PixelCNC: The best image-to-Gcode CAM software around!

[Clyde]

Sorry for not getting back to you until now.Good news! the program works on my desk top machine which is the unit that runs my X-Carve. For some reason it still won’t work on my lap top but that’s ok since I don’t use it to run my machine anyway. Thank you for taking the extra effort to make it work right.It’s now on my “next things to purchase” list, if not this month than next month.

[Clyde]

Howdy- I’ve just have gotten the time to start working with the software and when I try to load the samples i get a “image exceeds free trial” message at the bottom of the screen. My screen size is 640 x 480 and I’m using windows xp. thank you.

[CharlesVanNoland]

Hiya Clyde,

When you run PixelCNC what’s it say on the very top of the window? Does it just say PixelCNC or “PixelCNC (Trial)”? The examples included with the trial should all work just fine for the trial. The samples included with the full version will not work with the trial version of PixelCNC. The latest version right now is v1.18a, so if you don’t already have that version go ahead and grab it real quick: http://deftware.itch.io/pixelcnc/

EDIT: The free trial is limited to ~65,000 pixels on the image input, which means about a 256x256 (or any dimensions which multiply to 65k or less) will work on there.

[Clyde]

It was the trial version, I downloaded the version you listed and got the same message. When I shut down the program I got the following message:.Thank you.

[Clyde]

Her is the screen shot of the program running:

.

[CharlesVanNoland]

Thanks Bill,

There is no installation as of now, you just unzip the PixelCNC zip file into a folder and the PixelCNC.exe file in there is the actual program you run to use it. This is only because PixelCNC is still in alpha stage of development and it wouldn’t be convenient for users to have to do a full installation process for incremental updates (which are-a-plenty during alpha stage). A one-time installation will be used once PixelCNC enters beta, and updates will be automatically downloaded by PixelCNC. In the meantime the focus of the alpha is on adding features and functionality.

That being said, I’m in the middle of working on the next big release for which I’ve devised some changes to have a more robust rendering pipeline for dealing with older hardware. These days most developers wouldn’t even bother supporting hardware more than 10 years old, because it’s such a huge headache of labyrinthian compatibility issues, but in PixelCNC’s case I know from first-hand experience that a lot of the target user audience is running on older hardware that is otherwise perfectly capable given some coaxing.

At any rate, your laptop’s GPU is limited to OpenGL 2.1, which is sufficient for what PixelCNC needs to do (huge vertex buffers), but drivers from the 2.0/2.1 era are very iffy with how they implement the functionality they claim to support, because there had yet to be a solid naming convention for everything, so finding certain functionality in the drivers can vary. As such I have been working on some changes to how PixelCNC interfaces with OpenGL to make it more robust. I’d already started on that work a few versions ago and it has resolved compatibility issues for most users with older hardware but obviously there’s still drivers out there that want things to be a very specific way in order for them to cooperate.

The big update I’ve been working on now adds quite a bit of functionality and so the newer robust OpenGL interface stuff won’t be released until everything else on the agenda is ready. At the rate things are being ticked off the todo list a conservative ETA for the update is about three weeks, so I’d suggest checking back around then. I’ll be sure to visit this thread to give everyone a heads up. Thanks again.

[CharlesVanNoland]

Heya Clyde,

I’m not thinking that machine is going to be very cooperative no matter what. Just the limited screen resolution alone is going to cause you grief, let alone the fact that with an empty project you’re only running at 18 frames-per-second, which is a good hint that once an actual mesh is on the screen you’re likely going to be experiencing something more akin to a slideshow. If you’re actually loading the 256x256 images provided, and it’s responding as though they’re too large, I’m not sure that’s a problem I can solve, or that I should solve, simply because the machine is clearly not up to snuff all-around.

Your screenshot is 800x600, which is a step-up from the 640x480 you mentioned. Even at 800x600 though this is what you’ll be dealing with due to the interface being laid out and setup the way it is:

You can see how crammed it all is, which maybe I can adjust a few a few things to make some more room, but even that aside if you’re running at 18 frames-per-second with an empty view then throwing half a million triangles on the screen is going to be hell on your CPU/GPU and your experience altogether.

I know I can sit and track down every little thing and somehow make it all work (aside from the poor performance, PixelCNC is already extremely optimized on the mesh-rendering side) but that’s at the cost of time/energy adding actual functionality and capability to PixelCNC that makes it more usefu to both existing and potential customers.

At the outset I had originally decided to not support hardware more than a decade old, but I’ve opted to at least meet some older hardware halfway because that was only a few hours’ of work to make that happen. After that it really becomes a diminishing-returns sort of thing. I’m still adding a little more to the renderer pipeline to make it more robustly handle the strange landscape of OpenGL implementations that existed back in the 2.0/2.1 era, and that’s the very last that I’m willing to do on that front, because there’s still a lot of other work to do before PixelCNC is where I aim for it to be.

Thanks for working with me to try and get it going, but this is where I have to call it quits. I am still interested in determining why it’s not able to access the user documents folder for being able to write a configuration file, that is important, and another one of those things where different Windows versions demands that the program do different things, but if an XP machine is capable hardware-wise then I want users to be able to use PixelCNC as designed.

Thanks again.

[Clyde]

Thanks again for your efforts. I’ve been looking into purchasing another used computer to run windows 7 on since there are so many great design programs that would be invaluable for cnc. I’m must confess that I am not that great when it comes to computer hardware so if you could give me a some guidelines on what I need to look for when purchasing a computer to run your (and other) programs I would greatly appreciate it.Again, thank you for your help.

[KellyD]

Hey there, hope this is the right forum for support on PixelCNC. I’ve just purchased (couldn’t find a trial anywhere to download) but unfortunately the software does not open. I first get a message saying “Search for App in App store? You need to install an app for this task”. Whether I click Yes or No (Yes doesn’t take me to an App store or anything) I get the Early-Access Disclaimer. Clicking “Yes” here begins to load the software but an error pops up - “sys_glfunc: function not supported: glActiveTexture”. Clicking OK the software seems to continue to load but then just disappears completely.

This happens whether I try to launch v1.20 or v1.21. Where can I look for assistance trying to get this software to run for me?

[CharlesVanNoland]

Hi KellyD, this is not the PixelCNC forum, this is just a topic/thread about PixelCNC on the Inventables forum. The PixelCNC forum is at the bottom of the same page that PixelCNC can be bought from, which is the same page that the free trial demo version can be downloaded from.

Just below the free demo downloads is the section for the PixelCNC community/forum.

At the bottom of that is a button for posting to the PixelCNC forum.

If you have any questions/comments/concerns about PixelCNC, please post them there.

[KellyD]

Thanks and wow…prime example of poor web design (and poor web user habits). The page had me at “Hello” and I didn’t bother scrolling through the entire thing - I just wanted to get to work!

Thank you for pointing me in the right direction!

[BryanBoratynec]

Any timeline for exporting to Marlin gcode?

[CharlesVanNoland]

Allowing users to output the proper G-code for their specific machine(s) is most definitely a part of the core PixelCNC vision, and is fast approaching as one of the remaining features left to implement. Hopefully post processor support will be in by the holiday season, at the latest.

The support for CNC posts, which will include a built-in post editor for users to create posts if they are not included for their specific machine, has been on the development roadmap since the beginning as a ‘primary feature’, of which there are several others. Once all primary features are implemented PixelCNC will enter the beta phase of its development.

As these primary features are added so increases the value of PixelCNC as a whole - which is reflected in the price increases that come with new features being implemented. Customers who buy at the earlier prices get all new added features for free, at no extra cost. New customers pay full price. This works two ways: it shows appreciation to early customers for their support while simultaneously giving potential customers an incentive to buy now, spurring sales in the near-term.

Also, I’m debating integrating an automatic post-sharing system alongside the post editor where users can submit their custom posts to a central 3rd-party post library. The library would then be browsable from within PixelCNC so that users can seek out unique posts to install without having to hunt them down on forums or create them from scratch themselves. This is an example of a ‘secondary feature’, where supporting posts at all takes precedence and this would be a sort of bonus that likely would come later, depending on the complexity of the solution that is devised to make it happen. It would definitely make it a lot easier for users to share posts.

[LarryM]

The support for CNC posts, which will include a built-in post editor for users to create posts if they are not included for their specific machine, has been on the development roadmap since the beginning as a ‘primary feature’, of which there are several others.

This is a bad idea and one that would kill any interest I might have in your product.

Your product is useless without a post processor (or its equivalent) for a given machine. Fortunately, there are lots of other CAM programs that are customer friendly to choose from.

Most of the post processors I deal with have a built in post processor editor. It’s called, Notepad.

[CharlesVanNoland]

Thanks for the feedback. I never said that posts wouldn’t be simple text files, just that there’d be a built-in wizard for quickly modifying/creating posts in addition to post processing support.

I’ve been looking at adopting an existing post format, instead of coming up with a new one from scratch, so that all of the existing posts in that format for various machines would be usable with PixelCNC.

The format I’m just about settled on supporting is Mecsoft’s, for which they have a page that posts for many machines can be freely downloaded from: https://mecsoft.com/downloadposts/

Unless you know of a better post format that would be more ideal?

Yes, you can always just use other CAM software, but only to do what other CAM software does. You would not be able to do the things PixelCNC uniquely enables users to do, however, which is the very reason I felt the need for it to exist in the first place.

Have a nice day

[LarryM]

Unless you know of a better post format that would be more ideal?

No.

And I think I should wait a day before responding to posts.

And I did have nice day, Thanks.

[CharlesVanNoland]

Actually, there was another idea I had a while back. I was thinking of coming up with a very simple post format, more template-oriented the way that PostHASTE situates theirs, and would just write a converter utility that enables me to convert a library of existing posts into the simple format. This would be my fallback if there were any sort of legality issue utilizing an existing post format. Personally, I believe there should be an open post format, so that they can be interchangeable across all CAM packages that support the format.

Anyway, thanks again

[LarryM]

I believe there should be an open post format, so that they can be interchangeable across all CAM packages that support the format.

Yes. Let me say that again. Yes a common format would be great.

Unfortunately, many vendors still try to lock their customers into proprietary formats.

My comments above came from trying to modify an Autodesk post processor. They have a build program that they use to change their PP.

[TimStevens]

winder/Universal-G-Code-Sender

[CharlesVanNoland]

Hi Tim, can I ask why you are showing me UGS?