WEBVTT 00:00:06.720 --> 00:00:12.480 wow this look inside the Apple Vision 00:00:10.080 --> 00:00:17.600 Pro enables us to learn things that would otherwise be nearly impossible 00:00:14.759 --> 00:00:22.560 without destroying the device but the Neptune CT scanner that we just got lent 00:00:20.199 --> 00:00:27.480 from Luma field does so much more than just spit out a 2D image like this one 00:00:25.000 --> 00:00:33.559 it can stitch together hundreds or even thousands of 2D x-rays into a full 3D 00:00:30.720 --> 00:00:39.320 model come over and take a look oh look at this here's all the little magnets 00:00:36.440 --> 00:00:43.879 for the optional Zeiss lenses here's the worm drives for the automatic ipd 00:00:41.600 --> 00:00:46.920 adjustment we've actually got a really cool one where we isolated those 00:00:45.600 --> 00:00:52.000 components so we can see the whole mechanism and if you look closely in 00:00:48.920 --> 00:00:54.480 this view you can see every single BGA 00:00:52.000 --> 00:00:59.480 solder joint on the chips on the board pretty darn impressive for not removing 00:00:56.640 --> 00:01:05.960 any of the screws which by the way are convenient much easier to find now if we 00:01:02.559 --> 00:01:07.400 wanted to take the iix it approach and 00:01:05.960 --> 00:01:11.960 depending on the tuning which we're going to talk about in more depth later 00:01:09.240 --> 00:01:16.840 we can even zoom in Far Enough check this out check this out look at this 00:01:13.640 --> 00:01:19.799 that we can see this message from our 00:01:16.840 --> 00:01:23.960 sponsor thorum check out their Rings watches and necklaces made from unique 00:01:21.840 --> 00:01:27.840 materials like gibian meteorite and ethically sourced antlers you can check 00:01:26.040 --> 00:01:33.640 them out at the link below and use code LT to get 20% off without looking too 00:01:31.600 --> 00:01:38.079 closely at the spoilers behind me let's do that again but this time with a Sony 00:01:35.720 --> 00:01:42.600 dual sense controller because it really shows off both the power and the 00:01:40.040 --> 00:01:47.520 versatility of the Neptune CT scanner like before we're going to mount our 00:01:44.280 --> 00:01:53.040 device to foam Chuck it in the machine 00:01:47.520 --> 00:01:53.040 close the hatch and enable the 00:01:53.880 --> 00:02:02.399 X-rays I don't think that's ever going to get old once all that's done we get 00:01:59.439 --> 00:02:07.439 this view you can think of this less like a picture and more like a shadow of 00:02:05.159 --> 00:02:12.599 the controller but instead of a light Shadow it's an x-ray Shadow so the 00:02:10.319 --> 00:02:17.360 lighter areas here are where the less dense materials like plastic allow much 00:02:15.120 --> 00:02:23.080 of the X-ray radiation to pass through to our sensor the darker areas like 00:02:20.200 --> 00:02:28.080 these ones are the denser usually metal Parts like these screws or these giant 00:02:26.280 --> 00:02:32.720 vibration Motors that much more effectively block our photons we then 00:02:30.640 --> 00:02:37.519 take hundreds of these images from different angles and upload them to Luma 00:02:35.239 --> 00:02:42.560 Field's Voyager Cloud platform which does a bunch of maths to spit out this 00:02:41.080 --> 00:02:47.560 and I don't think this is ever going to get old either this is so cool this scan 00:02:45.879 --> 00:02:52.519 would have taken a couple of hours and you can actually see here we can flip 00:02:49.720 --> 00:02:58.040 our way through all of the raw x-rays that made up our 3D image then by moving 00:02:55.920 --> 00:03:02.599 the attenuation slider we can look at the different densities of materials in 00:03:00.480 --> 00:03:09.239 the controller for instance if we slide it all the way over we can see the air 00:03:05.760 --> 00:03:12.080 around the controller and then going the 00:03:09.239 --> 00:03:17.680 other way it removes the plastic first and then we can keep on going until it's 00:03:14.519 --> 00:03:19.319 broken down to just steel and solder it 00:03:17.680 --> 00:03:22.720 offers up a number of different views that can be extremely helpful depending 00:03:21.040 --> 00:03:27.959 on what you want to see the one we're looking at here allows us to go through 00:03:24.799 --> 00:03:30.959 the controller layer by layer so we can 00:03:27.959 --> 00:03:33.400 see the exact density of the capacitors 00:03:30.959 --> 00:03:38.280 on the trackpad for example we can see the rolls in the battery something that 00:03:35.519 --> 00:03:43.200 you definitely can't examine safely under normal circumstances and we can 00:03:40.560 --> 00:03:48.720 even look inside the rumble motor to see how it's built that is so cool we've 00:03:47.000 --> 00:03:52.720 been having an absolute Field Day scanning basically everything that'll 00:03:50.680 --> 00:03:57.720 fit in the machine like this Buzz Lightyear from the mom and pop TCH shop 00:03:55.519 --> 00:04:02.519 oh look there's his two infinity and beyond voice box uh we also did this 00:04:00.480 --> 00:04:06.159 high heel shoe when we were trying to figure out how to mount an adorable 00:04:04.000 --> 00:04:11.400 scary mouth to the bottom of it at a ubiquity access point to have a look at 00:04:08.360 --> 00:04:13.480 their antenna configuration and this 00:04:11.400 --> 00:04:20.759 power supply to see if we can use the Neptune to non-destructively inspect 00:04:16.560 --> 00:04:25.040 solder joint quality the answer is sort 00:04:20.759 --> 00:04:27.680 of is but let's back up a bit aren't CT 00:04:25.040 --> 00:04:30.680 scanners for medical applications and why doesn't this one look like a 00:04:29.240 --> 00:04:35.199 doughnut well while the main principles are the 00:04:33.639 --> 00:04:40.560 same between the Luma field and a medical CT scanner there's one main 00:04:38.000 --> 00:04:46.360 difference between the two in the Luma field we rotate the object while in a 00:04:43.680 --> 00:04:54.039 medical CT scanner we rotate the scanner itself also those things freaking rip 00:04:50.560 --> 00:04:56.120 and for a good reason humans are not 00:04:54.039 --> 00:05:00.919 great at staying still for a very long time and this is especially true of 00:04:58.160 --> 00:05:07.199 their heart and their lungs so a human scanner needs to be able to do a full 00:05:03.840 --> 00:05:09.120 360° scan in a fraction of a second or 00:05:07.199 --> 00:05:14.520 the motion blur from your organs doing inconvenient things like keeping you 00:05:11.680 --> 00:05:20.199 alive is going to ruin the image another reason medical CT scans go so fast is 00:05:17.440 --> 00:05:25.400 radiation an x-ray machine creates ionizing radiation which can potentially 00:05:22.639 --> 00:05:29.720 lead to inconveniences like cancer if you're exposed for too long so doctors 00:05:27.560 --> 00:05:34.120 are usually willing to give up some AC accuracy if it reduces how long a 00:05:31.600 --> 00:05:39.000 patient has to stay in there finally since humans aren't particularly dense 00:05:36.919 --> 00:05:43.639 medical CT scanners also aren't particularly powerful the Luma field 00:05:41.479 --> 00:05:52.560 though by contrast is powerful enough to see through steel so uh do not your hand 00:05:49.000 --> 00:05:54.440 here ever at all non-living objects 00:05:52.560 --> 00:06:01.639 though like this screwdriver available on LTT Store.com don't move unexpectedly 00:05:58.199 --> 00:06:03.919 and can be exposed to x-rays pretty much 00:06:01.639 --> 00:06:08.759 indefinitely without concern and the easiest way to keep them in there is 00:06:05.759 --> 00:06:11.160 floral foam as the name suggests it's 00:06:08.759 --> 00:06:15.639 designed to hold flowers but it also happens to work great for industrial CT 00:06:13.280 --> 00:06:21.880 scanning since it's extremely low density makes it virtually invisible in 00:06:18.800 --> 00:06:24.080 our scans one note is that it's very 00:06:21.880 --> 00:06:28.720 important to place your objects at a slight angle because flat surfaces can't 00:06:26.919 --> 00:06:32.840 be seen if they're parallel to the x-rays now that we've got our object 00:06:30.919 --> 00:06:37.800 secured we can place it on the table and begin setting up our scan what we want 00:06:34.919 --> 00:06:43.680 to do is line it up as best we can with our x-ray emitter the x-rays leave here 00:06:41.000 --> 00:06:49.840 and they go across the machine in a cone shape so we get the highest detail very 00:06:47.039 --> 00:06:55.639 close to the emitter right here oh I pumped it with my shoulder 00:06:52.919 --> 00:07:00.039 jeez we get the highest detail very close to the emitter right here and as 00:06:57.759 --> 00:07:05.919 you guys just saw the table can be reconfigured it can move up down left 00:07:03.199 --> 00:07:11.400 right and closer and farther from the emitter the problem with being in close 00:07:09.000 --> 00:07:16.039 is that we sacrifice maximum object size so we could do something like this right 00:07:13.560 --> 00:07:21.520 around here but if we wanted to scan something as large as Buzz here we can 00:07:19.280 --> 00:07:26.520 but he's going to have to go way at the back of the chamber and we are going to 00:07:23.800 --> 00:07:30.840 give up some detail after the x-rays pass through our object they hit this 00:07:28.840 --> 00:07:35.840 piece of carb carbon fiber right here called the scintillator the scintillator 00:07:33.400 --> 00:07:40.240 turns our x-rays into visible light for two reasons the first reason is that 00:07:38.520 --> 00:07:46.280 x-rays are famous for going through things which includes image sensors I 00:07:42.840 --> 00:07:49.720 assume that's bad yes yes that checks 00:07:46.280 --> 00:07:53.000 out and two image detectors for visible 00:07:49.720 --> 00:07:57.120 light are cheap like budget phones have 00:07:53.000 --> 00:07:59.639 three of them cheap so ctil lator it is 00:07:57.120 --> 00:08:05.440 anyway we can guess that the screwdriver is going to look good right about here 00:08:02.360 --> 00:08:07.560 close the door and give her a look see 00:08:05.440 --> 00:08:11.720 what are the odds I nailed this oh I didn't but hey remember 00:08:10.120 --> 00:08:15.720 this neat now that we've ensured that we're 00:08:13.840 --> 00:08:19.960 not going to cut off our object over the entire rotational range we're going to 00:08:17.599 --> 00:08:23.919 go ahead and set our scan parameters to 10 minutes this is going to be a pretty 00:08:22.240 --> 00:08:29.080 rough quick one and then we click optimize and start scan now for most of 00:08:27.199 --> 00:08:34.599 the scans that we do here the machine is going to take 900 individual x-rays 00:08:32.440 --> 00:08:38.839 slightly rotating the object then pausing for each exposure super quick 00:08:37.159 --> 00:08:42.440 scans like the one we're doing right now though are going to keep the center 00:08:40.360 --> 00:08:49.880 table slowly rotating for the entire scan in order to save some time hey it's 00:08:45.560 --> 00:08:52.080 done which I guess raises a question um 00:08:49.880 --> 00:08:55.880 was it safe for us to be standing here while it was doing that aren't x-rays 00:08:54.720 --> 00:09:00.640 you know radiation okay well first I just want to 00:08:58.240 --> 00:09:05.800 clarify that although Luma field does produce ionizing radiation it's not 00:09:03.360 --> 00:09:10.240 radioactive so that means the second the beam is turned off there is zero 00:09:08.000 --> 00:09:14.959 radiation from both the machine itself and from anything that was inside it so 00:09:12.920 --> 00:09:20.800 the only real concern is preventing x-rays from escaping from the machine 00:09:17.680 --> 00:09:23.760 while it's on but as it turns out that's 00:09:20.800 --> 00:09:29.079 pretty simple you just uh cover the whole thing in lead so much lead in fact 00:09:26.880 --> 00:09:35.640 that we had to rent a special heavy duty for lift in order to unload this thing 00:09:32.000 --> 00:09:37.959 at the back is three 8in lead plating 00:09:35.640 --> 00:09:43.160 and then around all the other sides is a/4 in of lead all of which is verified 00:09:40.680 --> 00:09:48.079 during installation to have no radiation leaks there's in fact so much lead 00:09:45.880 --> 00:09:53.640 around this thing that while the scanner is on you can stand right beside it and 00:09:51.120 --> 00:09:58.880 experience a reduction in the amount of radiation that you're receiving because 00:09:55.560 --> 00:10:00.839 it will block radiation from the Sun the 00:09:58.880 --> 00:10:06.079 radio acve materials in the ground below you and even that banana that your 00:10:03.120 --> 00:10:11.560 colleague is eating my devious plan to poison Linus is foiled again and also 00:10:09.360 --> 00:10:17.480 his insides wow all things considered the 00:10:14.399 --> 00:10:19.440 quality of this is pretty okay like 00:10:17.480 --> 00:10:24.640 obviously it's not as good as the longer scans but between the Reconstruction and 00:10:22.440 --> 00:10:29.320 the raw x-rays themselves there's a good chance that even a quick scan like this 00:10:26.920 --> 00:10:35.800 can tell us what we want to know like we we can see that oh a bit is missing so 00:10:33.399 --> 00:10:40.600 we managed to spot that but there's also a lot that this scan can't tell us like 00:10:38.360 --> 00:10:46.360 you can see the plastic around the bits here is just kind of a mess and there's 00:10:43.600 --> 00:10:50.399 no real detail in the ratchet mechanism so let's try this again but this time 00:10:49.040 --> 00:10:55.160 we're going to mess around with the manual settings first the main reason 00:10:53.399 --> 00:11:00.600 that the plastic in our first scan looks so strange is a phenomenon called beam 00:10:57.639 --> 00:11:05.079 hardening the X Source in our Luma field produces polychromatic x-rays which is 00:11:02.959 --> 00:11:09.800 to say that we have a range of low energy and high energy photons that are 00:11:07.440 --> 00:11:14.560 getting shot across the scanner the problem is the low energy photons get 00:11:12.120 --> 00:11:19.040 absorbed first while the high energy photons get through meaning that the 00:11:16.920 --> 00:11:22.560 first layer of material is always going to show up darker than the last bit of 00:11:21.200 --> 00:11:27.320 material which can lead to some of the weirdness that we just saw there's two 00:11:24.880 --> 00:11:31.760 main ways around this the first is to use monochromatic x-rays where all of 00:11:30.120 --> 00:11:38.079 the light that is emitted is at the same energy level but unfortunately achieving 00:11:34.360 --> 00:11:40.279 this requires a synchrotron or a linear 00:11:38.079 --> 00:11:46.639 accelerator why do those sound so expensive because they are so instead 00:11:43.680 --> 00:11:50.920 Luma field simply gets rid of most of the low energy photons by using one of 00:11:48.920 --> 00:11:58.360 the copper filters that you're going to see ah there we go on this wheel right 00:11:54.160 --> 00:12:00.880 here they range from .5 mm to 6 mm 00:11:58.360 --> 00:12:05.839 depending on how hard you want the beam the trade-off though is that the thicker 00:12:02.800 --> 00:12:07.480 the filter the less exposed the image is 00:12:05.839 --> 00:12:11.760 going to be meaning that you need to increase the scan time it's actually 00:12:09.720 --> 00:12:16.800 pretty similar to how an ND filter works on a film camera which if you think 00:12:14.720 --> 00:12:20.880 about it can actually be said for all of the settings on the Luma field let's set 00:12:18.839 --> 00:12:27.600 up the screwdriver scan again but manually this time exposure time 46 00:12:23.920 --> 00:12:29.959 seconds 2 and 1/2 mm and 00:12:27.600 --> 00:12:32.519 scan scan whatever it doesn't really matter 00:12:31.040 --> 00:12:39.000 because we're going to movie Magic this thing one cooking show Trick later we've 00:12:35.440 --> 00:12:41.399 got this there is still a little bit of 00:12:39.000 --> 00:12:45.240 weirdness in the plastic but our better exposure settings have given us a 00:12:43.079 --> 00:12:49.880 dramatically better look inside especially the metal portions of the 00:12:47.079 --> 00:12:54.639 screwdriver we can now see inside the ratchet and we can tell how far the 00:12:52.199 --> 00:12:59.480 Press fit portions are inserted we can also see exactly where the magnet ended 00:12:57.199 --> 00:13:04.279 up in the shaft and we can even see the individual teeth on the ratchet these 00:13:01.560 --> 00:13:09.480 are all things that were not as apparent on a quick 10-minute scan with that said 00:13:07.399 --> 00:13:14.360 for good results you don't always need to spend 12 hours irradiating apart the 00:13:12.480 --> 00:13:19.120 screwdriver just happens to be a particularly difficult example thanks to 00:13:16.760 --> 00:13:23.440 the steel and plastic that are very close together for an entirely plastic 00:13:21.600 --> 00:13:29.120 part you can get excellent results in half an hour with aluminum parts taking 00:13:25.519 --> 00:13:30.639 more like 4 hours and steel okay yeah I 00:13:29.120 --> 00:13:36.680 anything with steel in it is probably best left overnight but this raises a 00:13:33.800 --> 00:13:41.800 big question besides just looking really really cool what is the point of all of 00:13:39.000 --> 00:13:45.560 this and who needs one well the most obvious one is looking at the inside of 00:13:43.800 --> 00:13:52.279 an assembled part for things that you can only see while it's assembled like 00:13:48.760 --> 00:13:53.920 for example the squish of an O-ring or 00:13:52.279 --> 00:13:58.320 how much your battery pins are getting compressed once the cover is closed it's 00:13:56.560 --> 00:14:02.440 also great for validating injection molding or c casting like you could 00:14:00.320 --> 00:14:06.680 check for bubbles inside an alternator casing and you could see the deviation 00:14:04.680 --> 00:14:10.759 between the finished part and what was originally designed in the software 00:14:09.120 --> 00:14:14.720 there is one thing we should clarify though although it's unlikely that we'll 00:14:13.160 --> 00:14:20.079 damage electronics by looking at them in the Luma field it is possible after 00:14:17.880 --> 00:14:25.040 being exposed to ionizing radiation for a couple of hours bits can flip in flash 00:14:22.880 --> 00:14:29.639 memory and if that happens to the wrong bit whether it's the you know storage on 00:14:28.000 --> 00:14:35.560 your phone or the BIOS chip on your GPU whatever that is 00:14:32.560 --> 00:14:37.240 is going to be no worky fortunately Luma 00:14:35.560 --> 00:14:41.519 field has done the research on this and they've got a helpful article on how to 00:14:39.000 --> 00:14:48.040 calculate safe exposure levels to ensure that our phones and our gpus come out as 00:14:44.279 --> 00:14:49.839 happy as clams as for exactly what we 00:14:48.040 --> 00:14:55.839 are going to be using the Luma field for well I don't know guys the sky is kind 00:14:52.639 --> 00:14:58.040 of the limit here or well roughly the 00:14:55.839 --> 00:15:03.680 size of a basketball is the the real limit oh and also materials um gold it 00:15:01.519 --> 00:15:07.240 turns out is just too dense to blast through but neither of those limitations 00:15:05.680 --> 00:15:12.440 have stopped us from making good use of it already things we've scanned so far 00:15:09.759 --> 00:15:17.680 include this bread sorus prototype that we looked at to validate the factory 00:15:14.079 --> 00:15:19.680 glue application this massive Phillips 00:15:17.680 --> 00:15:24.160 head bit that we printed on our ridia metal 3D printer the Luma field allowed 00:15:22.079 --> 00:15:29.680 us to verify both that the dimensions are correct and also that the part is 00:15:26.040 --> 00:15:31.519 mechanically sound outside and in oh we 00:15:29.680 --> 00:15:36.680 used it to look at the hinge mechanism of the pixel fold we scanned a handful 00:15:33.800 --> 00:15:42.360 of computer mice which often you can't see the inside of without destroying and 00:15:39.399 --> 00:15:47.319 this scan of a GPU is already live on the LTT Labs website giving us a great 00:15:44.720 --> 00:15:51.800 look at the processor die size the memory chip layout and the cooler design 00:15:49.880 --> 00:15:56.639 without needing to remove the factory heat sink which could invalidate future 00:15:54.240 --> 00:16:02.240 thermal testing what's really turning heads though is not the functionality CT 00:15:59.440 --> 00:16:10.000 scanners have existed for decades it's the price now it is a subscription I 00:16:06.240 --> 00:16:12.639 know I know but the Neptune starts at 00:16:10.000 --> 00:16:17.040 $75,000 a year and includes their Voyager software that handles all of the 00:16:14.920 --> 00:16:22.199 Reconstruction that sounds really expensive because it is but it starts to 00:16:20.199 --> 00:16:25.920 sound a lot more reasonable when you consider that a lot of industrial CT 00:16:24.519 --> 00:16:30.639 scanners are going to require a full-time technician to run them who's 00:16:27.959 --> 00:16:36.160 going to put you over 7 $5,000 a year before you actually buy a quar million 00:16:33.040 --> 00:16:38.160 doll machine also I've said tons of 00:16:36.160 --> 00:16:42.360 times in the past that while I don't like subscriptions I don't mind them as 00:16:40.480 --> 00:16:49.160 long as they give me the option to buy the machine outright and Luma field does 00:16:45.560 --> 00:16:51.000 the price is uh well if you have to ask 00:16:49.160 --> 00:16:55.079 but it's intended for clients who are working with sensitive information and 00:16:53.000 --> 00:16:59.360 who want to do all of their processing in-house that actually feels like it'd 00:16:57.199 --> 00:17:04.199 be a really cool future video a GPU F to freak process this thing 00:17:01.959 --> 00:17:07.199 anyway it feels a little strange saying we're going to have a link below to 00:17:05.640 --> 00:17:12.480 where to get one of these for yourself but some of you for sure work in R&D or 00:17:10.520 --> 00:17:16.679 in manufacturing and are probably already thinking about how much time 00:17:14.959 --> 00:17:20.240 this thing could save you and then I guess for everyone else I don't know 00:17:18.199 --> 00:17:23.600 we'll link some of our scans so you can check those out and in the comments 00:17:22.000 --> 00:17:27.120 below let us know what you want to see us scan next just like I'll be scanning 00:17:26.120 --> 00:17:31.440 The Horizon for our sponsor 00:17:29.400 --> 00:17:35.559 thorum thorum has been in business for over a decade handcrafting elegant bands 00:17:33.679 --> 00:17:39.440 and rings out of rare materials with rings made out of titanium Ironwood 00:17:37.640 --> 00:17:43.840 World War II rifle stocks and even ethically sourced antlers they've got a 00:17:41.280 --> 00:17:47.240 ring for every taste don't eat them though each ring comes with a free 00:17:45.400 --> 00:17:51.200 thorum silicone activity band and an American Walnut Wood ring box their 00:17:49.520 --> 00:17:55.480 minimalist watches feature the same unique materials like Hawaiian coawood 00:17:53.440 --> 00:17:59.320 and gibian meteorite you know like the kind from space and they even offer 00:17:57.559 --> 00:18:04.000 necklaces now to match the same minimalist and premium Vibes best of all 00:18:01.880 --> 00:18:07.360 they ship to most areas worldwide and all of their products are covered with 00:18:05.320 --> 00:18:11.640 their lifetime warranty with over 10 years of experience and over 5,000 happy 00:18:09.720 --> 00:18:15.720 customers thorm is there for you whether you need a watch a wedding band or just 00:18:13.880 --> 00:18:21.960 a cool looking ring check them out at th.com and use code LT to get 20% off if 00:18:20.440 --> 00:18:25.600 you guys enjoyed this video check out the last time we got a super expensive 00:18:23.720 --> 00:18:31.360 piece of test equipment by watching our Emi chamber video actually we've got 00:18:28.840 --> 00:18:34.280 some upgrades to share on that front uh sometime in the near future you probably 00:18:32.720 --> 00:18:37.960 don't recognize any of that gear on the table outside it