Raspberry Pi Cameras - The BIG Picture

today in the workshop we're taking an in-depth look at the Raspberry Pi camera we'll see how the camera works what models are available and we'll learn how to use them we'll also do a performance test to see which one looks best we're getting the big picture today so welcome to the workshop [Music] well hello and welcome to the workshop and today we are going to be working with the Raspberry Pi camera now every Raspberry Pi produced in the last seven years has had a CSI or camera standard interface connector on it and this has allowed you to connect the small camera to the Raspberry Pi now unlike a USB camera the Raspberry Pi camera is connected directly to the Broadcom GPU on the Raspberry Pi board and therefore it doesn't need to use the CPU to process the camera data and it is much much more efficient and in the seven years since they first added the port to the Raspberry Pi the Raspberry Pi foundation has released a total of three cameras including the very recent Raspberry Pi high quality camera in addition there are a number of different cameras you can get from different manufacturers that are suitable for use with the Raspberry Pi so today what I want to do is cover everything we can about the camera I want to see how the camera works how you use the CSI port I want to see how you can use the camera both at the command line and in your Python programs as well as with other applications and I want to take a look at a number of these cameras including the new Raspberry Pi high quality camera and see how it performs and so let's get started by learning a bit about Raspberry Pi cameras the Raspberry Pi camera was first introduced in June of 2013 the camera attaches directly to the GPU using a 15-pin connector on the Raspberry Pi board this allows the data to stream directly through the GPU bypassing the CPU and improving performance the GPU encodes h.264 video and JPEG images there are many compatible cameras available for use with the 1510 connector the first Raspberry Pi camera module version 1 was released in 2013 it was available in two models one with an infrared filter and an Noi R noir version that has no infrared filter module v1 used an omni vision ovie five six four seven sensor this provides a resolution of five megapixels the sensor has an area of three point seven six by two point seven four millimeters camera module version one has now been discontinued by Raspberry Pi however this module is available from several other vendors including variations that include different lenses camera module version 2 was released in 2016 like version 1 it's also available in an infrared filter mode and a no infrared filter mode version this camera used the Sony IMX 2 1 9 sensor this sensor provides a resolution of 30 to 80 by 24 64 pixels which is 8 megapixels the sensor area is nearly the same size as version 1 3 point 6 8 by 2 point 7 6 millimeters this camera is still available officially from the Raspberry Pi foundation for 25 US dollars there are also versions of this camera available for many other vendors the Raspberry Pi HQ cameras released in 2020 currently there is only a version with the infrared filter installed although there are instructions for removing the infrared filter for those who don't want it this camera uses a Sony IMX 477 sensor it has a resolution of 40 56 by 30 40 pixels which is 12 point 3 megapixels the sensor area is quite large at six point two eight seven by four point seven one two millimeters the camera sensor retails for fifty US dollars the high quality camera comes without a lens it requires a lens that uses a seamount the camera is also provided with a CS mount adapter to allow additional lenses to be used currently the Raspberry Pi foundation has a second party manufacturer who can supply six millimeter and sixteen millimeter lenses for the camera the module has a quarter inch tripod mount note that the version one in version two cameras used fixed lenses all modern raspberry pies of a 15-pin connector for the camera you'll note the location on both the Raspberry Pi 0 and Raspberry Pi for this connector used the CSI or camera standard interface electrical protocol this is a protocol that has been used for cameras and cell phones and tablets the PI 0 has a smaller connector and you will require an adapter cable to use a PI 0 with the camera the camera cable is balanced and therefore does not require any shielding in its basic form a camera is simply a box that consists of a lens that focuses light onto a sensor any image that the lens can see will therefore be reproduced on the sensor the sensor consists of several elements called pixels these pixels have individual micro lenses to improve their sensitivity early electronic camera designs used CCD or charge coupled device sensors these days most cameras use CMOS based sensors as they are more efficient than CCD sensors internally they use filters or layers of silicone to separate the red green and blue colors now not all the colors of equal sensitivity to light and it should be noticed that the red is also sensitive to infrared unless it is filtered the cameras internal circuitry compensates for these color differences the pixels in the camera are arranged on a grid this grid is typically called the frame a larger frame equals better performance and more pixels will give you higher resolution now the data from the camera must conform to standards for image standards we use the JPEG standards video standards are determined by the h.264 standard the camera can provide images in the JPEG GIF bitmap PNG raw and other formats for video there's only one format raw h.264 video the h.264 standard also specifies the frame rate and it specifies the size both the horizontal and vertical dimensions for 1080 video we have the size of 1920 by 1080 which requires a sensor of approximately 2.1 megapixels 4k video requires a sensor that can do 30 480 by 2160 and this requires at least an 8 megapixel sensor now as both the version 2 and high quality cameras have enough megapixels to do 4k that might mean we can do 4k video with our Raspberry Pi correct incorrect in the Raspberry Pi the camera data is processed by a Broadcom video core for graphics processing unit the video core for GPU only supports h.264 encoding and the maximum resolution for video is 1080p at 30 frames per second in order to do 4k video you would need to use an H dot 261 coder but as the encoder is embedded into the video core for GPU this can't be changed another feature supported by the sony IMX for 7/7 sensor in the new HQ camera is HDR or high dynamic range the camera actually provides two signals to improve performance unfortunately HDR is not supported by the video core GPU hopefully the next iteration of the raspberry pi will use a new GPU with improved performance now there are many camera manufacturers who have taken the version 1 and version 2 modules and added lenses to them the lens can greatly affect the quality of the image the infrared filter on some models improves visible light performance but of course precludes using the camera for security applications in the dark now when you're purchasing a Raspberry Pi camera make certain to get an actual Raspberry Pi camera and not one intended for the Nvidia Jetson there are now several cameras available for the Jetson for which there are no drivers currently for the Raspberry Pi so now that we've learned a bit about these cameras let's take a look at some of them now here are a few cameras that you could be using with your ass berry pie and there are several more models of cameras that you could choose from as well now this is the original camera the version 1 camera of the Raspberry Pi the 5 megapixel version and although the Raspberry Pi organization no longer sells us on their store it's still readily available on eBay and Amazon from a number of different manufacturers and it's quite an expensive as well now this is another version one and this one has been made for the Raspberry Pi 0 and you can tell because it's got a very small connector on the end as opposed to the connector on this one here because the PI 0 uses the same connections but it uses a smaller connector otherwise this is the same camera now this is a very interesting little one for the PI 0 that I got from PI Moroney and this is about as small as you can get it's made exclusively for the PI 0 you can't adapt it to the other ones unless you have an adapter on this side but the idea of this of course is to make it very small and so they've got the camera module itself over here it's the same five megapixel module but they've got the electronics over here which is really cool they've put it right on to the ribbon cable and so this is about as tiny as you can get for the PI 0 now if you want to move up a step this is the version 2 of the camera from Raspberry Pi and this is the noir version DN o our version meaning there's no infrared filter on oh that was like that name because of course NIR is noir which is the French word for black and so I think it suits very well and so this is the 8 megapixel camera from the Raspberry Pi corporation now this camera over here is some argued cam and it's really nice I just picked this up recently this is the same sensor that the Raspberry Pi version 2 camera uses the same Sony sensor the 8 megapixel one but it's got a type CS lens on it over here and so you can interchange it with other types es lenses that's one of the tooth types that you'll find for security and instrumentation cameras the other type being type C and so this is very nice it's basically an improvement upon this camera with a better lens now speaking of both cameras with lenses of course this is the Raspberry Pi high quality camera or at least this is the camera module itself you can see it comes with the module and the module has a thread on the bottom of it so you can use a tripod which i think is wonderful and if you take this off over here you can actually look inside if you can see that at the sensor in the camera I don't want to actually touch that but you need to add a lens of this camera you cannot use this camera just as it is right now and so up over here I've got a lens for it this is one of the ones that recommend it's from a company called Cpl electronics and this is the the bigger zoom lens and it's a 16 millimeter lens and this is a type CS adapter now this camera actually uses a type C adapter but the camera also comes with a C to CS adapter that you can use to adapt this lens interestingly enough since it's the same type of lens I could probably swap it on to the argue cam as well to get an idea of what the difference is between the twelve point three megapixel sensor on this one and the eight megapixel sensor on this one but there you go a few selections of Raspberry Pi cameras for you so now let's go and start to work with them now installing the lens onto the Raspberry Pi high quality camera module is actually pretty simple the module has a cap on it right now to protect the sensor so you'll need to unthread that first take that off and try not to get your fingers or anything into there and try to keep dirt out of that as well now this lens that I'm using over here which is the 16 millimeter lens it has a zoom function on it is a type CS lens and this is threaded for a type C but they come with this C 2 CS adapter what you need to screw in first for this lens of course if you're using a type C lens you don't need to do this and so you put that in again it as tight as you can and then on this side over here we just pry the cap off of it and we can thread the lens into it it's really just as simple as that and again you want to get that threaded all the way and there we have a Raspberry Pi high quality camera we can adjust the f-stop we can adjust the zoom and of course change the focus on it and probably get some great images out of it so now let's go and check it out so I've got my camera all assembled and it's ready to connect to my Raspberry Pi now the procedure here is the same for any camera and any Raspberry Pi before we do connect the camera however I'm going to change its ribbon cable because I want to be using this tripod over here and this isn't going to be long enough and you can get these ribbon cables and all sorts of links I've got a really huge one over here which is probably bit open overkill another long one here this one here looks a good size and I'm going to use that I also have some short ones and what I want to show you over here is this one over here this won't work in this application but this is specially for the Raspberry Pi 0 because the 0 has a smaller connector so in that case you'd connect the camera to this end and the Raspberry Pi 0 on this end but as I'm using a Raspberry Pi 4 I'm going to be using the standard camera now these connectors are very simple to work with really and all you need to do is pull back on a couple of tabs you can use pliers but you can also do with your fingers so if you just grab the tabs like and pull them out you can then just remove the cable and just note the orientation of the cable the blue side is the non insulated side and the conductors are on this side now except for the Raspberry Pi zero it doesn't matter which end of the cable you connect to the camera and it goes back in the same way you just sort of slide it into there and then press these down and then that secures it and on the Raspberry Pi board itself it's the same deal you pull up on this little tab here it's loosen it and in this case the conductive side is going to face the back or the microSD side card so this blue one will face the front where the USB ports are and then you just insert that down into the camera and then push those things down I don't have that in write do it okay there's oriented better and it's pushed down and that's it and that's all it takes actually to connect these cameras so they can be connected and disconnected very quickly so now that we have our camera connected to the Raspberry Pi I'm going to hook a video monitor and Ethernet and keyboard mouse up to it and then we will see what we need to do to actually make the camera work so now that we're all experts on the Raspberry Pi camera and we've taken a look at a number of camera modules it's time to start using the camera so what I'm going to do is I'm going to hook it up to a Raspberry Pi I'm going to show you some of the settings you need to change on the Raspberry Pi to accommodate the camera and then we'll see how we can use our camera both at the command line and within our Python programs so let's get started now after hooking our camera up to the Raspberry Pi there is one more step we need to do before we can use it and that's to enable the camera interface so on your asper a PI desktop go up through the start button and go into preferences and go to Raspberry Pi configuration when you're in configuration go to the interfaces tab now by default all of interfaces will be disabled I've enabled my camera interface by clicking the radio button here and that's what you're going to need to do as well after you do that click the ok' button and when you do you're going to get a message that says that the system will need to reboot for the change to take effect so reboot the system and get back to this desktop with the interface enabled once you've done that you can open up a terminal and in your terminal type in rasp I still - OH test jpg and what this is is we're using the rasp I still utility which takes still pictures and we're pumping the output which is the - OH - a file called test JPEG so we'll hit enter on that and we get a little preview window up and you can see me in the corner here and some of the tools on the pegboard and now we've taken our picture so if we go into our file manager now we'll see we have a test dot jpg file in the root of our PI directory and if we take a look at that picture it is indeed what it is that we saw in the preview so this is a very simple method of testing out the camera the rasp I still command is actually a very versatile command that it has a lot of different parameters you can add to it to control the taking of pictures with the Raspberry Pi from the command line now let's go and add a parameter to our line our line as you recall says rasp I still - oh test JPEG which means that we're sending the output to a file called test on jpg let's go and see if we can improve this now you'll remember that there was a small preview window and it was rather brief and didn't give me enough time to get the picture set up so we can improve upon that and increase the size of the preview window with a - T parameter so we can do - T and then give it a time in milliseconds 10,000 should be 10 seconds which is enough time to set up the camera and then we can give it to - oh and we'll send it to another file I'm going to call this one test weight JPEG and we'll hit enter and there we go and Here I am on the side of the picture and gives me time to sort of get set up if I want a picture of my ear over here I'll get it in camera and I got ten seconds to do this and there we go and if we look down in our file manager we have test weight JPEG with a glorious picture of my ear now there are many other parameters you can add there a spy still as well and if you want to see all of them but what you need to do is just go into the help screen now if you take a look at this help screen you can see the number of different parameters that you can add to a spy still to control the taking of pictures so there really is a lot you can do over here you can affect the quality of the pictures you can affect the type of preview you can affect the size etc there is a lot you can go through over here now if you want to take video there's a similar command called a spy vid and we can do a - oh and you'll notice that I ended it with a dot h.264 because this is H dot 264 video format so let's just do a quick test video and here we are with our test video ok a very quick video and if we look in here we have test 8 - 6 4 and we play it back and there's our wonderful video all right now H dot 264 is a format that not all video players can handle in that case you can add a utility in order to change these into mp4 and if you want instructions for that please follow the article in the drôme bot workshop comm website it'll give you details on adding a utility that will allow you to convert your h.264 videos directly into mp4 if you want to see all the parameters for a spy vid once again you can do help and you'll see that it has a similar long list of parameters as rasps I still did and so using these two utilities you can control your ass Barry PI camera directly from the command line now another way of controlling the Raspberry Pi camera is through your Python programs and this is a very powerful technique allows you to automate a number of different camera functions now I've got four different examples to show you today and you can get these examples if you wish by just visiting the article that accompanies this video on the drone bot workshop comm website and you'll find a link to that article right below the video let's take a look at our first example we're going to start off by using the PI camera library which is what we need to use to manipulate the camera we also will import the sleep function from the time library and this is just to provide a little time delay now we start off by defining an object called camera it's a PI camera object and we do a start preview we will sleep for 10 seconds and stop preview so all this function should do is basically give us a preview for 10 seconds we're not actually going to record any still pictures or video so let's just run this right now and here we have our preview fullscreen and you can see me and you can see my ear and it's pretty wonderful and that to go on for about 10 seconds and then disappear and it does okay let's go take a look at our next example now this one is very similar to the last example we start the preview for five seconds but then after that we do a capture and we send it off to an image file with your calling image JPEG and we've given it the whole path over here so we'll end up in our home directory which is the PI directory on this machine and so let's go and run this one and now we get our preview I can back up for the picture and there I have a picture and if we go into our PI directory we should see it there image JPEG and there's my picture ok my third example over here is actually what you would do if you wanted to take a series of pictures and in this one again we import the same libraries we start the preview and then we go and we step through a range of five and we sleep for five seconds and then we do a capture and we change the capture name every time we do this variable I is going to be replaced over here so we're going to get at image one two three four five and so let's go and run this program right now and here we are a preview for five seconds and a picture and another preview and another picture boy we're getting a lot of pictures of my head in my tool box here's a good one on my ear there you go okay here's a good focus on this homemade haircut of mine and finally I'll back up so we can get a picture of the tools over here and there we go and now if we go and take a look in our directory we'll see image 0 1 2 3 & 4 over here and these are indeed the pictures that we took in sequence and so this is a way that you can schedule taking a picture at a regular event now of course we can also do video with Python as well and here we're just importing the PI camera library no need to import the time library the sleep function and we're setting up an object called camera we set the resolution on this this time when we're doing 640 by 480 but of course you can change that we start the preview and then we'll start recording and that start recording function will record video we're going to pump this one out to video h.264 and then we're going to wait for 10 seconds so we continue wait recording will keep us recording for 10 seconds then we stop recording and we stop to preview and so here we go it's from that and you'll notice a different size now I'm 640 by 480 and we're going to take a little video of this and eventually it disappears and here's our video in 640 by 480 and there you go so as you can see there are many ways of controlling the Raspberry Pi camera through your Python programs now being able to use the camera from both the command line and from within your Python programs really expands a number of possibilities for the Raspberry Pi camera but it might be that you're not looking to use it within your programs you may just be looking for an application that you can use to take pictures or to take videos or to get the best performance out of your camera now as it turns out there aren't really that many dedicated applications for the Raspberry Pi camera and most of the dedicated ones were written many years ago when the camera was brand new and they really lack a lot of features however I did find one called Pi camera app that seems to be very full-featured and has been updated relatively recently so let's go and install the PI camera app on a Raspberry Pi and see what capabilities it gives us now the PI camera app was contributed by user Bill Williams 1952 on to github and you can just go to the github page for the PI camera app you'll find a link to that in the article company this video and once you're there just go over to code and go to download zip and click on that you'll notice down in the corner that you are downloading and now have downloaded the zip file so we can close our browser right now we can go to our downloads directory and here is the PI camera app master zip and so we are going to do extract 2 I'm going to pick a location by clicking on this folder and I'm just going to extract it into my home directory over here so I'm going to take it into the home directory and do an extract then go into my home folder and here's PI camera app master and if I go into that I see another folder called source now right click on that folder and do open and terminal and this will open your terminal already in that directory and that's a handy way of opening up the terminal in a specific location from the file manager now what we're going to do is type the following sudo python 3 sudo python 3 PI camera app PI in order to run this and as you can see up in the corner we've got kind of a a monitor on our camera and it's not actually I don't know why it's in the corner actually I'm sure it's not meant to be like that but I've noticed that on a couple of raspberry PI's I've run it on a game this is very old software and I can move the camera around and you can see here I am with the camera now let's just take a look at this GUI over here as you can see we have a number of different things like this is this the basic paddle over here and we can set all sorts of parameters but look at things we can do exposure we can do advanced parameters over here this is quite something you got a number of different menus for photos taking a picture obviously saving a video taking videos toggling video on and off so what I've noticed about this is that most of these features are essentially the same features that you will find if you do a help on the screen when you are taking pictures manually with the Raspberry Pi at the command line and so these are basically the same functions brought out onto a GUI for you so if you like GUI interfaces and you don't mind having this little preview window up in the corner this pie camera app is something you might want to give a try now here's an interesting situation I have a Raspberry Pi version 2 camera another one of the noir ones and it has a defective sensor what's actually defective about it is the sensor is working but the lens has gotten all scratched up over here so the camera isn't really very useful as it is however it's not a throwaway you can buy the modules the module on the camera is actually just this little piece on the front and you can get it off with just a knife and pull it off of its connector and as you can see you can connect another camera to it and there are several sources of getting cameras eBay and Amazon or good places for the camera modules and here's one that I've picked up now this came from Amazon it's made by saying smart and it's the exact same sensor it's the same Sony 8 megapixel sensor except this one as you can see has some additional lensing on it it has an adapter for an m12 type lens and this is actually a fisheye Glen's a 160 degree of view lens that they've got already that comes with it but you can add extra lenses because of the fact that these m12 lenses are pretty standard and commonplace and these are very easy to put on now what I'm going to do first is I've got some isopropyl alcohol over here and I'm just going to clean this off a little just where the residue is where the other sensor was stuck and then as you can see on the back of this since you're here they've got some adhesive hopefully that'll come off okay I don't really have much for fingernails there we go okay and what you want to do is you want to get the connector on first you get that down and then this press this down and there you go that's really all that there is to it and now I've resurrected this 8-megapixel Raspberry Pi camera the version 2 camera with a addition of a better lens on it so we can add this to the cameras that we're going to be testing today all right now we come to the fun part of our video not that it hasn't been fun so far but this is the part where we get to compare the images and the videos taken with a number of these different Raspberry Pi camera modules now in order to give ourselves a reference point I'm going to bring another camera into the picture as well and it's going to be this one over here this is a Canon EOS m100 and this is a mirrorless camera so it gives pretty well the same quality of image as a DSLR but in a much smaller package and it allows you to use interchangeable lenses as well now obviously this is a proper camera and you would expect it to outperform the Raspberry Pi camera but I'd like to see how close the PI camera can come to this especially the new 8 Q camera to see if we can get anywhere near the same image that we get on the Canon camera now in order to do my tests I've spared no effort in creating some very scientific camera test targets I'm joking of course what it did is I slap a few pieces of wood together and I made this target over here which I've printed out and this will test some colors and on the other side of my piece of wood I've got this target over here that I have printed and this will test resolution etc now obviously the color test and resolution test are only as good as the printers that I printed those charts on too but again since we're comparing everything to this we should be able to get an idea of how the different cameras perform related to each other another thing I want to do is do a video test and for that video test I'm going to be the subject and along with myself I'm going to use this and I think I've showed you this before in my light hour video this is a test target that I use when I'm setting up my video since I have to be both the camera person and the person in front of the camera and I can't be at two places at once I put this at the spot I'm going to be standing and I focus on it and that way my camera should be relatively in focus and so we're going to use my target friend and myself to test videos with these cameras so let's go and compare a few raspberry-pi camera models and see how they perform now here are the cameras were going to be using in our test today and we've seen all of these cameras already of course here's the Raspberry Pi high quality camera with the 16 millimeter lens the argue cam this is the 8 megapixel camera that also has a CS mount lens on it and it's a fisheye lens on this one here the Raspberry Pi 2 camera the noir camera the Raspberry Pi one camera the 5 megapixel one the Raspberry Pi camera that we just modified to have the M 100 lens and of course my Canon EOS m 100 camera now the Canon camera and the Raspberry Pi high quality camera can be mounted on a tripod which I want to use for my tests the other cameras can't by themselves so I've created this little gig over here I hope you can see that it's got a series of standoffs on it and I can either mount a Raspberry Pi 1 or 2 camera onto here or I can mount the argue cam onto it and on the bottom of it over here I have a quarter inch thread this is actually just a GoPro adapter that I've screwed down to this perf board that I've got these pieces on so this is how I'm going to test everything out now another thing that I've got over here is a desk lamp over here and this is one of these lamps is also a magnifier but it's an LED one it's not a fluorescent one and the lighting in the workshop is also all LEDs so for the lights on test I'm going to leave all the lights on in the workshop now these are supposed to be 4000 degrees Kelvin lights at least that's what they are advertised as when I bought them I believe that's also the same color frequency of this LED light and this is this going to be shining on the target just to illuminate it a little better and every camera will be working underneath the same conditions now for the low-light conditions I'm going to turn off all the workshop lights but I'm going to open the doors to the workshop and let the ambient lighting from the other room come in and that will test to see how the cameras doing a little light so let's get at our tests I'm going to go and do them right now and I will show you the results now the first test is the color test with full lighting and this is the Canon EOS m 100 our reference camera under those conditions now for those of you who are into photography I had set the ISO at 400 because there's quite a bit of light over here and they also set the f-stop at 5.6 for this and it is focused manually now let's compare it to our first camera now this is the Raspberry Pi high quality camera and as you can see it produces a very vibrant image the colours are very nice the focus is very good the whites are a little bit different than the Canon and we can see that over here where we compare the two but otherwise the colors are equally vibrant the hues are a little different but it's not to say that one is better than the other all in all it's a very nice picture from the high quality camera now this is the argue cam the 8 megapixel Rd cam and this camera has a bit of a fisheye lens and so that's why you're seeing the distortion on the screen over here also what's interesting about this camera is compared to the other ones that picture is upside down in relation to the way that the cable is inserted into the camera obviously that's not a big deal because just inverted and software now the picture quality is pretty good the colors are pretty nicely balanced remember this is using the same eight megapixel sensor that the Raspberry Pi version two camera uses but it's got a type CS lens on it and so the optics are different and here we compare it to the base camera the canon eos camera and as you can see the colors are pretty good i notice on the lens with the bend of it that the edges the focus is a little bit off but it's very good in the center so all in all the colors and everything seem to be quite good with this camera and the light level seems to be pretty good too now this is the Raspberry Pi version one camera and right away you'll notice that this image is not as bright as the other image it's focused nicely in the center but around the edges it doesn't focus too well the black isn't quite as dark as I'd like it nor is the white as bright but let's also remember that this is a five dollar camera when we compare it to the canon camera so you could literally buy a hundred of these little modules for the cost of the canon camera but when you compare the two you can see the colors aren't quite as vibrant and as I said it's the blacks aren't as dark but for a lot of purposes this will be just satisfactory now this is the Raspberry Pi version two camera and this is the one with the eight megapixel Sony sensor and as you can see it's a very vibrant picture it tends to stay and focus throughout pretty well the whole camera range the blacks are reasonably black and the whites are pretty good as well when we compare it to the original reference one with the canon eos you can actually see that it is more vibrant but it also is a tiny bit grainy er but that is to be expected of course because we're comparing an 8 megapixel sensor to a 24 megapixel sensor and finally this is the version two camera that we modified by adding the same smart module and lens on to it so it's the same 8 megapixel sensor with a 180 degree field of view lens and you can see the fisheye on the lens it really gets a lot of stuff in what I really find intriguing about this lens is even when you look at the edges it still seems to be in focus it also has a very nice white and dark balance so this one seems to perform pretty well and compared to our reference again the blacks are pretty well just as good they might even be a tad better and the whites are perhaps even little brighter so this is performing excellently and of course now that it has an m12 lens on it we can exchange it for different lenses if we don't want this fisheye effect so those are the color balance tests now let's move on to the next one so now it comes time for the low-light test and for this I just turned all of the workshop lights off and the only illumination is coming from the two rooms adjacent to the workshop I left the doors open so the light could bleed in so this is about as dark a situation as you'd ever want to use a camera except perhaps for security purposes now here's our reference to Canon EOS m 100 I have to change the ISO in order to allow it to see things at this low light level so the ISO is set to 6400 right now on this so let's take a look at our first camera the Raspberry Pi high quality camera actually performs relatively well at low light as you can see you can still make out the color differences and it's quite easy to read the RGB values on each of the colors the ArchiCAD mate megapixel also performs pretty well in low-light it spoke the same as the Raspberry Pi high quality camera as well and again the RGB numbers are quite legible the Raspberry Pi version 1 camera didn't perform all that spectacularly on low-light it's very hard to actually read the text or the RGB value and of course part of that is because the edge of the screen is still a bit out of focus and it you can discern the colors but they are very very faint the version to camera did a lot better as we might expect since it had such a vibrant picture and you can indeed make out the RGB numbers although they're on the edge of legibility the colors can still be made out quite easily here the version to 1 that we modified with the lens actually didn't perform too well in low-light I'm not certain if it has something to do with the lens because of course it's the same sensor as the last camera we looked at you can make things out but it's very difficult to read the RGB values and it does seem that the lens is cutting down on a bit of light our final still picture test is the focused test or the black-and-white test and so here's our reference with the Canon EOS m100 now as you can see the Canon tends to produce a relatively cool white picture I should notice also that I didn't intentionally set the white balance of the camera to the target so I may have been able to get a clearer picture had I done that but since I wasn't able to set white balance for the other cameras I didn't think it would really be fair now here's the raspberry pi HQ camera and as you can see it produces a warmer white and otherwise the focus and everything is just great with this the picture is not grainy or anything it seems to do well with this level of light now when we compare it to the Canon we can see the difference in the white balance between the two different cameras but as far as performance goes on the chart it does seem to work very good all the small little gradients that you see on this chart are equally reproduced with both cameras now this is the art do cam eight megapixel camera and one thing about this camera that I noticed right away is that the center area of the area that's the boast and focus also seems to have a little bit of a bluey green tinge to it as compared to the other area which seems to be a warmer white now as I noticed before with the type of lens they have on it right now which is a wide-angle the picture around the edges tends to be a little bit out of focus and you can plainly see that on the focus target now when we compare this to the canon camera again you can see the difference in the white balance and you can see how it changes through the RG kam camera now this may indeed have something to do with the lens the Raspberry Pi version one camera a gain you'll note is comparatively a darker image even though there is quite a bit of light on our image also at the edges it tends to go a bit out of focus although it's not terrible and certainly usable it is also gravitating a bit toward the blue and the green end and you can see the difference between the brightness on the image when we compare it to the Canon image it's quite apparent over here now the Raspberry Pi version 2 camera produces a darker image it also seems like it would like a little bit more light however the image otherwise is quite nice the focus is very good across the image and the color does gravitate a little bit toward the red end of the spectrum but not too much it's not bad and you can see the difference when you compare it to the Canon camera over here it's not quite as bright and vibrant but otherwise it's a nice picture and it's a fairly clear picture and you can see all the little gradients on the chart so it's a very good focus picture as well and finally the camera we modified with the 8 megapixel sensor in the M 100 fish eye lens and as you can see with 160 degree lens it really is a distorted picture but you would expect that what is interesting though again is that even at the edges the focus seems to be quite good and the white balance seems to be pretty good right across the image the image does tend to have a darker and a bit more tending toward the red on the whites and that's similar to what we saw with the 8 megapixel sensor in the version 2 camera and that makes sense is the same sensor and we can compare that with the Canon camera and see the difference there so now that we've seen some still pictures let's take a look at how our cameras performed in a video test now this is the image that's the reference image the one from the Canon m100 and as you can see this is pretty well the image that you would normally see when you're watching me in a drone BOTS workshop videos so you're pretty familiar with the background I would say if you've been watching the channel for a while now we're going to look at the other videos keep in mind with different lenses we're going to get different fields of view than what we have in this reference video now the first video is a Raspberry Pi high quality one and as you can see it only gets a small segment of the video and it's very very clear picture as you can see from my smiling face and the piece of the test target that you can see now the argue cam of course distorts things at the corner but you don't see it as much because of the resolution we're using here right now there's a little harder to focus but otherwise the picture wasn't too bad the PI version one camera of course is a little bit off in color but actually this picture isn't that bad when you consider while looking at a five dollar camera I found it very hard to focus the version two camera and of course it's very bright keep in mind this is the Noy our version of the camera so that may be the reason why the picture is so vibrant and so finally the version 2 camera with the fisheye lens and as you can see it really brings in quite a lot of the image because of the 160 degree field of view lens and that's a comparison of the Raspberry Pi cameras in video mode ok well that wraps up our look at the Raspberry Pi camera for today I hope that you enjoyed it and that it's inspired you to start using the Raspberry Pi camera in some of your own designs now if you need a bit more information about the camera or if you want to grab the Python code that I use today check out the article in the drone bot workshop comm website and you'll find a link to that article right below this video you'll find another link below the video as well and that's the link to sign up for the newsletter so if you haven't done that yet please do it this is my way of keeping in touch with you and letting you know what's going on here in the workshop there's even another link below that yes we have lots of links today and that is - the drone bot workshop forum and you can discuss the Raspberry Pi camera Arduino zsp 32s or any other technical subjects with a bunch of great people who have similar interests to you so the forum's a great place to hang out and finally if you have not yet subscribed to this YouTube channel I would be very very appreciative if you would do that just please click on the subscribe button and also when you do that click the bell notification and that way you'll be notified every time I make an video so until we see each other the next time please take good care of yourselves please stay safe and we'll see you again very soon here in the drone bot workshop good bye for now [Music]

Channel: DroneBot Workshop

Views: 373,382

Rating: 4.915884 out of 5

Keywords: raspberry pi camera, raspberry pi hq camera, raspberry pi, pi camera, raspberry pi camera python, raspberry pi camera review

Id: MVgr302PNwY

Length: 52min 25sec (3145 seconds)

Published: Sat Jul 11 2020