In PcbNew (V5.1.10 and Windows) 3D view using screen-shot or File-Export Current View as PNG I can get a picture about 1200x900 pxl.
Assuming printer resolution 600dpi that means I get about 2"x1.5" picture even my PCB is 5"x4".
Is there a simple method to get that view 2x bigger in each direction?
I can do 4 screen-shots but cutting them and merging will not be very easy (al least for me).
May be there is a method to make the 3D viewer thinking that screen has more pixels that it really has.
What you can do is drag the 3D window to the lower right corner of your screen, and then grab the corner of the window and drag it to the upper left corner of your monitor. This makes the 3D viewer bigger then your monitor, and you can make higher resolution screenshots.
I just tried this and then: 3D-Viewer / File / Export Current view as .jpg and it resulted in a file with a resolution of 5967*4733.
My 9 year old PC does get pretty sluggish though when the 3D viewer has to keep track of that many pixels.
On my linux box I can also make screenshots of a window, regardless of that window is fully visible on the monitor or not.
Never heard about it. Will try later (I have my PC with KiCad switched off before I sow it nad it is enough late (21:13) to go home).
From what you have written I’m not sure if it works in Windows pr only in Linux.
Now you drag that up…
I have vague memories of windoze preventing a window from growing beyond the actual desktop size, but it’s been many years since I last used windoze.
There used to be a “virtual desktop size” thing, in which the desktop has more pixels then the actual monitor, and you could scroll over it. I don’t know if that is still a thing though.
Another workaround is to set your videocard to a higher resolution, let your monitor scale it back and then make a screenshot.
I also find it a bit strange that you are apparently working with a monitor on a 1200*900 pixel resolution with KiCad. That would drive me crazy and I would certainly buy a bigger monitor within a week.
Monitor is 1920x1200 but you can’t fill the whole screen with PCB. It is enough for me (it is much better then when I have designed my first PCB with CGA and 12" B&W TV converted to monitor).
I have checked - don’t works for me (Windows 7) the way you have written or I didn’t understood.
I’m on Ubuntu but if I up my monitor resolution to 3840x2160 and leave canvas scale to automatic, the 3D view render looks nice. At 1920 x 1200 there is a lot of aliasing ( jaggies ) on the edges of the render. If its rendered to a big image then you can scale it down in Gimp or Graphic Converter which will get rid of aliasing too.
I found place where I can set 9 resolutions starting from 800x600 but 1920x1200 is the highest.
At that moment I will not buy a bigger monitor.
Some time ago (about 10 years) I have written a program to merge screen-shots of segments of topographical map (scans of real maps) I found in internet. I just didn’t know that you can get right rectangles of such map and merge them. Now I know that you can but still don’t know how - I am an electronic and not computer specialist. I merged even 8x8 screen-shots to get one big map and then print it with 200dpi in original map scale (they were scanned 200dpi).
My program works well if source are something like that scans or photo where each pixel have really different color even if they look the same. But, when I tried it with city-plan when one color regions had the same color of all pixels it didn’t worked so I suppose it will also have a problems here.
My idea was to find the exactly the same 3x3 pixels to find how to put one piece of map on another. I will give it a try tommorow.
But the best would be just to lie the 3D viewer to make it thinking it has a bigger window than a screen.
Long time ago I used Hugin a few times, and I was very impressed by the results. Hugin is a program for stitching photographs together and make big panorama shots. Normal photo’s always have distortion because they capture a 3D image in a 2D field and Hugin can sort of “undo” that distortion to fit separate pictures together. You do not have to be very careful with making the input for Hugin, You just need about 20% overlap between the pictures and enough contrast in the overlapping area’s to give Hugin something to calculate with. It can also automatically correct for taking pictures under an angle. Hugin is also capable of stitching multiple small screenshots (without adding any of the “corrections”).
There are many programs for stitching panorama’s. When I last used Hugin (8+ years ago) It was a very flexible program with lots of ways to tweak it, but not very user friendly nor easy to learn. If you do a general search for programs to make panorama’s from multiple photo’s you’ll probably find quite a few.
There is even a thing called photogrammetry First time you see it, it’s pure magic: You take a lot of photographs from a 3D object, all from different directions, and the software figures out what the position of the camera was from all the photographs and reconstructs a 3D model from the pictures. This is not very relevant for you, but it may give you some Idea of what software can do with pictures these day’s.
In Windows 7 if i make un-maximized window as big as a screen and then move it down right or down left to be able to resize it I am able to increase its size only few mm to left/right and top and something blocks further magnification.
The problem with these pictures is not very critical. I just wonted to know if I can simply get it with higher resolution. Now it looks that I have to:
wait until I will have bigger monitor resolution,
switch to Linux (would be possible if I had separate PC for KiCad, but is not currently in my plans),
search for program to merge some screen-shots,
modify my own old program (working at endless map is something different than here - there is no idea to try to find the same white rectangle (top-left 3x3 pxl corner of top-right screen-shot to be merged)).
It had a chance to work as I was selecting 8 small rectangles at each picture border and searched them (more then one match was not allowed) in picture to what I want to attach that one.
So I checked it now. It does not work. The reason - when I move 3D view to get 4 screen-shots it is modified to show different perspective so 4 screen-shots I got are not 4 parts of the same picture.
Each part is ‘seen’ from different point. I think there is no idea in merging such screen-shots. Even if some program will do it you will see that something is wrong with that picture.
Others have made beautiful renders of PCB’s with Blender, and that can very likely be done in much higher resolutions. I never looked into how that is done and do not know how much effort it takes.
After I have written I found it. I was searching that function in menus but it seems that it is only in toolbox.
My program correctly connected only 2 of 4 views (top-left + bottom-right). But it was not written for such kind of pictures. It assumes that each pixel has close to unique color what here is not true.
My C++ Builder 5 program has only 180 lines source code (.cpp) so it should be not very difficult to remind me what I have done in 2011 and may be modify it a little.
But I’m not sure if isometric perspective is what I wont
An year ago I tried to run renders (as I remember there are two there) in FreeCAD but I can’t make them working. I am not surprised - I have 0 experience in such subjects - I don’t understand what someone writes to me.
I don’t need pretty picture. I just wanted to add to documentation a little higher resolution PCB view but prefer orthographic projection over isometric.
I think I will be happy with what I have.
