Is it possible to have something scroll across the screen really smoothly. I understand my Formula might not be the Best . However I have tried several things all with the same result . I see it as being a bit jumpy . It looks like a 1/10 second Jump . I wonder what happened to the third digit of the MS times . I guess as soon as you start doing maths with these numbers they get trimmed to save power .
Any Help will be appreciated.
.
.
(-180+(abs((((([UTC_TS]))%7200)-3600)/10)))
.
.
(((([SEC_MSEC])*60)%900)-450)
1 Like
Are you seeing this on the watch or just in the emulator. It may just be the emulator doesn’t perform as well.
If it is on the Watch I think it is time to report a bug with a support request. Be sure to provide an example project so they can duplicate it.
Ron
Samsung Developer Relations
1 Like
Thanks Ron . Appreciate you response and also thanks for the link . I will set something up that demonstrates the Issue .
@russellcresser , if i not wrong you are try to change x, y position, if i remember correctly x , y position is to 1/10th decimer, so what ever the result of your formula is it the movement will only be up to 1/10 of 1 position. Maybe that why u see in jumpy
E.g fomula by time = 10.1
Next ms = 10.2 etc
Best u can do is the jump is happens quicker so in negligible to the eyes.
1 Like
I created a simple rectangle and had the rotation and X position set at zero and the tags [SEC_MSEC] *10
And it kind of tumbled across the screen fairly smoothly but I think the rotation was a bit smoother than the position.
Ron
Samsung Developer Relations
1 Like
Yeah . Thanks so much . @r.liechty_SDR and @Knightwing . Yes I see the Error as 1/10 sec . And yes it does not show that much . We have a similar Problem on the Other Platform with one of their timers that includes Milliseconds . My issue is if the resolution is not three decimal call it Deci Seconds and be done with it . We know it is possible to have at least two decimal as the Smooth second hand is Great . I understand it is about Power Budget . It could be that there is a Limit in the number of Timers you can us with 2 or 3 decimals . Like the 15fps of the sequences . Ha Ha .
This is the Fake ZIP of the Test face I sent Samsung . Not heard back yet .
.
.
.
MY_SCROLL_TEST (2).ZIP (193.1 KB)
Using this tag
(([SEC_MSEC])*6.01) on the rotation it comes somewhat close to matching the animation hand but it seems to update every 30 milliseconds and it is a noticeable
Interesting.
Ron
Samsung Developer Relations
1 Like
Ha Ha . Oh what fun . Working out what the backlash on the Cogs is . I see what @Knightwing was saying about x and y position being .1px resolution . I am surprised in a way that we don’t have access to a really smooth Scrolling on X and Y but then again I am not as all these thing represent a load on the processor and the Battery . I have taken the App of the other Platform off my Watch as it is annoying in a WFS Development Cycle as a sync from WFS forces that App to Reboot on my watch which is very annoying as it takes ages . So the simple answer is to dump it . I have already reproduced my favourite Personal Faces . Well as far as I can .
Any way the point is the WFS Face I am running at the moment uses half the power during the Course of the day than the Other one does . As my Battery begins to Fail this will be my go to way of running it .
Hi, the ms is 1/1000 resolution, its correct but its the rotation/position resolution that is the problem.
As i explained… so what u getting without any multipling is every 100ms u move . 0.1 points.
If you want smooth a feature request is that for example i can enter something like this in x or rotation 10.001 degrees or x = 10.001px
or at least 2 decimals px or degress
saying that it could be WFF or API limitation? what i know is photoshop for x and y its allows up to 2 decimals px, so i think its possible.
There also [MSEC] tag = milliseconds in wfs
1 Like
Yes . If you open the Zip I posted you will see I have been playing with everything even UTC_TC or whatever it is . you have correctly Identified that it is rounded to 0.1 .
The story Continues . Thanks so much for your continued Input .
Ciao @russellcresser prova a vedere se quanto sotto in allegato può funzionare. Io mi son fatto una WFS con il testo scorrevole.
Hi @russellcresser try to see if the attached below can work. I made myself a WFS with scrolling text.
russellcresser.zip (152.8 KB)
Sesto stormo Ghedi.zip (6.1 MB)
Hey. @Mayalona . Thank you for sharing those Tests with me . The second one you posted needs a Password so I am not sure which one yours or mine . I will leave that one . You will see if you inspect my Test that I have 3 text messages ( hidden ) nearly touching for the development of continuous Scrolling across the screen . My problem is that it is not smooth . The Milliseconds attached to the Seconds that you have used as well is not Milliseconds . It is actually 1/10 of a second . I have a Ticket with Samsung and await a Reply although it is probably a Google issue . Who Knows these days . Thank you so vey much . We can work together to see if we can make something better .
To examine my TEST download the file and change ZIP to WFS .
.
.
EHI. @Mayalona. Grazie per aver condiviso questi test con me. Il secondo che hai pubblicato necessita di una password, quindi non sono sicuro quale sia la tua o la mia. Lascerò quello. Vedrai se controlli il mio Test che ho 3 messaggi di testo (nascosti) quasi toccanti per lo svolgimento dello Scrolling continuo sullo schermo. Il mio problema è che non è liscio. Anche i Millisecondi associati ai Secondi che hai utilizzato non sono Millisecondi. In realtà è 1/10 di secondo. Ho un ticket con Samsung e attendo una risposta anche se probabilmente si tratta di un problema di Google. Chi lo sa in questi giorni. Grazie mille. Possiamo lavorare insieme per vedere se possiamo fare qualcosa di meglio.
Per esaminare il mio TEST scarica il file e cambia ZIP in WFS.
For cleaner not random precision…
Do something like this
(floor([MSEC]*10)/10)
This will remove the made up values u spoke about @russellcresser
1 Like
Thank you . Sorry about our time zone Gap . I chat with guys over there and one in Taiwan . The window is only open for a while . I will try that very soon . Good call at lest it will not be stalling . It will still not be millisecond smooth . Thank you very much for staying with this . I platy a lot with this stuff on another other platform . Their Sec Mills is broken as well , but there we can use their smooth second hand angle as a timer for Smooth Scrolling .
Ciao @russellcresser , ottimo test il tuo. Ho provato a guardare la tua WFS, in effetti non si riesce a migliorare il movimento dandogli linearità . Ho provato con varie formule ma non ci sono riuscito. Proverò in questi giorni di festa a fare qualche test. Intanto ti allego copia della mia WFS aperta senza password. Per esaminarla scarica il file e cambia ZIP in WFS.
Hi @russellcresser, your test is great. I tried looking at your WFS, in fact you can’t improve the movement by giving it linearity. I tried various formulas but failed. I will try to do some tests during these holidays. In the meantime, I attach a copy of my WFS opened without a password. To examine it download the file and change ZIP to WFS
Hey . Sorry I missed this earlier I am in bed now . I will check you WFS in the morning . Thanks very much for opening it .
.
.
EHI . Scusa se me lo sono perso prima, ora sono a letto. Ti controllerò WFS domattina. Grazie mille per averlo aperto.
In your example watch face
use this formula for the rotation
(([SEC_MSEC] *360) / 60)
it matches the 15 mhz perfectly
Ron
Samsung Developer Relations
1 Like
15 mhz ???. By all the Gods that would be smooth . Thanks for your Input Ron I will check that out . Sadly that Formula , Ron has the same effect as (([SEC_MSEC])*6) . I think Basically because 360 / 60 = 6 .
It is not anyone’s maths at Fault is is because . MSEC updates randomly approximately every 1/10 second . It would actually look smoother if it was exactly 1/10 sec but it is not . My test it two Identical seconds hand . One is White and running at 15HZ as per WFS default . The other above is black running the tests . It is very obviously Glitchy . This it the test I have posted above If you want to check it . Also included my scrolling tests which raised the question .
Ps they are Hidden but it is not too difficult to switch them on . It is just difficult to know which is on and which is off if you have a few Similar layers . I still await a call back on my Ticket from Samsung . I bet it has run out of time .
I tried it in your example. (([SEC_MSEC] *360) / 60) removes the fractions and is smoother.
Ron
1 Like
I will show you a video of how it looks on my watch . I see you say the refresh rate on GW6 is 60HZ what is it on GW4 .