The ramifications of Zbrush 3.5 and beyond

[Clappy71]

After giving Zbrush 3.5 a full workout, it made me wonder what things will look like in CG down the road.

I have been around long enough to "never say never". When I first started out Nurbs were all the rage. The vast majority of artists disdained Poly's for their inaccurate results. Then along came MJ Poly Tools & Polyboost and the rest is history.

Next revolution is Zbrush and organic high poly modeling bites the dust. Hard surface poly modeling survives.

Than along comes Zbrush 3.5...

Here are some possibilties going forward:

1) the death of the big 3D app. Why would anyone (or studio) spend $3500 per license on a program that your only using for one specific task? As it stands, Maya just became my rendering plugin for Zbrush. I could see there being 2-3 smaller & less expensive apps which combine to do everything.

2) the death of poly and box modeling. Zspheres2 are way quicker than building a base mesh with polys. And with the new planar brushes you could probably model something like a car, or a building, as fast as poly's, once you got the hang of them. Future releases will probably make this stuff even easier and faster. Poly modeling could be just for retopology.

3) No more bones. Zspheres (or something like them) could be the future of rigging.


Just a thought.

MC

[AVBN5000]

I think you are right on target with your statements! I bought Zbrush because of its features and the new anticipated release of 3.5 and the upcoming 4. What I liked about this application was how much is cost as well. In a 3D world...this is a steal even though it is pricy....but nothing like its big brothers. Zbrush minimizes time and gets quicker results. I have nothing against the other applications though. I was trained in Maya and still think its a joy to work with. Zbrush has just opened up a lot of doors much quicker then Maya could based on its structure of a program.

I do think that Zbrush is and will be the future to come. Just think how nice it would be to take the Zsphere concept and just plop the controls into Face Robot and get that rig up and running! I have no doubt that Zbrush will be injected into every studio and will be a key asset to the CG Media Industry if it isnt already.

[splodge]

Immediate future: NURBS - NURBS modeling combined with sculpted displacement maps. There'll be no need for retopology.

Distant Future: Voxels - All 3D engines will eventually be voxel based. Again, no need for retopology.


Just maybe! :)

[slocik]

Come on,zbrush is great softwere for ogranic modelingand simple hard edge, but thats the only thing its good for. zbrush 3,5 didnt introduce anything new, just little updates to make stuff easier for the user, but the zspheres are a joke if you weant to create a complex model, even a base mesh. You would be NEVER, EVER, NOT EVEN AFTER SPENING A YEAR MAKE A CAR IN ZBRUSH !! IF you started with box modeling you schould know that.
1, why would they ? But what about, uvs, animations, rendering, rigging, and so on, 3d aps arent only about modeling, and when your working on somthing bigger you dont want to switch betwen 20 smaller softwere aplications every few minutes. Even Modo that startyed as mostly modeling soft, has animations and a complex rendering system now ;f ame goes for zbuesh, they are adding more and more stuff, so you dont need other aps, it will eventually end being the next maya.
2. Polymodeling will never die, especialy box modeling, you are still thinking only about organic models not the bigger picture, monsters and human models isnt the only thing thats needed in cg industry ;f Try doiung a retopology on somting that will divide nice wirth hard edges...
3. But zspheres for bones ...... are still bones ;D You saying we will stoip using bones for animations, because we will start using bones ?;f

I dont see any changes in the next 10 years coming, the only one is either zbrush dying, and soft like 3d coat taking over, or zbrush moving to real 3d with voxels. I bet that when zbrush 5 or 6 comes out, i will be still using modo 805 for modeling ;D

[John Strieder]

As you mention car modeling, take a look at this sculpt from this great artist:

http://crashmgn.deviantart.com/art/furai-from-zbrush-109008168

(Done with ZBrush 3.1)

[Darukin]

I agree, Zbrush has a huge potential for the future, if you ask me they've had a very humble beginning and they've only just begun! Not to mention they created a new style of going about 3d completely, Zbrush is very unique.

I dunno about the fate of 3d technology when it comes to polies or voxels (it also depends on computer speeds in future) but Zbrush is already an integral part of many pipelines, and also for personal works, see any of the best images on all major CG / 3d art sites, and u will see that over 90% of character arts use Zbrush (at some point at least).

With 3.5, hardsurface seems to have improved a lot and they're still pushing

As far as ZSphere bones, yes there is amazing potential, check out my thread to see this potential:

http://www.zbrushcentral.com/zbc/showthread.php?t=74724

I believe the future of CG is bright!
Cheers!

[jamespthornton]

1) the death of the big 3D app. Why would anyone (or studio) spend $3500 per license on a program that your only using for one specific task? As it stands, Maya just became my rendering plugin for Zbrush.

That's a stretch based on your use of those apps. Most modelers and studios use bigger 3D apps for WAAAAAAAAAAAY more than one task, most of which ZBrush would be silly to try and replace. Never.



2) the death of poly and box modeling. Zspheres2 are way quicker than building a base mesh with polys. And with the new planar brushes you could probably model something like a car, or a building, as fast as poly's, once you got the hang of them. Future releases will probably make this stuff even easier and faster. Poly modeling could be just for retopology.

For many types of hard surface models where freehand is useful, the planar brushes ARE a replacement. Now that's an accomplishment! It's brilliant! I haven't even used them, I don't have to, I can already tell!

Yet for so many other types of models, they will remain the far less useful tool. For instance, an architectural visualization artist will not be using planar brushes to accurately build the interior and exterior of structures. Nor will an engineer accurately designing complex parts. In many cases, true modeling is the definite answer offering the control you need. So I still gotta throw out a Never.

It's good that ZBrush is heading in it's own direction, and focusing on their strengths. This will keep them ahead of the game in their niche department, and keep them around in the future. Trying to replace bigger modelers and renderers would be obviously counter productive for them. Specializing in the brush based artistic tools and keeping the price within everyday artist's reach is what makes them so awesome and successful. They're doing a hell of a job at it too! I'm a huge fan! :tu:

Still though, these apps are just tools on a shelf. Grab the one that works best for the task at hand. Don't hammer a screw just because the hammer is incredibly cool.

[Clappy71]

:cool:I realize studios use apps like Maya for many things. But those tasks will dwindle if they don't start producing. For example, the most cutting edge apps are Zbrush, Vue, Hypershot etc..all of the innovation is happening outside the walls of Autodesk.

Eventually you should be able to model/sculpt, texture, rig, animate, and render, in a combination of 2-3 smaller apps which do a better job than the big clunkers, for cheaper. We just need an app to do for animation what Zbrush has done for modeling. ZMotion??

And call me crazy but I can see a day where even industrial designers and archetects sculpt, or 3D paint, their innovations instead of crappy CAD drawings. They are artists, they don't want to pull verts all day. I am sure Pixo or some other company will come up with a super accurate way of scuplting for design work.

It is early in the game yet. Can only imagine what a couple of years will bring about.

Ok, enough procrastination. Back to work.

Cheers,

MC

[jaystein777]

None of the above.

None of the tools that are slated for demise will disappear.
It's like throwing away a hammer when you get a nail gun, there's just some times that you will need a hammer.

Sorry but the mentality of this thread is a bit backwards, all tools are tools and are needed for specific situations.
In fact the new ZSpheres are just Metareyes but properly evolved.

[splodge]

jaystein - metareyes, lol...

splodge - nurbs are just for defining certain types of surfaces, nurbs and Maya's subDs don't render -not without undergoing a conversion to polys. That's because a (poly)triangle is the most economical surface definition possible in 3D, it's the equivalent of a square (pixel) in bitmap space.


Of course NURBS are converted to polys. I never said they weren't! And I was talking about the creation side, and not necessarily the final output. :)




And sorry but you can't have a voxel-based 3D engine. Voxels don't exist in 3D, so even if a futuristic ZB would somehow incorporate animation, you'd wind up with a voxel-based sprite-engine. And such an engine cannot emulate what can take place, over time, in 3D.


You can have a voxel based engine. You don't have to convert everything into polygons if you're rendering using raytracing rather than rasterization. :p

There may come a time when 3D scenes will become so highly detailed that polygons no longer offer any real benefits. But this is still long way off.


Still, I admit to leaving a lot of blanks in my earlier post, so I kind of expected to be misunderstood. No worries! :)

[Clappy71]

jaystein - metareyes, lol...



Clappy -

And it's for the same reason why no one with an inkling in 3D would ever commission a ZB car anymore than a ZB teapot. Your disdain for cad like tools doesn't indicate a superior aptitude or intellect.

-G
Not sure where this is coming from. Not claiming to have any superior anything. :D. Just having fun predicting the future. Have a nice day dude.

MC

[sirquadalot]

I would like a a haptics

[dustinbrown]

I agree with guringo. Further, at the risk of sounding like I don't care (I do), I'm not terribly concerned about what my software will be like years from now.

When it comes down to it, I don't choose the software I invest my time and energy in. The job market makes those decisions for me. If the industry news I read mostly tells of Maya, Max, and ZBrush being used in the field, then that's where I dedicate my time - period. If that changes, and five years from now most of the jobs I want require a working knowledge of LightWave, 3D Coat, or Mudbox, then I'll bate and swich so fast it will make your head spin. This is the singular reason that, a year ago, I choose to invest my time in learning ZBrush instead of Mudbox - self-marketability, pure and simple. ZBrush was the bee's knees and Mudbox at the time was a newcomer...I don't even believe it had been acquired by ADSK yet.

I don't even care if Mudbox is the superior product. I may care on principle, but it won't show in my actions. I'll always go with the thing that stands the best chance of landing me the job I want.

I have zero allegiance to a software package. It's just a tool. What about personal preference? My personal preference is to be employed.

[Clappy71]

hey MC, this isn't personal...


in no particular order;

1. polys = ugly, heavy, disdainful, inaccurate, soon to be extinct...

Very common, especially in this forum, but it's just the opposite, the poly is the most economical surface possible in 3D, so it's way more likely that future improvements to hardware and software will allow us to handle vast amounts which in turn may cause us to abandon voxels -but it's even more likely we'll just keep using both for what they're actually suppose to do.. Either way voxels can never substitute polys, and I can say 'never' because a poly already is a 3D voxel. So no 'death of polys', no 'death of hard-surface or hi-poly modeling'...

I did not say that Polys are currently inaccurate. I was referring to when I started out (about ten years ago), Polys were an inaccurate tool to work with: you could not an insert edge loop, there were no loop or ring select tools, there was no sub-d preview. People did not use them, gaming was not what it is today and most of the work was in film. They used Nurbs (lofting curves). But things changed. and poly's got more user friendly. And things will change again, as they always do.

What I mean by "death of Poly modeling, is the death of extruding edges and pulling verts. Into something more like drawing. Not the actual format of the geometry, or what something is convertied to at render time.


2. cad tools - crappy, disdainful, ugh, soon to be extinct...

You wouldn't pass a job interview based on that. You're expected to know and use the right tools for the job, not opt for fun ones. Perhaps worth mentioning is how rewarding it is to do so. And how excruciating it can be when you just think you do.

Would never be in an interview for a CAD job. But my colleagues who are toy designers and architects dream of more creative feeling tools. But point taken.

3. Maya - render plugin w/ dwindling tasks, soon to be extinct...

That Maya is now no more than your render plugin just screams 'I DON'T KNOW JACK'. At least to those of us who know both jack and ****.

And fyi, studios are not using Maya in some diminished dwindling task/chore capacity. Nor is anyone worth his salt "pulling verts all day", in fact, that would pretty much epitomize a clunkhead...

You may be great at what you do, but you definitely need to break away from your partially informed biases and to that end, I'd suggest catching up on some technical lit'.

Wow. This is actually my fault for not further explaining what I meant. When I first started out the studio I worked for used Maya for everything but compositing. There was no Zbrush, no Vue, no UVLayout, no Motionbuilder, just Maya, Max, and Softimage. These days, very few people I know are using one package for everything because there are smaller programs which do a better job at specific tasks. Maybe your experience has been different. So I think as programs like Modo and Zbrush mature and other apps pop up, there will be little need for the "all in one package".


As for me, (started out as an animator) at this stage of my career I am focusing on character modeling. So with a full understanding of Zbrush I really only need Maya to create my shader networks and to render. I used to have to rig them as well, but with the advent of Transpose Master, I can demonstrate animation friendly rigs without having to spend days on a fully scripted rig.

Don't get me wrong I am longtime user of both Maya and Max and will continue to be so until it is no longer necessary.



At the end of the day I don't really care one way or the other. As long as the art looks good. I was just enjoying looking down the road from my own perspective, even if it is uniformed and biased.

Cheerio.

Matt

[slocik]

But you started a discussion even without basic knowlage about how 3d soft works, you schould care, and you can learn some important stuff on the way ;f You can create mindblowing sculpt in zbrush without knowing what a voxel and polygon is, but its always better to know this stuff if you want to predict the future of the industry ;] On other hand you will never graps maya fully, im in 3d for many years now, and i still know very litlte ;]
Pologons will never die, since without them you cant create any 3d model ;]

[Disco Stu]

Hardsurface Subdivision with polygons will never die.
Zb3.5´s hardsurface brushes make this pretty obvious:p

[splodge]

splodge,

you're still under a false premise or two.. As there's no right time, side or reason to consider nurbs for organic modeling.

It's a common misconception to think of nurbs as 'smart' smooth surfaces without the hassle or alleged penalties of polys.

Apparently you can also know various facts about nurbs and still misconstrue them. ;)

Nurbs are heavier than polys for all the additional functionality it accomodates, none of which happen to be applicable to organic freeforms.
The heaviness you may not necessarily feel but nurbs,as opposed to polys, do require you to toe the line (or seams).

Now I wouldn't say that box-modeling is always the optimal approach for organic forms. I know it's not. You can, for instance, loft a leg from nurbCircles, but you 'output' polys and not nurbs.

The reason, in one word; 'polysubDs'.

You don't need to polysubD necessarily but doing so would of course emulate nurb tessalation or whatever 'smoothing' you care for.

PolySubDs are nurb patches without the 'patching' or nurb-irks, such as seams, uvs, retopologizing etc... and because most nurb operations/tools have a poly equivalent, its possible to polysubD a car with almost the same ease as with nurbs but the reverse is not true when it comes to organic forms, which you probably knew, as well as how both nurbs and polySubDs max out at the 'plastic-organic' look...

The polysmooth function was always available in Maya. So was connecting the outmesh to the inmesh attribute between two poly shapenodes. And likewise resolving 'shells', the equivalent of stitching nurb patches, using polyUnite and polySewEdges commands... Except no one really cared at first, nurbs were considered cool, in spite of their limitations, while polys were considered plain, tedious, etc....

Dirk Bialluch, back in 99, uploaded his CPS toolset to hiend3D (connectPolyShape), one can think of his polySubD as an early ztool in true 3D... but again, due to misconceptions, it took a while to get the ball rolling... Alias honored him with a Maya Master in '01... and yet it took a number of versions before Maya incorporated/legitimized 'proxies'...

For a brief while regular subDs were in mode (Jerri's Game) but it didn't take, and as far as I'm concerned, rightly so. I'd be hard press to come up with something that I would prefer to contend with in subD than polysubD or nurbs or 'voxelize'....

This more or less covers why at no point in time can nurbs make a comeback and why, in spite of better options, even pros used to model characters with nurbs...

It's not a question of raytracing vs raster. Voxels simply don't exist in 3D, that's why you can't have a voxel-based 3D engine.

If voxels could exist in 3D - they'd be called polygons.

If you want to capture 'voxel data' and have it stand in for actual geometry in actual 3D, try normal maps.

And lets say normal maps could evolve into voxel-maps, they wouldn't behave/respond as polys. So you'd need a substantial amount of underlying polys for cloth simulation and sub-skin deformation etc, and it would still be considered a cheat of limited application/value and one collosal headache to integrate..

Bear in mind that pros/studios don't sweat the polycount, as they optimize for articulation/rendering purposes, and this means, among other things, that things that don't need to be 3D geometry, aren't, so no one's waiting for voxel-maps nor can some stupendous futuristic polycount warrant collapsing 3D to 2D....

And for what its worth, I'm still under certain misconceptions about odd things and then probably some I'm even oblivious to.. as such I really embrace the opportunity to stand corrected. And I don't hold your misconceptions against you, but if I were your supervisor, I wouldn't take kindly to "I didn't say that" and "I know that!".....

all the best, -G
I don't really agree with most of what youv'e said, but I'm not really the argumentative sort. So I think we'll just have to agree to disagree and move on. :)

[Clappy71]

Clappy,

fair enough, for starters you sound a lot more grounded than in your first and second post... So you do know some jack and your points are valid within your context...

I thought I had a decent understanding of Maya till I started learning mel and I'm still a couple of years away before I'll be ready to tackle the API...

Sometimes you just need to know that you need 5 years in maya before tackling mel, to not burn out on it beforehand or waste 5 additional years before rolling up the sleeves....


cheers, -GCheers to you too. :cool: I suffer from thinking I am communicating everything clearly, when in fact it is all over the map. Probably why it took me 6 years to finish university!! I am also prone to drama, i.e. using words like "death".

Type in "3D Sketching" into Google and you will find some pretty amazing things. You can see where the hardsurface and design stuff is headed. Albeit, it can take ten years from university lab to an actual product.

Check this one out. It does not create 3D geometry (yet), but looks promising:

http://www.baekdal.com/future/Interaction-Design/3d-sketching/

Take care,

Matt

[10000hours]

:cry:im so sorry:cry:

i wont ask anymore,okay ?

[slocik]

just my curiousity...dont get offended mkay ?
i feelofended by your question,i hope you apologize !

Clappy71 you seriously lost me there, i think you dont know what hard surface modeling means, it has nothing to do with drawing curves in 3d, you could do that for years now in every 3d package ...

[10000hours]

:DHahaha, you got me. That's almost how i felt lately. I think it is a good advice making make moderate goals and try your best without rushing and making it too perfect. I don't mean to impose anything from the username i used. I used it because it motivated me to use the best of my time, efforts and sacrifices to do what I'm doing right now. Thanks man.:idea:

[Burchigb]

Personal thoughts only:
Seen some amazing things come from the CG world. Many films, graphics, 2D print and more.
Amazing that the talent that goes in to what I have seen so far.

MAYA with the price tag of $5,000 is WAY too much for a person like me. The tools are sluggish and there is a real steep learning curve. Not saying it could be done but it would take too long to figure out something simple and should flow with the artist/tech illustrator. HINT

With that being stated there once again is a cost factor. An average student or some who uses this as a hobby cannot afford the buy in and the repeated updates at high cost.

I have been using Poser/Lightwave/ZBrush for a couple of years now and I am using it as a hobby. My work flow across these programs have been harsh to the point of throwing the computer out the second floor. Other times I hit it on the head.

ZBrush and those who contribute to it are by far one the best forums that I have seen. The people who generated and maintain this program are amazing. The keep generating tool sets that for me are based off of “Wow you can do that!” . The amazing art that has been generated is beyond words.

The path that ZBrush is going towards I hope will be based off of industry standards so that people like me can follow and contribute. OBJ since they are easily exported and can/Should be easily imported by all 3D programs. Based on common ground so that rapid generation of a creation can/should be possible.

My flow is to use Poser to generate the basic figure and exporting it to Lightwave, setting up the bones and basic rig so that I can bring life to my creation or inspiration is quite neat. Then bringing it into the ZBrush environment and putting those finer touches to bring it to a specific level of my expectations then export to Lightwave for final processing (when it works and multiple times) is something I strive for.

Doing this in one day or a couple of hours would be great but something that detailed and that meets my inner perfectionist takes time and being able to focus on the end expectation. PaintStop which is a very addictive program which has consumed much of my time hopefully will be integrated into a better work flow so that I can paintstop something and bring it to the 3D environment rapidly then port it over to a program such as Lightwave for the render/motion would be amazing.

ZBrush has come a long way since version 2. I hope they go forward and keep the pace up.

With the new tools that ZBrush has came out with the ability to generate muscle (definition) and other types of movement is very possible in the near future. CG world has never to date been able to do that naturally.

Ideally I should (when tech comes of age) put on the goggles, gloves, boot up the computer and start building my character/environment. Would be cool to work in the VR environment. Would be amazing to see what would/could be generated.

I keep using the word amazing a great deal because to me this is like having an ice cream on Sunday. A real treat and it gets better every week once I learn more.

Sorry got long winded. Just ZBrush is going towards something hopefully towards what the artist needs and can use.

[MudboxRules]

My take is that the industry is becoming more left-brained. Things like base mesh topology and absolute rigid control over vertices is a thing of the past. Zbrush turned modeling into sculpting. Whether it will turn maya into a rendering program is still up for grabs. The significance of 3.5 is that one can create a mesh, normal map, and texture straight in zbrush. the key to sealing the deal will really be GoZ, I think that GoZ will present a very serious ramification to Autodesk, if GoZ is what pixo says it is. And with Z4's spotlight and lightbox looming on the distant horizon...

Edit: :o Right brained MY BAD! lol

[Intervain]

most of what I'm thinking has already been said here but it's quite interesting to see how you boil the industry down to modeling. I would like to see you attempt particle simulation or simple matchmoving in Zbrush LOL.

It's interesting that Modo is given as an example of moving away from 'all in one' packages. I think the development of Modo is proof to the contrary - as it's slowly growing to be one of the so called 'clankers'. The same can be said for some other lower end modelers.

Having all those small packages like Zbrush or topo gun or unfold 3d around is great but I feel a clancker with the abilities of all of them would be even better. Having to export and import objs continually and dealing with issues that it sometimes causes can be a pain in the patella in the production pipeline - especially when you're dealing with riggers waiting for you to correct stuff or texture artists who like to use their own way of unwrapping which may prove to make the model impossible to import back to Zbrush all of a sudden, because cutting the mesh changed the point order 0_o...

So yeah clanckers can be annoying and are certainly more difficult to learn but they do have their advantages... I love my xsi and find it totally not clanky to be honest...

[Clappy71]

Guringo: I did notice that Autodesk is coming out with a similar tool. So much for my theory! Thanks for the productivity tips, I am definitely going to use the egg timer method. I actually started a group with a bunch of artists which is about meeting deadlines for your own projects. We meet every two weeks and present our progress. The week in between you have to upload your results to a google group we created. Has really helped me progress with my modeling portfolio. Being accountable to others is a great motivator. I am going to present some of those ideas you mentioned at our meeting next week (with full credit to you of course). :cool:

Intervain: I must say that XSI users always seem to be the least bitter. I think I will have to check it out once I have landed my next gig.

Glad this thread sparked some lively and informative discussion. Even if it was born out a rant.

MC

[MudboxRules]

most of what I'm thinking has already been said here but it's quite interesting to see how you boil the industry down to modeling. I would like to see you attempt particle simulation or simple matchmoving in Zbrush LOL.:D You're right hehe, I just reread my post and it seems like I need to stop posting late at night. All in one programs like max and maya most likely won't be replaced. You're right, they have hair, particles, scenes, animations, shading networks, etc. I myself am learning maya right now. It's great to see some of the amazing stuff when the two have been used in conjunction.

[Intervain]

Intervain: I must say that XSI users always seem to be the least bitter. I think I will have to check it out once I have landed my next gig.
that's because it's a really clean and well organized software. Has it's problems but for a modeler and especially one using Zbrush a lot it's perfect, since you can import really highres meshes [ I mean I imported level 5 meshes - more than 1 subtool] into xsi and it handles them easily [in comparison to Maya which chokes badly on a mid res mesh]

[maddox.br]

If you know what you're doing in Maya and did your retopo homework, there's little to no reason to have a level 5 mesh in there. Actually I wonder what such a case would be, care to enlighten? :rolleyes:

[Disco Stu]

Prettyness in the viewport :D

[maddox.br]

How can a super-dense mesh be pretty in the viewport? Well, maybe if you never use the wireframe mode I guess, but it's still useless from a practical standpoint.

Anyways, level 3 is also quite pretty, except if you're modeling from a standard sphere heh :D

[TezZy]

but isn't it just nice knowing that your 3d program of choice can handle lots of polygons :D......just saying

[taemee]

ZBrush is a very good sculpting "tool" that has done some improvement in its hard surface modeling capabilities, recently. Nothing more than that.full stop. :cool:

[Intervain]

well in xsi for one and it actually often takes less time to render the high res mesh than the displacement maddox.br .. and yeah I made my homework thnx

my choice of method that's it - I'd rather render the high res mesh because I dont' bother about a displacement seams not working and often it looks better anyway :P

[maddox.br]

well in xsi for one and it actually often takes less time to render the high res mesh than the displacement maddox.br ..I do admit Mental ray's dispmaps are not only slow but kinda suck as well but that's *definitely* not the case if you use renderman, its displacement just flies.

Of course anyone's entitled to use anything he wants/likes, and don't take me wrong, I'm actually a big fan of XSI and wish I could wrap my mind around its workflow and endless menus. All I'm saying is that I find it hard to believe anyone actually wants to animate a 2 million polygon object, but whatever fills your cup.. ;)