@Stef_Weicks in my program the time to write to the card the entire universe (80323 bytes) takes sometime (way more than displaying the leds) hence the possible skip of frames that is why I hoped that the use of core 0 would help.
I could write each universe one after the other but if one universe was missing then the entire frames would be messed up. I could write this implementation and see.
@Yves_BAZIN no on the esp8266 only 4 universes till now… but with those it worked as far as i know.
ok i understand what you mean…
unfortunately to the line artnet.resetsync() i also know no explanation. i could try it without this line?
@Stef_Weicks the explanation is the following
Jinx is pushing at around 25 FPS so if you has to push 4 universes you have
- plenty of time between frames
- it’s fast to write
When you have 16 universes you have - less time between universe 16 of frame x and universe 0 of frame x+1
- more data to write
Solution :
1)I will create a most basic implementation hoping that you never looses universes - I will create a saving with a buffering that will allow more flexibility
And yes please try without artnet.resetsync() in case of 
@Yves_BAZIN i think you understood this system pretty well and it sounds like a great plan!
i tried it but now recording didnt start anymore.
@Stef_Weicks can you show me the output ?
@Yves_BAZIN
WiFi connected.
IP address:
192.168.1.101
Starting Artnet nbNeededUniverses:16
Ready to record…
@Stef_Weicks and if you add artnet.resetsync() it restart ?
@Yves_BAZIN if i add it it starts recording by starting the output in jinx.
output in jinx closed and i put in the sdcard it stands at Ready to record…
if no sd is in the slot it says : Recording stopped …
the recording starts only with artnet.resetsync() in the code.
@Stef_Weicks I am not sure I understand
With artnet.resetsync() does it record
What do u mean by if I put the sdcardin it stands ready to record ?
yes its recording with the line artnet.resetsync() 
@Stef_Weicks oki then
@Stef_Weicks
hello here it is
first new Artnet library
in your program to read from the wi-fi
line 105
artnet.begin(NUM_LEDS+160,UNIVERSE_SIZE);
replace by
artnet.begin(NUM_LEDS+160,UNIVERSE_SIZE,1);
new program to save
@Yves_BAZIN ok thank you! so i tried your new library with the new code to record and the modification.
but there are also jerks in the animations I do not know if they are different than they were before .
i did a short video of it so you can see … it seems as if the jerks are in a kind of rhythm.
missing/deleted image from Google+
@Stef_Weicks could you record a longer pattern ?
compteur<2000 for instance and show me the result ?
@Yves_BAZIN
this video is the middle part of a animation done with compteur<2000.
These jerks move through the whole video in this rhythmic sequence
@Yves_BAZIN isn´t it possible to save an animation on the sd card …without transmitting the data over wifi. so we could rule out whether it is the recording or the playback-program.
@Stef_Weicks yes that is possible but we need to change the program to read serial and change the output of your jinx to transmit via serial and not wi-fi.
@Yves_BAZIN ok if there is no other way to do this? yesterday i informed me about this. However, I am not sure which output device I should choose … artnet is only possible via ip address … there is also an option to send over tpm2 to a com port … or to save each universe as a * .out file …
If I have a look at the tpm2 format that should be possible.
There format are more or less just a question of headers.
You have the possibility to save the animation on a .out file ?
That is maybe the best solution then.
Could you create one store it on the sdcard and read it using the program ?
There is also something I noticed it’s like there half a frame read only when it skips …
But could you try WiFi with a longer recording ?