Do I need photoshop to create a beautiful GUI?

[keyman_sam]

I've developed a simple plugin and its working. Now I'm onto the GUI part. I want to create a beautiful GUI (sort of like UAD, Waves, propellerheads). I can get artists if I need to. But I want a beautiful GUI. Do I need to get photoshop or can I just work with freebie tools?

The main thing is a waveform display of incoming audio (sort of like the soundcloud display of your song). Can someone point me to the right resources/free tools? What's the industry standard?

Much appreciated!

[Regnas]

No, you don't need photoshop.
There are plenty of free tools that you can use to create beautiful GUI, e.g Gimp, Inkscape etc..
If you want 3D, you can use Blender..

[Sendy]

A waveform display can't be made in a graphics editor, it has to be coded, since it's generated in response to the audio by it's very nature. The Gimp is the standard de facto freeware Photoshop replacement, and it's pretty good. Even I've managed to use it to do a few editing jobs.

[arkmabat]

No, you don't need photoshop.
There are plenty of free tools that you can use to create beautiful GUI, e.g Gimp, Inkscape etc..
If you want 3D, you can use Blender..

There's all sorts of tutorials out there, this one is for a big shiny button:
http://mygimptutorial.com/round-web-20- ... metal-ring

Also, if you want, this will probably prove useful for knob animating:
http://www.kvraudio.com/product/knobman-by-g200kg

[Phil999]

real-time waveform display has nothing to do with graphic editors, as Sendy has already pointed out. I don't know the industrial standard, sorry.

[mystran]

Also, if you want, this will probably prove useful for knob animating:
http://www.kvraudio.com/product/knobman-by-g200kg

This animated GIF is NOT the best example of what knobman can do.

[aciddose]

I get a lot of use out of paint.net

The new version is coming with hierarchical layers support and it already has great planes/plugins/formats support. If v4 doesn't have quality primitives it will hopefully be possible to implement new brushes via plugins.

http://blog.getpaint.net/

[tony tony chopper]

My advice is: don't use bitmaps.
The first sin in GUI-making is animated knobs IMHO. Those eat precious memory while you can always do a better job by fully coding them, or at least without using animation strips.

But the main reason to do a vectorial GUI is: every little tablet (& Macs too, & now real tablet PCs like Surface Pro) now has 2x or 3x higher PPIs than a desktop PC monitor. For how long do you think PC users will keep accepting this?
It's not impossible to go bitmaps for a pixel density-independent GUI, but it's hard.

[aciddose]

I've always been a fan of vector graphics, but unfortunately the processing power still just isn't there.

If you want to use hardware acceleration you're forced to live within the limits of the interface you use, which is usually not an option but rather "my way or the highway" on a system like IOS.

Memory isn't a concern even on the least powerful devices until you start to really waste it like mad. The few 8 or 10 megs required to manage all the bitmaps on a full-screen GUI filled with countless animations and frilly effects is really insignificant.

I do render my own background elements in software and I also mix my own bitmaps because the hardware and interfaces aren't capable of doing what I need short of using gl shaders, but it simply isn't possible to get the same sort of results you'd get from bitmaps rendered in a tool like blender. Not a chance.

[aciddose]

But the main reason to do a vectorial GUI is: every little tablet (& Macs too, & now real tablet PCs like Surface Pro) now has 2x or 3x higher PPIs than a desktop PC monitor. For how long do you think PC users will keep accepting this?

I will be shocked if 4k displays ever become accepted as a replacement for desktop monitors unless they go significantly below 30". We are years from that happening at this point, the smallest 4k display you can get under several $1000s is still 32".

The technology to produce these panels simply doesn't exist and neither does the demand for them. If I could get a 20" panel in 5200x4000 for $1000 I would have ordered it the moment I found out it existed. (In fact get me the same at 1920x1200... I wish...)

Unfortunately it doesn't, and neither do the systems for transmitting these high-resolution signals (HDMI/displayport/etc) nor the graphics processors (nvidia, etc) for handling rendering the frames in the first place.

A 32" display is uncomfortable at a desk unless the monitor is set a significant distance away. People simply do not have the eye-sight capable of resolving detail at that resolution and in many cases 1080 already supplies far more than necessary.

With a small display like I describe, being very near-sighted myself and having much higher resolving power while wearing glasses (30/20) than most people at distance, and without wearing them being able to clearly resolve characters 1/5th mm across which would require a high powered magnifier for most I still doubt such a high-res display would be limited to within my resolving power.

I do not see people using 60" televisions mounted on a wall in a studio, as much as that might look like the bridge of the enterprise or dr.whatever's evil underground layer in some 1930s sci-fi flick, it just isn't practical.

[tony tony chopper]

I've always been a fan of vector graphics, but unfortunately the processing power still just isn't there.

it's there enough for me, in software.
The problem is that you shouldn't limit yourself to generic libraries. I use AntiGrain, which is great, but it's generic and slow. That's good for controls, but when you have a lot of data to draw, it's always better to use optimized versions of what you wanna do. A good custom library combined with AntiGrain (wich IMHO could be a lot faster & still generic) works pretty well.

But ok: I've done bitmap-based GUIs and if you asked me if I could redo those by coding, I'd say no, not a chance. But it doesn't mean that a good GUI can't be done vectorially.

This the GUI the next FL will have
https://drive.google.com/file/d/0B6Cr7w ... sp=sharing
It's good enough for tools, the question is, is it good enough for plugins? Possibly not, as people still seem to be expecting photorealistic ones, but it really depends. FabFilter's GUIs aren't photorealistic and they seem well accepted too.

I used to believe in hardware 2D acceleration, now I'm not sure anymore in practice.

Memory isn't a concern even on the least powerful devices until you start to really waste it like mad. The few 8 or 10 megs required to manage all the bitmaps on a full-screen GUI filled with countless animations and frilly effects is really insignificant.

it's insignificant for a plugin, but the plugin isn't alone. And it's even worse when plugins don't share GUI resources accross instances.
The loading time is not neglectable either.

I do render my own background elements in software and I also mix my own bitmaps because the hardware and interfaces aren't capable of doing what I need short of using gl shaders, but it simply isn't possible to get the same sort of results you'd get from bitmaps rendered in a tool like blender. Not a chance.

I would in fact more believe in a great 3D-accelerated GUI than 2D-accelerated. I mean with 16x oversampling you can get nearly pixel-perfect quality, and it's not like photorealistic GUI's were still "in" anyway.

[tony tony chopper]

A 32" display is uncomfortable at a desk unless the monitor is set a significant distance away. People simply do not have the eye-sight capable of resolving detail at that resolution and in many cases 1080 already supplies far more than necessary.

I don't know, I'm using a 19''.. which covers most of my viewing range. Why use a large monitor if it's to place it far anyway?
By using a large low-PPI, but distant monitor, you're doing the same as using a close, high-PPI monitor. That still requires a scalable GUI, because you need bigger GUI than I do on my screen that's closer to my eyes. I also imagine that system settings will move away from PPI & use PPD, which combines the PPI & the viewing distance.

I just can't imagine that by 2020 we will still be using <100PPI monitors.

[aciddose]

Unfortunately I can imagine that.

The reason you'd be forced to use a 32" display is that the fabs available for these panels are still varying panel size by resolution. It's a simple matter of producing a fab that creates the same layout and density of pixels and chopping the "bad pixels" introduced in the process to produce smaller panel sizes.

That is why you'll find 1920x1200 can only be had at 24", and is impossible to find at 20" or 18", which I would absolutely love to be untrue. If anyone can ever point out a <$1000 quality panel in this size/res combination I will be grateful.

I'm still waiting on these >100PPI displays and they're just not coming. I swore they should arrive last year, and still we're here without them. Better luck this year?

Edit:

Fact is though PC sales are dropping like crazy, which explains why all the effort has gone into panels designed solely for hand-held devices rather than desktop monitors. If things continue this way we'll never see them. We'll no longer be able to get any of the stuff we've become accustomed to.

[tony tony chopper]

well, today's LCD-manufacturing problems aren't very relevant anyway, as by 2020 we will all have OLEDs. I mean, they're there, they're amazing & they're >$10000, so in 6 years everyone has one.

I believe the #1 reason, by far, for PC monitors to still be low-PPI, is software compatibility. *I* would never buy a high-PPI desktop monitor today, available or not, because I know how poorly (mostly) everything will scale.

And yes PC sales are dropping, but I believe it's mostly because PCs have stopped evolving, someone's 3-years-old PC is still nearly as good as a recent one. Also possible that desktops are less used, but in our domain, I doubt it (I mean, let's agree you can't do anything serious on an iPad).

[aciddose]

LED fabs have similar issues. While maybe possible, I believe they layer the LEDs directly on the same silicon platter to produce the panels rather than gluing/soldering discrete LEDs.

The higher DPI, the more likely you'll have a defect near the center of the panel. The more defects, the more wasted panels and the higher the price. Only way to cope is to be able to saw platters into variable panel sizes so that you can build an optimized tiling pattern after measuring defects on each platter.

I'm not sure if they actually do that or not or any of the details, but for whatever reason they all do seem to come from the same process (same pixel geometry, density, etc) and panel size is strictly correlated to resolution.