Debunking Tony Northrup.

[juanm]

So, a user in a different thread brought to my attention a certain video blogger who makes videos explaining basic concepts of photography. He's just plainly wrong in many aspects, and while I'd usually just let it go and not care, he has lots of views, and people are taking him seriously, so I feel like we should explain where he's wrong.

Here's his channel:
https://www.youtube.com/channel/UCDkJEEIifDzR_2K2p9tnwYQ

He mixes concepts like bokeh and focus breathing, and light gathering ability and depth of field, so it's rather important to set the record straight.

Use this thread to discuss him and his wrong ideas, or any other concept you don't fully understand, and we'll try to explain it. Collective intelligence at its best.

 

[The Bad Guy]

To be honest he comes across as a bit of a douche and I've never really paid any attention to what he has to say (I already know how to take photos and don't care for his tech descriptions).

But damn…have you seen his wife? And how great is that head of silver hair and cheesy smile of his.

 

[juanm]

To be honest he comes across as a bit of a douche and I've never really paid any attention to what he has to say (I already know how to take photos and don't care for his tech descriptions).

But damn…have you seen his wife? And how great is that head of silver hair and cheesy smile of his.

Yes, that's a fine piece of glass he shows in his videos...

I'd never heard of him until yesterday, but he appears to have invented a conspiracy theory of sorts in a parallel universe where aperture changes with focal length, and I some people are falling for it... That grey hair in some way must give him credibility and create a knowledge distortion field.

 

[kingalexthe1st]

Yes, that's a fine piece of glass he shows in his videos...

I'd never heard of him until yesterday, but he appears to have invented a conspiracy theory of sorts in a parallel universe where aperture changes with focal length, and I some people are falling for it... That grey hair in some way must give him credibility and create a knowledge distortion field.

Um, aperture does change with focal length. It's defined as the focal length divided by the diameter of the lens opening. So if you keep the opening the same and increase the focal length then the aperture increases. It's why a lens will be f/3.5-5.6, for example, over a zoom range.

Alex

 

[Janichsan]

That's the problem in this age of YouTube celebrities: anyone can spread his nonsense without necessary expertise…

 

[simonsi]

I take it you mean his assertion that aperture changes with crop-factor-multiplied focal length? Yes nonsense.

 

[juanm]

Um, aperture does change with focal length. It's defined as the focal length divided by the radius of the lens opening. So if you keep the opening the same and increase the focal length then the aperture increases. It's why a lens will be f/3.5-5.6, for example, over a zoom range.

Alex

Good, that's why I created this thread. That's a good example of a concept misunderstood.

You're starting from a wrong example, and I, on the other hand, should clarify what I said:
Yes, the physical aperture changes of course with the FL. For instance, a 50mm at f/2 will have a diaphragm opening of 25mm, and a 200mm at f/2 will have a iris opening of 100mm*. that's why f/numbers exist: do give a constant value across a different range of focal lengths.

However, that's not related in any way to your example of a variable aperture over a zoom range, since the variable iris opening would be already taken into account in the f/number.
Let's take an example with easy numbers (f/4)
A 70-200 f/4 zoom at its maximum aperture will have an iris opening of 17.5mm at the wide end 70mm (70/4) and an aperture of 50mm at its 200mm end (200/4). See? the f/number remains the same (and thus the exposure remains the same), even though the iris still physically changes its opening diameter.

The example you mention (a zoom with a variable aperture) is because a zoom with a constant zoom necessarily will be big in diameter, so they compromise by making the aperture comparatively smaller at the tele end.

Let's do some math with an extreme example: a 18-300 f/3.5-5.6
18mm/3.5=5.1mm iris opening (one fifth of an inch, more or less)
200mm/5.6=53mm iris opening (a little over two inches)

A constant f/3.5 aperture would get us a theoretical optical iris opening of 85mm (over three inches) and the lens would have to be exceedingly large, once you take into account the necessary mechanisms around the lens to focus and move the iris blades, so they just make the maximum relative aperture smaller at the longer end.

*Caveat: most consumer manufacturers will "cheat" when creating the optical design of a lens and engineering it, and the iris opening diameters I mention can only be considered nominal.

----------

I take it you mean his assertion that aperture changes with crop-factor-multiplied focal length? Yes nonsense.

Yes, nonsense (among other things) and yet beginners will believe him, pushing them towards buying more expensive gear than they should need, to compensate for his imaginary aperture conversion factor. That's why I created this thread, although in all fairness, someone should get him to take offline his videos, edit out the false assertions, and issue a formal apology.

That's the problem in this age of YouTube celebrities: anyone can spread his nonsense without necessary expertise…

Since the FBI won't do anything about it, I think it should be corrected.

 

[Bending Pixels]

urm...why spend time on this topic? It's like debunking Ken Rockwell.

 

[Meister]

So, a user in a different thread brought to my attention a certain video blogger who makes videos explaining basic concepts of photography. He's just plainly wrong in many aspects, and while I'd usually just let it go and not care, he has lots of views, and people are taking him seriously, so I feel like we should explain where he's wrong.

Here's his channel:
https://www.youtube.com/channel/UCDkJEEIifDzR_2K2p9tnwYQ

He mixes concepts like bokeh and focus breathing, and light gathering ability and depth of field, so it's rather important to set the record straight.

Use this thread to discuss him and his wrong ideas, or any other concept you don't fully understand, and we'll try to explain it. Collective intelligence at its best.

You seem to attack Tony Northrup personally instead of "debunking" anything.
Please explain where you think he is mistaken.

 

[juanm]

urm...why spend time on this topic? It's like debunking Ken Rockwell.

funny.

At least Rockwell often provokes on purpose, partly tongue in cheek, and even though his reviews are not serious on any level except maybe lens sturdiness, and his artistic taste is dubious (his kids at Legoland) he used to be some kind of engineer and you won't find a technical fault in his explanations. This guy on the other hand is 100% convinced what he says is true

----------

You seem to attack Tony Northrup personally instead of "debunking" anything.
Please explain where you think he is mistaken.

I've only watched a few minutes of his videos (the one relevant to the MFT light loss conspiracy theory discussion), but another example would be when he talks about focus breathing to sound like a pro, for instance. He means bokeh. Focus breathing can only be appreciated in video, since it implies "variations of bokeh over time" so to speak.

He knows the terminology, but he uses it wrong. And judging by his replies to comments in his videos, he's not trying to make it simple so people understand it, he just doesn't know what he talks about.

 

[AlexH]

On the whole I like his channel. I think he does a lot of stuff right, and I think it's an especially helpful channel for beginners. He may not be perfect, but he does have some good real world experience to back up is channel, and I appreciate that.

 

[juanm]

On the whole I like his channel. I think he does a lot of stuff right, and I think it's an especially helpful channel for beginners. He may not be perfect, but he does have some good real world experience to back up is channel, and I appreciate that.

Well, some of the things he teaches are false, and he pushes people to overspend. That said, I'll admit his show is well done and convincing (and like The Bad Guy said, his wife is hot)

 

[kingalexthe1st]

Good, that's why I created this thread. That's a good example of a concept misunderstood.

You're starting from a wrong example, and I, on the other hand, should clarify what I said:
Yes, the physical aperture changes of course with the FL. For instance, a 50mm at f/2 will have a diaphragm opening of 25mm, and a 200mm at f/2 will have a iris opening of 100mm*. that's why f/numbers exist: do give a constant value across a different range of focal lengths.

However, that's not related in any way to your example of a variable aperture over a zoom range, since the variable iris opening would be already taken into account in the f/number.
Let's take an example with easy numbers (f/4)
A 70-200 f/4 zoom at its maximum aperture will have an iris opening of 17.5mm at the wide end 70mm (70/4) and an aperture of 50mm at its 200mm end (200/4). See? the f/number remains the same (and thus the exposure remains the same), even though the iris still physically changes its opening diameter.

OK, we need to just clear something up. I'm sorry, I'm a real stickler for this. You put "See? The f/number remains the same". The 'number' in your sentence is the aperture. I was just typing that saying that "f/number remains the same" isn't right, as 17.5mm does not equal 50mm, when I realised what you were talking about. It's an annoying thing that f-number and f/number are similarly written and mean different things but such is life. The '4' in f/4 is the same however, which is what you're getting at. Also, saying that the aperture is 50mm at the 200mm end is incorrect too because, as we know, it's the '4' in f/4 that's the aperture. Apologies for nit-picking, but I didn't get your next bit (as you'll see) and getting on a level playing field will help me understand what you're saying.

The example you mention (a zoom with a variable aperture) is because a zoom with a constant zoom necessarily will be big in diameter, so they compromise by making the aperture comparatively smaller at the tele end.

Let's do some math with an extreme example: a 18-300 f/3.5-5.6
18mm/3.5=5.1mm iris opening (one fifth of an inch, more or less)
200mm/5.6=53mm iris opening (a little over two inches)

A constant f/3.5 aperture would get us a theoretical optical iris opening of 85mm (over three inches) and the lens would have to be exceedingly large, once you take into account the necessary mechanisms around the lens to focus and move the iris blades, so they just make the maximum relative aperture smaller at the longer end.

And that's where you lost me. I'm getting the maths, just not the explanation. 300mm / 5.6 is 53mm. Now if that were the maximum diameter, then 18mm / 53mm would give aperture = 0.34, which is unheard of so something is obviously going amiss. I think this is what you're getting at with the maximum relative aperture stuff, could you go in to a bit more detail?

Alex

 

[juanm]

The Voigtlander you are refering to acts like a nifty-fifty would act on a FF, but for 10 times the price. Fast M4/3 lenses are way more expensive than fast FF lenses.

Another misconception made in Northrup. Lenses for MFT are usually cheaper, in that the focal lengths are shorter, and usually shorter lenses are cheaper. I'll admit he seems so convinced when he says it, a beginner could easily fall for it. For instance, If I want to shoot a game fifty meters away, and on a Full Frame Body I'd need a 400mm lens to get good shots, with an MFT system I'd get the same FoV with a 200mm. And a 200mm 2.8 is about three to four times cheaper than a 400mm 2.8. (of course, you'll get a much shallower DoF with the FF and the 400mm if you shoot both cameras at f2.8)

So to sum up, in other words.

A 200mm acts exactly like a 400mm would on Full Frame body in terms of :
- Field of view
- Low light performance (that's what a fast lens is)

BUT:

Since depth of field is inherent to a focal length and its aperture (basically, the entrance pupil diameter) and to get the same shot, on a Full frame body you'd have mounted a 400mm instead of a 200 yes, you'll always get MORE depth of field (less blur) on a smaller sensor. But that's a byproduct of the process of choosing a shorter focal length for the same field of view.

However, that fact doesn't change a thing: MFT lenses at 2.8 are just like FF lenses at 2.8, the only difference being a shorter Flange focal distance and a smaller image circle (easier to manufacture, so cheaper, by the way)

That's why some lens makers like Rokinon/Samyang/Walimex have it really easy to make MFT lenses: they take their Full frame lenses and build in them the flange focal distance adapter.

Back to your point: IF the only thing you're after is shallow depth of field, then YES a MFT system would need faster lenses and hence would be more expensive. However, a given lens, used with adapters on different sized sensors, will only change its field of view, not its low light performance, there's no "aperture conversion factor", that's only in Northrup's head.

Caveat: I'm leaving out circles of confusion on purpose.

 

[paolo-]

I don't care for him either. I think he's a poor instructor. Instead of properly explaining basics, he jumps to conclusions and leaves viewers with a skewed understanding.

 

[juanm]

OK, we need to just clear something up. I'm sorry, I'm a real stickler for this. You put "See? The f/number remains the same". The 'number' in your sentence is the aperture. I was just typing that saying that "f/number remains the same" isn't right, as 17.5mm does not equal 50mm, when I realised what you were talking about. It's an annoying thing that f-number and f/number are similarly written and mean different things but such is life. The '4' in f/4 is the same however, which is what you're getting at. Also, saying that the aperture is 50mm at the 200mm end is incorrect too because, as we know, it's the '4' in f/4 that's the aperture. Apologies for nit-picking, but I didn't get your next bit (as you'll see) and getting on a level playing field will help me understand what you're saying.



And that's where you lost me. I'm getting the maths, just not the explanation. 300mm / 5.6 is 53mm. Now if that were the maximum diameter, then 18mm / 53mm would give aperture = 0.34, which is unheard of so something is obviously going amiss. I think this is what you're getting at with the maximum relative aperture stuff, could you go in to a bit more detail?

Alex

It's good to be a stickler and try to understand what happens in a camera. Maybe I should upload a video, it's easier to explain than in writing...

Basically, although it's not very intuitive, in an optical system (and I'm simplifying this maybe too much, it's just for the idea) the amount of light that would reach the sensor through a hole 5mm in diameter but 50mm away from the focal plane would be the same as if the pinhole were 30mm in diameter but 300mm.
In both cases the F number would be the same (f/10 since "f" stands for "focal length", and f divided by ten would give us 5mm and 30mm respectively). I try to use simple f/numbers, it makes it more intuitive.

In a zoom lens like the one I talked about, for instance, the F number (which is the same as the f/number, just a different notation) can open up to F3.5 and close up to F22 at the wide end and open to F5.6 and F40 at the tele end.

Maths:
at the wide end 18mm, when we open the diaphragm at F3.5, it will be in theory about 5mm big, and when we close it at F22 it'll be 0.81mm big.
at the tele end 300mm, when we open the diaphragm at F5.6 it'll be 53mm big, and when we close it down to F40 it'll be 7.5mm large.


Where I lost you, is because zooms don't maintain the same physical diaphragm opening, but the same aperture as in "opening relative to the focal length".
For instance, if you put your camera at F8 which is a value available trough out the whole zoom range (to make things easier) the physical theoretical aperture will change during zooming from 2.25mm to 37.5mm.

Likewise, on a 70-200 f/4 zoom, the maximum aperture -f/number- is constant but the physical diaphragm diameter is not. It'll go from 17.5mm to 50mm. Of course, if you look through it, the light beams you'd see will bend all along during zooming by going through all the glass, so it might not be easy to see.

Of course, modern zooms are very complex, with lots of lenses moving front and back depending on the focal length, so this is only in theory, and you'll often find that exact F-stops are not always respected, often varying by as much as 1/2 stop.

 

[Designer Dale]

People need to stop thinking of aperture as a measure of the size of the hole. It's a measure of the amount of light let in by that hole in relation to the focal length of the lens. A 200mm would need a bigger hole than a 50mm to let in the same relative amount of light simply because of the length of the tube and number of glass elements involved. f/2 is f/2 in terms of light only. Most folks here know that.

Dale

 

[FWRLCK]

Yeah, I wish there were a way to tell YouTube that I'm not interested in Tony Northrup videos, but AFAICT, there's not.

 

[kingalexthe1st]

People need to stop thinking of aperture as a measure of the size of the hole. It's a measure of the amount of light let in by that hole in relation to the focal length of the lens. A 200mm would need a bigger hole than a 50mm to let in the same relative amount of light simply because of the length of the tube and number of glass elements involved. f/2 is f/2 in terms of light only. Most folks here know that.

Dale

Right, which is where I think I'm getting confused. juanm seems to put units on the end of his aperture calculations every now and again so I don't know if he's referring to diameter or aperture. Eg:

Where I lost you, is because zooms don't maintain the same physical diaphragm opening, but the same aperture as in "opening relative to the focal length".
For instance, if you put your camera at F8 which is a value available trough out the whole zoom range (to make things easier) the physical theoretical aperture will change during zooming from 2.25mm to 37.5mm.

That's not where you lost me. I really need to point out that the aperture does not change from 2.25mm to 37.5mm. That's the diameter; the aperture is the ratio (8, in this case). And this is where I am getting confused because I think you're equating aperture to diameter, or something similar.

On a 70-200 f/4 zoom, the aperture -f/number- is constant but the physical diaphragm diameter is not. It'll go from 17.5mm to 50mm. Of course, if you look through it, the light beams you'd see will bend all along during zooming by going through all the glass, so it might not be easy to see.

OK, yes in this case the aperture does not change with focal length, and instead the diameter does in order to keep the aperture constant. I did not realise that it would open up more at f/5.6 when f=55mm than at f/3.5 when f=18mm. I could have done the maths before I posted but it wasn't that important to me at the time. I think I would like to re-phrase my original comment in this thread and say that for a given diameter, the aperture does increase as focal length increases, which is how I read your original quote:

...but he appears to have invented a conspiracy theory of sorts in a parallel universe where aperture changes with focal length

I obviously read 'diameter' somewhere in there in my head

Where you lost me was this:

The example you mention (a zoom with a variable aperture) is because a zoom with a constant zoom necessarily will be big in diameter, so they compromise by making the aperture comparatively smaller at the tele end.

...which, when I replace 'aperture' for 'diameter' in the last sentence starts to make sense. (I also think that when you put 'zoom with a constant zoom' you meant 'zoom with a constant aperture'?) However, if you make the aperture smaller at the tele end, then you make the diameter even bigger which caused me to scratch my head no end.

Alex

 

[Doylem]

A mistake on the internet? Say it ain't so, Joe, say it ain't so...

 

[juanm]

That has nothing to do with the topic of aperture/focal length conversions.
Did you merely make this thread to rant about this Northrup character or will you explain to me where he (and therefore I) is wrong.

----------

You are wrong. The total light collected by a m4/3 is less then the total light collected by a 35mm sensor. With the change of FOV also changes the total light gathering capability. And this change occurs exactly according ro the equation:


Once you apply the repective crop factor to both the focal length and aperture.

That formula doesn't apply. Ask around.
You still don't understand, or do not want to admit you're wrong...

Of course the total amount of photons that reach the sensor is less. But the area is also proportionately smaller. Same exposure, all other things being equal. Big news. The same amount of light reaches your house and your neighbour's.

One thing that's true is that a lot of light is wasted when you put a Full frame lens on a cropped sensor, but that's why speedboosters exist, and why they allow us to gain up to 1.5 stops. For instance, a Full Frame Nikon 50mm 1.8 with a Speedbooster on a GH4 could get as much light as a f/1.1 lens with the drawback (or advantage) of seeing its focal length reduced to 35mm.

 

[Meister]

I've only watched a few minutes of his videos (the one relevant to the MFT light loss conspiracy theory discussion), but another example would be when he talks about focus breathing to sound like a pro, for instance. He means bokeh. Focus breathing can only be appreciated in video, since it implies "variations of bokeh over time" so to speak.

He knows the terminology, but he uses it wrong. And judging by his replies to comments in his videos, he's not trying to make it simple so people understand it, he just doesn't know what he talks about.

That has nothing to do with the topic of aperture/focal length conversions. I haven't watched the entire video either, but I saw the part where he explains the crop factor conversion and I agree with him. According to the laws of physics he is correct.
Did you merely make this thread to rant about this Northrup character or will you explain to me where he (and therefore I) is wrong.

----------

So to sum up, in other words.

A 200mm acts exactly like a 400mm would on Full Frame body in terms of :
- Field of view
- Low light performance (that's what a fast lens is)

BUT:

Since depth of field is inherent to a focal length and its aperture (basically, the entrance pupil diameter) and to get the same shot, on a Full frame body you'd have mounted a 400mm instead of a 200 yes, you'll always get MORE depth of field (less blur) on a smaller sensor. But that's a byproduct of the process of choosing a shorter focal length for the same field of view.

However, that fact doesn't change a thing: MFT lenses at 2.8 are just like FF lenses at 2.8, the only difference being a shorter Flange focal distance and a smaller image circle (easier to manufacture, so cheaper, by the way)

t: IF the only thing you're after is shallow depth of field, then YES a MFT system would need faster lenses and hence would be more expensive. However, a given lens, used with adapters on different sized sensors, will only change its field of view, not its low light performance, there's no "aperture conversion factor", that's only in Northrup's head.

Caveat: I'm leaving out circles of confusion on purpose.

You are wrong. The total light collected by a m4/3 is less than the total light collected by a 35mm sensor. This is why smaller sensors have worse low light performance. With the change of FOV also changes DOF and the total light gathering capability. And this change occurs exactly according to the equation ..

... once you apply the respective crop factor to both the focal length and aperture.

----------

That formula doesn't apply. Ask around.

That formula is the aperture. Once you claim that a 25mm on m4/3 "acts like a 50mm on a FF body" you have to also claim that it's speed of 0.95 on m4/3 "acts like f1.9 would act on a FF body", because the formula is the f-stop. Those are the laws of physics and there is no need to ask around.

 

[phrehdd]

What a rather hot and lively conversation here.

I'll just say that people ought to be using terms the same way or rather, have working definitions of terms such as

f/stop
t/stop
diaphragm
aperture
iris
impact of glass within a lens on measure of center of lens to back of lens etc.

I'll just add that back in the days of using view cameras, I could take a landscape scene with the bellows not drawn out too far then immediately take an image of a flower placed in the scene up close. The bellows draw would be considerably more and an adjustment made for the change of the amount of light hitting the film ("f/stop" or leaf shutter).

Then again, much of the challenges of potential shifts of light really don't matter because people use cameras with built in metering. I often think the exercise was more about having fixed ASA/ISO of films and trying to get consistency in exposure when lenses were switched.

Zoom lenses have come a long way and as glass moves within a zoom, there are changes made in the relative* size of the aperture in the diaphragm to represent an exposure value referenced as an f/stop.

Discussion like these are interesting and certainly food for thought.

 

[MCH-1138]

You are wrong. The total light collected by a m4/3 is less than the total light collected by a 35mm sensor. This is why smaller sensors have worse low light performance. With the change of FOV also changes DOF and the total light gathering capability. And this change occurs exactly according to the equation ..

Image

... once you apply the respective crop factor to both the focal length and aperture.

I'm sorry, but I think this is where the analysis goes south. Yes, a 35mm sensor has more area and therefore can potentially collect more total light than a smaller (APS-C or M43) sensor, but the amount of light hitting the "crop" portion of the 35mm sensor is exactly the same (assuming all else is equal -- i.e., same lens, same settings, etc.).

I think it is easier to think about this portion of it by not getting into crop factors, effective fields of view, etc. If you shoot the same scene with the same lens (say 50mm 1.8G) at the same settings (say 1/250, f/2.8, ISO 100) on a D800 and D7000 and then crop (not resize) the D800 image to 16.2MP, the exposure should be the same in the resulting images. Yes, the D7000 sensor would have collected less total light than the D800 sensor, but the light that is relevant to the final image (i.e., the whole image on the D7000 and the cropped center of the image on the D800) is exactly the same.

 

[juanm]

That has nothing to do with the topic of aperture/focal length conversions. I haven't watched the entire video either, but I saw the part where he explains the crop factor conversion and I agree with him. According to the laws of physics he is correct.
Did you merely make this thread to rant about this Northrup character or will you explain to me where he (and therefore I) is wrong.

----------

You are wrong. The total light collected by a m4/3 is less than the total light collected by a 35mm sensor. This is why smaller sensors have worse low light performance. With the change of FOV also changes DOF and the total light gathering capability. And this change occurs exactly according to the equation ..

Image

... once you apply the respective crop factor to both the focal length and aperture.

----------

That formula is the aperture. Once you claim that a 25mm on m4/3 "acts like a 50mm on a FF body" you have to also claim that it's speed of 0.95 on m4/3 "acts like f1.9 would act on a FF body", because the formula is the f-stop. Those are the laws of physics and there is no need to ask around.

You do not change the focal length. A 200mm lens remains a 200mm lens no matter what you put behind it: a 5d, a gh4, or a banana.
The crop factor is just an easy way to compare fields of view to a widely known standard (24x36mm film).
When someone claims a lens "acts like" they refer only to the field of view. People like you and your grey haired guru just don't know the basics and confuse different concepts.

And no, smaller sensors have worse low light performance because they usually try to pack too many pixels and since the surface is smaller, they make them smaller. A cropped sensor with the same pixel size, pitch and technology on a hypothetical "Sony A7S DX" would have the same light sensitivity as the current A7S