Joined: Thu Nov 10, 2016 11:29 pm Posts: 114 Location: Brisbane, Australia
Hi everyone, just joined up a few days ago but I've been following this board for a while now.
It seems every couple of months, there's a new thread about writing an updated app to design gems, with awesome, long feature lists and big ideas but then the thread dies and no app comes out...
I come to you today with a different approach. First, I've already got a functional prototype, with all major technical hurdles out of the way. I wouldn't be wasting your time if I wasn't 100% confident this will see the light of day. I write software tools for a living, this is not my first project
What I'd like to discuss is establishing a minimum feature set. Part of the problem with making a new design app is you're obviously going up against GemCad/GemRay. Those represent YEARS of work so they're pretty feature-packed. To compete, you'd either have to work silently for years and pop up with a complete replacement, or work your way up to their feature set over time. You can guess that I've chosen the latter.
My main goal is lowering the bar to entry, and to make the process more fun, more akin to actually faceting! This is achieved by a single app with a visual interface and immediate results. What I mean by this is you start cutting a facet, you see the result on your plan view as well as a raytraced output, even the cutting instructions. You change angle, depth, index or symmetry and you see how the cut would look before committing to it. You change the stone R.I. or colour and you see the final rendering change as you slide the value. You tweak tangent ratio, not by learning which angles work best for each design, but by visually seeing the result on the stone you'd be cutting.
Hopefully the following paragraph has sparked some excitement within some of you, I know I'm pretty excited about it!
I also want to make easier for people to share designs, straight from the app, maybe even directly to the Gemology Project (admins, interested? I just read the steps required to post a design here from GemCad ) Of course that raises questions about the "morality" of potentially a whole bunch of people flooding the place with designs that haven't been proof-cut but that's another discussion.
Now I've only been faceting for a few months, and even less time with GemCad/Ray. I've learned both apps well enough to use them but I'd like to put a call out to the pros around here, or at least the people who work with GemCad/Ray a LOT. (paging Dr Ahkavan, Mr Voltolini, Gem Ranger who I believe lives only a few KM from me, and anyone else I might have missed) I'd like to call on your experience with the software to narrow it down to the steps you ALWAYS take when doing new designs. If you had to cut out 75% of what those apps do, what would remain?
Hopefully the question isn't too open-ended and you see where I'm going with this. I'm not ready for an open beta yet but if any of the people mentioned above would like to see what I've got and try it, send me a PM. I would have bugged you privately myself but: new user = can't send PMs . At the very least you can then vouch for me to prove I'm not lying, without me having to reveal my whole deck just yet.
I don't want to exclude anyone either. I'd appreciate any input on the subject. Looking forward to your feedback. Thanks, and greetings!
Post subject: Re: Another "new gem design app" thread
Posted: Tue Nov 15, 2016 12:20 pm
Valued Contributor
Joined: Fri Nov 01, 2013 3:45 pm Posts: 289
Hello Rej,
I might know your wife: Clothilde! We have know each other since we were young.
She told her friends FB that her husband was faceting. So when I saw your name Rej and look where you live, I asked myself if you were not her husband.
Anyway, you have joined a great forum, as you will soon find out.
Post subject: Re: Another "new gem design app" thread
Posted: Tue Nov 15, 2016 5:06 pm
Valued Contributor
Joined: Thu Nov 10, 2016 11:29 pm Posts: 114 Location: Brisbane, Australia
thomas.adamas wrote:
Rej, I suggest you code this app such that it will run multi-platform (osx and linux), not just ms windows.
Yes, it's already running on Windows, Mac, Android and iOS. Linux involves running through WINE for now but the app doesn't have many dependencies so it might be possible to make a proper install package. Running through WINE is pretty simple though and I assume if you're on Linux that you're familiar with it.
I know some will recoil at the idea of an app made for mobile AND desktop but I promise I can make it work. (and designing on the iPad is really cool!)
Post subject: Re: Another "new gem design app" thread
Posted: Tue Nov 15, 2016 8:19 pm
Gold Member
Joined: Fri Oct 19, 2007 5:04 pm Posts: 1642 Location: Walnut Creek, CA
For designing, gemcad has pretty much all you need. It would be nice to add features for translating the cursor at specific angles, a drag&drop feature to move the tiers on the list quickly, highlighting the tier when clicking on the respective tier on the list below, and such small improvements. A feature missing that could be interesting would be the possibility to set the meetpoints as "nodes" and drag them in a virtual 3D snapgrid so they can only be set at integers for the given index wheel. A necessity? Definitely not. A nice feature? Yes. Complicated? Probably.
As for previewing/optimize the cut, my wishlist would be to have something with the capabilities of BOG (maps with merit figures and patterns for different angle ranges tangent-ratioing pavilion and crown). In addition to adding the classic "random" "iso" and "cosine" pattern I'd add a colored rings/hemispheres environment to better see patterns/head reflection issues. See e.g. here: viewtopic.php?f=8&t=18705&start=30
No real need for photorealistic renderings, even if they are nice, especially animations, to predict the outcome of the stone. I do like them, tho, but the good ones require proper libraries, setup, and a decent machine too. I won't run them on an iPad.
Post subject: Re: Another "new gem design app" thread
Posted: Tue Nov 15, 2016 8:31 pm
Valued Contributor
Joined: Thu Nov 10, 2016 11:29 pm Posts: 114 Location: Brisbane, Australia
maialetto wrote:
... In addition to adding the classic "random" "iso" and "cosine" pattern I'd add a colored rings/hemispheres environment to better see patterns/head reflection issues. See e.g. here: viewtopic.php?f=8&t=18705&start=30
Oh, I've done my research Marco ..
Here's your "Squilliant" in coloured rings.
In fact, you'll be able to add lighting models yourself by dropping new images into a folder.
Post subject: Re: Another "new gem design app" thread
Posted: Tue Nov 15, 2016 9:50 pm
Moderator
Joined: Fri Jun 25, 2010 10:06 pm Posts: 2267 Location: Chapel Hill, NC / Toronto, ON
Rej wrote:
I've already got a functional prototype, with all major technical hurdles out of the way.
Well damn, that's the right way to go about this! Have a prototype first, THEN ask for details.
Quote:
What I'd like to discuss is establishing a minimum feature set.
What are the minimum features necessary for a functional gem design app? 1. Ability to create the facet pattern easily, with the facets being the desired shape and position 2. Ability to scale the design easily 3. Ability to quickly see what the design will look like in different materials with different scaling factors.
With GemCAD, I can make an outline for a design pretty quickly and easily. For regular or high-symmetry designs, I can fill in the pavilion easily such that it establishes the girdle outline, then I can do whatever the hell I want with the crown. GemRay is also easy to use for scaling and rendering, although it's more time consuming than it has to be.
Quote:
Part of the problem with making a new design app is you're obviously going up against GemCad/GemRay. Those represent YEARS of work so they're pretty feature-packed. To compete, you'd either have to work silently for years and pop up with a complete replacement, or work your way up to their feature set over time. You can guess that I've chosen the latter.
This makes sense...but don't forget that people are going to be divided into two camps, those who want something easy/low functionality, and those who want powerful more complex tools. You'll miss out on the second segment with this approach, but as long as you continue to add the remaining features post-beta release, it's fine.
Quote:
My main goal is lowering the bar to entry
Excellent!
Quote:
and to make the process more fun, more akin to actually faceting!
So cheesy BUt I totally understand the impulse. Then again...GemCAD basically works like faceting. How would you make your app more like faceting in real life?
Quote:
This is achieved by a single app with a visual interface and immediate results. What I mean by this is you start cutting a facet, you see the result on your plan view as well as a raytraced output, even the cutting instructions. You change angle, depth, index or symmetry and you see how the cut would look before committing to it. You change the stone R.I. or colour and you see the final rendering change as you slide the value. You tweak tangent ratio, not by learning which angles work best for each design, but by visually seeing the result on the stone you'd be cutting.
I cannot begin to overstate how important the single-interface system would be. You'd still need the ability to branch out into more powerful tools from the same app, eventually, like BOG's render map, but that doesn't need to be a "now" thing. Immediately being able to see the effect on the end design is critically important, particularly for designs that run into the head-shadow problem, as fixing that in stubborn designs is a huge pain in the ass.
Quote:
I also want to make easier for people to share designs, straight from the app, maybe even directly to the Gemology Project (admins, interested? I just read the steps required to post a design here from GemCad ) Of course that raises questions about the "morality" of potentially a whole bunch of people flooding the place with designs that haven't been proof-cut but that's another discussion.
This is an excellent idea. There already exists a mostly-good database, FacetDiagrams.org, which is run by Robert Strickland, who created GemCAD. It has its downsides - poor searchability, for one, and lack of renders and/or test cut pics, but it's otherwise excellent. If you could somehow integrate and/or obsolete it (I'd HIGHLY recommend talking to Robert first, because he's awesome and would probably help you), maybe set up your new app such that you can easily upload to a database with proof of test cut... just an idea. I have to admit, when I was starting out I uploaded a batch of maybe 50 untested designs, and people got super-pissed. The majority of them were really good, but some were just shit. Don't do this.
Quote:
I'd like to put a call out to the pros around here, or at least the people who work with GemCad/Ray a LOT. (paging Dr Ahkavan, Mr Voltolini, Gem Ranger who I believe lives only a few KM from me, and anyone else I might have missed) I'd like to call on your experience with the software to narrow it down to the steps you ALWAYS take when doing new designs. If you had to cut out 75% of what those apps do, what would remain?
Looks like all three of us are responding back to back Here's what my design process normally looks like if I'm writing a new design based on specific rough. It's a disaster most of the time. 1. Make a girdle outline. 2. Make a minimum of 3 girdle-defining pavilion facets using vaguely appropriate angles. 3. Build off of those girdle-defining facets and the pre-existing girdle outline, while trying to stick to reasonable angles for light-reflecting facets, and not-too-steep facets for girdle-defining non-reflecting facets (i.e. barion facets). 4. Randomly dick around on the crown with vaguely appropriate angles until it looks decent. 5. Randomly trial-and-error crown and pavilion scaling in GemCAD using the pink-green test and looking at "Random" vs COS and SC2, because that gives me the best visualization of windowing and contrast respectively. 6. Randomly throw in or take out pavilion facets to fix the problems I see in the rendering. 7. More random test rendering with different scaling factors.
It's way more different if I'm going after a specific "look", like a tessellated design, or a high-dispersion design, or something with lots of bars.
I have to agree with Marco on the additional features: 1. Translating the cursor at specific angles. GemCAD allows you to move at right angles with the shift button, but I do a lot of triangular checkerboards, rhombi, and even pentagons. 2. Drag-and-drop for tier rearrangement 3. Contrast mapping, but with the ability to use custom ring-colour variants, including the two standard ones used in gem labs, as well as at least one built-in one with a smooth gradient. 4. Assigning meetpoints as "nodes" with drag-and-drop (holy shit this would be hard but amazing) 5. BOG-style render mapping across a range of scaling factors for crown and pavilion 6. Optimization for dispersion 7. The ability to easily generate a performance map for light output as seen here 8. Better support for identifying windowing.
Post subject: Re: Another "new gem design app" thread
Posted: Tue Nov 15, 2016 10:49 pm
Valued Contributor
Joined: Thu Nov 10, 2016 11:29 pm Posts: 114 Location: Brisbane, Australia
Thanks for the in-depth answer Arya, I've added some responses inline, hopefully without creating a quote mess.
Faceting 101 by Arya wrote:
but don't forget that people are going to be divided into two camps, those who want something easy/low functionality, and those who want powerful more complex tools. You'll miss out on the second segment with this approach, but as long as you continue to add the remaining features post-beta release, it's fine.
Agreed, I'll never tout this is a total GemCAD replacement for the hard-core users, well maybe one day if this works out but right now my focus is on getting the 5% of features that get used 95% of the time, working.
Faceting 101 by Arya wrote:
So cheesy BUt I totally understand the impulse. Then again...GemCAD basically works like faceting. How would you make your app more like faceting in real life?
Maybe cheezy, I hate marketing so that's the last of that, but what I mean by that is when you're faceting, you can set up you machine, touch the stone to the disc and check what's happening. Sometimes you can even correct mistakes if you haven't cut too far. My app is more like a real machine in that you "lower" the mast and see how your cut progresses as you do that. And of course because it's virtual, raising the "mast" again grows your rough back so you can adjust your settings without having to rely so much on the undo feature. With GemCAD it's set up and commit to the cut. Plus there's the dreaded "X facets missed the stone" message why I always get and grinds my gears for some reason. That ties into lowering the bar for beginners. Warning popups contribute to a bad user experience.
Faceting 101 by Arya wrote:
I cannot begin to overstate how important the single-interface system would be. You'd still need the ability to branch out into more powerful tools from the same app, eventually, like BOG's render map, but that doesn't need to be a "now" thing. Immediately being able to see the effect on the end design is critically important, particularly for designs that run into the head-shadow problem, as fixing that in stubborn designs is a huge pain in the ass.
Exactly, I've imagined how some of those things could be done but I'm not going to touch that until I've got a first release ready with the basic "build" steps completed.
Faceting 101 by Arya wrote:
This is an excellent idea. There already exists a mostly-good database, FacetDiagrams.org, which is run by Robert Strickland, who created GemCAD. It has its downsides - poor searchability, for one, and lack of renders and/or test cut pics, but it's otherwise excellent. If you could somehow integrate and/or obsolete it (I'd HIGHLY recommend talking to Robert first, because he's awesome and would probably help you), maybe set up your new app such that you can easily upload to a database with proof of test cut... just an idea. I have to admit, when I was starting out I uploaded a batch of maybe 50 untested designs, and people got super-pissed. The majority of them were really good, but some were just shit. Don't do this.
I'm not sure how cooperative Mr Strickland would be to a potential competitor... Honestly I suck at everything web, which his why I'm hoping there's an API out there that I could plug into from the app. It would suck to have yet another design database website but if someone out there has these skills and would like to team up on this... I think one of the people proposing a new design app had web skillz.
Faceting 101 by Arya wrote:
Here's what my design process normally looks like if I'm writing a new design based on specific rough. It's a disaster most of the time. 1. Make a girdle outline. 2. Make a minimum of 3 girdle-defining pavilion facets using vaguely appropriate angles. 3. Build off of those girdle-defining facets and the pre-existing girdle outline, while trying to stick to reasonable angles for light-reflecting facets, and not-too-steep facets for girdle-defining non-reflecting facets (i.e. barion facets). 4. Randomly dick around on the crown with vaguely appropriate angles until it looks decent. 5. Randomly trial-and-error crown and pavilion scaling in GemCAD using the pink-green test and looking at "Random" vs COS and SC2, because that gives me the best visualization of windowing and contrast respectively. 6. Randomly throw in or take out pavilion facets to fix the problems I see in the rendering. 7. More random test rendering with different scaling factors.
Lol, wow that's a lot of randomness. Maybe I should just make a big "I'm feeling lucky" button and see what comes out! But seriously this is good info. I wondered if people had a gem in their head and tried to get it on screen (i'm sure that happens too) or if there was more experimenting involved. This reinforces my "real-time" aspect as you'll be able to instantly see any changes you make affect your resulting polished gem.
Faceting 101 by Arya wrote:
It's way more different if I'm going after a specific "look", like a tessellated design, or a high-dispersion design, or something with lots of bars.
Ah, there we go.
Faceting 101 by Arya wrote:
I have to agree with Marco on the additional features: 1. Translating the cursor at specific angles. GemCAD allows you to move at right angles with the shift button, but I do a lot of triangular checkerboards, rhombi, and even pentagons.
I have to admit I have no idea what this means or how you'd use it. Don't worry about it for now, I've got plenty to go on.
Faceting 101 by Arya wrote:
2. Drag-and-drop for tier rearrangement
Easy™
Faceting 101 by Arya wrote:
3. Contrast mapping, but with the ability to use custom ring-colour variants, including the two standard ones used in gem labs, as well as at least one built-in one with a smooth gradient.
I'm not familiar with the standards but like I said the lighting models list is open to adding your own environments. As long as these two standards can be mapped as "half-dome" maps, they can be added in my raytracer.
Faceting 101 by Arya wrote:
4. Assigning meetpoints as "nodes" with drag-and-drop (holy shit this would be hard but amazing)
Indeed, it would be a good challenge...
Faceting 101 by Arya wrote:
5. BOG-style render mapping across a range of scaling factors for crown and pavilion 6. Optimization for dispersion 7. The ability to easily generate a performance map for light output as seen here
Yes, I need to do more digging into these, I'd imagine them as "plug-in" style tools.
Faceting 101 by Arya wrote:
8. Better support for identifying windowing.
I've got support for changing the background colour as well as the head shadow color. This combined with the stone color gives you a pretty good indication of windowing but I could always add another color for leakage like Gemray does. Or is there another solution?
Faceting 101 by Arya wrote:
Also I sent you a PM.
I see that because I got an email notification about it (and another I got from another user) buuuut ... "You are not authorised to read private messages." Maybe they'll show up once I hit 10 posts, maybe they're lost forever.
Actually, Gem Ranger sent me a message via the forum's email feature and I got it, so maybe try again using that instead ?
Last edited by Rej on Wed Nov 16, 2016 1:36 am, edited 1 time in total.
I would like to be able to change the angle of one facet tier, and not have the rest of the stone tangent ratio.
And what would your expectation be re the other tiers and meetpoints?
My expectation would be, if I was working with say the pavilion of of a SRB, and had the first tier at 43 degrees, and the second at 40, if I changed 43 to 44, that the tier at 40 would not become something like 41, but be able to remain at 40. The meets would stay the same. Of course there could be conditions where subsequent meets would not work out, but for many designs they could. I manually do this all the time with GemCAD, but it involves deleting and then re adding the other tiers.
Post subject: Re: Another "new gem design app" thread
Posted: Wed Nov 16, 2016 12:34 am
Valued Contributor
Joined: Mon Nov 08, 2010 9:18 pm Posts: 155 Location: Northern California
Precision Gem wrote:
Caedmon wrote:
Precision Gem wrote:
I would like to be able to change the angle of one facet tier, and not have the rest of the stone tangent ratio.
And what would your expectation be re the other tiers and meetpoints?
My expectation would be, if I was working with say the pavilion of of a SRB, and had the first tier at 43 degrees, and the second at 40, if I changed 43 to 44, that the tier at 40 would not become something like 41, but be able to remain at 40. The meets would stay the same. Of course there could be conditions where subsequent meets would not work out, but for many designs they could. I manually do this all the time with GemCAD, but it involves deleting and then re adding the other tiers.
I see how it could be useful in some cases. In the general case you'd potentially be cutting off existing facets or opening up uncut surfaces where the volume bounded by the facets is very large or even infinite, exceeding any reasonable rough dimensions.
Doing it via some sort of drag interaction would be nice too.
Users browsing this forum: No registered users and 34 guests
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot post attachments in this forum