Hey,

Just a heads up, i noticed this as well on our last shoot, but couldn't put my finger on the problem and the shoot was to busy to trouble shoot, so i compensated with manual adjustments. We have another shoot next week and i will have a look out if this happens again. Also, you don't seem to be the only person noticing this... It almost feels there is something "off"(?) with the gyroscope.

Hi!

How do you make compensation of drifting?

I think most problem is postion dirft rather then rotation (gyroscope) cause if we powering off celing area - drift noticable near 0 (if expose outut pin of game pose node in realtime we get rotation drift in 1\100 of deegree)

What i did was to plug a "Smooth transform" and a "Hold Transform" node and used them as needed.... we where lucky and didn't use much "live movement" and did a more traditional cut to cut tv style shoot. The main reason for the above nodes is also the very noticeable "jitter" with long lenses and antilatency.
We have another shoot this week in a smaller studio with a smaller ceiling, while i didn't noticed the drifting just yet (first day shooting is tomorrow) i noticed the jitter again, so the two transforms are back in the composition and i switch them as needed.

Great Thank you for the explanation. I will check it in our area and write back results here

Same problem here with jitter (looks just like your reference video). We also doubled the amount of trackers with no noticeable improvement. Sadly, this was a costly, but futile exercise. We do see moderate improvements when adding the smooth transform node in Aximmetry, although it's not a true fix for the problem (it's like telling a person with a broken leg to walk on a padded floor). I'm hoping that someone finds a solution.

Hi Steamshow,

Are you using a wired or wireless socket for the tracker?

Can you do about a minute-long recording of the jitter to see how it compares to others? (and do not move the camera while recording)
You can easily set up the following logic in the flow for recording the tracking:

Warmest regards,

Hello, thanks for your response. We are using a direct connection to the alt trackers via USB-C. We reinstalled the entire ceiling and doubled the tracker count and still have the same level of jitter. We’ll add the suggested node and record the jitter results for your review. I’m hoping that it’s a simple fix to correct, I can see that it’s not an isolated occurrence and that a solution would benefit many Aximmetry users that also use Antilatency tracking systems. Thanks again for you help on this!

Hi

Same issue here with antilatency jitter, exactly as the video you uploaded.

Also wired directly with USB-C and was also considering doubling ceiling makers, but seems it does not help.

The solution proposed by Eiffert in another post was a step in the right direction, but really needs a lot of tinkering and ideally more complex filtering  solutions by the Aximmetry developers would be really helpful :)

Hi, here's a link to the test and example files that should help you to see the Antilatency tracker jitter issue. The tracker was completely clamped down and not moving. It looks very similar to what the other users are experiencing. Let me know if there is anything else that I can do to help out with this issue.  Looking forward to a fix for this!

Link to files: https://spaces.hightail.com/space/6uCFAYXTxk

Thanks!

Scott

Alexander Zhivykh

OMG this jittering is the same as cheep HTC Vive tracker that we are using at the moment. We order Antilatency a couple of days ago and waiting for the package. We bought it to get rid of jittering and from what I see in your video, it is exactly the same... It doesn't make any sense... 

We started with the Vive system as well… And just like you, we switched to Antilatency as a next evolution step. After dealing with the Vive, I must say that Antilatency is far superior in so many ways. We just need to get past this glitch in the system. Our original Antilatency “test” floor setup didn’t have this issue and was very stable. We’re hoping that this is is just a momentary software glitch or an issue with incompatible versions of the firmware or software. The teams at Aximmetry and Antilatency have been quite helpful and I feel confident that a solution will be found soon. Fingers crossed!!

Looking forward for the solution too. I've tried the smooth camera compound as a temporary solution, but the drifting still remains. 

It’s important to distinguish between jittering and drifting. Jittering is when the virtual camera make small position shifts at a high frequency but stays at the same average coordinates over time. Drift is when the virtual camera slowly shifts its position from the absolute positioning in relation to the physical camera.

@Eifert, is the trackkng data smoothing applied to the raw high fram-rate tracker data or just at the project frame rate “tick’? The data should really be smoothed independent of the project frame rate and then read on tick, in lockstep with the project frame tick. 

Drifting will occur if your IR markers aren’t exactly (I mean EXACTLY) lined up. You will always get some amount of jitter with any tracking system, especially when zoomed in. Clever filtering might help hide it, but responsiveness will suffer for fast camera moves. Depending on how Aximmetry handles the smoothing it can be more or less effective and with more or less side effects. 

Hey guys! I'm Artem from the Antilatency team. The jittering/drift problem can be caused by various reasons and requires a personal approach. I suggest you write to our support or discord server so that we can solve all your problems.

Hi all,

Somethings to consider:

If you are zoomed in, any jitter/drift will be more visible.

There are no perfect tracking systems, there will be some level of jitter in every tracking system.

Putting smooths and filters will make the jitter less visible, but it will also cause lag.

And as Artem said, there are various reasons and approaches for jittering/drift. For example, we found that putting the tracker further away from the center of the camera and from the camera's stand gave it a bigger line of sight and hence gave a bit better accuracy. There are probably many other tips that our friends at Antilatency can provide if you contact them on their support or discord.

Warmest regards,

No drift issue and a smooth transformation of 0.3 temporarily solves the problem of jitter.

I hope Antilatency could look into this issue, seems like many people encounter this jitter issue. 

Hey everyone. Did someone find a workaround for this problem? We still have severe jitter even thought we were told by Antilatency that this is considered normal. It works sort of OK for wides but basically unusable for the closeups.  We tried removing lights from studio ceiling with no effect. 

https://youtu.be/qeHfocJAQaw

Hi,

The latest version of Aximmetry (2022.1.0) comes with a new feature to stop jitter when the camera is not moving.
Note that this won't fix constant drifting. And if the jitter is visible in normal usage we recommend first to ensure that your tracking system works correctly.

In short, enabling Tracking Dejitter will lock the camera to a fixed position while the camera doesn't move more than Dejitter Pos Thresh and doesn't rotate more than Dejitter Rot Thresh.

These settings can be found in every Tracked camera compound in the Inputs:

Related Aximmetry's Version History reference: "TrackedCam: added Dejitter options which allow suppressing the noise of the tracking system when the camera is still."  https://aximmetry.com/learn/software-version-history/#%E2%96%BA-latest-2022-1-0 

Warmest regards,

Thanks so much for adding this great new feature, looking forward to trying it out!

Hi, in my case, the camera tracking shakes considerably even on a tripod. Tried dejittering but makes it even worse. It seems that i have to move the camera super slow.

https://youtu.be/PseugznXrO4

Great software anyway, and glad that is free.

Hi Carlos,

Warmest regards,