My uncle Richard is one of the inventors on Honeywell’s early phase‑detect autofocus. Patent US4333007A, which figures out both the direction and amount the lens needs to move instead of hunting.
Modern systems like Canon’s Dual Pixel AF in bodies such as the EOS R5 are very direct descendants of that idea, just implemented on‑sensor with far more processing power.
Every time I see an article such as this, I beam with pride. (Pun intended).
This article presents the inventive solutions Canon has found to shoot beams of light into the camera's viewfinder in order to light up individual autofocus points. Six different approaches are shown using six Canon DSLR models between 1994 and 2009.
It's absolutely wild to me that this function would be so important as to justify all the time and effort to create not just two, but six entirely different ways of solving it.
The first one seemed perfectly adequate to me. But I guess that's why I'm not a Canon engineer.
I haven't used all of them, but I have used both the 9 point and the 45 point types, and the difference is massive. The 45 point was far, far more tactile and responsive. I don't mean speed of autofocus, but the actual way that the points sit over top of the viewfinder and light up, it's hard to explain. I'm sure part of that is software, but owning an older model and then trying out a newer one in the camera store in like 2013 really was eye opening, it blew my mind. The 9 point feels like a toy.
On the other hand in actual usage i don't think that they are really that different. It's useful for sports/wildlife for focusing closest moving target (and speed of AF on these cameras was not that quick so you were hunting focus anyway). Otherwise selecting some offset autofocus point is pretty niche. With more static subjects majority of photographers would use the single middle point, focus and then recompose.
It's only with advent of smart focusing of mirrorless cameras with people/faces recognition where there is a big difference.
Huh, that definitely explains it. I wonder how many people know that. In that case it's particularly unfortunate to downvote the OP simply for filling a field in the submission form! Sigh… I guess it's another case of LLMs having made the world a little worse for everybody.
Yeah, but before LLMs we didn't have "reads like LLM output" as a downvote reason. In 2022 nobody would've had qualms with the phrasing of the comment.
But the articles are well-researched, very high quality (the illustrations in particular are incredible), and in every way prime HN material, and there's no money involved as far as I can see. I definitely don't have any qualms about self-promotion of this sort of stuff.
Maybe in some niches, but for most modern ILCs (e.x. Sony E mount, Nikon Z mount, etc etc etc) it is typically some form or another of On-sensor Phase Detection coupled with Contrast Detection for finer tuning.
The biggest advantage compared to older SLR designs are that Phase Detection can now work with full light (vs whatever got split off on an SLR pentaprism for the dedicated AF sensor) and can work in conjunction with the contrast Detection for fine tuning focus.
Then, of course, all the predictive stuff added in the last 10ish years as far as processing sensor output to detect eyes/birds/motion/etc.
The only AF systems that use LIDAR are the one on the newest Hasselblad medium format mirrorless cameras (since DJI owns Hasselblad and can leverage the tech from drones/cinema cameras) and possibly some phones.
I suspect it's fairly challenging to implement since the LIDAR sensor doesn't operate through the lens, so you'd have to continuously align the depth map with the image to account for parallax; plus it's only useful for close-ish distances (since the lasers can't be too powerful) and can cause unwanted focus behavior with windows or reflections.
No. Modern cameras usually use a combination of contrast detection (pure image analysis checking the contrast of the region you want in focus), phase detection (an optical system where you split the income image in two and then compare them) and sometimes help of some sort of assist lamp.
And just mentioning for people unfamiliar with this stuff, that's not a camera and not even an accessory for a camera. That's an accessory for an accessory. :D
It's a LiDAR that follows a subject and gets distance measurements, and then sends them to an additional accessory which is typically used to control focus in cinema cameras. That second accessory has a motor and is attached to a cinema lens that has certain threading or grooves where the accessory can grip and change the focus.
In cinema, the camera operator (usually) only moves the camera, but not the focus. For that, there's a 'focus puller'. A person who finely operates the focus, sometimes at a certain distance, using some sort of specialized control.
You can get a Canon 5D II or a Canon 6D for that. Older cameras for sure, but full frame with excellent sensors. And there are a ton of inexpensive used EF lenses available.
Compared to moderns systems the main difference is the autofocus and video capabilities. Modern mirrorless have cosmically better tracking, eye detect etc.
Also speaking of older models I think it is important to repeat that the pixel count is not what define the quality of the image but mereley only how cropable the end result will be and it is only really useful if the higher pixel count isn't made of garbage. Accutance of the end result is in most cases much more important.
For instance, human eyes can't perceive the difference between a 12MP and a 50MP image printed in a poster format from a typical 1.5-2meters viewing distance and 8MP is usually good enough for most large prints.
So I would advise choosing a second hand model taking shutter count, general state, lenses quality, autofocus speed and image stabilisation efficiency as more prioritary parameters than sensor pixel count.
On one hand, you have to remember that huge MP doesnt do much if the glass can't resolve well enough.
On the flipside, I have to note that switching to high MP full frame makes it a lot easier to do good, clean crops. Sometimes I might care about a small portion of the frame but for composition reasons (e.x. can't get closer for one reason or another) I at least can lean on cropping more.
Full frame sensor = full frame lenses = heavy and expensive. You need something light that you want to carry with you. Otherwise what's the point of having the best medium format camera + lens that you have at home, collecting dust.
That means a camera APS-C or micro four thirds sensor might suit better to someone who is new to photography.
I've always thought i am that APSC person and I ended up not taking the camera that much with me because it was small difference to compact camera i was carrying in my pocket. I thought it was skill issue because my photos just didn't have that bite/detail/look.
The moment i got older full frame i realized that "look" has so much to do with the sensor size. In beginning i only had cheap 30euro manual soviet lens and even with that the quality of the photos just shot up.
Full frame is heavy yes but it can be pretty affordable (lenses from china are becoming extremely competitive).
To add: Hit https://en.wikipedia.org/wiki/Canon_EOS_5D_Mark_II and scroll down to the very end, then click "show" to expand the "Canon EOS digital SLR timeline". This is an incredible collection of information that I can't figure out how to link directly.
I recently finally got my first fullframe (i've always used apsc crop sensors). I would say you want fullframe - to me it's really major difference in the look of photos. I would get apsc only if you really need portability. But for 200eur even old fullframe is tough.
I've tried to get Canon 5D/6D but they are becoming pretty sought after (because of the availability of lenses). Another problem is when they are sold cheap they have huge shutter counts often way after their spec lifetime. Not sure about Nikon maybe there are some cheaps. You can probably get 6D mark I with pretty high shutter count for around 200e.
Best usable 5D mark III deals i could find were around 450eur (thats camera from 2012). I ended up getting Lumix S5 (mark 1, from 2020) for like 500eur that is very different beast of a camera while having L-mount which is becoming only "open" camera mount (third party lenses are being "disallowed" by most manufacturers now).
I've been hunting 6D for a while in eu for 200eur you get like 400 000 shutter count. But 6D is also in that range where people don't care about it and often sell it as some junk. So one might get lucky.
My uncle Richard is one of the inventors on Honeywell’s early phase‑detect autofocus. Patent US4333007A, which figures out both the direction and amount the lens needs to move instead of hunting.
Modern systems like Canon’s Dual Pixel AF in bodies such as the EOS R5 are very direct descendants of that idea, just implemented on‑sensor with far more processing power.
Every time I see an article such as this, I beam with pride. (Pun intended).
This article presents the inventive solutions Canon has found to shoot beams of light into the camera's viewfinder in order to light up individual autofocus points. Six different approaches are shown using six Canon DSLR models between 1994 and 2009.
It's absolutely wild to me that this function would be so important as to justify all the time and effort to create not just two, but six entirely different ways of solving it.
The first one seemed perfectly adequate to me. But I guess that's why I'm not a Canon engineer.
I haven't used all of them, but I have used both the 9 point and the 45 point types, and the difference is massive. The 45 point was far, far more tactile and responsive. I don't mean speed of autofocus, but the actual way that the points sit over top of the viewfinder and light up, it's hard to explain. I'm sure part of that is software, but owning an older model and then trying out a newer one in the camera store in like 2013 really was eye opening, it blew my mind. The 9 point feels like a toy.
On the other hand in actual usage i don't think that they are really that different. It's useful for sports/wildlife for focusing closest moving target (and speed of AF on these cameras was not that quick so you were hunting focus anyway). Otherwise selecting some offset autofocus point is pretty niche. With more static subjects majority of photographers would use the single middle point, focus and then recompose.
It's only with advent of smart focusing of mirrorless cameras with people/faces recognition where there is a big difference.
It’s impressive the optical complexity some of these systems use. Custom and complex prisms and mirrors to show the AF points.
Arguably more complicated than the anutofocus optics. The engineering of electromechanical cameras fascinates me.
@ExAr I suspect your comment was downvoted to oblivion because people didn't realize you're the OP and thought you were an LLM summarizer bot :(
The comment reads like an LLM summary. And we don’t typically get an OP summary on a new post. Just post it and let people read it.
HN appears to encourage it, because it shows a text box which becomes a top level comment when you submit, although it isn't obvious that will happen.
That box doesn't become a comment, does it? It goes into the description that shows under the link.
Huh, that definitely explains it. I wonder how many people know that. In that case it's particularly unfortunate to downvote the OP simply for filling a field in the submission form! Sigh… I guess it's another case of LLMs having made the world a little worse for everybody.
I don’t think it’s because the comment was submitted. I think it’s because it reads like LLM output.
Personally, I’ve only ever provided a summary if I felt the headline wasn’t clear enough.
Yeah, but before LLMs we didn't have "reads like LLM output" as a downvote reason. In 2022 nobody would've had qualms with the phrasing of the comment.
I only realized they weren't a bot when i went to look at the other comment history.
Thanks for pointing this out! I had no idea something like this would happen by filling in a comment box in the original post window.
That's perfectly ok, they're using HN almost exclusively for promotion. See comment/submission history.
But the articles are well-researched, very high quality (the illustrations in particular are incredible), and in every way prime HN material, and there's no money involved as far as I can see. I definitely don't have any qualms about self-promotion of this sort of stuff.
Does AF use LIDAR nowadays?
Maybe in some niches, but for most modern ILCs (e.x. Sony E mount, Nikon Z mount, etc etc etc) it is typically some form or another of On-sensor Phase Detection coupled with Contrast Detection for finer tuning.
The biggest advantage compared to older SLR designs are that Phase Detection can now work with full light (vs whatever got split off on an SLR pentaprism for the dedicated AF sensor) and can work in conjunction with the contrast Detection for fine tuning focus.
Then, of course, all the predictive stuff added in the last 10ish years as far as processing sensor output to detect eyes/birds/motion/etc.
The only AF systems that use LIDAR are the one on the newest Hasselblad medium format mirrorless cameras (since DJI owns Hasselblad and can leverage the tech from drones/cinema cameras) and possibly some phones.
I suspect it's fairly challenging to implement since the LIDAR sensor doesn't operate through the lens, so you'd have to continuously align the depth map with the image to account for parallax; plus it's only useful for close-ish distances (since the lasers can't be too powerful) and can cause unwanted focus behavior with windows or reflections.
No. Modern cameras usually use a combination of contrast detection (pure image analysis checking the contrast of the region you want in focus), phase detection (an optical system where you split the income image in two and then compare them) and sometimes help of some sort of assist lamp.
https://en.wikipedia.org/wiki/Autofocus
I think only in certain cinema rigs. See DJI's products eg https://store.dji.com/product/dji-focus-pro-lidar
And just mentioning for people unfamiliar with this stuff, that's not a camera and not even an accessory for a camera. That's an accessory for an accessory. :D
It's a LiDAR that follows a subject and gets distance measurements, and then sends them to an additional accessory which is typically used to control focus in cinema cameras. That second accessory has a motor and is attached to a cinema lens that has certain threading or grooves where the accessory can grip and change the focus.
In cinema, the camera operator (usually) only moves the camera, but not the focus. For that, there's a 'focus puller'. A person who finely operates the focus, sometimes at a certain distance, using some sort of specialized control.
hijacking the thread to ask: under 200€ DSLR to get started?
You can get a Canon 5D II or a Canon 6D for that. Older cameras for sure, but full frame with excellent sensors. And there are a ton of inexpensive used EF lenses available.
Compared to moderns systems the main difference is the autofocus and video capabilities. Modern mirrorless have cosmically better tracking, eye detect etc.
Also speaking of older models I think it is important to repeat that the pixel count is not what define the quality of the image but mereley only how cropable the end result will be and it is only really useful if the higher pixel count isn't made of garbage. Accutance of the end result is in most cases much more important.
For instance, human eyes can't perceive the difference between a 12MP and a 50MP image printed in a poster format from a typical 1.5-2meters viewing distance and 8MP is usually good enough for most large prints.
So I would advise choosing a second hand model taking shutter count, general state, lenses quality, autofocus speed and image stabilisation efficiency as more prioritary parameters than sensor pixel count.
Two tricky things when it comes to pixel count...
On one hand, you have to remember that huge MP doesnt do much if the glass can't resolve well enough.
On the flipside, I have to note that switching to high MP full frame makes it a lot easier to do good, clean crops. Sometimes I might care about a small portion of the frame but for composition reasons (e.x. can't get closer for one reason or another) I at least can lean on cropping more.
Full frame sensor = full frame lenses = heavy and expensive. You need something light that you want to carry with you. Otherwise what's the point of having the best medium format camera + lens that you have at home, collecting dust.
That means a camera APS-C or micro four thirds sensor might suit better to someone who is new to photography.
I've always thought i am that APSC person and I ended up not taking the camera that much with me because it was small difference to compact camera i was carrying in my pocket. I thought it was skill issue because my photos just didn't have that bite/detail/look. The moment i got older full frame i realized that "look" has so much to do with the sensor size. In beginning i only had cheap 30euro manual soviet lens and even with that the quality of the photos just shot up.
Full frame is heavy yes but it can be pretty affordable (lenses from china are becoming extremely competitive).
To add: Hit https://en.wikipedia.org/wiki/Canon_EOS_5D_Mark_II and scroll down to the very end, then click "show" to expand the "Canon EOS digital SLR timeline". This is an incredible collection of information that I can't figure out how to link directly.
The direct link is here: https://en.wikipedia.org/wiki/Template:Canon_EOS_digital_cam... You can get to it by pressing the `V` in `V T E` at the top left of any tmeplate
Direct link: https://en.wikipedia.org/wiki/Template:Canon_EOS_digital_cam...
Not sure why it is a "template".
I recently finally got my first fullframe (i've always used apsc crop sensors). I would say you want fullframe - to me it's really major difference in the look of photos. I would get apsc only if you really need portability. But for 200eur even old fullframe is tough.
I've tried to get Canon 5D/6D but they are becoming pretty sought after (because of the availability of lenses). Another problem is when they are sold cheap they have huge shutter counts often way after their spec lifetime. Not sure about Nikon maybe there are some cheaps. You can probably get 6D mark I with pretty high shutter count for around 200e.
Best usable 5D mark III deals i could find were around 450eur (thats camera from 2012). I ended up getting Lumix S5 (mark 1, from 2020) for like 500eur that is very different beast of a camera while having L-mount which is becoming only "open" camera mount (third party lenses are being "disallowed" by most manufacturers now).
Get a used 6D or 70D and a 50mm f1.8 lens.
With the 200eur budget you wont make the 6D but you might 70D maybe even most of the lens. I would pay a bit more and get 6D though.
I got a 6D from facebook marketplace for ~125 USD, but it was in the US, which has a more 'liquid' second hand electronics market in my experience.
I've been hunting 6D for a while in eu for 200eur you get like 400 000 shutter count. But 6D is also in that range where people don't care about it and often sell it as some junk. So one might get lucky.
Go to your equivalent of estate sales.
Is the SLR part a requirement? Would a mirrorless camera work for you?