Open Source Ecology Development Team, 21 Aug 2018

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Development Team Meeting - Aug 21, 2018

Description

See notes for Aug 21, 2018 in the meeting history of https://wiki.opensourceecology.org/wiki/Development_Team_Log

Agenda: Getting ready for OSE Boot Camp - 18 3D printer builds in one day coming up this Saturday - then a month of Immersion Training - first ever program to absorb developers to work with OSE full time.

1. vBOM for OSE Modified Prusa i3 MK2 Extruder
2. Photogrammetry with Sara Bajor - [[Sara Log]] on wiki
3. Arduino power supply with Miles Ransaw - [[Miles Log]]
4. Extruder documentation - Jon Takacs - [[Jon Log]]
5. Power Cube + 3D scanning - Abe Anderson - [[Abe Log]]

----------------------

What you see here at Open Source Ecology is an ambitious program based on a volunteer effort. To help us reach the goals - please consider joining as an OSE Developer in 2017-

http://opensourceecology.org/wiki/OSE_Developers

Take a minute to subscribe to our email newsletter (updates, workshops, etc): http://bit.ly/1LtcM44

A

Starting recording, so this is our August 25th first meeting OSE developer team: this is the the week before game day. The game day is the 25th. That's Saturday we're starting the immersion program boot camp. So that's a lot of activity. That's gonna be going on here and preparing doing mostly inventory right now we're pouring the floor in fab lab for the merging program, as we currently speak right now and looking at.

A

Let me share my screen.

A

And, at the same time, these cantaloupe zarkon's and growing in our aquaponic greenhouse, so I'm gonna, a back I'm Way outside I'm I'm here, okay and when they get right, they drop off to the floor and I eat them, we're working on an OSC marketing. Here we are working on the logo or the 3d printer, open, the micro factory and all the machine. So it's a modular design.

A

What else? What else can I report on I'm just getting ready for the for the immersion I wanted to find out? Let's see so this is photogrammetry here. I wanted to find a picture now, you've seen the hydraulics picture, that's getting buried by concrete right now and we're gonna have heat so we're preparing for your own operation looks like with immersion and our growth. We're gonna, be probably building starting to build again on-site at open source ecology here in terms of additional facilities to make a world-class research and development center for open product development.

A

So, let's see I got plenty of people joined up okay, so when we go right ahead with actually yeah, let's actually roll today, a part of the deal is that we're getting van up and going and actually got an appointment to get in the van inspected today, so I got to cut out like quarter to quarter to three so just a little short today, but but that's my my my agenda appears been doing some documentation work. This is John is doing work on a extruder of the 3d printer.

A

This is something I did as a little visual ability rows for the extruder itself. I think the extruder itself is perhaps one of the most complicated parts of the tiny printer compared to anything else, I mean the rest of. It is pretty simple. The extruder itself, I would say it's much more complicated and it's something we're borrowing from Prusa i3 mk2, it's a modified version. We modified it slightly for a larger distance sensor, so we can stay further away from the heat heated bed, so we don't crash into it as easily.

A

So that's a design geared for scalability in terms of larger print surfaces. Like say we want to do a 1 meter area, 3d printer. We can do that. We have less risk of hitting the bed with a wood extruder or with the sensor itself, all right. So let's continue. Let's, let's have some updates, maybe Sarah. Since you were excitedly talking about updates on the photographs. Maybe you can continue next see what we've got for your progress on 3d scanning, with a digital camera.

B

A

B

Yeah, the idea is, take some pictures of an object and get it into a forum that could be 3d printed, so Harman had kind of defined or found out a process or software to do that, the first step being to use Cole map we've run into a lot of hardware issues using Cole map to the full extent, but I was able to do to build a Hartman's work is using a Cole map, I generated, what's called a sparse point cloud and then using another tool called open.

B

&Amp;, Bs I made a dense point cloud and once you have that you can take it into a mesh lab to actually generate a mesh that can be converted into something that this is what I'm not right now. But the idea is it'll give us the final product that we're looking for that could be used to 3d print the object back out. This.

A

Call not have you're saying we're giving up on Colma because it's got hardware issues like it's not it doesn't work great with processing processor reserve well,.

B

What I found is without any special hardware, you can use it for first step of the process, but yeah right now. For me, as someone who doesn't have the specialized hardware, it's a three step process. It's pretty involved, so there are ways to simplify it, but from what we know right now, it's pretty Hardware dependent. So.

A

Are you saying that with Cole math it sucks up so much power that you need a GPU to run it.

B

Yeah the feature that generates the dense point cloud, which is like the second step in the process right, you go sparse point, dense point, mesh and I cannot generate a dense point with Cole map because it requires a CUDA enabled GPU. So these are video brand, but.

B

And still wasn't able to get it to work, so the claim is different from our experience.

A

Okay, do you have any that you can share or anything.

B

A

I can show you maybe send a link or share share your screen.

B

B

A

You could share your screen, meshlab, don't think we have meshlab yet and osc linux we're working on an up next, oh I see Leonard Linux release too. So we can have all the new software like for CNC circuit, milling and as well. I was at the photogrammetry right within OST Linux.

B

Can you see my screen yep, okay, so Harmon actually put together photos a photo set of the cordless drill, but I went ahead and used a test data set so that I could be sure that images we're actually working properly and.

C

B

This is some castle and what you're seeing here is the dense point cloud in mesh lab yeah.

D

B

My next steps are going through some tutorials to figure out how to refine this and complete the texturing of it.

A

B

Think that meshlab can do all of that. So I think I'm gonna in a good spot again to make progress and the.

A

You didn't try this with the drill. Cordless drill, beta I have.

B

Not yet, but once I know what I'm doing with, but that's all look at that. Okay.

A

D

A

Great, so what what are your main learnings in this adventure so far.

B

Yeah main learnings I was really amazed by all the difficulties with hardware and installing a Windows machine. I was really unfamiliar with so I had to build open MVS for that program and that took so long. It was just visual studio took like half a day to update, and then there are all these requirements.

D

A

So I mean there's tricks to everything. Yeah.

D

A

Wonder how those guys within was there any comment made by, for example, when Prusa he made that reconstruction of that statue minute video he was so mad. He didn't mention anything about GPUs and and stuff like you didn't raise that like, if you have a GPU or any thoughts on that.

B

I'm not sure it's all about video, ok,.

A

Yeah, it's time to call my page on the OSC wiki.

B

hmm I'm not sure.

D

Yeah, okay, yeah.

A

Cuz cuz, uh when I saw calm at when I sucked. Oh no, it's called the pages called open source photogrammetry on the wiki, which I'm pulling up right now. I can show my screen.

A

Yes open so photogrametry has this nice video there there you have a statue and then you have a three print and it looks like oh no problem so I guess and that there were no warnings about some of the difficulties with hardware, which is it's kind.

C

Of always the case.

A

You know when, when you see something on the internet, like you know, does it really work is always the question of it. You know we always run into issues and trying to address that one by one. Okay. Well, that's good a good progress. Maybe you can teach all of us if you have any more time before immersion. If you can teach us some something about that during immersion, that would be good, yeah.

B

And I'd also definitely like to try out the pipeline on Linux. Instead, we'll see how that goes, yeah.

A

Definitely definitely try that and if you haven't seen, though, have you seen an open-source program a tree page on the wiki? Yes,.

B

A

Yes, there's that a video there but yeah yeah, let's see, let's see what we can do cuz during the program. Perhaps we can. We can see how, for example, it might look to three Scantlin levar, say well, the whole cream printer, for example, or maybe just a part of it where you printer, perhaps is a test exercise but yeah, let's see what what it looks like the idea there being. If we have this on Elysee Linux and the toolchain like the software, we know it works. Everything works the weekend of people during our design.

A

Jams incorporate that as part of our normal workflow that we can teach anybody in a in rapid time to do so. That's that's division so that once again democratizing this process, making it accessible to everybody all right. So, let's continue on a meeting meeting.

C

Agenda here so Eric any updates.

A

There I see there's a tonight.

E

I'm going to the local makers loose, you know looking to see if there's a workspace, that I can continue on I'm back from training.

E

A

Thank you shall I move on to John, let's see John, can you speak up or can ya? Okay go ahead, get yeah.

C

So yeah we've got a lot of stuff going on working at home. Why not simply do this up evening, but I've got a bit more time recently, no I kind of changed. My sleeps good fill rounds to get some stuff done. Do you think so it's able to.

B

C

Of bit of bill on the extruder.

C

So yeah it went pretty well so definitely what you've done with the film that's here, you know getting the exploded, part diagram of building materials and require and screws is great. That's kind of considered. My pictures, no I was super helpful. The kind I you know started assembly expressions. These parts you need this is where they are gonna, go.

B

C

Super helpful, so nothing no real hit pitches on the construction, major ones. So, if I can pull my photos on it, biggest issue is really the 3d printed abroad that goes on the 3d printer filament idler got ended up being a little bit too large, so I had the file down, so it was a large tissue. I had, and just you know well they're minor print issues that you are. You can see pretty plainly like.

C

Had it a little bit too much of stick egde to the back of my heated bed, yeah kind of cosmetic but I, don't think that's really gonna affect it functionally.

E

C

Yeah, you know next steps: I got some of the one-inch pipe cut. Have a secure fit I was able to locate the correct size screws that will go through the one-inch pipe and screw into the universal axis, so it kind of goals. Moving for you get a bill materials, more fine cleaned up, get those parts together, build the thing and started running it hopefully soon within the week. I should have find the time, and so, let's sit. Let's see this thing run.

C

Yeah I see what week so the first thoughts I'm PVC with this larger bed sizes I'm, seeing really a pronounced bending of the tube. So okay I'll up these on out and there expose all this heat made it even better at least start to bend after a while. So just storage of the materials for this, it's gonna be sporting, a flat surface. So.

C

That will we'll see how that ends up affecting the prints, but uh you know I'm looking forward to get somewhat uh spread SATA there, any test prints are necessarily using for calibration I know, there's plenty of those on online that can use for like metrology, you make a prince here's. My defying dimensions here are my results: yeah someone to get into that side of it and you can sort of any knowledge or experience you have in that regard on how you know we can set up a sign. You know science method of hey.

C

This is what we expect. Are we getting the quality required and just defining what kind of quality you expect out of this right is sort of next steps, because you know it was just a matter of getting materials and they had the materials I just need what a couple hours really to get this stuff built I mean extrude. It took me, you know two hours, press is just fine and documentation getting stuff organized.

C

You know the nobility of 3d printer. It should only take me two or three hours when I find the time that I have all my materials together right. Hardest challenge is getting this stuff. You know one place yeah and uh finding the correct screws and measuring stuff yeah.

A

That is a big challenge and I would say: yeah like that. Extruders got so many different screws, and so once again this is not our design. This is barred from the Prusa 3d printer, which is the most popular 3d printer in the world. But this.

C

A

Very very complicated look at the number of so if you look at slide number six look at the number of different screws and ya.

C

A

It's crazy, like just a huge amount of hardware yeah, and to think that, from our perspective, what we do is we do always the concept of part reduction so make a little modification and force that screw like, for example, cover and three and m25, and now yeah.

C

A

Piece they're, making the same, modify the design and making the same like make it fit with the intent of minimizing part count as one of your design principles well, yeah.

C

And that could mean that was not considered in.

A

This bill, that's like people might say: oh well, we're gonna optimize it for elegance. Well, we want to optimize it for replicability. Replicability means that you can.

C

A

As possible, yeah.

C

Definitely there's definitely room. You know building this and you built this as well. There's definitely plenty here, Mac up this CAD drawing yeah and iterate on the extruder, so yeah.

B

So that's when my plans.

C

Down the road yeah furnace, definitely but after we would listen, let's just I'm trying to make a personal goal to get some prints this week, so yeah, hopefully I can get there. Yeah.

A

C

This so can other than that I just had a oh but I'm.

A

Saying this thing can work pretty well.

C

We can use the volcano nozzle with.

A

This as well the larger, much larger nozzles like one point, two millimeter, which are huge and I printed, it so much faster. So that's still, but we're doing just one step at a time right now we're just doing the one point: seven five millimeter filament we can use. Yes.

C

Three millimeter I.

A

Know that Prusa does not do it and there's controversy worries whether you can make.

A

It's easier to make the filament if you're, making three millimeter film and as opposed to 1.75, it's just shorter distance. You have to make and irregularities, don't matter as.

C

Much so you want to go larger.

A

C

Yeah and you know, I've I've run the Airwolf extensive, the three millimeter on a 1.75 or you know smaller nozzles of larger filament. It does kind of wear it down a little bit. You get some of not sufficient heating. You get some rotation but yeah. Let's get some basic prints out of those guys. Let's get some metrology information and get our data collection done on this, and then let's roll immediately into how can we reduce part count on the extruder and how can we increase our throughput and output of filament?

C

You know one thing at a time if.

B

C

Can get part reduced, yeah, okay, it runs fine part reduce next, let's add larger nozzles, maybe larger throats right. Things like that and Susan get out of this extreme. That's Rob, so yeah.

A

So, okay, so on page number, six and a current document, there's I added the photos folder. What I did is piece by piece, I strip that extruder, apart from the complete assembly, complete assembled right here step by step, you can trace it. I took out the idler screws. I took out.

A

The burying I took out the nozzle step by step every single step, so that's actually valid for annotation as a built procedure like if you study that you can see that each consecutive picture has just got like one more part exposed until we go to the very end where the last very very last picture has every single component exposed and that's what went actually into the visual Bill of Materials.

A

So John, maybe I can ask you if you have the time as a as a priority for now, because that will help during the workshop now in the workshop will have a sample. Extruder I have all these parts. Well. If we can get a step by set step sequence like this, that would help for a build instructional. Oh yeah,.

C

A

C

A

They'll be great and idea: is this I like this, these kinds.

C

A

If you, if you look at the extruder of visual beyond Bill of Materials, it's in four super animation best Cleavon, you have an instructional. They show you step by step and.

C

A

Follow blindly what you're doing, but we like to do, is try to to show people the comprehensive pictures of it and actually do do this. It makes just much more sense. I know, for example, if you look at the instructional, that's step by step, I mean I, don't really like it, it's it works, but it's not as information rich. It doesn't give you enough perspective cuz. It's really like designed to be kind of like dumb down to just okay, do the stuff in this stuff without ever really having the perspective of what you're doing overall.

A

So if we can get that perspective into our work, let know me I would help people yeah.

C

Yeah it'd be awesome, me I just said you know it parts out in front of someone's they can see exactly what I need. You can lay it out and then make it like a Lego build instructions like kids Lego set each thing, that's language agnostic and it's really easy to see how it all comes together to a final piece. It's that's pretty awesome. Actually.

A

C

A

You could do that. Saturday is when we would need it if you can get it in time. Otherwise, I need that yeah. You need that. So the start of that would be just start, a start, a Google presentation and pasting. Those pictures consecutively and, we can add, add annotation as far as okay now take out these two screws, etc. So forth and you've done it actually.

C

A

Experience definitely.

C

Andy, so each each slide, I should yeah, let's put the arrows in each slide and then, if I, what the part is you gotta be awesome and I, and also what I'd like to do in these things do is take the my calipers as much as I can and show measurements for things. I think that's helpful too. Yeah.

A

C

People, so we can get all that work and exactly.

A

They'll be good and for the actual build you know what actually helps a lot if we simply take a sheet of paper and draw out the outline profiles to scale for all these ones. So when you're actually taking the materials from the inventory area, you're pretty much matching each piece of material to its to its contour unsimplified sheet of paper.

A

That would be an effective way to to do that to make sure you've got all the parts and depends how you do it like: we'll have different teams working on different the printer during the build itself and by the way guys 20 bills in a single day. So that's what's happening outside about 20 people that are pretty much participating in the boot camp, the first day at least so a lot of builds. The what's worked for us before when we did some 3d printer pills is when you for a small company.

A

So when there's many many small parts, it's useful to just print out a sheet of paper white sheet of paper on a printer and just put the contours of all the parts labeling, the parts so that when you pick the parts from the part part boxes you're like matching to a visual cue, so you know you're, not just working from a list. You have a visual cue to know that all this is actually in the correct part. Could you see its contour on a page? So that's something we can do as well.

A

Okay, all right! So that's great great work there John! So it's kind of cool to see like the John you're. You know you're doing in parallel, you kind of building the same thing so we're in the remote locations we're actually collaborating on on getting the developments on that. So that's good! That works. Okay, all.

C

Righty all.

A

Righty next do we have miles miles? Do we have you here.

A

Go ahead so miles so you're working on power supply, so some electronics work, so kicad is one of the things we're gonna cover with in the mercial program and a boot camp. Here we have some old sample design some miles. Take it away.

F

It's input and we got a projector.

F

You know, isn't drying that French the voltage dropped across the the first ambassador will increase and so will increase. So the Zener just makes sure that it doesn't go about nine.

A

Zener diodes are are labeled by like, at which point they they drop the currents so like you can get one for like nine bulbs or ten volts or twenty volts. Is that so or they come in certain values. So.

F

I found a bug in a pencil rip it up in photography and then succeeded them all.

F

And now there's also a reference I'm, the senior as well.

F

A

So working on a basic basic little power supply, so we're rectifying alternating current.

A

You know: can you explain this for for a sucker in electronics? How do you do this? So you get AC rectified, so we can understand, rectification that you can google that Google a rectifier. Anyone can do that. An English major can find that out how to do that. They can understand it. So we got rectification happening. Tell us about sakai fuse in your diagram there. So I'm, looking at page, 11, okay, sure you've got too much current. You want to blow up you, so you don't you don't blow anybody up. You got a switch.

A

What's the deal there, the the RC RC deal before the rectifier. What is that.

F

A

That's a filter, that's like smooth smooth is the signal. What are we doing.

A

Okay, okay, so that capacitor there gets you down to the required voltage. Is that the deal.

F

A

So, are you going there from your drop in there from 120 volts to like six volts or 5 volts or something or what are you doing there.

A

To 10 volts, ac, still, okay, so passing 10 volts.

C

Ac, your rectifying it I.

A

A

Charges this thing: what if you didn't have that resistor there keep this charged.

F

It might slowly discharge through this unit as it simmers about the first value. That's going to go live all right.

F

A

A safety feature right, whoa.

A

Yeah, okay, that's cool, oh, and what's going on with c2. What does that do? This is.

G

So we're filtering.

A

D

D

So during the okay.

G

So that's: okay,.

A

Now, that's so tell us about that there. So is this cukes the open source, quite universal circuit, simulator.

F

Yeah, that was the graph was created by absolute spice, and so it yeah they use this ng spice or it can use any spice to be like a but I use at least twice.

A

F

A

Spice tell us more about this: I mean I, don't use cubes before so you're saying you got ng spice, which is a software that does what tell us more here, go some perspective.

F

Relation so six, that is the design and the simulation they have a simulator called.

F

A

Did you take this? Was your workflow there to go from so I'm seen he can? He can correct by the upper diagram yeah? Okay, so we got kicad you're generating yes, sir circuit, which then you can actually go into a layout which you can actually mill on our CNC circuit mill. So you got. Are you actually in putting the the kicad design into spice or ng spice recuse.

G

Yeah had remade it, they might be a way to do that, but I don't.

A

I wonder if that's I mean it should it should be able I mean we should so we don't have to redo things twice right.

D

Okay, so I mean this.

A

Is pretty high-tech stuff people here, so we got an actual simulation miles. Do you think you can create a little video on how you obtained the that amazing graph I.

A

Mean that's that's getting to some real analysis there. That's you know they don't teach you that in kindergarten, but it's something we should we should be stable. Can you explain a little bit so you've got what's the difference between the transient reference versus v vo? What is that.

F

So he had left the left, accesses voltage and then the right axis is this current. So the blue one is a is represented by the right axis and the other two are on the left. Aha, so.

A

Saying you got like 0.25 of current and you've got like nine volts or so on the voltage. That's how you read that.

A

So it's transparent, yeah I mean it's. If you can read that graph yeah, that that makes sense this or that or the English major, the red line is the voltage graph and.

A

For the botany major we've got the blue one as current.

G

And then the black ways the records.

A

And the reference voltage is what like got like four vaults reference.

G

F

Your simulation would be about four point: seven, are you can put in a voltage to New York? You know it uses it as a reference and be more accurate than then that's internal reference, but it might not actually be.

A

Okay, well, it will be really cool if maybe like, for for when I don't know for a winner cuz we got immersion I, don't know we're gonna have our next meeting people to think about it, but next week the boot camp is here: I can't I, don't think I can really skip. Unless we do like a really short meeting, we could we have a lunch time, so maybe we can possibly do that I mean the parm is pretty full throughout the whole next five weeks, but lunch is available.

A

Would people be okay to meet maybe like at lunchtime like noon time for the next meeting, because we got to continue this people? This is it. This is history in the making. So is everyone? Okay without it next week, you know: okay, I will.

F

F

A

Well, since we relying on you for do you think you could you could do that little video of the how you did the simulation for 12:30 next Tuesday?

A

Okay, let's call a meeting for next so so lunch is 12 to 1.

A

Let's have a quick like a pretty quick, powerful check-in meeting next week on all the progress at 12:30 I'll send an email out to everybody on that, but that's good. So what we're seeing here on this Arduino power supply pages as a Corson in circuits, but what we're seeing is pukes the open source circuit, simulator, meaning you draw the circuit, I'm dependin, stop like alternating currents or any kind of a switching power supply like for an induction furnace. The Holy Grail to me is the induction furnace people. Is it's gonna be a derivative of this.

A

What we got here, that's gonna, be you know this plus other elements, but if you've got an induction furnace, for example, you definitely want to simulate it in a circular circuit simulator, and then you can mill parts with our CNC circuit Mel. You can design the CAD milling and with in kicad. So there's a whole open pool chain that we're developing and spreading to the public, but this is cool. This is really cool. I learned something here: let's see what more we can do so Myles, let's do that next week at 12:30, they'll be awesome.

A

Well, thank you for this excellent little tutorial here. So that's good I'll. Just make a note here this. This is cukes. It's called q, uq UCS, which is quite universal circuit simulator.

A

This is what this result comes from. These are all accessible tools if we know how to do them to use them, then it's pretty powerful to actually be productive.

F

Nation was done, which is.

F

A

For spice integration.

A

And you said: spices, proprietary, but ngspice is open-source yeah.

A

Okay, very good, excellent. Okay, next person, it looks like aids got a lot of cool stuff there on the power cube there, hey you and I'm continued on that.

H

Yeah remember where I last updated last time so right.

H

That I did update a cat before I could have merged everything from the large power cube and frame, and all that adjustments to the smaller keep frame and but I think those are pretty well up to date. Now and I've been reviewing a more different CAD design stuff because with get and everything, I've been looking at that so I'm trying to figure out how to do my CAD files better. Is that that forget way of doing that?

H

I think it broke it down more with more files and smaller parts and not to do more assembly and sub assembly, but that would work better with git and things like that, because I haven't even really been using the git repository features I just been uploading the files in a simple manner, so I'm figuring some of that out of it gone Cole map I, don't have.

C

H

Either cuz it takes an NVIDIA GPU, that's pretty good I, don't know who those I know what most most laptops, because they don't have that Nvidia. It's an older CUDA framework for acceleration, a lot a lot pest another computer, you don't have a fairly decent, discrete graphics card and a desktop, and it's probably not gonna- do work well, so I just use the CPU and mostly because it takes so long.

H

In some cases, I just started doing tests trying to figure out how the photos you know different effects of the photographs and quality of the backgrounds and so on, but effect of the speed and and what works for getting more points. I was trying to get more points in the cloud leave it on the sparse and I saw someone tutorial, but it was usually they demo taking a pictures of rocks and from what I can tell that's, because he's you need a lot of texture kind of on your part, smooth parts.

H

I, don't know how well they work unless you take really high-resolution photos, I think no I was using a wooden part for that most, but I tried different things like putting masking tape on it. As you know that would paddle. In that example, a lot of times it took photos and when I start doing the masking, like that, the photographs tried to not at all the rock.

H

They said that you have to leave a certain amount of stuff around the object, but you can erase a lot of the background and that eliminates alyssum of those wasted. I, it's wasted, compute. All the background points, because.

C

You don't even want those.

H

Points in your cloud, you just want points on the object, although I was looking fees about whether some some points around it might help it compute or triangulate, because it is trying to triangulate like the shape of the object from specific texture points. It won't point. I tried this grid background and in what I got from that was just a bunch of points that look like a grid, so I stopped doing that.

H

I think what we're best, maybe especially for like this wooden object, that I've tried to test with, is getting closer shots, which is higher resolution, but closer shots that overlap more and getting some of the grain up closer. What you get I think a little better, more points on the object, so I may test more with that, but mostly I'm, just trying to figure out what works fastest, maybe speed some.

C

Of the processing.

H

On that have a GPU, because I know that.

E

There was a lot of spaced.

H

Points and I did I did lower some of my stuff into like my slob and figure some of that out, but I always tended to have very few points. Even when I did photographs outside on a cloudy day, it was a lot better, but it it's still what's kind of kind of lacking. So I think resolution model is a little more than I thought and just getting closer shots and making sure that they overlap for small objects. You probably would need to focus up real close and try to get grain.

H

You know on the object, but if first I didn't think that resolution matter too much I'm, not it does, but it just needs to be able to find features to match and I I'm still not sure, I understand exactly how it processes all that out. Even though I read through the core information on a bunch of that, it does say that there are issues with cameras.

H

I noticed that there was some other software listed on the photogrammetry page there that works differently, and it more specifically has a database with different types of cameras, because it wants to figure out information about the color space and I. Think the the lens and focal points and so on.

H

But oh and the other thing that I did was I was doing pretty much all video instead of just taking individual images.

A

H

Be quicker to do video and I think the thing that worked best was doing like us motion on my phone with.

C

H

Was 60 frames a.

C

Second, which.

H

Seemed odd, but it it it kept it from being too blurry a higher frame rate, helped it and then a short video about walking around the object and then in in kdenlive I just extract frames. Kdenlive has a function.

H

You just speed speed the video up to like a thousand percent, and then you just tell it to check the frames and it doesn't that, like in several seconds and then you've got like 50 to 100 photos or whatever you set it to placed on speed, and then it can process those photos like that, so that that seemed to work pretty well for extracting photos from video yeah.

A

Okay, that sounds pretty good, so keep going because yeah team up with Sarah and then Alex have we heard anything from Alex at all regarding this I think he might be on vacation is one of the emerging people.

B

A

Yeah all right that sounds good, so I I need to get going here on my side here, direct register in advance. So we can people pick people up on Friday evening, but yeah. This is good. Some pretty powerful updates this week, we're continuing, there's going to be a lot of activity, happening, lots of hubbub and and progress happening out of the boot camp itself and out of the immersion, because part of that will be doing some of the builds and consumed development and learning and updates, but yeah.

A

Let's do let's do the 12:30 next week and I think we'll probably want to do that for the next few weeks during the merjan program and that that's looking really good it's game day, I'm doing it inventory and really busy, though the details leading up to that but yeah great stuff.

A

So thanks everybody and we'll see you next next week any quick questions before I run out here any questions or concerns.

A

Okay, so that's good and just a note, we are trying to get the the agenda up like as soon as we're done today, because progress happens throughout the week, it's nice to as soon as you have something to actually put it into the development meeting agenda.

A

If you have a good result so that you'll have to scuttle for that on Tuesday like before, our meeting feel free to do so asking John who keeps these meeting agendas to kindly post them as soon as possible after the Tuesday meeting, so that we have the next one already prepared. So whenever we have any updates, we have a venue. We have a placeholder where we can put that in instead of trying to go, go back to our former results on Tuesday. So that's about it! Okay!

A

Well, thank you. Everybody game day is Friday, we'll continue meeting next week and see you soon. Thank you.