3DEqualizer4 [featurette] - Release 5: What's New

hi and welcome to an overview of the latest features shipping with the exciting release 5 of 3d equalizer for this release is packed with new functions and scripts so we picked the best ones to demo them be prepared and let's start with the updates to the 3d for UI release 5 was downloaded and is waiting for us so let's start it the first thing we notice is the general overhaul of the UI it got a bit darker and cleaner than before text fields option menus tabs handles most UI elements were moderately redesigned to preserve the typical 3d for look and feel epic topic during the redesign process was increasing space for content the separator bars of all paints are now just half the previous size resulting in a noticeable increase of space in both paints and display area further scrolling bars are invisible if the paint content fits into it it only appears if needed so with a smaller pane with we get a bigger display area next selecting got an exciting new feature this feature will be shown with all these points on the wall and three orientation top view and Co there are so many reconstructed points in this project but assume we like to select the points forming the curved wall alone traditionally pretty annoying you in release 5 3d features a lasso selection tool holding keys ctrl + alt it is possible to draw a selection shape selecting everything inside of it now it is easy as pie to select structures like this wall this will save a lot of time great all points of the wall are selected this tool isn't bound to three orientation controls we can use it everywhere items can be selected of course in every other controls like lineup and manual tracking as seen here and in editors like curve editor and deviation problem all right paints got an interesting update as well to be more precise which paints we can add to display area as usual a typical pain like curve editor now it is possible to add Python scripts as pain as well no more windows floating beside or beneath the main window all your favorite Pacha tools and many others can be organized to write within 3d for of course not all scripts make sense as a pain those which opened their own window not blocking 3d for can be selected here to get your script ready to be added as pains simply a single line of code has to be added to existing scripts please refer to the latest Python documentary in Section tech talks at 3d quad so calm ok there's a nice setup containing several script paints of course such an arrangement can be saved as an environment as usual in window configured we can save a pain arrangement environments now became more individual and personal than ever before we can set one for any possible step in our workflow so easy and so fast father as known from previous releases when closing 3d for the latest arrangement will be saved independently from the latest set environment after starting 3d for all pains appear again from one big workflow update to another even bigger one we added so many scripts we hardly can see the content in this play area especially if there's a huge 3d model or many 3d points it can be a bit frustrating seeing only a bit at once if we do not like to zoom out drastically introducing hide paints in release 5 a simple click on this button and boom all paints disappeared and we have a clear view at our content that was fast ok so again with paints and with a clear view add the content to make it even simpler by default shortcut tab was assigned to this button so with a single key press height show/hide of course all functionalities of the open panes like delete from point browser are not deactivated and still accessible from the hidden paints that's it with the exciting new interface of Breed equalizer for release 5 redesigned UI less of selection script paints environment and high paints all these new features help to work faster than ever before this chapter is about improvements of the 3d solver say hello to dynamic point waiting here we have a project with one not so perfectly tracked point messing up the otherwise great reconstruction quality in a certain range having a closer look at this point we clearly see the cause of the peak and deviation a bad 2d tracking in sad range of course we could simply refract the point in these frames but let's assume for this demo there is no chance to improve 2d tracking simply disabling the point in the given frame range can cause an unstable and shaky camera movement if they aren't enough other points to compensate the sudden lack of tracking information 3d 4 release 5 provides an option to handle such problems have a closer look at updated parameter point weighting and welcome the new option dynamic let's see what changed and what we can do now first obviously static weight cannot be edited anymore in point browser the point is labeled as dynamic so good so far but where can we now set values for dynamic weight the answer is the aviation browser let's open a second browser for a better overview good now hide average deviation curve next select show point weight curves selected points only a yellow line appears it's look is pretty the same in those of previous releases but from now on we can edit this curve like in curve editor having a look at deviation curve we see three CVS are needed to weight the curve one CV with value zero and the peak frame and two additional CVS at frame positions before and after the peak expand the lower weighted frame range great or more precise positioning along axis y let's enter the CV position values manually our middle frame should be exactly zero and the other two you might guess it one so they're matching with the other points weighting value we noticed in release five it's possible to add at position values of multiple points script edit curve keys was redesigned from scratch and now works on all curves in curve editor as well as deviation browser the project and here we go perfect we got a clean deviation curve the advantage of a dynamic weighting curve compared to simply disable the point within the badly track frame range is the soft transition as mentioned earlier simply disabling a point might lead to a shaky camera movement further if they aren't enough other tracking points available it is recommended to set the lowest cv slightly above zero to prevent this point from becoming passive and not contributing to the camera reconstruction all right next up on the future rollercoaster forced calculation what do we have here it's a nodal shot of a green screen filled with markers well to be precise as we can see by the yellow rod it's a mere nodal shot nevertheless they are tracking points available so let's call it the project deviation value looks good like the curve itself well until the last frames it seems 3d for couldn't calculate anything in the end let's expand point browser and have a quick look at the clock status of each point there are some points which only could be triangulated by a 3d force solver when selecting them we see these points are all tracked in the last frames of the sequence so we have the cause of our problem release 5 introduces an exciting new option to force the solar to calculate a point on school right in the points attributes we can find a new parameter called triangulation behavior here we can set this behavior to forceful point calculation now after calculating the project we can see this calc status was set in part browser and more important according to deviation curve the solver was able to reconstruct the camera in the last frames so let's change the triangulation behavior of all remaining points to give the solver no other possibility then take each point into account perfect we got a nice curve that's it with the two best 3d solver improvements dynamic point weighting and force calculation give us powerful options to handle problematic projects welcome to improvements for object Bank Group's release five has impressive new features for handling object tracks let's start right away with the demo this shot shows the handsome guy holding a blue box everything was reconstructed correctly as we can see at deviation curve and white geo lines on the box so let's switch do with reorientation this demo is all about the point group path more precisely how this curve is depicted in curve editor with the point group selected let's open set editor and select one of the position curves pause set for example we see the translation curve in z direction now if we scale the point group we quickly see that point row path moves in that direction and even changes its shape accordingly to the transformation after releasing the mouse button we notice it changed path but the path and curve editor remain unchanged these curves are normalized object point group moves relative to the camera transformations so scaling an object in global space won't affect the curves in curve editor in release 5 it is possible to show curves as they really are now we can edit them as they appear in global space select view 3d rot path curves show curves in global space a new curve appears and as we notice it is now much closer to the curve we see in display area this time the point group is scaled all transformations are depicted correctly and immediately in curve editor it doesn't matter if the path or even certain series are modified in curve editor or within display area both curves are synced and behave identically okay now we're getting to an extremely interesting feature optimizing the z-depth of point groups yes 3d for finally provides a way to handle distance related errors of object point groups at first let's have a quick look at this shot we see a blue box in front of a green screen filled with markers which was lifted and put back down by a mysterious person we could see as well everything was already reconstructed let's have a look at the reconstructed box in 3d space hmm something's looking strange we remember the box was standing still on its chair but nevertheless there's a noticeable tremble shake it like a Polaroid picture this tremble is a common problem for object point groups even more if the object is far away from the camera and therefore really small in the footage the deceitful aspect about this travel can be seen back in lineup zoom in for a closer view at the box okay to be precise we see this aspect by seeing nothing when looking through the camera the tramble is not noticeable since it's really small such a small movement towards it away from the camera is really hard to see nevertheless the data in 3d space speaks a different language there is an annoyingly shaky box that should be fixed release 5 introduces the completely new object pong grooves at distance tool by the way beside opening this tool as a pain we can find it as well and menu edit and as a button in display area ok back to the tool if a point group is selected we can see a curve depicting the Z distance towards the camera select another view option to frame the entire path good so what's the curve telling us in the middle there's the part where the box is lifted having a closer look at the beginning or ending part where the box is supposed to stand still on the chair we clearly see the shaking ok so now we have a great and precise possibility to visualize the boxes trembling the best thing about this visualization we don't have to worry about the point groups individual rotation position curves this curve Somaly shows the z distance towards the camera so editing it keeps the other rod pulse curves consistent now theory let's start editing and see what magic happens moving as CV down leads to a box movement to watch the camera moving it up you might guess it translates the box further away everything is clearly noticeable here so let's switch back to line put the entire box into display area okay then move a CV we see that we see nothing as mentioned these movements are simply too small to be seen in with a human eye luckily with this grip we can optimize the curve for a perfect result one quick tip if the curve was modified the originally unedited version of the curve is displayed as a thin line get back to the original status okay in order to reduce the treble let's filter the curve select all CVS belonging to this part then enter a filter value and click filter Seavey's nice we got a smooth curve let's check the box in 3d orientation the curve structure already indicated the trend will change to a box well slightly swimming through 3d space so we know this box doesn't move since it stands still on the chair delete all selected civvies great now everything is fine check again in line up marvelous now hard numbers and visual appearance are on the same page no movement in any view before we move to the middle part let's quickly fix the first part the same way select all series and delete here as well no trembling after the edits and in three orientation perfect as a last step let's check the middle part here the curve suddenly leads at a certain frame have a closer look at the box in global space around this frame as indicated by the curve the box really jumps from one frame to another instead of moving smoothly like in the footage in this case a tiny tracking error was sufficed for this behavior nevertheless obviously this entire bunch of CVS has to be placed above its current position on axis one instead of doing that manually off checkpoint groups at distance tool provides a cool feature to seamlessly integrate these CVS into the rest of the curve with all needed CVS selected click Edit patch CVS and boom perfect finally since we still see some trembling across the entire movement filter all civvies a lower value might be better then have a look at the shot at first year in free orientation we fixed it the trembling at the beginning and end is gone and on top the overall movement of the box is more calm and moves like in the footage to complete this chapter have a look again in lineup here is well nothing to complain everything's smooth and fine all edits didn't affect the visual quality of the reconstruction even more the movement is now correct in every view we've seen - amazing updates to object point groups global space and the incredible news at distance tool will be a great deal when working with object tracks release 5 brings a huge update - 360 degree shots and changes the way vr projects will be soft in 3d for VR you ready let's quickly get to know them at first have a look at the ship we see two guys performing some badass karate moves everything's filmed in 360 degree I wouldn't want to meet them in a dark alley obviously the first thing to mention here is three for release five can read and play 360 degree sequences instead of only as still as in release four we need parameter group long lat conversion for the next step button long that conversion changes the view like we would see through a normal non VR camera resolution sets the footages resolution in display area focal length controls the focal length of the virtual camera the possible range is very wide but in this demo F value of hundred is the right choice next is weight there are two parameters missing pan and tilt are gone okay now really gone they were moved to a different place with a very good reason open curve editor scale it since will focus on the curves in the next step here we see a new entry called long lead conversion curve expanded and great here all the parameters previously located in attribute editor so why were they moved here now it is possible to create dynamic pan and tilt curves yes we can animate the camera in 3d for simply create one or more CVS to adjust the cameras rotation such an animated camera is vastly useful for object tracks for example if the karate guy should get another face so to demo this exciting feature let's say we like to follow this guy and make sure he is always approximately in the middle of the image to track his head later to follow the actor the camera has to pan so the recently created CV on curved camera pan should be around 30 degree in the beginning good then around let's say 100 we add another CV and set it to a value of roughly 150 degrees our guy is back and frame nice in such a short time we animated the camera a quick tickle ong the path yep the camera follows him perfectly of course the second kicking master shouldn't be ignored so the camera should move back and follow him until the end of the shot a little adjustment to reduce the panning speed at second CV cool have a look at the camera movement as if we would have been on the set and captured the footage as we see it nice well now we do have a camera movement but how do we track features to get the camera in 3d space and 3d point positions of course we can search a feature and track it throughout the shot release 5 awaits with new features to help us here be prepared to be amazed please welcome converting points from regular to long let converted switch back to the regular view of the 360-degree shot then choose any features and place a 2d tracking point on it and I really mean it any feature considering the current camera movement we concentrate on the left half in a world record time we got a thousand of points ok we have some points but back in long that converted view these points of course do not stick to their positions we chose in regular view with a single click this can be faked select edit convert regular to long lot conversion and three to one awesome isn't it all points snap to their correct positions and stay there throughout the entire shot all points got correct tracking curves again with a single click due to the curves in curve editor 3d 4 has precise information of the camera movement therefore is able to calculate the GD tracking curves of each point tracking a VR camera was never easier of course this function works flawless if the footage was filmed with a fixed camera position if the 360-degree camera rig moves within the scene features have to be tracked as usual in long Latvia there's nothing else to do so let's call it the camera set a fixed camera position constraint then cackle oh well the deviation doesn't look good what went wrong obviously the tracking quality isn't a problem maybe the current lens causes this error like in tab lands we also find here a parameter called horizontal angle keep the 100 degrees in mind then switch to tablets as we can see both parameters have completely different valleys the virtual camera was set to an angle of 100 degrees therefore the footage seen in this play area effectively was like film would 100 degrees like solving a normal shot if the lens does not match with the material we get an incorrect solve but here in VR and elaborate adjustment procedure is necessary we already have all infos we need this parameter simply has to be set to the same 100 degrees as in temp camera then Calculon Wow water result there isn't any calculation error or inability to count the shot if camera and lens are matching we really get a deviation of zero the perfect result let's check the result in through urination nice points and camera movement were reconstructed correctly for a final check let's add a 3d model choose a point on the floor then let's add a cube and snap it to the selected point switch to line up and play the shot so with just a few clicks in like two minutes we successfully animated a VR camera and reconstructed 3d points and the cameras position in global space now we will be ready to track the actors face this chapter is about how to correctly use both convert point scripts in case the camera movements should be modified we see a camera animated with the new camera pan curve in curve editor further a point rect on the actors face this point was tracked in mote long lat converted the advantage of tracking a point here instead of the 360-degree regular view are the distorted edges of such a shot in long not converted there is no such distortion and the karate guy always stays undistorted in frame after tracking we suddenly realized the camera movement isn't satisfying the ceiling should be more visible but changing the camera obviously will make our tracking point obsolete instead of tracking again and this point was really mean to track 3d for release 5 provides a new and easy way previously we had points in regular and converted them to long not converted this time we do it the other way around and convert the points from long that converted to regular select edit convert long led conversion to regular and here we go the point is back on the actors face within the entire range this point was originally trapped now we can edit the camera as wished since the ceiling should be more in frame curved tilt needs to be filled with some Seabees one at the beginning then let's say around 50 the camera I should look up a bit good back to regular the point is still where it belongs and regular view changes of the camera of course do not affect the points the cameras animation is ready we can convert it back to long not converted switch back to long I converted and when playing the sequence we see our point is where it belongs a big advice if we should have altered the camera movement without first converting the points to regular there is no chance for 3d effort to convert this point correctly later it always uses the camera curves in curve editor for the conversion in each way always make sure all points are converted to regular view before editing the camera we are staying with topic vr this chapter is about camera rigs we have a project containing several cameras on a VR trick all cameras have tracked points they are synced and each has an individual lens attached to it and since this isn't a complex enough there are some 2d points just seen in a single camera but in the end everything could be reconstructed as we can see of course there is no impressive feature without a big problem zoom in and centre the cameras to have a closer look at this problem we can see all camera movements are not really steady to each other each camera moves individually not something you would expect having a rig with fixed cameras in it to prove the visual appearance numbers are always a good shot at script differences to display area a quick introduction this script calculates an average camera of all existing ones in the project the values display position and rotation of the three axis x y and z the rod pass values of each camera are in relation to the average camera so have a look at camera objects and values all numbers are going wild live the cameras in global space it's so shaky that it's not deniable this shot wasn't solved yet since there shouldn't be any movement how do we fix the current case release 5 introduces script adjust piggyback camera constraint and it will be very helpful this script uses the calculated average camera to force a virtual constraint in each selected camera this virtual constraint adjusts the position and rotation curves in relation to the average camera so with all camera selected open the script in the appearing window we see the path rod offset values of the selected cameras the position and rotation of the average camera is displayed in the curve widget here we see the average cameras movement throughout the entire shot like in all curve widgets all series are editable all right we like to adjust translation and rotation offsets of the selected cameras as set by default button ok transfers all parameters to parameter adjustment so let's open it seven cameras with six parameters each well this will be a lot of work for the solar adaptive all and adjust to start the adjustment after some time we got a result transfer all values wait say hello to 3d for disco light patent pending have a look at the camera group in display area all cameras were adjusted in position and rotation last the project has to be calculated again to adjust 3d point positions great pick one camera to center it if we now play the shot we see everything is steady all cameras do not change their position towards each other finally we have a correct rig that's it with VR pretty big update right camera animation converting points to and from long land and piggyback these new tools will help so much solving VR projects only with release 5 this chapter demonstrates the new features and enhancements of orientation controls let's start with 3d models at first we need such a model more precisely a big one like this car with a count of over half a million polygons when navigating around the model we see everything smooth no matter how fast we change the view just the look of the model changes during navigation ready for release v can handle large 3d models way better than ever before when importing big models fully for automatically detects the size of 3d bottle and if the new option performance rendering should be activated to demonstrate the old performance deactivate performance rendering we see everything's lagging not so pleasing limb previously so better activate this feature again the performance depends on the render capabilities of the built-in graphics card here you can control the quality of the model during navigation so even if a slow card is installed a fluid workflow will be possible working with large marbles was never more comfortable than in release 5 next we're moving on to texture mapping for a better demonstration a different model will be imported 3d for automatically detected the model size and let's performance rendering deactivated in this case due to its small size release 5 introduces a new option for texture mapping beside the already known budget projection now we have the possibility for texturing with UV maps we don't have any map manually imported for this model what can be done with botton UV map so the human is rendered with a solid gray color further the 3d model we imported contains no information about UV coordinates well we cannot do anything about the texture right now but about the coordinates in menu 3d models we find new script generate UV coordinates here in two simple steps one for setting a projection orientation and one for the scale we can generate simple UV coordinates for the selected 3d model if no texture was imported a checkerboard is mapped by default looking good but we won't stop here the famous script Lambert Iser downloadable from script TV at 3d kawazu tom was rewritten from scratch and implemented natively which drastically enhances its speed so for those who aren't familiar with the script what does it do it creates a Lambert shading effect for the selected 3d model in the upcoming requester we can adjust the appearance of the effect okay to apply the effect nice the model is looking more vivid and reveals much more details than before thus enhances the possibilities for a visual check of reconstruction quality the next demo shows a scenario there described functions will be helpful here we have a mocha project already solved showing a reconstructed moving head talking something as seen clearly in display area until now we only could project the footage on the mesh created from reconstructed 3d points so to apply the things we learned previously set parameter texture mapping to UV mapping UV coordinates were already generated and we don't have a texture imported for this model either so we see an accurately map checkerboard so good so far checking the model in lineup we see well the checkerboard and at least that it moves pretty accurate but we cannot see anything from the face which isn't very helpful for checking reconstruction quality so back to 3d orientation controls then to Lemur tiser you might have noticed beside contrast and brightness it is possible to set an alpha value exactly what we're looking for so let's reduce the value to increase the faces visibility back to lineup perfect both model and actress face can be seen we can now check not only the reconstructed 3d points but the over movement of the mesh so where the point selected have a closer look it's looking great the square smashing precisely with the face movements the solve therefore is proven to be correct in a way nicer visual way than before that's it with orientation controls more speed more mapping more possibilities for visual checks working with 3d models was never more comfortable and faster this chapter can be summarized with a single line be prepared to be amazed 3d for release 5 introduces some amazing new features for 2d tracking points and modeling cel fasten your seat belts and let's start with tracking here we have a project containing several reference frames and the set within the frames should be reconstructed which means a tedious hard work by placing many many tracking points throughout the entire bunch of frames as usual let's start with frame 1 and search a feature worth tracking like this pattern on the ball so let the hard work begin we'll be back in a few hours ok you didn't really buy it right me neither wouldn't it be nice if we don't have to search manually for this vague pattern in all frames but instead let 3d for handle this it would so let's do it simply click on track the magic begins through the e4 release 5 can find set patterns and other reference frames automatically isn't that great the tracking is done on the GPU so its speed depends on the Kart you have it stopped and as we can see it's really precise it is impossible to find these exactly matching 2 D positions manually that quickly let's choose another one and just it's tracking box by the way the outer search area was abandoned and released 5 point tracking reference frames this pattern will be searched within the entire image instead of just the search area so this box was redundant start the second round again this pattern will be searched in all frames completely automatically we see this pattern was interact in all frames there are two gaps within these frames it seems the 2d trekker was unable to find the SAP header let's examine the two and start with the first one when zooming out it's getting off is very quickly the pattern cannot be seen within this frame it is hidden by the lamp the 2d tracker is clever enough to skip placing a tracking point in such a case in the other frame well here the pattern can be seen but the 2d tracker somehow still couldn't find it in such cases 3d for release 5 provides a new option for controlling the 2d tracking behavior have a look at per meter deep tracking in most times the default value of zero is more than enough but in some cases the tracker should spend more time on finding a pattern with an increased value the two detractor looks deeper into the image thus increasing the chance of finding the set better ok then track again bar weight as we remember there are only two untracked frames and one of them is redundant due to the pattern hidden by the lamp so instead of using button track which would try tracking both remaining frames we speed things up by tracking the next frame alone all track frames are key frames so frame 9 can be set as the new reference pattern by starting tracking with this frame then use button track next frame as mentioned a deeper tracking comes at the cost of performance but here we go the pattern could be found let's quickly check it with the help of the reference pattern great this point is finished as an alternative way in cases a pattern cannot be found in a specific frame and also as an additional help if even deep tracking set a1 won't work release 5 provides a different approach remove the key frame to restore the old status of no point in this frame although the 2d tracker wasn't able to find the pattern with deep tracking set to zero we can identify the pattern obviously it must be here okay approximately here of course that's not really precise but the rough starting point is all we need three four release five introduces updated script align point to previous next keyframe since version two in addition to sequences it can be used with reference frames as well the script is executable here in menu tracking or with button aligned to key in display area either way it works great with a new tracking engine simply select the script and here we go the 2d point was positioned precisely in contrast to the 2d tracking seen previously with this script the engine only uses the area around the set point instead of searching within the entire image therefore a pattern can be find much faster if we can set it to the point and it's approximately correct position some more points will be needed so continue tracking features to reconstruct the wall okay we have several features track throughout all reference frames let's talk all to see the result best in 3d orientation controls looks good all reconstructed 3d points and cameras are positioned correctly alright this result leads directly to the next great feature coming with ready for release five modeling like and release four we can use the reconstructed points to create image well there are it done in polygons so let's clean the model up a bit good now with footage projected onto the 3d model we quickly see this model isn't very precise in the areas between the reconstructed 3d points the footage on it translates happily around especially when having a closer look it's not hard to see that this simple mesh is far away from an accurate reconstruction of the wall so how do we get there may be tracking many more points for more vertices may be at least if you have way too much time three four release five introduces refine polygons this new feature deforms the selected 3d model based on information as sequence or reference frames provide without any additional tracking data in its requester we have some options to control behavior sources and speed of the refining process one option to be mentioned here is use auto tracking masks for this demo we'd like to reconstruct the ball alone but as we can see there are a big table with a monitor and some lamps standing in front of the wall they shouldn't be considered in the refining process so they are masked out in auto tracking controls in the requestor of refine polygons this option ensures the masks are being considered okay we can leave all other options at default so let's start there refinement based on your graphics card this might be fast or a bit slower nice the 3d model changed dramatically with dramatically better results we quickly see the new model is way more detailed and shaped like we'd expect from the curve wall in the footage the attributes from the source model are assigned to the new one so footage projection is already activated have a look at the new result great the translation seen previously are gone all patterns and markers do not move proving they're good refinement of course the difference is considering blurring and brightness are coming from the different reference frames the wall is now reconstructed accurately so having this great 3d model wouldn't it be nice to have it accurately textured as well when scrubbing through all frames we see these annoying black areas with no image information that would be really nice alright then let's do it previously we generated UV coordinates alone due to the lack of images now we generate a true texture map with textures and coordinates for this model with generate UV texture map this is done fast and easy first we set a file name for the texture let's say great wall texture then define the resolution there is no upper limit it is possible to create really huge textures besides standard 8-bit format this function also provides 32-bit open e XR with a high dynamic color range last of course all cameras available should be included into the texture but no K starts the texturing process after a short while a texture was created and we instantly see the black hours are gone thanks to all the images taken into account showing the wall from different angles having a closer look we obviously don't see any transformation it is a texture switching back to a footage projection we can compare the quality of the texture to be sure again in a different frame very nice result one more thing in the end you may have noticed generate UV texture map also provided an option to consider the auto tracking masks to exclude table monitoring lamps from the texture with the footage projected on the wall we see this huge lamp which of course does not belong there back to UV mapping poof it's gone magic well not really but it feels a bit like it right ok we have a detailed modeled and textured wall the last step is exporting to Maya the first thing we notice here is the updated exporter so many new parameters its usual setup on it next we can set which cameras should be exported then for the first time 3d models are available for export again we can choose which should be considered and it doesn't matter if it's a static or animated mesh all models within a scene in 3d four are suitable for export further UV texture export is now supported as well we have such a nice texture so yeah we want to do this and overscan won't be needed in this demo we're done everything's ready for the Maya artists our job is finished ok that's it with the amazing new features for tracking and modeling only with release 5 it is possible to reconstruct even complex geometry without the need to track thousands of points and even if that would be necessary in a reference frames project track was never easier and faster

Channel: 3DEqualizer

Views: 15,868

Keywords: 3de, 3de4, 3dequalizer, 3dequalizer4, 3dequalizer 4, science d visions, sdv, tutorials, tracking, motiontracking, motion, matchmoving, matchmove, match moving, visual effects, motion capture, 3d effects, visual arts, 3d, film, 3d film, camera reconstruction, lens distortion, distortion, tutorial, python, script, best software in the world, best software

Id: 7BvGbyg_7rI

Length: 50min 40sec (3040 seconds)

Published: Wed Oct 26 2016