PDA

View Full Version : Opinions: Lightwave vs. Modo.



Mr. Limpet
01-07-2014, 10:04 PM
Hi guys, just a general question. I'll state my circumstances first.
I have had Lightwave since 2000, but have not used it for much more than occasional flying logos.
I am a video editor by trade with some graphic skills. I don't spend a whole lot of time in 3D.

But now I want to ramp it up and become much more proficient in 3D.

About 4 years ago I started using Modo for commercial work. What I liked about it was that you could turn on a basic global illumination and it looked good.
Not much noise and reasonable render time. Prior to that I had difficulty in LW with high noise levels with GI and I gave up on it.

So, I am going to diligently train in 3D but I don't think I have the time now to do both Lightwave and Modo. It's one or the other.

Obviously LW is a more fully featured 3D app. So I would like to go that direction.
But how do the current versions of those apps compare in terms of render quality and speed.

I have Modo 601 and the most current version of LW as I have been fairly diligent at keeping things up to date whether they get used or not.

Thoughts?

Thanks for any input.

JohnMarchant
01-07-2014, 10:21 PM
Hi,

Hope this is not going to start a conflict here :):). Personally i would go with LightWave. If you have Modo 601, then you have a very good modeler in that package, but as you have already said LightWave is more fully featured than Modo and this is not going to change in the near future. LightWave falls behind in my opinion over Modo in the modeling aspect, LightWave needs some real modeler TLC and hopefully this will come soon.

I think LightWave still has the best out of the box renderer going and you can have it on as many machines as you want.

So really you have all you need but as far as learning and where to put your money in the future i would go with LightWave. I would also look into Blender as well, the UI for me is not great but its because im not used to it. I really have to hunker down with it and learn. Its great for fluids which is where LightWave lacks and of course its free. Check out Dan Ablan's tutorials on LightWave 11 and onwards, they really helped me as LightWave 10 and 11 are quite different.

These are all my opinions and of course others would feel differently :):):)

Mr. Limpet
01-07-2014, 10:57 PM
Thanks for your reply John.
Let me just say that I didn't intend to start a dust up.
I have both products and will keep using both.
Just looking for which one would be better career wise to tackle first since I can't really do both at the same time.

I am leaning towards LW since it is more fully featured, but am open to suggestions.

JohnMarchant
01-08-2014, 03:15 AM
Career wise, well if you will be doing it yourself and working freelance then its up to you. Big VFX houses tend to use Maya, XSI,Houdini, ZBrush and a few others. LightWave is used allot in TV VFX work as its quite easy to get good results fairly quickly. The previous software mentioned of course require a bigger budget. What you have is a very good mid range setup and i would concentrate on the LW aspect, Modo for some modeling and of course this setup requires allot smaller budget.

Look at some plugins for LW, i can suggest LWCAD, a must for the short comings of modeler and not just for archviz work. If rigging then RHiggit v 2 is a very powerful plugin. If you need multipass rendering for compostion then Janus is about the best there is for LW, i have it and its great.

There wont be any dust up as people tend to avoid getting into the one software versus another software threads which could explain why you dont have many replies although nearly 90 people have viewed it.

50one
01-08-2014, 04:56 AM
Depends what you do, try both. That's all I've got to say, anything more and this thread will be deleted, just like the previous one we had on MODO...

Darth Mole
01-08-2014, 05:25 AM
I used to own both, now I only own LW. Just never did get on with modo's weirdnesses and constant crashing.

Phil
01-08-2014, 05:55 AM
Opinions are ten-a-penny. Honestly, I find character work far more enjoyable in modo, given the immediacy of weight painting tools and the awesome addon that is ACS. I've not really had as much fun with character work in a long, long time. The same goes for particles and dynamics (in 701), due to the clever implementations of both.

Up until 701, for me, modo was a distinct secondary tool that had lots of potential, however they have made significant performance improvements to the point that I can actually do a growing amount of my primary work in there. They have made several orders of magnitude of performance improvement with tiled images, for example - and that's no exaggeration. In 601, these image formats were basically unusable. infiniMap on LW still has an edge of roughly 10-25% at times.

They each have their own strong and weak points.

As for render quality and speed, LW makes it terribly easy in my experience. You can rely very heavily on the adaptive sampling to clean up noise. As such, set your recursion limits to meet your needs in case of artifacts. Set shading and lighting samples to ~low numbers (1 or 2 per pass in most of my cases). Set the minimum AA level to a low level (e.g. 1-5); set the maximum level to some large number (e.g. 512 in numerous cases for me) and tune the AS for noise clean-up.

In modo, though, it's far more involved. I had to run several days worth of deliberate experiments to compile a table showing me how the various parameters interact. It's nowhere near as simple to approach and that's evident from several training series that have sprung up for this topic. It's extremely easy to misconfigure modo so that you get a long-running render that is very noisy and that can also consume a lot of memory. There's a lot of complexity in there that needs to be understood.

On the distributed rendering side, there are also clear differences. Buckets vs. full-frame. More interesting, to me, was about the approach. LW takes a content directory approach so that you can work from a central location and then submit jobs to the render farm against that location. That means that you are getting the nodes up and number crunching as soon as possible and they each get the data they need as fast as your network can deliver it.
With modo, though, you have the interesting issue that, even if you work from a network location, when you commit a render to the network from within modo, it will re-send that whole dataset across the network to each system. That's been a challenge for me to handle because it tends to mean that the transfer time is dominating the render job. By way of example, I have a 16 GB+ earth map here, along with large cloud maps and others (as tiled EXR files). The overhead of transferring this data prior to it being streamed by the ndes is non-trivial in modo. In LW, with the common network location, that's not been a factor.
I can mitigate this by using Deadline to drive the modo render. I still need to have the files in a central location (similar to modo) and have to make the file folder structure on the nodes match the working machine (just like LW) to make the image, etc. references work, but it's faster to get going.

So, rendering is a mixed bag. It takes a long time to comprehend it in modo, just like it used to require significant work in LW prior to universal sampling, and there are some infrastructure issues to be aware of. Once you get familiar, you can get equivalent performance in my experience, with various cases where there is a clear winner and loser.

hrgiger
01-08-2014, 12:57 PM
Honestly I think we will be hearing this year about Modo 801 and LW12 so I might wait if thats an option for you to see which one might better suit your needs.

Incidentally I also hope we're on the cusp of a new Zbrush version as well. More excited about that at the moment.

cresshead
01-08-2014, 12:58 PM
both are cool and both are really annoying!

just depends on what YOU are personally doing with 3d and what you need from a toolset and workflow...
I can't answer for you ...personally I like lightwave and Modo...and zbrush and ...3dsmax

each offer some great and utter rubbish things...

Greenlaw
01-08-2014, 01:49 PM
both are cool and both are really annoying!

Agreed! :)

I use both. LightWave is still my primary for both animation and modeling, but I like to use Modo for certain types of modeling. It took me a long time to understand how best to use the Modo interface (forget the buttons, use the hot keys,) but once I got it down, it's really fast to work in. There are a lot of cool tools in Modo for which there is no equivalent in Modeler. Surprisingly though, there are still a few tools in Modeler that work better than in Modo. (Try doing the Rail Cloning hair technique from the Worley website in Modo and you'll see what I mean.)

Don't get me wrong, I like both programs a lot, but neither one is the perfect tool for every situation.

G.

Olivier D
01-08-2014, 01:51 PM
I wanted to answer to the previous thread, but I think it is closed. There were a lot of Modo fans, but, as a Modo user and trainer, I think some things that have been said in this previous thread needed corrections.
I use both Modo (since 2007) and LightWave (since 2000). I just wanted to say that features of the docs don't say all. Modo has particles, LightWave has particles, but LightWave's particles work far better. Modo has dynamics, LighWave has dynamics, but again, LightWave's are far better. When I have a simulation to do, I do it in LW and export the result in Modo with a mdd file. So, you can't only compare the two soft with the docs, you have to test them and work with them.
The interface of Modo is (in my opinion) better and more intuitive. The modelling part is what I prefer in Modo, with the rendering. I don't like LW Modeler. I usually use Modo for all my advertising job, and use LightWave for animation and VFX. I have been very enthusiast with LW 11.5 and 11.6. They were great versions.

So, I don't know if "Modo vs LightWave" is a good title. I'm happy to have and use both, sometimes together.

Phil
01-08-2014, 04:21 PM
I wanted to answer to the previous thread, but I think it is closed. There were a lot of Modo fans, but, as a Modo user and trainer, I think some things that have been said in this previous thread needed corrections.
I use both Modo (since 2007) and LightWave (since 2000). I just wanted to say that features of the docs don't say all. Modo has particles, LightWave has particles, but LightWave's particles work far better. Modo has dynamics, LighWave has dynamics, but again, LightWave's are far better. When I have a simulation to do, I do it in LW and export the result in Modo with a mdd file. So, you can't only compare the two soft with the docs, you have to test them and work with them.

I have. I find modo's dynamics to be great fun to use and, indeed, to work well, in particular because there's a full set of constraints in there so that you can really make dynamics-operated systems, with cantilevers, motors, etc. The last time LW had something like this was with Dynamic Realities' Impact. It's been quite disappointing that the dynamics system hasn't so far been completed. Likewise, particles. Being able to sculpt trajectories and to generate curves from those very same particle trajectories for all manner of purposes is just outstanding. Being able to tune a running simulation in a single frame, particular for particles, is also great. I'm a massive fan of modo's dynamics. It's just a huge boon to productivity and I've been pounding on this for a while.


The interface of Modo is (in my opinion) better and more intuitive. The modelling part is what I prefer in Modo, with the rendering. I don't like LW Modeler. I usually use Modo for all my advertising job, and use LightWave for animation and VFX. I have been very enthusiast with LW 11.5 and 11.6. They were great versions.

So, I don't know if "Modo vs LightWave" is a good title. I'm happy to have and use both, sometimes together.

The interface question is an interesting one. modo's interface is certainly very well conceived, but it seems to occasionally suffer redraw failures and, from an addon perspective, I'm not at all keen on the XML glue that needs to be tackled to get an addon or script to show up in the interface. It's less tricky in LW, it seems. modo continues to suffer from poor SDK documentation, which doesn't help.

From a user perspective, there's an awful lot to like about modo's UI. I just wish it was less prone to redraw failures. I've even come to tolerate the icons. :D

Olivier D
01-09-2014, 02:41 AM
I have. I find modo's dynamics to be great fun to use and, indeed, to work well, in particular because there's a full set of constraints in there so that you can really make dynamics-operated systems, with cantilevers, motors, etc. The last time LW had something like this was with Dynamic Realities' Impact. It's been quite disappointing that the dynamics system hasn't so far been completed. Likewise, particles. Being able to sculpt trajectories and to generate curves from those very same particle trajectories for all manner of purposes is just outstanding. Being able to tune a running simulation in a single frame, particular for particles, is also great. I'm a massive fan of modo's dynamics. It's just a huge boon to productivity and I've been pounding on this for a while.


Yes, but sculpt trajectories work well with 300 particles, not 3000, and dynamics work with 300 parts of a broken wall (if you succeed in stopping the jitterring parts on the floor), not 3000. With 3000 parts, Modo just crashes or asks hours to simulate, when LW do it in minutes. That's why you can't see dynamics simulations of great real projects on the luxo (foundry ;-) ) forum.
I had a character that hitted a wall, or brake a pillar : I have had to simulate it in LW. I'm a great Modo fan (as I am a LW fan) but dynamics need a lot of improvements. I hope they will be done in the future, because it is, as you mentionned, a very interesting way to use particles and dynamics.

As I said, I use both applications, and try to use the better of both ;-)

Here are some tests I've done with LW dynamics and Modo rendering :
http://vimeo.com/73702925
And here, an animation done in Modo (all smoke and fire effects done in LW with TurbulenceFD) cloth simulation is done in Marvelous Designer :
http://vimeo.com/77921708

Hail
01-09-2014, 05:48 AM
[QUOTE=Phil;1361991]I have. I find modo's dynamics to be great fun to use and, indeed, to work well, in particular because there's a full set of constraints in there so that you can really make dynamics-operated systems, with cantilevers, motors, etc. The last time LW had something like this was with Dynamic Realities' Impact. It's been quite disappointing that the dynamics system hasn't so far been completed. Likewise, particles. Being able to sculpt trajectories and to generate curves from those very same particle trajectories for all manner of purposes is just outstanding. Being able to tune a running simulation in a single frame, particular for particles, is also great. I'm a massive fan of modo's dynamics. It's just a huge boon to productivity and I've been pounding on this for a while.


After giving modo a spin for a while, I also found the particle system to be useless for simulating anything above 5000 particles, same goes for dynamics.
Lightwave's is by far more usable and reliable

prometheus
01-11-2014, 06:29 PM
[QUOTE=Phil;1361991]I have. I find modo's dynamics to be great fun to use and, indeed, to work well, in particular because there's a full set of constraints in there so that you can really make dynamics-operated systems, with cantilevers, motors, etc. The last time LW had something like this was with Dynamic Realities' Impact. It's been quite disappointing that the dynamics system hasn't so far been completed. Likewise, particles. Being able to sculpt trajectories and to generate curves from those very same particle trajectories for all manner of purposes is just outstanding. Being able to tune a running simulation in a single frame, particular for particles, is also great. I'm a massive fan of modo's dynamics. It's just a huge boon to productivity and I've been pounding on this for a while.


After giving modo a spin for a while, I also found the particle system to be useless for simulating anything above 5000 particles, same goes for dynamics.
Lightwave's is by far more usable and reliable

usable..under the same situations? do you compare the same particle setups, or do you expect modoīs more advanced operators to work as smoothly as lightwave does more normal perfomance operation?
interesting to know anyway, I guess the modo team have to work on both particle handling and recoil, though it looks utterly cool when they sculpt the mesh path to conform to a face, then convert back to partices and
have the particles flow along the face surface.

I like what I see in the voxel section in modo, with the option to use a mesh or model geometry to make it completly volumetric and not only on vertices or particles, that is something Ivé been shouting for the Lightwave team to implement since many years ago, so maybe at 2005 I suggested they should work on something like that, and now the modo team did it did get there first, I found that the modo team seem to implement more of what I want than the lightwave team does, that however doesnīt mean the lw team isnīt doing a great job, I like many stuff
implemented by the lw team, and generally I like the feel of lightwave more, at least for layout and scene setup.

Regarding modo voxels and hypervoxels for lightwave, the blending mode between particles or vertices/nulls in modo, is much smoother than the lightwave blending mode for volume voxels, the lightwave team did implement the blending mode for volumes not long ago, but I think it isnīt looking as it should be looking, in fact it is quite poor that is, how it looked like in the old dynamite plugin for lightwave is how it should be, or as it is in modo..so the lightwave team needs to redo that one, and of course listen to what I said and what modo did when making it appliable on a mesh as a volume item.
Modoīs smoothing seem to be smoother att end thickness too, thickness in hypervoxels with gradients hasnīt been working as it should, local density not working
so you have to use dissolve/local density to smooth out, also it has light propagation scattering and forward scattering options, and also shadow maps to render faster.

heres a modo head in full volumetric mode, then next image with texture on it, though it was just a 1minute setup from a guy at the modo forums, and the texture appliance isnīt anything special, but the option to use
object as full volumtric objects will make it a better tool to use for more realism in clouds for example, than just using vertices or particles, it is similar to cloudfx tools in houdini..which might be a snap better though.

119232 119233

Have yet to install the demo and test it, but I will soon.

Mr. Limpet
01-12-2014, 08:47 PM
Guys, I just wanted to thank you for all the input. I think for myself the best route would be to go with the Lightwave training for now. I think I will skip the Modo 701 update and look at it again when 801 or whatever they call the next release is available.
I have Dan Ablan's training for LW11 and will get the 11.5 upgrade. Also will keep an eye out for other LW training.

One question for those here who have used Modo. The most difficult concept for me to get my head around are the multiple levels of axis in the Modo modeler. Does anyone know of what training might be available that helps make the effective use of that clearer?

Thanks for all your comments and help.

Kind regards,
Mr. Limpet

Pensart
01-13-2014, 03:05 AM
Transform tool ;)