Trainerday Android application - split screen and floating window

Android 11. It mainly concerns the application used on the tablet, although the smartphone can also be used with a split screen, although the problem is even greater there and the application is practically unusable in this mode.

On the tablet, you can turn on the use of the application on a split screen and / or in the form of a floating window. Unfortunately, bad scaling of the application causes that, apart from full screen, part of the user interface (mainly what is at the bottom) is truncated and the application cannot be used properly.

Similarly, the application as a floating window does not scale properly and the problem is as above.

It is a pity, because on the tablet I would like to have an application running in the upper / lower part of the screen (actually 3/4 of the screen), and a movie, e.g. youtube, displayed at the bottom or at the top.

A correctly displayed floating window would be useful when Chromecast is casting the screen of a mobile device to a TV screen. Alternatively, you could think about the Chromecast interface that allows you to display training data over, for example, a Netflix movie, of course, control still from the mobile device.

If improving the split screen and floating window mode is not possible at the moment, it may prevent the application from entering these modes?

Hello Leszek, good to hear from you again. I don’t really know much about this but I have a feeling this would be a lot of work to getting split screen working. Are you suggesting if we can’t do this soon (which we can’t) we should prevent split screen?

Here is something I have considered, is Xert does some web page you can view your data while riding the app. If we allowed that and maybe we make the web page on lower part of the screen would that work for you? I not sure what is technically possible regarding locking a window at the bottom of the screen, but even small browser that you manually put there? Or maybe we allow you to open netflix or other windows in the top and put our iframe at the bottom… Just exploring ideas.

I too use the split screen function on my phone because I have Trainerday app on the top at 3/4 of the screen and Fatmaxxer (DFA-a1) on the bottom 1/4 off the screen.
Trainerday can’t be scrolled in split screen, goes back to library and pauses the workout when anything else happens on the phone or in the Trainerday app. When you take it out of split screen for example to check the trace, or when there is an incoming phone call or …
That is annoying, but if nothing special happens and you start the workout in Trainerday when it is already in split screen, it does the job for me.
I would be very pleased if Trainerday was made fully compatible with split screen but I can understand if that is not a priority at this time.
The reason I’m working like said above, is that this is for me the most convenient way at this time, to see my DFA-a1 real time. I do have my Garmin running at the same time, but Fatmaxxer is still unable to send the a1 value to the Garmin Edge.
I set both apps in split screen and first start Fatmaxxer. At two minutes in (the time to get your first a1 value), I start Trainerday app and my Garmin at the same time. This allows me to align all data… A bit cumbersome for now, but I’m sure things will improve in the near future.

I used Fatmaxxer with a smartphone, I didn’t think to use the tablet in split-screen mode. I did tests with the IQ DFA Alpha 1 field with the Edge 1030 - the results are very similar, provided that both the Smartphone and Edge use a BT connection. If the smartphone uses BT and Edge ANT + are the differences, BT is definitely better in this case.

For me, Fatmaxxer sends messages about changes a1 to Garmin Edge / Fenix. Quite readable, although I would prefer some IQ field to display the current state calculated in Fatmaxxer. The calculating widget A1 is quite accurate, but must be run for 2 minutes to display the correct condition

Alex, I think if a feature is not working properly it is probably better to turn it off. In this case, operation on a split screen, although looking at MedTechCD, for example, it may leave it to the user’s decision with the information that we do not recommend using it on a split screen due to problems with scaling the interface. Unless it would be possible to set the switch in the (advanced?) Options enable / disable work on a split screen. Users would not be surprised that the application allows you to work in this mode, but not quite as it should.

The ideal, however, would be to be able to operate on a split screen or in the floating window mode (this is rather a tablet-specific function, I only have a split screen on the smartphone).

Maybe it’s due to some simple app UI scaling bug?

Adding the ability to work on a Youtube / Netflix movie would certainly be well received by users. Please think about something like that;)

For now, I have been excluded from training for several months, I have a spine surgery next week, then a gradual recovery, it will take at least a few more weeks before I get back on the bike, maybe you will do something by then;)

If not, then possibly a tablet / smartphone will be used;)

The existing IQ field for dfa-a1 only runs on Edge 1030. Anything else hasn’t got enough processing power and I’m on a Edge 830.
I used the notifications for a while but I find them hard to read and quite intrusive on my Edge screen. It blocks the view on almost half the screen.
I suggested Ian from Fatmaxxer to emulate a new sensor on the phone app and then find someone to develop a receiving IQ field. Ian emulated a HR sensor that sends a1*100 for testing purposes but although it can be used on some equipment, the Garmins see a new sensor but can’t connect to it for unknown reasons. On Github, you can find a version with this BLE emulated sensor.
Things seem to be stuck at this point…

Interesting about dfa-a1, I had not heard of that metric. But it sounds a little contrived, meaning saying your aerobic threshold is at 75% of your full HRV score/swing. Is there proof of this in any reasonable studies? That sounds very individual.

I find when using SMO2 (which I would assume is more accurate than a random 75% figure) that my aerobic threshold is actually at around 70% of HR max consistently so once I know that if I just go 68% I am close enough. I am way more out of shape than Seiler and most that can get that up closer to 80% :slight_smile: Although Dr Maffetone also an expert claims it should be closer to 70% for most individuals and he has done thousands of tests. I guess my main point is it seems more stable than requiring on going paying attention, I guess if you are trying to always target 1 BPM under… but it would seem just dropping 3-5 would be super safe.

Maybe I am confused on something here as I said dfa-a1 is totally new to me (just reading up today) but since I build the second app in the iOS app store HRV+ 8 years ago I know a bit about HRV (long before all these new comers :slight_smile: ). I don’t know any where near as much the experts like Marco from HRVlogger or Simon from iThlete…

There is already a bunch of studies and the dfa-a1 metric is hot these days.
I suggest you read up on it at this site:
AI endurance, Fatmaxxer, Runalyze and HRV Logger are apps to calculate the a1 value, some in real-time, others in post-analysis. It is getting accepted as a very good estimate for VT1 really fast.
There is some debate around the VT2 being at a1=0.5, but that seems much less accepted.

That is absolutely not what it means. The a1=0.75 is a point where the detrended fluctualisation analysis becomes random. At higher a1 values, things still fall in patterns. Well, it’s quite an advanced mathematic analysis…
A lot of people are getting results where a1=0.75 is very close to the VT1 determined in a lab test. Very close being 5bpm or less. The advantage is that it can be displayed in real time with more or less standard equipment; thus giving you an oportunity to witness fatigue, decoupling on the fly.
I’m hooked on it!

Ok very interesting. Yes I read too fast and just tried to lightly understand. With Maffetone and following low HR for running I was able to eliminate cardiac drift up to 1 hour. But this does sound interesting. I will have to play with it some time. Now I see drift after about 5 minutes :slight_smile: at 50% of FTP :slight_smile:

I should also ask are you saying they are getting a semi-reasonable prediction of LT1 from dfa-a1? This is something that interested me very much in my SMO2 testing, it seemed possible but if this works from HRV I agree this is a breakthrough concept and I will have to understand it better.

a1=0.75 is a really good estimate of LT1 and the advantage is that it follows your state of fatigue, recovery. That is the primary use of it. Having it displayed in real time is very nice during endurance rides and keeps you from riding at too high intensities.
Further more, once you know your normal a1 value during a standard warm-up, it also gives you a good idea if you are ready or not for a harder session. Lower then normal a1 values indicate incomplete recovery.
This page will get you going with the concept:

Exactly - Medtechcd introduced a question that is quite well documented not only on (but this blog is a knowledge mine …) Just throw in google on this subject and we get a lot of valuable information .

In brief:

  • DFA Alpha 1 Might Prove to Be a GroundBreaking Advancement to Correctly Set Our Training Zones and Track Fitness. IT Only Requires You to Wear A Heart Rate Monitor That Tracks HRV (Polar H10 Recommended). This Way, You Will Be Abe to

  • Determine your aerobic and anaerobic thresholds. For Example, Your Aerobic Threshold Sets The Upper Bound of Your Endurance Training Zone. It is Important This Bound Is Set. Correctly, this Properly Polarize Your Training, and Actually Take Your Easy Activities Easy For Constructive Training Progression.
    Avoid Pitfalls of Common Test Protocols. If you do not have access to a Physiology Lab, You Can Currently Assess Your Aerobic Threshold as the Pace / Power Where You Can Comfortably Hold A Conversation. Since This Method is obviously Not Very Accurate, DFA Alpha 1 Presents an Accurate But Affordable and Non-Invasive Alternative. For The Anaerobic Threshold, Time Trial Fitness Tests Are Hard to Pace and Lead to Significant Fatigue. On The Other Hand, Ramp Tests Which Take A Percentage Of Your Max Power / Pace As Your Anaerobic Threshold Depend on the Ramp Speed ​​[…]


Currently, the best source of high-quality HR data is Polar H10 / H9 connected BT. As I mentioned, the EDGE 1030 DFA Alpha 1 widget estimate as well as FATMAXXER provided that it will be connected by BT. Ant + does not give such a quality data on which the author of the blog and what I confirmed by registering data and then analyzing in Kubios.

Estimation of zones (mainly aerobic threshold area) in real time is very useful for polarized training. The HR drift does not mean significance, because the estimation is carried out from HRV data. The fatigue of the body, etc. is also taken into account. A very interesting and cheap alternative to laboratory measurements.

Hence my request (in another thread) about the possibility of checking training on a trainer based on heart rate. At the beginning of a simple, but you could think about advanced training polarized with heart rate and power control, even estimating the real-time DFA Alpha 1 - FATMAXXER code is available (Apache 2.0 license).

I love it guys… I feel like I have been living in the dark. I would love to build this into our app. I have used Polar H6/H7/H10 for a long time but in the past comparing Wahoo tickr I was getting nearly identical HRV values. I wonder if there is something else critical about H10 that required.

1 Like

This is incredible. Fatmaxxer is open source how nice. At the moment I would say for sure we will build it into TrainerDay and maybe even allow automated power control via dfa-a1 in addition to hr targets… I need to keep researching and test this out for myself. Good luck with your surgery, I wish you a speedy recovery.

That sounds absolutely fantastic and would, for my use case, eliminate the need for the split screen functionality.
If you have questions, please reach out. I have been following the project since a very early stage and did the dutch translation for the Fatmaxxer app.
I first touched the subject months ago (feb this year):
and have ever since started to use it in some way. First post analysis with the Colab math sheet from Marco Altini, then real-time with Fatmaxxer.
Regarding Polar straps:

  • They are widely known and used as the most accurate ones for RR intervals. Thus developpers easily adopt them
  • Fatmaxxer has some advanced artefact detection and records a 10sec ECG strip in csv when such an artefact is detected. Polar H10 can do that, don’t know if others have that functionality
  • I’ve used results from my Garmin Dual HRM, recorded by Edge 830 HRV option, in the Colab calc sheet and results were similar when artefact number was low. Unfortunately, I didn’t know at that time that BLE connection was way better than ANT+ for this.

Here are 2 examples of artefacts from my endurance session yesterday evening:

I think they are harmless APC’s but will check with a cardiologist. Sure thing is that one such artefact throws off the a1 value for all windows where it is present. That’s why Fatmaxxer filters them out.

Fatmaxxer is open source and Java. David has integration of it on the intervals todo list. Post analysis in that case.

Ok yes, regarding Polar. I always recommended them for HRV back in the day too, but it seems others have come along too just they are the gold standard. 100% for sure I will reach out to you if we decide to move forward with this. I need to test it myself and get a better grasp on it. Probably would start it in the late winter, early spring would be my guess.

Yeah I actually bought an EKG/ECG device a long time ago when I got obsessed about HRV and HR in general. Not hospital version… but I realized quickly that stuff got very complex :slight_smile: But yes your readings look good, not that I know what then mean :slight_smile:

Yes my HRV app filtered out errors too. They occur. What would almost be funny / interesting is real-time FTP estimates but I would not want to do that :slight_smile:

Thanks again. After and or during my testing I will follow back up.