Must watch video on full frame vs crop cameras. "Full frame look" covered.

[andy lee]

this is what happens when people get obsessed with comparing everything to Full Frame - why bother!

They dont shoot Hollywood movies full frame any more Vista Vision died years ago. Super 35mm is King.

 

I don't ever remember when I shot film (16mm and super 35mm ) saying I wonder what the equivilent light levels of this lens is on Vista Vision ?

 

I personally do not like the full frame look at all (yes I've shot alot of pop videos on Canon 5D, I'm just not a fan of it )

 

Just do the best with what you have got , having a head full of conversions to full frame and now all this extra nonsense that this video has drawn up achieves nothing useful at all.

 

Know how to get the best from your system - that's all that matters.

[Inazuma]

Still can't believe Northup put up this video. The only thing I really agree with him on is that these MFT lenses are fricking expensive and that they should really make faster zooms.

[utsira]

Dishe, Andrew, again appreciate the discussion thank you.
I am not arguing for the sake of arguing but so far I do not understand how what I am stating is wrong.
But I will gladly learn something and stand corrected.
 
I do understand that sensor pixel size is to be taken into consideration when talking about exposure of the sensor, larger sensors generally having larger pixels than crop ones for instance. And architecture as far as gaps between pixels and the amount of pixels. 
 
In order for me to understand your arguments, can we proceed step by step?
 
Can we begin solely with light intensity? Regardless of sensor specs.
 
What I don't understand in your arguments is regarding the intensity of the light, focused by speedbooster or not.
 
For the sake of understanding here is a lens that has an aperture that lets through 4 rays of light:
Is this correct or is this wrong?
 

But in the full frame example the 4 rays are spread over a much larger surface. The crop sensor is just that, a crop. It doesn't make the cropped area brighter or darker. That is why almost everyone on the planet does not multiply aperture by the crop factor.

[Jacek]

Relax people! Stop throwing arguments and try to understand each other.

 

In general you are both right, so all your new arguments and examples from both sides will be correct.

 

You are just looking at the same thing from 2 different perspectives:

 

• EOSHD side is looking at parameters (ISO, light, f-stop) by DENSITY perspective (which is real definition of these well-known parameters)
• Video guy side is looking at all parameters by TOTAL perspective (which is not real definition of these parameters - it is just a new perspective which he tries to present)

 

Video guy said it clear (was very clear in ISO section, but maybe not so clear in F-stop section), that his 'total' perspective is not the established definition of parameters, but the perspective, which he thinks is more appropriate for different sensor sizes comparison.

 

Which perspective is better to compare sensor sizes is up to you, but the calculations from both sides are in general correct (except for nuances like pixel-level sensor production or same lens construction in different scale possibilities).

 

Now I'm ready for rocks from both sides :), but I'll just pass them to the other side.. :lol:

[Jacek]

Video guy presentation is clearly inspired by SpeedBooster capabilities.

 

He realized, that it is wrong that price of FF 24-70/2.8 is similar to m43 12-35/2.8 price.

 

Why is it wrong? Because simple SB for m43 will make FF 24-70/2.8 much better lens than m43 12-35/2.8 lens - for the same price (except one [still overpriced] SB for all your lenses).

[jcs]

Oh man, that's just silly. It completely disregards A) megapixel levels. Is the fullframe 36mp D800, 22mp 5d or 12mp a7s? Is the crop sensor 16mp?  And b ) light levels. What are the actual light levels? In iso 100 you could argue that a 24mp a6000 takes sharper images with more dynamic range than a 22mp 5dIII. How does that fit in with those silly analogies? It doesn't.

 

The circle/particles/area comment was to show that a larger diameter bucket will capture more water per unit time vs. a smaller diameter bucket. Pixel density doesn't matter with respect to the math and real-world results when using the math to predict what will happen with various lenses, sensors, ISO, and f-stop settings. ISO or gain is calibrated to the sensor- this is why the details of the sensor design don't matter. If ISO is increased from 100 to 200, we've doubled the gain and image brightness in a known way, regardless of anything else because ISO is calibrated to that sensor. Is there variance in camera manufacturer ISO calibration? Sure, however his examples match the math.

 

Using any of items 1-3 above as shown in the video, is there a real-world example that shows the math is wrong?

[jcs]

He realized, that it is wrong that price of FF 24-70/2.8 is similar to m43 12-35/2.8 price.

 

Why is it wrong? Because simple SB for m43 will make FF 24-70/2.8 much better lens than m43 12-35/2.8 lens - for the same price (except one [still overpriced] SB for all your lenses).

 

Right- a m43 12-35/2.8 on m43 will have the same bokeh and light gathering performance as a FF 24-70/5.6 on a FF camera (not very fast, not much bokeh compared to F2.8 on FF). He demonstrated this effectively in the examples in his video.

[jgharding]

I agreed with the point of "if they're going to use a 35mm equivalent focal length they should, possible legally be forced to, show the equivalent aperture"

 

This is because the lens is a 2.8, yes, at it's actualy physical focal length, but as soon as you're listing equivalent focal lengths, not listing equivalent aperture is bullshit, simply because f-stop is a ratio of focal length. 

 

All the rest of it, i think he manages to make everything more confusion by trying to be controversial.

 

The stop system works fine.

 

I think perhaps there should be a legal requirement to list sensor gain in dB as well in the manual next to each ISO

[richg101]

 

 

I think perhaps there should be a legal requirement to list sensor gain in dB as well in the manual next to each ISO

 

 

I think this is where his arguments are stemming from.  He is saying that the iso number is not tangible since manufacturers can use exactly the same sensor, yet gain them differently for the same iso level they display on the lcd screen.  - the fuji example - their iso levels are gained differently yielding cleaner high iso's - even though you need to pull them up by 1/3rd of a stop in order to obtain the same exposure as on a canon at the same iso. 

[jgharding]

Yes I think there are two good points in the piece, the rest is a bit muddled.

 

I also agree with Andy on the obsession with full-frame in video.

 

S35 (nearly APSC) is full frame for video really.

 

Photographic full frame, or 135 film is kinda different.

 

Still, the world is changing and both are converging, but there's still a lot of confusion out there because of all these crop factors going about.

 

I have a feeling it will all settle gradually.

 

Of course, all the problems here are the fault of sales and marketing people, bending the truth or just flat-out lying in order to sell a product.

 

That's the real problem we face.

 

Reality is not a spec sheet or a presentation, it's using the tool to make something real.

[KarimNassar]

But in the full frame example the 4 rays are spread over a much larger surface. The crop sensor is just that, a crop. It doesn't make the cropped area brighter or darker. That is why almost everyone on the planet does not multiply aperture by the crop factor.

 

If that is true then demonstrate to me using a graphic such as mine how a speedbooster provides more light and increases exposure.

[KarimNassar]

I do not see how it is possible for a full frame lens to provide the same amount of light to a crop sensor as it would to a full frame.

Unless you are using a speedbooster.

 

Now we are talking amount of light only, not sensor sensitivity.

 

Saying a f4 full frame lens becomes an f2.8 with speedbooster on a crop sensor is in my opinion false.

 

We all know for fact that speedboosters increase the amount of light on the sensor.

 

But how?

 

I say : you lose light on crop sensors to begin with and speedbooster makes up for the loss by focusing the lost light back on the crop sensor.

 

 f4 full frame = f4 on crop with speedbooter

 

You say : there is no loss to begin with, and speedbooster increase the light on the sensor compared to full frame.

 

f4 full frame = f2.8 on crop. I don't see how that works, please demonstrate via graphic how that works

 

This is how I understand it, please do correct this if it is wrong:

 

 

 

 

 

 

[jcs]

It's a common misconception regarding the Speedbooster and F-stop: the Speedbooster increases the T-Stop not the F-Stop.

 

F-stop = focal-length/aperture-diameter: the Speedbooster doesn't change this. It simply scales down the image circle.

 

The T-Stop increase is relative to the same lens without the Speedbooster on the same camera (e.g. the regular Metabones adapter).

 

Notice how the bokeh doesn't really change when comparing the same lens on full frame to Speedbooster on S35 (size change is due to 1.1 crop (Speedbooster isn't quite full frame)):

[utsira]

If that is true then demonstrate to me using a graphic such as mine how a speedbooster provides more light and increases exposure.

Your own graphic demonstrates this (the first one you did). It is the density of the rays, not the number of rays, that determines the brightness of the image. So yes, 4 rays are hitting the full frame sensor vs 2 for the crop, but those rays are spread over a much bigger area. The speedbooster makes the image brighter because the density of the rays is increased. 

[Inazuma]

It's the same concept as lighting something with magnifying glass. The magnifying glass captures light from a wide area and condenses it into a much smaller spot, thereby increasing the aomunt of light and heat that spot receives.

[Guest Ebrahim Saadawi]

I watched the video but didn't have the time to go through the discussion here. But as far as I am seeing, all he's saying is that marketing teams always apply the crop factor to the focal length and not the aperture, or ISO. 

A 12-35mm lens is in fact a 24-70mm f/5.6, not an 24-70mm F/2.8. 

We all know that in terms of depth of field, that's correct. But his theory is that it also applies to the exposure value. 

That a 12-35mm F/2.8 is not the same brightness as a 24-70mm f/2.8, but it's the same as 24-70mm f/5.6, and smaller sensors have to apply more gain to compensate for that exposure difference, thus have more noise. 

Example:

Two pictures. 

1- 35mm - 2.8 - 800 ISO on M43s
2- 35mm - 2.8 - 800 ISO on fullframe 

1 & 2 don't look similar in terms of field of view, depth of field, or ISO

so to match field of view:

1- 35mm - 2.8 - 800 ISO
2- 70mm - 2.8 - 800 ISO 

Now 1 & 2 look similar in terms of field of view, but they don't look similar in terms of -ISO -Depth of field. 

so to match depth of field 

1- 35mm - 1.4 - 800 ISO on m43s
2- 70mm - 2.8 - 800 ISO in fullframe

1 & 2 now look similar in terms of both angle of view and depth of field, but they don't look similar in terms of - ISO (1 is two stops brighter than 2)
 
so to match ISO 

1- 35mm - 1.4 - 200 ISO
2- 70mm - 2.8 - 800 ISO

now 1 & 2 are exactly identical in terms of field of view, ISO noise, and depth of field. 

_______________________

So as far as I understand his theory is that we should also apply the crop factor to the ISO value. That a smaller sensor at 200 ISO = 800 ISO on full frame, and he takes noise levels as his proof. Thus says that smaller sensors don't have noisier ISOs. 

_______________________

Not suggesting he's right or wrong, I have no idea. 

but it's interesting. 

[KarimNassar]

Your own graphic demonstrates this (the first one you did). It is the density of the rays, not the number of rays, that determines the brightness of the image. So yes, 4 rays are hitting the full frame sensor vs 2 for the crop, but those rays are spread over a much bigger area. The speedbooster makes the image brighter because the density of the rays is increased. 

 

I think I understand what you mean, I was focusing on light transmission only, without acknowledging final light density on the surface of the sensor.

 

So my graphics are correct as:

 

- A crop sensor without speedbooster receives less total light as the full frame sensor

- A crop sensor with speedbooster receives the same amount of light as the full frame sensor

 

But:

 

- Since the crop sensor is smaller, the same amount of light on a smaller surface = higher density

 

correct?

[jcs]

The concepts from the video and the math are helpful. A GH4 with a Voigtlander 25mm F.95 at ISO 200 will look the same as a 50mm F1.9 ISO800 on a 5D3, with about the same amount of noise. The same scene and settings requiring ISO6400 on the 5D3 will need ISO1600 on the GH4: perhaps similar low-light performance (Northrup cleaned up GH4 ISO12800 footage with Neat Video (pool shot in 5D3+GH4 video): looked decent on youtube).

[AndrewM]

There is a lot of apparent disagreement here but not much actual disagreement, I think. Is this a fair summary?

 

If I am buying a camera BODY, there are some numbers that I should know:

        - sensor size

        - pixel count

Now all other things being equal more pixels is nice (up to a point), but as Andrew points out at the beginning of every camera report, more pixels on a smaller sensor probably means worse pixels all other things being equal. But bottom line is that, especially for video, we have more pixels than we need. Really, it is PIXEL SIZE, which is derived from the above two numbers (roughly - Andrew's point about global shutter circuitry and other technologies like backside illumination must be taken into account) that matters more.

 

However, all other things are almost never equal. So if we want to know how good the actual image is, everything is in the details. If I want to know the details, then I rely on review sites that take pictures under carefully controlled situations with the best possible lenses, look at how ISO 800 pics on different bodies compare, in terms of noise and resolution and so on. If there is a fudge factor in ISO, as the OP video claims, it will show up in the quality of the images at this point, so this does not concern me. Crop cameras should look worse by the factor he claims, and if so then... they will look worse, and we will all be able to tell.

 

Now what if I am buying camera LENSES?

 

First, we all know that there are enormous variations in lens quality - both in technical quality (MTF etc) and in aesthetic quality. Put that aside for a moment and assume all these things are equal. A lens can be described by three numbers (ignoring zoom, minimum focus, etc):

        - focal length

        - maximum aperture

        - image circle

 

If I have my camera body already, then I have to make sure I buy lenses with a big enough image circle. If I have a full-frame body and I buy APS-C only lenses, I am going to be really disappointed by what I see, and will be using crop mode all the time and might as well have bought an APS-C camera. But I can put a larger image circle lens on a smaller sensor body (assuming I can physically attach it) and it will work fine, giving me images (again, all other things being equal) equivalent to a crop of the larger sensor body the lens was intended for. Of course, because of the crop, I may need to reframe for the same image, and because of that, depth of field and perspective will not be the same. By multiplying the focal length by the crop factor, I know how much I need to reframe for the same image (or conversely, I know which (different focal length) lens to chose so I don't have to reframe).

 

Ok. Now let's talk about about what you get for your money. Suppose I am comparing two lenses that will render the same field of view on their native sensor size and roughly equivalent "quality" - say, a 50mm f/2.8 full-frame and a 25mm f/2.8 m43. While similar in many, many ways, the fundamental difference is that that the full-frame lens renders a larger image circle, which means that it is a more complex piece of engineering - it maintains an adequate image over a 4x larger area. That means more glass (and a heavier lens), tighter tolerances etc. It is much harder to keep that wide aperture and still resolve over that large image circle. Also, if shallow depth of field is your thing, you are not going to get as shallow with the m43 on native body.

 

If I put that full-frame lens on a crop body it will work fine, but I am wasting an awful lot of engineering by doing so, because much of the image circle is just being ignored. Also, as Andrew points out, I may lose an awful lot of the character of the lens, which comes from its fall-off, distortions, etc.. But this may not worry me.

 

The magic of the speedbooster is that it takes all the engineering, all the character, of the larger-format lens and squeezes it down to fit onto a smaller sensor, so you are not just throwing that away. It means you get the lens as it would be on the larger sensor.

 

Now here is the part where I agree with the video that started all this. If you look at the lenses and their prices, full-frame lenses look like much better value. That 50mm f2.8 FF and that 25mm f2.8 m43 cost close to the same, but if you think about the job they are doing, the full frame lens is way more impressive. (It is also way heavier (because of the job it is doing)). And on a native body, it will do things the other won't, like a shallower depth of field. If that is what matters to me, then yes, I need a 25mm f1.4 native m43 lens. If that doesn't matter to me, and if other things matter more (like size) then I may not care about the full frame comparison.

 

I'm looking at getting a new body and new lenses. And I am facing this issue every time I try to decide. I look at the GH4, and I look at the native lens selection and what you pay for them, and what they do compared to full-frame or APS-C lenses and what you pay for those, and the value seems off. I get much more for my money with the larger-format lenses. But I pay more for the body that handles those lenses, and the body may not be as good for video anyway. It is, after all, the combination of lenses-plus-body that produces the actual images.

[sunyata]

What the hell is going on here? I've totally lost track. Bring back film!