MVR exports into external 3D pre-vis software

[Nathan1fb]

Im currently in my final year of uni, writing my dissertation on the accuracy of pre-vis software used in the events industry.

 

A lot of designers choose to draw their rig and produce plans in Vectorworks Spotlight, then MVR export into external software like Depence, WYG or Capture, or use VWX Vision.

 

My question is, how accurate do you find pre-vis to real life? Does pre-vis save time, or do you end up tweaking that many positions, colours, beam focus etc that actually its not too beneficial? 

 

If you use pre-vis in your workflow, what software do you choose and why did you decide to choose this over other software available? Are the renders you produce and the accuracy of them relevant to clients or not?

 

Any help on this topic would be greatly appreciated!

Thanks

[VIRTUALENVIRONS]

HI Nathan,

Not Sure where you are going with this.  Below is a Trailer to a movie.  At the end of the Trailer is a link to the full movie, but I think you will get the idea on this short Trailer. Everything except the characters was modelled in Vectorworks and then exported to CINEMA 4D.  This is an area I have used Vectorworks for, for over 25 years.

 

 

 

[Jeff Prince]

4 hours ago, VIRTUALENVIRONS said:

HI Nathan,

Not Sure where you are going with this.  Below is a Trailer to a movie.  At the end of the Trailer is a link to the full movie, but I think you will get the idea on this short Trailer. Everything except the characters was modelled in Vectorworks and then exported to CINEMA 4D.  This is an area I have used Vectorworks for, for over 25 years.

He’s researching the accuracy of specific workflows for theatrical lighting visualization, it’s highly specialized.

[jcogdell]

 

13 hours ago, Nathan1fb said:

My question is, how accurate do you find pre-vis to real life? Does pre-vis save time, or do you end up tweaking that many positions, colours, beam focus etc that actually its not too beneficial? 

In my own experience when I was still working directly on shows and events, previs was essential for show and especially moving light programming.
Frequently there just isn't enough time onsite for more than minor updating and corrections in your console programming. As long as your original venue plan and lighting plot are accurate, previs saves an enormous amount of onsite time and stress.
Saying that there will always be some tweaking and changes that have to be made on site, its very rare that a venue model will be fully accurate and there will almost always be minor differences between the CAD plan of the rig and how it is actually built by the on site crew, especially in small and medium scale touring scenarios.
It isn't a black and white paradigm, its dependent on the sort of show you are programing (touring, theater, event, etc...), the size of the lighting system, the complexity of the cue's and looks, and the available lighting fixtures.
 

 

 

[VIRTUALENVIRONS]

11 hours ago, jeff prince said:

He’s researching the accuracy of specific workflows for theatrical lighting visualization, it’s highly specialized.

Thank you for the clarification Jeff, I appreciate that.  I am clued out sometimes on current terminology.

 

Stage or movie still requires extensive lighting and interior lighting needs to have "global illumination" for some level of accuracy.  Having said that, the amount of horsepower required to get realism for a stage venue like "American Idol" for example is not going to be found on your Mac or PC (M/P).  A M/P could produce a fairly accurate single frame perhaps over a few hours using maximum settings for Global illumination.  Vectorworks has a built in limit I believe. To get the full motion, on/off colour change etc, it could only give you a good reference point in an animation.  Obviously the lighting staging and cameras etc. are going to be accurate enough and I would imagine that the stage technicians can make the jump from Pre-vis to stage in their minds, based on what they are seeing on screen.

 

There are a lot of things that can affect lighting.  For example, if it is a large venue and very warm inside, it can affect light.  Is there is smoke in the room, etc.  How reflectinve is the room itself.  Vectorworks Global illumination would only have minimal light bounce (~4 per sec), where in the real world millions per second.

 

Having done similar visualization for large venues, anything that can help the client visualize is a bonus, but if the question is "how real"....still a way to go.

[Jeff Prince]

24 minutes ago, VIRTUALENVIRONS said:

Thank you for the clarification Jeff, I appreciate that.  I am clued out sometimes on current terminology.

 

Stage or movie still requires extensive lighting and interior lighting needs to have "global illumination" for some level of accuracy.  Having said that, the amount of horsepower required to get realism for a stage venue like "American Idol" for example is not going to be found on your Mac or PC (M/P).  A M/P could produce a fairly accurate single frame perhaps over a few hours using maximum settings for Global illumination.  Vectorworks has a built in limit I believe. To get the full motion, on/off colour change etc, it could only give you a good reference point in an animation.  Obviously the lighting staging and cameras etc. are going to be accurate enough and I would imagine that the stage technicians can make the jump from Pre-vis to stage in their minds, based on what they are seeing on screen.

 

There are a lot of things that can affect lighting.  For example, if it is a large venue and very warm inside, it can affect light.  Is there is smoke in the room, etc.  How reflectinve is the room itself.  Vectorworks Global illumination would only have minimal light bounce (~4 per sec), where in the real world millions per second.

 

Having done similar visualization for large venues, anything that can help the client visualize is a bonus, but if the question is "how real"....still a way to go.

Are you a Chatbot?

[VIRTUALENVIRONS]

Again modern terminology.  I am assuming that is not a compliment.

[Jeff Prince]

2 minutes ago, VIRTUALENVIRONS said:

Again modern terminology.  I am assuming that is not a compliment.

No, it is a legitimate question.  I see you comments popping up all around the site and they have this uncanny quality to them, like an AI Chatbot.

[VIRTUALENVIRONS]

17 minutes ago, jeff prince said:

No, it is a legitimate question.  I see you comments popping up all around the site and they have this uncanny quality to them, like an AI Chatbot.

Hi Jeff, my apologies for the tardy response, I had to find this video and upload.

 

I think if you checked on comments per day, it is an area you might win.

 

But, I guess you are talking about advanced modelling.  Below is the Theatre of the National Arts Centre Ottawa.  I did this in 1998 , which is why the video quality is low.  I think it is 480 x 360.  It is a complete reconstruction of the venue.   There is an associated TV news bite on it and Vectorworks did an article on it.  I can provide if you like.

We did this to test exactly what the topic is about. Although I have little or no knowledge of current Pre-vis software, I do understand the underlying technology behind it and lighting.

kind regards....Virtual

 

[TomWhiteLight]

Have you searched the forum for similar threads? @Nathan1fb

 

[mjm]

14 hours ago, jcogdell said:

 

In my own experience when I was still working directly on shows and events, previs was essential for show and especially moving light programming.
Frequently there just isn't enough time onsite for more than minor updating and corrections in your console programming. As long as your original venue plan and lighting plot are accurate, previs saves an enormous amount of onsite time and stress.
Saying that there will always be some tweaking and changes that have to be made on site, its very rare that a venue model will be fully accurate and there will almost always be minor differences between the CAD plan of the rig and how it is actually built by the on site crew, especially in small and medium scale touring scenarios.
It isn't a black and white paradigm, its dependent on the sort of show you are programing (touring, theater, event, etc...), the size of the lighting system, the complexity of the cue's and looks, and the available lighting fixtures.
 

 

 

I recently did a piece with a vertical wall of fifty Mini-B wash heads (which I am in love with): Five rows of ten fixts each.

Sent the bare minimum Cad information(fixtures correctly placed and the deck only!) to my Capture person, and a sheet of ten focus palettes because I knew there would not be time to build from scratch. My onsite Console Op was so grateful because all he had to do was:

• Understand the concept of each of those ten foci
• tweak till tasty. 1.5 hours instead of 4 hours easy.

some of the tweaking was more about fitting the concept/theory to reality than correcting focus per se.

Very important to make sure all involved are CLEAR on say where the tails go and what pan / tilt accommodations I required (everything pans to CL in this case, for ex.)

 

Only thing I did wrong was not do MORE of that for that show. Coulda Shoulda had Capture Guy build  my EFX as well. Now I know. 

[Jesse Cogswell]

I bought Vision in 2017 shortly after they were officially acquired by Vectorworks specifically to help pre-cue a production of In the Heights that had a relatively short tech period.  I went with Vision at the time because of the relative ease of getting a model out of Vectorworks and into a pre-vis software without having to build everything twice.  The plot had relatively few moving lights (18 total, a mixture of VL 3500s, SolaFrame 1500s, and SolaWash 2000s) and a lot of conventionals (380).  Unfortunately, there were several bugs with the fixture profiles for the SolaFrames and SolaWashes (SolaFrames had a bug with the color wheel where it constantly jump between two colors unless you put the wheel in Continuous Spin mode, SolaWashes had the iris inverted and didn't cast shadows), and the file ran really slow when running cues because of the sheer number of fixtures.  Vision's speed very much hinges on the number of emitters and whether they cast shadows.  Rocking a plot with 300+ emitters, all casting shadows (very important for theatre), meant that watching crossfades between cues turned into a slideshow.

 

So, I was not able to use pre-vis for the intended purpose of pre-cueing the show, but I did find that it was fantastic for setting focus points.  That production did a two weekend run-out to another venue with no additional tech time outside of a single run-through rehearsal.  I very purposefully made my plot 1:1 for both venues, and built all of the moving light presets out focus palettes that matched my conventional acting areas.  Between venues, I removed fixtures from palettes that they didn't need for the show, so when we hit the second venue, I just had to update the palettes and the show was 97% of the way there.  Moving light focus took all of 20 minutes.  But being able to build those focus palettes at home and correct things like flip direction saved an IMMENSE amount of focus and tech time.

 

Moving forward, I've used Vision a lot for building palettes and base presets before tech and much less often for pre-cueing.  I've pre-cued small theatre shows with low fixture counts, but recently run into problems with how Vision renders color temperature, making a N/C Source 4 look like it has R09 dropped in it.  This necessitates manually adjusting the color temperature of every Source 4 in the plot and bumping it up 1800K just to get "realistic" color output.  The other thing to keep in mind is that things like lens focus will NEVER be accurate.  I'm not sure if this is something that has been solved in Capture, L8, or Depence, but in Vision, if the Edge parameter is at 50%, the edge will be sharp regardless of which gobo wheel is on, or the iris is in, or the shutters are in.  In real life, these all occupy a different part of the focus train in terms of sharpness.  This means that you can't really do things like layering gobos or animation in Vision and have a realistic sense of what it will look like.  Things like gobo size and shutter placement are also not entirely accurate to "real life" and will always take some adjustment.

 

One thing that pre-vis is super great for is music and event shows using beams as "eye candy."  Vision encourages turning off the Cast Shadows option for the sake of performance and framerate, which is a difficult ask for theatre work but easy if you're only worried about building patterns with beams.  I don't do a lot of that kind of work, but during the pandemic I spent some time timecoding moving light shows to music and can absolutely see the value in pre-vis for that purpose.

 

Since coming back from the pandemic, I went back to lighting and programming primarily theatre, dance, and opera.  With the advent of Eos 3.0 and Augment3d, where the lighting console now knows where the lights are in space, I don't think I've opened Vision more than updating it to the newest service pack.  For a big musical with 70+ movers and 60 focus points, I can now use the console to rough in my focus points in 20 minutes by putting in XYZ values instead of taking a week manually focusing in pre-vis.  That said, I would recommend Augment3d STRICTLY as a programming tool and not as a pre-vis software.

 

Also keep in mind that most of my experience is with Vision.  Vision has come a very long way since 2017, but based on the output that I've seen from Capture, L8, Depence2, and even WYSIWYG, I would say that it lags pretty far behind.  And with all of those software using the MVR import, it no longer has the advantage of easy Vectorworks export.

 

As for renderings and clients, I haven't found Vision to be terribly useful.  On an opera I was working on in the fall of 2018, I had been planning to use Vision to create renderings to show the stage director and design team and was hoping to have them ready before the big design "show and tell".  Unfortunately, the movers that I had specified (SolaFrame 1500s and Vipers) still had issues with the fixture profile in Vision.  After reaching out to the Vision team, I was told to sit tight and wait until Vision 2019 launched, since it would fix those issues.  The launch date was after the design meeting, but admittedly the jump from Vision 2018 to Vision 2019 was HUGE in terms of visual quality (integrated normal mapping and bump mapping, a more accurate illumination model, better handling of curved geometry, and fixes in the fixture profiles), but the software had come out so late that we were already in tech, so those renderings never got shared with the team and were made as more or less "practice" for me.

 

Renderings created in Vision 2019.  Some of the normals are off (SR doors) due to how VW exported normals back then, but a big step up from Vision 2018

 

 

There was another opera that I co-designed in 2021 that we used Vision heavily to pre-cue.  Tech time was at a premium, so we used Vision to cue the whole show, which was moderately successful and did save us time in tech, coming in with very rough cues.  Because of the pandemic, we were not able to have the singers kiss in a scene, so we had to find a way to "fake" it.  The opera was filmed, so we opted to use shadow play to show the kiss by having the singers on different planes but filming their shadows as they converge.  For rehearsal and to see if the idea was even feasible, we had to determine the exact placement of the light (8" fresnel), the camera, the singers, and the projection surface (a big moonbox).  I mocked this up in Vision first, but wasn't happy with the way it displayed curved geometry and couldn't really get it to look good, so ended up doing the rendering in Vectorworks anyway.

 

Renderings from Vision:

 

Renderings from Vectorworks:

 

 

While Vision was MUCH faster to setup and render, we were ultimately happier with the Vectorworks rendering, and that was what we shared with the team.

 

Final Shot:

  

If I did more music and event work, I imagine rendered videos and such would be VERY valuable to share with the client.

 

To answer your question about why I use Vision, it was initially because of the ease of getting models out of VW and into the software.  At this point, most of the software has an easy path to do this, so that is no longer an edge.  At the moment, I don't need pre-vis enough to drop money on another piece of software (lighting design is already an expensive profession, no need to spend more than you have to), but if I were starting my pre-vis journey today, I would likely look toward using Capture (looks relatively easy to use, with lots of neat features like animated crowds with cell phones), L8 (really good output with reflections and animated models, can handle water, laser, and smoke effects), or Depence2 (expensive and looks fairly complicated, but really nice output and looks well optimized for working in real time).  It might also be worth waiting for a couple of years to see what happens with the Unreal Engine.  It being very powerful and widely available, and now with DMX control options, means that it's only a matter of time until we get a feature complete software out of it.  Actually, it looks like we might already be there: https://www.carbonforunreal.com/

[Dave Donley]

@Jesse Cogswell Thanks for this post.  It's valuable and appreciated!

[mjm]

@Jesse Cogswell—thank you very much for your experiences Jesse. In 2017 I lit a world premiere for a Ballet company north of the border. They'd invested heavily (for a dance company) in Vision: the software and a massive PC, massive graphics card etc, solely dedicated to running Vision and VWX. The previz experience was terminally painful. Conventional fixtures could not be reliably shuttered. Any number of issues with the software freezing etc. All we were trying to accomplish was focus points. I don't remember the output onstage matching much of what we'd managed to previz. I subsequently stayed away from all previz for years. It's only been recently that I've occasionally moved back toward previz as a tool due to the budgetary realities of the dance world: first thing cut: time. Unfortunately things like focus take time and sending my mover focus to previz when it can be afforded is now my preferred route. No client I have can yet afford a full on previz from focus through cueing. Only upside there is the joyful tactility of making light in real time and space.