My name is Paul R. and I am a visual artist living in The Netherlands.
I am new to do software and want to know a couple of things.
1. Has anyone done larger prints for gallery work ? What are the challenges when it comes
high res print (larger then 7500X5500 300 dpi), colors, is the output RGB ? Is everything still razor sharp. Of course any examples will be welcome.
2. Is there a way to add several images. So they become all sources and layers ? And can you work on them as a layer. Or will the next images be the new source and is the old one just gone ?
I think we're all in agreement that we need more videos, tutorials, etc. We used to ship 8 hours of video tutorials with the program in the V2 and V4 eras when we sold a box with physical discs in them, and in the back of my head i'm always assuming everyone has watched those, but that is only true for old timers.
Paul, your selection suggestions are great. I like the rim idea.
I do want to point out that you can route any IpOp effect into the selection buffer. And the Composite operations should still work for that. So i think you can already apply an IpOp effect to an existing selection mask with the appropriate compositing option.
Someone should try using Interactive Adjust to manipulate the current selection back into the selection buffer. Try Smart BW Threshold and i think you are somewhat there for what you want to do Paul.
And i guess you could use Interactive Warp for similar ideas. So try something like a Scale warp again routing the output back to the Selection again, and you'd have to muck with the Composite options to get what you want to get out of it.
I hope I don't get too boring here.... :)
So, trying to experiment following your advice. Using a flower image which is a very tricky one from a masking point of view, by using the lasso tool. The flower has plenty of separated purple coloured flowers on a background of various tomes of green, but the flower has elements of black with tiny white hairs.
My objective is to see if I can achieve acrylic strokes on the background only whilst leaving the purple petals alone. Some of the petals are (deliberately) out of focus, others tack sharp. Also, some of the background is out of focus, others (such as cobwebs) tack sharp.
So, it seems to me that selection based on colour range is the way to go. Theres nothing in the colour selection tool that's going to help. In PAS (Acrylic Impression), stepping through but I can't tell what the 'Autopaint' steps are based on (in order to see if I can play with anything recognisable as helping to create a selection). So, stepping through in Editor mode takes me to the Paint Synthesizer - paint fill apply has the Composite but nothing I can see for selection.
Please don't misunderstand me, I love some of the interesting results but I just can't control where its doing it just within SA.
I've tried other presets in the PTSynth. but haven't spotted anything that seems its going to help setting up the selection.
Next I've tried several ImOps including _IP Source/Region Selection and Selective Colour Matte but get mostly a black screen or crashing (even after I managed to change Palette and the memory blocks - its all very much less than obvious) - for e.g. if I select one colour I get either a mostly or all black screen, if I then try to select a different colour, it crashes.
I've tried several other elements, such as Composite output to 'Replace Region Selection' (presuming that means the ImOp will apply to the selection made with the selection tool, but that doesn't get me to selecting what I want in the first place.
I have no idea how to generate more selection effects, nor have I spotted anything like a 'flood select'. To repeat, I'd like to learn a lot more, but not being able to get beyond basics is dispiriting - I'm either doing something wrong/missing something/wrong place or, SA cannot (yet?) do what I would like to see, but maybe its design is such that it cannot?
So, I may have to save a lot of sessions (to cope with the crashes in this respect), save more canvas', perform selections in (e.g. Affinity Photo) with several round trips back into SA, return to AP to recompose when the painting of each selection has been done, return to SA for resize, back to AP for soft proofing and output to printers (etc).
I really don't like being in this 'nagging mode' as it may seem negative - that's not it at all, just that its currently a clunky area currently and if you have a 'Selection Mode' it seems a shame not to take further advantage - if that makes sense? :)
Lets see if file upload is straightforward for me (... :) ). Its just an example of where I have applied an effect (Via PAS) overall, taken output to Affinity Photo to do the masking: load original img as separate layer to do colour range masking (then refine it). Reapply the output from SA to another layer in AP and play with blend and opacity on the masked area (i.e. in this instance to achieve a subtler version of the SA effect where I want it), merge to a new layer. Take that back into SA to apply any further effects (e.g. in this instance I had stopped the PAS before it finished all steps, then completed the final step on the combined image). Save session, flatten, resize. Back into AP for final touches and soft proofing before output to required format required by printers (or myself, as the case may be).
Do I like the result? most certainly. Do I prefer it to 'other software' I've used? Undoubtedly, and then some. Could I gain from changes - yup, if its possible technically, but I can live without them :)
Blast, can't upload the file as it exceeds the limit. I'll reduce the quality.
I have since tried both ways on a number of images (but all of the same genre, so others may vary) and I can't see any difference when I've output them to a large screen. So, for me, it seems simpler just to complete the painting on a 30x20 canvas, flatten, then resize to 300dpi and ensure size is set back to 30x20.
The alternative approach I tried was to size canvas to 120", load the image, then resize canvas to 300dpi at which moment SA reduces the canvas size back to 30" (well, almost - it does seem to vary that marginally, i.e. by and inch or so). However, the paint process is very slow indeed, so I've only tried it once!
I've also just tried doing a single run of the vectorizer (Acrylic preset) but with No. Colours set to the higher level number I was using with composition set to Edge1 (followed by smart blur) - the result was totally different from putting the vectorizer on its own layer and adjusting the composition (and opacity) from there. I find it quicker from there as I can experiment with different blends. Its a shame that the composition tool in SA's modes aren't either the same or can be listed somehow as equivalents - but then, I suppose, its another level of experimentation :)
We are taking this thread very seriously here at Synthetik, so our main preset designer is running through our factory presets putting together an analysis of what re-renders well from a lower res working canvas to a higher res final output canvas (and what doesn't).
They have identified several places where if we enhance our re-render code to deal with re-adjusting certain IpOp effects typically used in the factory presets, it can make a big improvement in the re-rendered high res output. That is on the plate for adding to the program in V5.5.3 development, which probably starts tomorrow.
In addition, we're looking into adding a new feature to the Vectorizer to do the 'vector-up-res' process directly in the Vectorizer when you are running it. As opposed to how you would have to do it now, which would be to work in a lower res canvas, then output to a svg file, then import that svg back into Studio Artist using the File : Import : Import SVG as New Canvas... menu command, where you specified the higher final resolution in the canvas resize dialog that comes up when you run that menu command.
I can't promise that will be in the final V5.5.3 build (since i'm still working through everything involved in making it happen), but it is something we are very serious about making happen sooner rather than later. We also ran some tests comparing working this way to using the best neural net interpolator you can buy right now to upscale that lower res vectorizer raster image output, and our 'just keep it vector' approach for upscaling to high res output is far superior.
Not to slag neural net interpolators (they are cool), but if you are already working with vector information it really makes more sense just to keep it vector and rescale as necessary there before converting to raster. Unless you enjoy the artifacts the neural net interpolator introduces into the image.
Thanks for all the interesting and detailed feedback. Much great stuff to come !
I was also hoping to have the solution for putting 300 dpi in the software and getting the same 300dpi back. Without trying to solve something is broken. so I rather would not do 72dpi and enlarge again and try to have the same quality back for large prints.
I there a reason the software has to go to 72DPI once you have put in a 300 DPI file.?