Camera merged to vantage scene has wrong values

[fixeighted]

E.g. my camera has a fixed focal length of 38.959. This was set in Vantage, exported to Max and came in with no problem.
That same camera 'opened' in Vantage comes in with a length of 0.175.
The camera 'merged' into the scene comes in with a length of 17.164 mm
My Max scene is metres.
I haven't tried with any other scale as this is a live project.

[Alokpatil]

I have the same problem . Additionally, my camera target gets moved in vantage at random times. Usually it moves when I turn on/off the lens correction.

[fixeighted]

Today, with nothing at all changed (apart from a forced, annoying Win10 reboot when I was sleeping) this behaviour has stopped in the current shot, so I have no idea what is going on. I have a number of other shots to do today so we will see if it reappears. I have a feeling that it may have something to do with what is 'selected' when you import the camera....I think it may steal those current values and apply them to the new one, though can't be sure.
What is clear is that the process is a bit flaky and unpredictable for the most part.

Admittedly we are using this in a way that was maybe not predicted when developing the software, so we struggle on in trying to get all our shots completed.
I think the guys are busy at the moment so feel free to search my other posts on this and other bugs/temp fixes which may help you, especially regarding animation and the way it handles sequence length. Some of it may be useful, although some may appear contradictory, as it is confusing to work stuff out at times

One thing I do on every camera import to Vantage is to make sure to 'reset' things by going to the focal length spinner and upping it by some value, then reverting to what it should be. This can seemingly then 'set' those values again.

I can't say anything about the tilt correction as we are not using it, however we did notice that enabling it alters the full window display and so that needs to be reset with the maximise/minimise button to get the bottom of the window back. Tested on both my monitor and that of my colleague 200 miles away, at exactly the same time

I love Vantage but it's of course unfinished and very buggy with regards to several areas which I'm sure will be fixed in due course

[vladimir.nedev]

I can't say anything about the tilt correction as we are not using it, however we did notice that enabling it alters the full window display and so that needs to be reset with the maximise/minimise button to get the bottom of the window back. Tested on both my monitor and that of my colleague 200 miles away, at exactly the same time

Can you please start a separate thread with a screen-shot about this ?

I love Vantage but it's of course unfinished and very buggy with regards to several areas which I'm sure will be fixed in due course

You mean the camera FOV/focal length issues and the extra frame in the animation or other stuff as well ?

Greetings,
Vladimir Nedev

[fixeighted]

Actually I just discovered that it's not down to any one particular parameter; it occurs when you expand all the dialogues when the animation editor is open (or they will all fit), so as you expand the last/bottom one (happens to be tilt correction) then the window goes beyond its boundaries and the only way to get it back is to make it not full screen, then maximise it again I'll post the topic with screens.

Anyway, the issues so far, trying to do what we are doing for this project are:

Focal length glitchiness which needs that 'value reset fix' to make it work correctly. We can overcome this by me taking the cameraman's Vantage camera and me re-exporting it so that he has a now glitch-free camera to use.

The apparent sequence length mismatch/frame clipping, for which we accidentally found the solution; although of course it's not ideal, at last we now can match overlays etc.

The recently found scaling issue when merging cameras is an odd one. When a 120mm focal length cam is merged, it comes into a fresh Vantage tree scene as 30.224.
No idea why. I'll try to find time to test this more and post a scene.

Since we switched to using focal length/f#/focus distance, rather than using fov/aperture and f# conversion it seems to be behaving in our actual scene, so will have to say that appears to work correctly for now. YAY

So, we do now have a pipeline which works for us. I think the other things you probably now know about, such as;

The ability to set a render sequence other than the entire timeline, either as in/out points that can be accurately placed/snapped to frame, or better as a dialogue function.
If it's the in/outs then they need to where the animation ranges are, as it doesn't make sense to have them where they are now....I actually forgot they were even there, as because they can't be accurately placed I never use them. Also they are only in timecode rather than frames.

The ability to have it save settings, samples # is one I noticed, which I keep having to set each time I do a sequence. Not sure atm if it saves others...

Is aperture size there purely to have it tie in with other DCCs? Max doesn't have this parameter, so that was initially confusing our particular workflow.

I think there are other little things but for today I have to get on with the job I'm afraid, so I will have to note them as I find them again.


As I say, I still love Vantage entirely and continue to plug it, despite all the screaming headaches it gave me, plus we have it doing our bidding and can deliver the shots on time now

[vladimir.nedev]

The ability to have it save settings, samples # is one I noticed, which I keep having to set each time I do a sequence. Not sure atm if it saves others...

I think this is possible from the "Save defaults" button and the Preferences or do you mean something else ?

Is aperture size there purely to have it tie in with other DCCs? Max doesn't have this parameter, so that was initially confusing our particular workflow.

I think 3dsMax does have it too, in the Render Settings. It's used when you don't have a (V-Ray) physical camera. So we wanted to support this case too.

I think there are other little things but for today I have to get on with the job I'm afraid, so I will have to note them as I find them again.

Thanks for listing the issues.

Greetings,
Vladimir Nedev

[fixeighted]

Hehe, it's funny what you can just not see despite looking....thanks for pointing out that handy save preferences option

I never have really paid attention to the aperture in the render dialogue before - but now that I look at it, it is not changeable when set to HDTV preset, which I commonly have used for a long time, only when using 'custom'.
Oddly, when I just did a little research I found that apparently the aperture setting in the render dialogue is the same as film gate, though rather confusingly with the Vray camera the Aperture setting is relating to F#, which is what I thought it was. So in Vantage I naturally assumed it was simply an alternative method to achieve the dof if for some reason one didn't want to use the F#....so I remain confused a little.
Also, just out of interest, if Vantage needs Vray gpu to work, then how would anyone use it without it? .....or maybe just when using a standard camera I guess, for whatever reason...

I've been awake since 4am so maybe I'm just tired and not thinking clearly

[vladimir.nedev]

I never have really paid attention to the aperture in the render dialogue before - but now that I look at it, it is not changeable when set to HDTV preset, which I commonly have used for a long time, only when using 'custom'.
Oddly, when I just did a little research I found that apparently the aperture setting in the render dialogue is the same as film gate,

I don't think V-Ray for 3dsMax uses this parameter, and I am sure it's not exported to the .vrscene.
In case other people are reading and don't know what we are talking about:

though rather confusingly with the Vray camera the Aperture setting is relating to F#, which is what I thought it was. So in Vantage I naturally assumed it was simply an alternative method to achieve the dof if for some reason one didn't want to use the F#....so I remain confused a little.

Yes, that's what it is - a way to specify the DOF amount independently of the focal length and f-stop.
Internally focal length and f-stop get translated to "aperture size", when the "aperture size" checkbox is not checked.

When loading a .vrscene, the aperture size comes from this parameter in V-Ray for 3dsMax, but only if you don't have a (V-Ray) physical camera.
So by having this "aperture size" override in Vantage, we support this workflow where you are not using a physical camera.

Also, just out of interest, if Vantage needs Vray gpu to work, then how would anyone use it without it? .....or maybe just when using a standard camera I guess, for whatever reason...

Not sure what you mean, use what without V-Ray GPU ?
Vantage needs V-Ray GPU in 3dsMax only for the live link - for the CPU code that exports and updates a V-Ray version (in-memory .vrscene basically) of the .max scene that Vantage can load and update.

Greetings,
Vladimir Nedev

[fixeighted]

Yeah, I get that now, thanks...I was tired so was a bit fuzzy.

Another thing that occurs to me that I'm not sure is possible....To import the cameras that I am sent, I need to switch to GPU each time, even though I am not actually using Vantage. So the question is, can that be decoupled so it comes in without switching from cpu? It's not a live link after all and Vantage remains closed.

I only ask because I am receiving cameras one by one and keep having to reset my render settings each time, which can be tedious. it keeps catching me out and I end up starting a net render with GPU enabled, which takes an age to quit out of due to the scene bitmaps
No problem if not...I'd rather have all the cameras in one go, so I can batch process them, but that can't happen due to logistics unfortunately.

[vladimir.nedev]

Another thing that occurs to me that I'm not sure is possible....To import the cameras that I am sent, I need to switch to GPU each time, even though I am not actually using Vantage. So the question is, can that be decoupled so it comes in without switching from cpu? It's not a live link after all and Vantage remains closed.

Why do you need to switch to V-Ray GPU ? Are you using one of the Vantage toolbar buttons ?
You can just right click in a viewport and choose ".vrscene exporter".

Greetings,
Vladimir Nedev

[fixeighted]

Ah no, sorry, I didn't explain properly.
I receive the cameras as Vantage configs, so I need to import them, then resend them as mergeable .vrsenes to my cameraman, so he can render them out.
He is doing the Vantage renders, as he needs to edit them quickly, whilst I am prepping everything and creating the Vray cpu render overlays.
We have to do it this way because of the frame sequence glitch and as he basically cannot use Max, this is our annoyingly clunky workflow.
In any case he only has Max 18 which is not fully supported.

Welcome to my over-complicated world

[vladimir.nedev]

Hm, yes, it does sound complicated and usually moving back and forth between programs creates these kind of issues.

I am prepping everything and creating the Vray cpu render overlays.

Is V-Ray CPU needed because of some missing feature in Vantage ? What is it ?
If you can render everything in Vantage, would that make the whole workflow simpler ?

Greetings,
Vladimir Nedev

[fixeighted]

Ah yes, well therein lies the major issue.
Although we have solutions now for creating all the shots needed, the reason I found the bugs is due to the
workflow we needed to use. I claim a major victory in that regard

The majority of this stadium, with a full audience and exterior landscaping, can be done in Vantage and is the game changer,
where we've saved many many 1000s in renderfarm costs, plus the turnaround allows for errors, reshoots and VFX.

The problem is that we also have a 3000 kids choir, all needing to be separately rendered on cards, matted as separate layers,
with a full orchestra done in the same way. This couldn't be done in Vantage of course, so that needed cpu rendering
due to the video being 6500 frames in total, and 150gb of data.
Also the client, who is a cameraman, cannot use Max, so he couldn't possibly have project-managed any of this really.
I had to teach him some very basic methods of exporting the main scenes and also teach him how to use Vantage.
LOL I must be mad

So a tough one to manage, but given that the longest single shot is 200 frames I have a pipeline to net render them with Backburner, on one luckily powerful pc,
which is the only tool that allows me to simply pause the render, which flushes the ram, allowing me to do around 40 frames at one time, with the pause/resume
taking just a second or two. And that is with the entire choir in the frame. It's quicker and simpler with fewer visible of course.

Vantage with the .ifl ability would have been good, though I'm perfectly happy with how everything else has finally worked out

Biggest thanks again for managing to implement the camera animation, as without it the shots would be a little duller.
The other fixes I can personally wait for, as I won't be doing anything like this for at least...never

[vladimir.nedev]

Vantage with the .ifl ability would have been good, though I'm perfectly happy with how everything else has finally worked out

So the lack of multi-frame/animated/ifl bitmaps was the only reason you had to use V-Ray CPU in this case ?
No need for render elements, shadow catcher, multimatte, etc ?
Because you mentioned "all needing to be separately rendered on cards, matted as separate layers", and I am not sure what this means exactly.

Greetings,
Vladimir Nedev

[fixeighted]

Sorry, translation is everything and I forget sometimes to be very clear. When I say cards I mean billboards

It's basically like the attached image; so we use a 5x4 grid of people, unwrapped to billboards (the smallest resolution possible) as an animated bitmap sequence, in 40+ separate materials, for the entire piece of music, for this orchestra and also the choir of kids, which means an enormous amount of frames in total...around 260000, roughly 150gb, which wouldn't fit into my gpu ram anyway.

So the main camera shots are rendered with Vantage, using a proxy non-animated material setup, so the cameraman can see where to focus etc.
Also we include those same cards in the Vantage renders, so we get all the shadows and I don't therefore need to increase cpu render times by including those in the matte/shadow (with dof).
Rendering only the cards then with Vray cpu, with everything else set to alpha -1, means really fast renders and (with the frame sequence length now solved here) a perfectly matched overlay.