I am still working on my router software, and have made a lot of changes. I have an image I rather like, but need to know if you recognize her. This one is 3k threads segments.
Please let me know! Also opinions on if it is pleasant to look at, a nice image.
Thank you so much.
a few days ago I posted a route image and asked if anyone could recognize who it was. One person did, and thanks for that. Now, I have a new router idea and need opinions on which route is better,1 or 2? Recognizably feedback also appreciated. Thanks so much...
This is the first run of the mystery woman in my other post, on the hardware. Also included, a brief tour of the hardware as I got a request to say more.
There are several problems l had today. There were a lot of pull outs of thread because I had a new canvas, new thread, etc. so I slowed the machine quite a bit to solve that for now, and am preparing a new canvas.
Also, after 1400 threads (the posted image of mystery woman is 2700 total segments) I just stopped because it became obvious the image was going to be useless, due to 2 factors. The first is I was using heavy (92wt) thread, which is thicker, and the second is that the image was very small compared to what I will be making. It is less than 1/6th the area I am targeting. As a result, the fat threads pile up on the small area, and, there you go: Black image. I'm confident the next effort will be something to look at!
Let me know if you have any question, post or pm would be great! Thanks for your time!!
4 years ago, I made 2 machines to make thread art, and used available software for the route. I nuked them because the images I was producing were not recognizable to my friends. "That's great! Who is that?" Crushing.
So now, I am making a new machine (which I've posted about) and writing the router software myself in Python. (It is a new idea where I poke canvas with a scary needle)
So before I torture my machine with its first image, I need to know if my router software was successful (enough).
Do you recognize who this is? If so, to the machine. If not, back to Python...
Lemmeknow!
And thanks so much.
(PS. ignore the red, that is left over diagnosis I forgot to take out....)
I ran 2 stress tests on my thread machine today of a couple random images for fun. One with 300 pokes, and one with 500, and it never dropped an stitch, which did dropped my jaw because I never expected such reliability. So pleased.
My portrait generation software is nearing completion, and is producing pretty good images now.
I hope to attempt a 4k-ish poke portrait presently.
Thanks for all the thumbs up on my first post on this machine, I appreciate them!
About a month ago I started a post about a machine that I was building that was unique in several ways. I was going to make a series on its creation. It got a couple hundred views and minus one thumbs up. The public has spoken....
Yeah, I need to stop posting on that topic and delete it.
BUT, today I successfully stress tested my beast and it is working. So I am posting anyway. Dang the thumbs down, full speed ahead.
I am excited and quite proud so far, there is no machine in the world doing what this one is, and I'm happy.
I do hope that someone, at least one, out there will appreciate and enjoy what I have so far!
I am still working hard on the portrait extraction software which is also unique in that I do not have commonly used perimeter pins, so the software is under development still. But the machine and the SW controlling it are up and running.
I am using Chat to help me learn and write c++ for a thread machine I am making. I use Arduino IDE, using Arduino helpers etc. I am putting coordinates for thread routing into a file and storing it in nonvolitile memory on the ESP32S3. The helper function/program I use is called LittleFS, which does the uploading and storage. Chat was making a test file with just 50 pairs to test my machine with. So it made the file and gave it to me, and said "Use LittleFS to upload"
I was punchy and said "Every single time I read LittleFS, my brain sees 'Little Finger' being a GOT fan".
The next line was:
I can see that, "Chaos is a filesystem".
Wow, a program made me lol. The future is now...
This is the second in a series describing the creation of my StringArtMachine I am making. I am developing the software, electronics, and hardware to make String Art, or Thread Art. See my profile for the other parts.
In this part 2, I am going to show the image manipulation progress I have so far. It begins with finding (pilfering) an image I want to make into a piece. Let's start here:
A fine place to start.... The first piece of software I have takes that, crops it according to parameters the program uses, and zooms similarly, and turns it to a grayscale (which is a GOT disease I believe) and outputs that for further considerations.
zoom, scaled, gray scale. Wowzer
The next thing I do is place hole locations just to eyeball and see if I can image what is getting coverage, and what isn't. I am using black and white threads, as Crum does, to get results I want, but I am going to poke holes instead of wrap pins. Here is one possibility,
At this point I am working on the routing, and so have no idea if these hole locations will be fine until the router is up and working.
I will post more when I have something to show there. I would add, that I am simultaneously working on mechanicals, and electricals, and will post those topics soon also.
Thanks for dropping in!
PS, can someone tell me how to scale the photos I drop in here. Yes, the irony is not lost on me....
I am starting a rather ambitious string art project. I have built 2 machines before, but have a new idea, new concept, and am excited, but intimidated. My last machines were 1) a typical circle using existing software for picture analysis and pin sequence, but all electronics and mechanical build by yours truly, and 2) a Cartesian machine moving x/y, same ownership.
They both worked, but I was disappointed with the brightness, and resemblance. This was mostly because of the software I "borrowed" from the internet. This time, I am going to write all the picture analysis and routing sequence software myself, as well as build the machine. I would add, there is much better software available out there now, than 6 years ago. But now, I will own/control/write it all myself. I hope it is better!
The big changes will be a rectangular picture instead of the standard now overused circle, random "pin" locations, and two color (for now) threads. When I say "Pin", I mean hole. I'm going to use no pins, but use a puncture technique I saw someone here using. The difference here is he had locations on the perimeter only. I will use locations everywhere. This is very difficult to work.
It is a most intimidating project, and I am very excited for the software, electronics, and mechanical aspects. I will be posting progress and information as time goes on, but I suspect this is a many many months project.
So far, I have written the original photo transformation program in Python. It takes a standard photo, in almost any format, crops it, resizes it, makes the canvas size I want, zooms, and transforms to grayscale (isn't that a GOT disease?) and places "pins" or "holes" or "locations" on that for routing. So far so good. Routing is next.
Meanwhile, I am staring the machine. Lots of t-slot extruded aluminum (or aluminium for those non US folks) and also started the electronics.
I will update (by adding new posts and referencing past posts in each) the progress as time goes on. Meanwhile, if you have any questions regarding MegaProject, let me know, I will share what you need to know.
I will start adding pictures in the next post of progress.
I hope someone enjoys my journey.
Cheers.
