
Full screen: https://www.youtube.com/watch?v=7dcOzligjgo
Again I cropped a little bit to make the 35mm like a 36mm, but being further back I think it has less effect. I still see the left side desk expanding and contracting.
The green fringe is gone from the lamp so I think stopping down helped with that.
Full screen: https://www.youtube.com/watch?v=7dcOzligjgo
morgan_moore
Major Contributor
It should be remembered that lenses are often 'off spec' so a 24 might be 24.9 - a typical marketing job.
I think practically the two sensor sizes can match. That is simple.
If one gets super technical a different size front element may have a marginal difference. But such theoretical stuff is far beyond the lenses one can buy.
Ive had a 50 on FF vs a 80 on a MF vs a 35 on S35 and.. all the maths basically lines up FOV and DOF match with the appropriate adjustments.
Simply FF is a great option if you want to dive into the legacy still lenses back catalogue.
S35 is great for zooms and PL glass.
S
I think practically the two sensor sizes can match. That is simple.
If one gets super technical a different size front element may have a marginal difference. But such theoretical stuff is far beyond the lenses one can buy.
Ive had a 50 on FF vs a 80 on a MF vs a 35 on S35 and.. all the maths basically lines up FOV and DOF match with the appropriate adjustments.
Simply FF is a great option if you want to dive into the legacy still lenses back catalogue.
S35 is great for zooms and PL glass.
S
I downloaded your clip and toggled back and forth between the two shots. What I think you are seeing is
1) the slight forward protrusion of the 35mm lens compared to the 24mm, which makes the shot physically a tad closer to the desk and chairs.
2) The "expanding" desk caused by barrel distortion on the 35mm lens.
Neither of these is caused by an inherent difference between full frame and APS-C capture. If you moved your camera back so that the light hit the aperture at the exact same spot, the chair and desk would not look closer on the 35mm. Also, two lenses with identical amounts of barrel distortion would eliminate the "expanding/contracting" that's happening in the corners. Also, I would say technically, the APS-C shot you provided renders a more accurate perspective, as it's not stretching the sides like the Full Frame version.
Are you saying you can notice this on TV programs shot on APS-C cameras? A tightening effect? Considering the APS-C sensor size is essentially the same as Super35, and that has been the standard for film and television for decades, I find it confusing that this somehow doesn't look right, or that full frame looks better somehow because of this. Which I don't even think is the case, except on a lens by lens basis.
I'm not trying to be patronizing. Sorry if it sounds like that. I'm just wondering if the difference is objective and whether your perception might be colored by the more commonly accepted differences between the formats, like shallower DOF.
Last edited:
Viscount Omega
Carbonite Member
After spending hours all day viewing, editing and grading 100% full-frame footage (2 1/2 years now) when I turn on the tv and flip through the channels I immediately see the difference. I don't think that's an outlandish claim.
Everything looks pretty much the same to me except for the full-frame stuff. That's the stuff that stands out to me, that makes me say: "Now that's a full-frame shot!" It's usually a movie on AMC or Turner Classics.
What I'm trying to figure out is: (assuming it's mainly a depth of field issue), if the camera guy shooting Super 35 on some tv crime drama simply opens the aperture a stop or two more than he regularly does, does it now look like full-frame footage? If not, why not? Is there something else at work here?
filmguy123
Veteran
Questions/guessing:
- For the same DOF on FF vs S35/M43, you have a sharper image in lens' sweet spot (which is not close to wide open) on Full Frame, since you are operating stopped down a few clicks. Yeah?
- What about differences in glass engineering? For example, I imagine a 17.5mm lens for a M43 system has to be made in a way that has more barrel distortion than a native 35mm lens that is used on a Full Frame, correct? Or is the glass curvature the exact same? I know the field of view will be equivalent, but what about the way the glass sucks in light? It just seems to me that a 10.5mm M43 lens would HAVE to cause more barrel distortion than a 22mm FF lens? Or am I misunderstanding something?
- Does the overall MORE light just help a Full Frame image? I mean, no matter how good or clean high ISO is on a smaller sensor, aren't you are fundamentally starting the entire process with LESS light? If light is good, more light is better, yeah? I mean, even if you blast a scene with light and have to stop down or use ND filters, isn't it true that either way, a larger sensor will absorb more light into the image processing? And woudn't/couldn't this change the "look" of the way it is rendered?
morgan_moore
Major Contributor
There are many 'advantages' of full frame - for example if you can only manufacture glass at certain LPI you can record higher resolutions
Even dust on your sensor is less bad news as it ruins a smaller portion of the image
But lenses have to be bigger/heavier so there are downsides too
None of these things affect the "theoretical" look though
Even dust on your sensor is less bad news as it ruins a smaller portion of the image
But lenses have to be bigger/heavier so there are downsides too
None of these things affect the "theoretical" look though
deltoidjohn
Veteran
How do you know that it is full frame stuff? Are you just assuming it is because it conforms to what you personally believe is the full frame look? What shows are you generally seeing it in? After watching it, are you going and researching the particular scenes and finding out exactly what they are shot with? I'm not familiar with those particular channels but a quick scan of their websites/programs shows films and programs which are mostly shot on 35mm (or even 16mm), Arri, or Red Dragon. So it's all s35, or very slightly larger in the case of Dragon. You might start to see more full frame content on AMC once the Red Weapon 8K becomes commonplace.
If you can pick the shots using FF so easily when they brush past in a matter of seconds on TV, why could you not do the same in the previously posted test where you scored about 50% (ie the statistical odds if you were to take the test blind) My guess is you're ascribing certain types of shots (UWA Architecture? Super shallow DOF dialogue shots?) to being shot on full frame, when in actual fact it is just good (or different) cinematography.
Viscount Omega
Carbonite Member
I take it you see no difference between full-frame and cropped stuff. To each his own.
To be honest, I don't either. I just work the remote control. There's a drunken midget I keep locked in a steel cage who calls out "Full-Frame!" as I'm going through the channels. We usually stop on a feature film and watch until he gets bored. I feed him whiskey, raw meat and whores. He also supervises the editing of all my stuff, yelling "Cut!" when he feels the scene is just "right".
As for the "test" you're referring to, if the shot is a close-up or wide shot with no context, and no comparison shot, who on earth can tell if it's APS-C or Four-Thirds?
Or it's like showing me a shot of a brick wall and asking if it's IMAX or a compact camera. Just guessing on most of them, 50% seems about right based on the ridiculous samples he used.
To be honest, I don't either. I just work the remote control. There's a drunken midget I keep locked in a steel cage who calls out "Full-Frame!" as I'm going through the channels. We usually stop on a feature film and watch until he gets bored. I feed him whiskey, raw meat and whores. He also supervises the editing of all my stuff, yelling "Cut!" when he feels the scene is just "right".
As for the "test" you're referring to, if the shot is a close-up or wide shot with no context, and no comparison shot, who on earth can tell if it's APS-C or Four-Thirds?
Or it's like showing me a shot of a brick wall and asking if it's IMAX or a compact camera. Just guessing on most of them, 50% seems about right based on the ridiculous samples he used.
Last edited: