So, as we probably know by now, the 1080p60 and 1080p96 modes use Pixel Binning instead of real post processing to save some milliseconds for the upcoming DJI Goggles.

Now, I know that DJI Ken and others keep saying that it's "not otherwise possible due to hardware limitations" but where exactly are those limitations?
The DJI Mavic has enough processor power to do proper de-Moire effect filters for the 1080p60 mode (like other drones have, as example the P4 or P3). Even if it's not a dedicated chip in the Mavic that's going to handle the processing then, it would probably be possible if some genius DJI technicians sat down and think about a clever solution


So, long story short: I want a full depth explanation (without the specs sheets, I know them already) on why it wouldn't be possible to implement such a mode even with the lack of a dedicated chip by doing everything on other processors inside the Mavic. We could then switch between "Latency-Free" or "Quality" mode with a switch in the DJI app if it's going to be possible or maybe even add an "Expert" mode to the app with a warning which lets the user freely set everything the Mavic's picture sensor has to offer.

Or is it simply because DJI wouldn't have a selling point for the P3/4 anymore then?

Btw. this thread is also here so I can spread the information across other forums currently discussing this topic, it's spreading like a wildfire and they're all (including me) waiting for a really technical, in depth explanation on this topic.

Saving milliseconds for the goggles makes sense as to why they would do it that way, Hadn't considered that angle.

From other posts i gather that the heatsinks on the bottom of the craft can get pretty warm.
It might simply be a limitation of the tightly packed hardware , Adding any further processing might just melt it..

If the hardware is capable then they should definitely have an option to switch to a higher quality sensor readout/debayering method. Or detect the use of the goggles and switch to the fast binning for that use automatically. This is essentially what the gaming mode on your tv does.. dissable all the image processing for less latnency.

Personally i don't understand the marketing angle of the Mavic to be used as an fpv racing drone. It's far to expensive and fragile to be crashing it into trees and warehouse walls etc. especially when you can buy or build something for that purpose for only a couple of hundred dollars.. But that's another topic.

i dont think its to save milliseconds, its to have a high fps, 30 fps in first person mode makes people sick. {:4_117:}

ferrari-458-ut Posted at 2016-11-7 07:37
i dont think its to save milliseconds, its to have a high fps, 30 fps in first person mode makes peo ...

The question isn't why is there a high frame rate. The way that it is processing the 60/96 frame rates (so it has less latency in transmission for FPV reactions) makes the footage unusable. This post is about why can't they have the ability to change the way it processes it so you can use those frame rates to capture video that comes out looking good.

At least I think thats accurate.

trance728 Posted at 2016-11-7 14:01
The question isn't why is there a high frame rate. The way that it is processing the 60/96 frame r ...

Yes, correct

Thats why, wont go away:

http://forum.dji.com/forum.php?m ... 0MjY5MTF8Njg4NDI%3D
http://www.sony-semicon.co.jp/pr ... roducts/imx377.html


In full context:
http://forum.dji.com/forum.php?m ... mp;page=1#pid543211


Nils

nils_heidorn Posted at 2016-11-7 20:14
Thats why, wont go away:

http://forum.dji.com/forum.php?mod=attachment&aid=MzE3MDZ8ZjA0MjQ2N2F8MTQ3 ...

... do you even read?

"I want a full depth explanation (without the specs sheets, I know them already)"

EpicLPer Posted at 2016-11-8 20:14
... do you even read?

"I want a full depth explanation (without the specs sheets, I know them alr ...

So lets see if someone takes the time laying it all down to you if you ask so nicely :-)

Nils

EpicLPer Posted at 2016-11-8 12:14
... do you even read?

"I want a full depth explanation (without the specs sheets, I know them alr ...

I want a pony, oh, and a new Mercedes!

I want, I want...... you sound like Varuca Salt.

Manners cost nothing.

Bonjour,
Le pixel binning est un moyen soit d'augmenter l sensibilité du capteur ( pour les anciens ccd ) ou de diminuer le taux d'information au détriment de la résolution. On regroupe les info de plusieurs pixels pour moins transmettre de données. Ici avec le taux de transfert de la caméra, qui n'est pas si mauvais compte tenu que l'on est en en 4:2:0, on peut imaginer que le jusqu'à 1920/1080 on peut aller à 60fps sans dégas. ( une XDCAM sony donne de débit en 1920/1080 à 30FPS mais en 4:2:2. Ensuite plus de fps, faut compresser et perdre de la résolution. Noter qu'au cinéma on est à 24 fps et que l'image est belle et fluide. Le shuter est lui à 1/50 presque toujours. L'intérêt des vitesses plus élevées est  le ralenti possible. Le rest j'ai des doutes, une image vidéo n'est pas une image de console de jeux; pas la même esthétique.
Essayez donc le 25 fps avec un shuter à 1/50 ; avec les filtres gris si il le faut.