View my explaination in this thread
1 Like
@russellcresser sorry i realise my formula should
Be floor([MSEC]/100)* 100 instead.
Try that when u can.
Btw
For [MSEC] = floor([MSEC]/100)* 100
For [SEC_MSEC] = floor([SEC_MSEC]*10)/10
1 Like
I got this the other day actually but it does not really answer the Question and does not work either .
.
.
Its around there 10hz which is = 100ms refresh rate. So they saying they not going to speed up the tag expression refresh rate.
1 Like
If it was 1/10 sec it would not look glitchy with the 10HZ hand sweep . So it is not even that .
Do you mean using the 10hz sweep without tag expression already glitchy? Meaning the hand sweep is also having issue?
1 Like
No I think the 10HZ hand sweep is good as far as I can tell . This formula on an Image of the same hand superimposed should be the same according to Samsung .
((floor(([SEC_MSEC])*10)/10)*6)
But it is exactly the same ( Glitchy ) as other tests I have made . Ron and Samsung Consider the Question Solved . So noting is going to happen from this point on .
Foot Note . The the 5HZ looks pretty similar with Floor 10 and SEC_MSEC*6 .
I think I am done with this .
Hi, i think its the lags between each change. As i explain probably tag expressions caculation is more laggy.
Then the sweep, so they need code optimisation i guessâŠless line of code to do the same thing
I have done tag expression and the 10hz sweep together there seem to be a minor difference.
Sudo code explaination (guess work)
-
sweep
Display image, sleep(100) , display image -
tag expressions
Get time , do caculation, display image, repeat
So you see more steps (or this adds to the âGlitchâ ) and who knows what the process time for get time
And do do the simple math
1 Like
Wonder why we cant pm each other . This help make this thread less noisy for people who doesnât want to read into the details
I just wonder how other Platforms manage to make the Scrolling of an Image and the rotation Tags Smooth to the Eye . I dare say it is not Millisecond . The Pixels have to be Drawn and Redrawn . Of course we know that nothing Moves . I am almost Tempted to Look at Watchmaker and see how they Do it . I can not Believe Samsung have enough Power Overhead to run their Heartrate sensor at 12HZ continuously but not have a MSEC counter that is no better than 10HZ .
1 Like
Hi, do they base on tag expressions? Maybe they have code optimisationâŠas a programmer i do that all the time, e.g my first code takes 20ms to run , but with optimisation technic the same task could be reduce to 10ms.
Go ask samsung to explain why the 10hz sweep vs 10hz tag expressions is different outcome, see is it a bug or as i explained⊠it could be both
Because tbh i making assumptions base on my coding knowledge and what i observe when i did a testđ€Ș hope u understand it not a fact untill we get explaination from samsung
1 Like
1 Like
If they Hide the Hand that shows the error there will be no Error ! So I think I Know that . I can not use the smooth ( 15HZ ) action of the seconds hand to animate any thing else on the WFS platform . Either they are a bit Thick or assume I am .
1 Like
@russellcresser is this githcy to uâŠ
I have used a red 10hz sweep and a white overlap tag expression hand. ((floor(([SEC_MSEC])*10)/10)*6)
Or vise versa
What watch u usingâŠon my watch doesnât seem as bad as yours video , my watch is a gw5 pro
The normal [sec_msec]*6
Only notices the strobbing here
Verdit if on my watch look smoother, it means its a hardware issue. And lag issues due to the watch cpu
1 Like
I must say that looks smother . I think the cruelest test is the white 10HZ below the Black MSEC . At this stage in the game I can not remember what I have tested and filmed . I will do another . I am quickly getting to a point where ther is no time for this as we are going away . Nice Videos BTW .
I am using a Galaxy Watch 4 Classic . One UI 5.0 Watch WOS4 .
@russellcresser Hardware or firmware or refresh rate of screen.
One of those. Coupled with all the other issues
Btw on wfs preview, its glitchy for all the test i did.
1 Like
One more @russellcresser without floor formula
Ha Ha . Jinks . Just done the same . So thank you so much . I see the Blog is champing at the bit to close this topic . So Bar Aliasing problems this actually looks a bit smoother than yours . Probably a resolution thing .
I apologise to you for my Petulance . I appreciate all you have done . I guess you have work published . I would love to look at your work and follow you if I can . Will you please post some sort of link . Sadly we can not DM on this Platform . With your agreement I will flag this Solved or perhaps leave it in case any one wishes to add more .
BTW I used . (((floor(([SEC_MSEC])*10))/10)*6) above a 10HZ Sweep second hand hand .
.
.
@russellcresser close this topic? Anyway no i dont publish my faces.
I design for personal used.
1 Like