Metric - Barbell Tracker 4+

I really wanted to give this app five stars because I loved everything about it, except for the apps inability to recognize reps. Time after time, no matter where I put the camera, it didn’t record reps and thus didn’t give me any output for them. Because of that I won’t be going on to use the pro version, but otherwise this app was pretty useful.

Agh, Jason, we were so close!!! Thank you for the review, and your honest feedback. We really want to keep improving and maybe there is a bug or an issue we can fix that will solve this issue. If you could, maybe send me through some example videos that had failed to track your reps? We can take a look at these to see if we can keep getting better and improve the tracking reliability. I'd love you to be able to get good use out of the app in your training. - Looking forward to hearing back Jacob. You can contact me at jacob@metric.coach

It works pretty well, but it would be nice if they could integrate the Apple Watch and LiDAR that are part of the iPhone Pro line. The Apple Watch is a point of contact with a built-in accelerometer, which is closer to the bar and the LiDAR to get actual distance and more accurate reading of velocity with the camera. The app does an okay job, but the measurements are a bit off. The integration of allowing coaches to see their athletes' data is incredible, but with other devices, it is Standard practice. I can put a device that attaches to the barbell for 500 dollars, comes with a companion app, and it’s way more accurate. The downside of those devices is that they require an iPad and a mat for the laser to reflect on. I also can’t have the athlete use it at their will. They need to either own the device or have it on hand, But if you think about it, an iPhone Pro 12 can be found for less than that and an Apple Watch. There are over devices that run a couple of grand that, again, if you think about it, between your newest watch and iPhone, you're already in about that price point. It’s a great concept I hope they can expand on the accuracy and use available hardware instead of focusing on stop gap measures like giving me a notification that “I’m improving”.

Hi Jbarah,

We have actually been recently externally validated against MOCAP and found to perform really well for accuracy.
We have considered using the Apple Watch and like the idea of LIDAR also, but in our research into incorporating these technologies found they did not improve accuracy at all. In fact, the apple watch was a major step backwards from computer vision - delivering very erratic and inconsistent results. Something that is a known issue with other VBT hardware devices and wearables that have utilised accelerometer data - it is just too inconsistent.

It seems a bit inconsistent from where the timer starts for the movement and stops for the movement, across different sets of the same exercise. On one occasion, it had been way off in tracking the bar path, evidenced by the line it displayed that went in the complete opposite of the bar path. Additionally, some times don’t logically match with the load/intensity I used (i.e. set 1 of 405 lbs deadlift measured 0.55 m/s while set 2 of 455 lbs measured 0.52 m/s while set 3 of 405 lbs measured 0.57 m/s, with standardized rest times). The concept of the app is great, but I don’t feel like I can trust the measured data.

Hey xelanart,

Agh, a few people reported this quirky bar path issue a couple of weeks ago, we should have hopefully pushed a fix for that with v4.3 & another that tightened up eccentric tracking with v4.4.

If it happens again do you think you could send us some sample videos of this weird bar path you are seeing? We are always hungry to improve the tracking algorithm, and bug report cases like these are really helpful at exposing weaknesses and limitations in the app!

In terms of the velocity differentials these numbers are very normal and that level of variation would be fairly typical no matter the tech you use to track velocity.

Every set will have its variations in technique, velocity and ROM, a spread of 0.52-0.57 for these 405 lbs deadlifts sounds like a pretty typical workout! Happy for you to send me the videos for these and I can explore further if you think the numbers are wrong though!

A spread from 0.5 - 0.65 would be concerning, but all sets between 0.05m/s is Metric doing a great job of delivering reliable data!


Jacob