Check out recent posts and join the conversation by asking your questions.
This post is from a suggested group
Is there a way to take an existing solved camera and then line it up to geo that isn't survey solver or geo tracker? Basically something that I could pin the geo like we do in geotracker but that is just adjusting the tracked camera like an orient camera node.
This post is from a suggested group
I'd love to be able to create more than nine workpages, thanks for considering!
This post is from a suggested group
I was wondering if we could get a direct /export undistort/redistort function inside our lens distortion preset modes. I know it's far better to solve shots but I'm trying to have a generalised lens distortion workflow for the studio that we can start when we get lens grids. Also not every lens is for a matchmove shot, lots are just comp but I wanted to centralise the work here in PFTrack.
Am I missing something very basic that already exists?
Hi Toby,
For a generalised lens distortion workflow, you don't need to solve a shot just to generate the ST-Maps, you can apply your preset to the chart clip in the Clip Input node, then connect a Scene Export node. Under the Distortion tab, enable the ST-Maps you want to export.
The reason this is handled through Scene Export rather than directly from the calibration panel is that the distortion can be adjusted while working on a shot. Scene Export ensures the most current matched distortion setup for the clip is used when exporting the maps.
https://pftrack.thepixelfarm.co.uk/documentation/pftrack_node_scene_export.html#st-maps
Regards,
Adam
This post is from a suggested group
Hello, Is there a way to have all users load a custom centralized set of macros, exports, and presets? I understand that we can load specific macros, presets, etc, by putting them in the user's folder here: ~/Documents/The Pixel Farm/PFTrack/ I'd like to be able to "append" an additional, version controlled package of macros,presets, etc from an arbitrary location. Is this possible? Thanks, -Jake
Hi Jake,
Yes, you can do what you need. All you need to do is set up a folder in a central location (like a shared network drive) that is accessible to any machines using PFTrack. Once you have copied and organised your scripts and presets into that folder, simply go into the Software Settings in PFTrack on each machine and point it to that new directory.
By mapping to a central location, you will only ever need to update or modify your scripts in that one spot, and the changes will be accessible for all users.
I hope this helps.
https://pftrack.thepixelfarm.co.uk/documentation/pftrack_preferences_application.html#file-locations
This post is from a suggested group
I have a plate that's pretty washed out, and I'm using an Enhance node connected to my User Track to up the contrast.
My Enhance looks like this:
but my User Track still shows the original washed out plate:
Should I expect to see the results of the Enhance in the User Track? This is another "I swear I remember it this way, but I could be wrong" moments :)
Hello Michael,
The way you have connected and are using the Enhance node is correct. We’ll take a look at this on our side and get back to you.
In the meantime, if you’re working with Log-encoded footage, you can use the OpenColorIO controls within the Clip Input node to transform the colour space and gamma of your footage. You can also make adjustments to Exposure and Gamma there.
You can find more information in the documentation here:
https://pftrack.thepixelfarm.co.uk/documentation/pftrack_node_clip_input.html#source
This post is from a suggested group
We are trying to map the IMU data we are getting from blackmagic ursa 17k rushes that we are getting as EXRs. But there are no examples of HOW imu should be entered.
This is a pretty constant issue because I have no control over how the post houses pack metadata for VFX pulls and I've had issues before just trying to map lens info that was there but not working when I tried to map it in my custom metatags.
SO can we get a full example of every possible metadata entry to help map things?
For example we have:
{exr.braw.motion.acceleration.x 0.0897826}
{exr.braw.motion.acceleration.y 9.69771}
Hello Toby,
All the supported tags you can use (such as frameRate, squeeze, focal, position, orientation etc..) are in the example metatags.xml file that ships with PFTrack, but it looks like you've got some raw IMU data from an accelerometer and gyroscope that cannot be mapped directly using the existing metatags.
Raw IMU data needs to be converted into position and orientation values that PFTrack can understand. This involves integrating the high frequency data from the IMU over time using calibration information specific to your device such as noise, bias, offset, drift and camera-relative orientations.
PFTrack can read the position and orientation values for ARRI and RED cameras, or from EXIF data generated by DJI drones because these devices have already integrated the raw IMU data using their camera-specific calibration before writing their media files. It looks like the Blackmagic RAW SDK (or the conversion to EXR) is just passing through the raw data, so we'd need a dedicated Blackmagic IMU importer to support this in PFTrack.
There may be some more information available in your EXRs or original BRAW file, but you'd also need all the original high frequency IMU samples for accurate results (rather than just one sample per frame) because IMUs for navigation typically operate at several hundred Hz.
If you can post a download link to your data then we can take a look and see what's possible. We are planning on looking further at the Blackmagic SDK, although there's no timeframe on that currently.
This post is from a suggested group
I've never seen this before, is there something I can do here? Solved camera with undistortion and a Geometry Track.
This post is from a suggested group
Hi all,
I'm solving four cameras together with shared user tracks, then piping all four into an Orient Camera node. Right now I have to choose a camera in the Camera Solver node with Current Clip, then switch back to the Orient Camera to view it, then back to the Camera Solver to choose the second camera in the Camera Clip, then back to the Orient Camera to view it, and so on. It wasn't always like this, was it? I remember there being a Current Clip pulldown in the Orient Camera, am I crazy?
Hello Michael, thanks for posting, we'll take a look into why that isn't appearing. In the meantime, you can use the keyboard shortcut 'C', which will do exactly what you need it to without the extra clicks.
This post is from a suggested group
We put together a deep-dive on metadata in modern matchmove pipelines, particularly the multi-source problem. Curious how you're handling mixed capture (cinema + drone + phone) in your own work.
https://www.pftrack.com/post/why-metadata-matters-in-modern-matchmoving
Hello Toby, would you be able to provide some more detail about your current workflow?