Wide 4K Tracking Without Zoom?

I’ve been going down a bit of a rabbit hole trying to find the right setup for recording full football matches, and Folocam looks like it could potentially solve a lot of the problems I’ve been having.

So far, I’ve looked at multi-camera setups with stitching, such as Veo Go with multiple iPhones. I bought an Insta360 and experimented with AI tracking/post-processing (since returned it), looked at the GoPro Mission 1 Pro but was concerned the lens/FOV wasn’t wide enough, and even had an Xbot Falcon on preorder before cancelling it.

More recently, I’ve been looking at fisheye lenses for my iPhone with the idea of simply capturing as much of the pitch as possible in 4K and then reframing/tracking the footage afterwards.

Then I came across Folocam…

From the footage I’ve seen, the tracking looks really impressive and potentially better suited to what I’m trying to achieve. However, I have a few questions about whether my particular use case is supported.

My main concern with most automated football tracking systems is that when the camera is zoomed in, occasionally the ball/play moves outside the frame. When the tracking catches up, the pan can also look quite sudden and unnatural.

For me, keeping the play in frame is more important than getting a close-up shot. I’d rather capture a wider view reliably and then crop/reframe it in post.

1. Can Folocam track the play without zooming in?

Could I have Folocam follow the general area of play while keeping the iPhone camera at its widest/native field of view?

What I’m imagining is almost a 180° tracking/panning mode: the gimbal simply pans left and right to follow the general direction of play, while the iPhone continues recording a wide 4K image without digitally zooming.

This would give the tracking much more margin for error. Even if the ball moves quickly, it should remain somewhere within the wider frame, and I could then produce a tighter version afterwards.

Is something like this currently possible?

2. Can I retain the wide 4K master recording for post-processing?

Rather than relying entirely on Folocam’s final tracked/framed output, I’d ideally like to retain the highest-quality, widest recording possible.

I could then take that footage into another application afterwards to crop, reframe or potentially run AI tracking over it.

Essentially, my ideal workflow is:

Folocam tracks the general play → gimbal pans left/right → iPhone records a wide 4K master → I decide on the final framing afterwards.

Is that workflow possible with Folocam?

3. What is the absolute best gimbal for Folocam?

If price wasn’t a consideration, what gimbal would you recommend specifically for recording football?

The COMITOK Mogo looks like the best option from what I’ve seen on the forum, but in some of the footage I’ve watched, the movement still looks quite jagged rather than being a smooth pan following the play.

I’m happy to buy whatever gives Folocam the best possible tracking and smoothest movement, so I’d be interested to know what the developer considers the best current setup.

4. Is the lifetime Folocam subscription available in the UK?

I noticed the lifetime subscription offer and, assuming I can achieve the setup above, I’d definitely be interested.

Is the lifetime option available to UK users?

My use case would be recording full 7v7 UK football (soccer) matches around twice a week, to uoload onto youtube, so Folocam would get a fair amount of real-world use.

Also, if the developer is interested in this particular use case, I’d be very happy to provide match footage, testing and feedback.

I’ve spent quite a bit of time experimenting with different approaches already, and I think there’s a really interesting opportunity for Folocam here. A phone + gimbal setup that can reliably follow the general area of play while capturing a wide, high-quality master recording could be a fantastic alternative to the much more expensive dedicated football-camera systems.

Would love to know whether the wide 4K + 180° pan/tracking + external post-processing workflow I’ve described is currently possible - or whether it’s something that could potentially be supported in future.

Thanks!

This may work for small fields, but not for 11v11 soccer. This is exacly why XbotGo Falcon fails miserably with tracking. The FOV of its main lens is perhaps 120 degrees, and it would be nearly impossible to track the ball at a distance with such a wide angle.

It might be possible to track using the main camera while recording video captured by the wide angle camera. This is something I may try, but at this time I’m focusing on the AI tracker, which can capture the 180 degree view, so it will never miss an action.

I would suggest Comitok Mogo with remote. Right now they’re clearing out their inventory so the price is really low. There isn’t much difference between this gimbal and other gimbals in terms of performance.

Yes, but you’ll have to pay for the shipping. I would suggest just getting the Comitok Mogo with remote since it is so cheap now, and buying the yearly subscription from the App Store.

Thanks, really appreciate the detailed response!

Just to clarify your point about the wide-angle camera, do you mean potentially using the iPhone’s main camera for the AI ball detection/tracking, while simultaneously recording the actual footage using the iPhone’s ultra-wide camera?

If so, that sounds really interesting and very close to what I’m trying to achieve. My priority is essentially to give the tracker the best possible chance of following the ball, while capturing a much wider master recording that I can crop/reframe afterwards.

That leads me to another idea I’ve been considering.

Would it theoretically be possible to use the iPhone + FoloCam + gimbal purely as the tracking system, but mount a GoPro Mission 1 Pro for on the same rig for the actual recording?

For example:

iPhone/FoloCam detects the ball → controls the gimbal → GoPro Mission 1 Pro mounted above the phone → records the same movement in 8K60 with a wider fov.

The Mission 1 Pro wouldn’t need to communicate with FoloCam at all. It would simply move with the iPhone/gimbal.

My thinking is that this could potentially give me the best of both worlds. The iPhone’s main camera gives FoloCam enough detail to reliably detect the ball at distance, while the Mission 1 Pro gives me a much wider 8K60 master recording. I could then take that footage into post-processing software and crop/reframe it much more aggressively while retaining more detail.

Would mounting a second lightweight camera above the phone be feasible from the gimbal/tracking perspective, or would the additional weight and balance cause problems?

I’m also curious about the longer-term direction of FoloCam.

Have you ever considered putting the AI/tracking hardware directly into a dedicated gimbal or pan head, so the iPhone wouldn’t need to perform the tracking itself and essentially any camera could be mounted to it?

Or is using the iPhone’s processing power actually one of the reasons FoloCam’s tracking appears to perform so much better than some of the dedicated systems, such as XbotGo?

I’m wondering whether products like Falcon are limited partly by the processing hardware they have available for real-time computer vision, whereas FoloCam effectively has a very powerful processor, camera system and neural processing hardware already available in the iPhone.

One other question, as I’ve never actually owned or used a gimbal before.

In some of the football footage I’ve watched, I can still see a little bit of shakiness or quite abrupt/jagged movement when the camera changes direction. Are some DockKit gimbals noticeably smoother than others for this type of slow football panning?

For example, would an Insta360 Flow 2 Pro or DJI Osmo Mobile 8 potentially produce smoother movement than the Mogo, or is the actual movement largely determined by FoloCam/DockKit, meaning there isn’t much difference between the gimbals?

For my use case, smoothness is probably more important than tracking speed. Ideally, I’d want the gimbal to gently follow the general area of play rather than make lots of small corrections trying to keep the ball perfectly centred.

I’m going to order a setup shortly and would be very happy to experiment with some of these ideas. I’m recording mainly 7v7 and some 9v9, so I have a few years before worrying about 11v11. :sweat_smile:

Thanks again!

This could work, but the AI tracker I’m building now should provide better tracking, mostly because it has a 180 degree view

This would not work – it is very difficult to mount anything else other than the iPhone on these gimbals.

The shakiness is usually not caused by the gimbal. It is usually because either the detection is too slow (especially on older iPhone models), or the ball detection model isn’t consistently detecting the ball.

1 Like