First of all: i HATE digital zoom. I never used it on Action 4 or any other camera i own.

BUT: Now the Action 5 has a 40 mp sensor so it may make sense using it in well lit situations IF the camera leverages on the full 40 mp resolution to crop the zoomed part of the image.

I did a quick test in Horizon Balancing Dewarp and RockSteady+ Wide, 100 ISO best light conditions (using a LED panel), 50 FPS and full bitrate horsepower.
Here you can admire my desktop daily mess: the original clip zoomed 100%, 150% and 200% in Davinci Resolve VS the clip shot at 1.0, 1.5, 2.0 zoom levels in camera (i know, 100% and 1.0 are worthless as they are the same clip but it is cool)



From what i can see with my eyes on the original footage (hope YouTube doesn't crap it all, i uploaded at maximum bitrate i could) there is a slight advantage on the Action 5 digital zoom side. At least, it looks more defined to me at 1.5, while in 2.0 some added sharpness is still visible but it starts looking artificial.

Hence, a few considerations arise:

• Is the Action 5 digital zoom already at its maximum performance, considering the 40 mp sensor or is there still some room for improvement?
• Would it be possible to stop the zoom at 1.5 so in quick situations one doesn't make distraction mistakes?
• Would it be possible to store the zoom level in the quick configurations so it can be recallable also with gloves in freezing conditions?
  

I think that working with a good sensor cropped 1.5 zoom would be useful in many situations to give some variety to the video.

It would be nice to have a 2x zoom as the Ace Pro , that seems to be much better than DJI's 2x digital zoom, which like you said , it's pretty bad and not very useful at this point .

I guess a digital zoom based on a quad Bayer sensor is heavily dependent on light conditions. Tried doing 2.0 in low light and it's awful because the sensor works in low resolution trying to gather the most possible light information. Working in well lit environments the digital zoom has a marginal advantage over Resolve zoom, but I don't know if the detail level is comparable to what can be achieved by the Ace Pro. I noticed that there is a slight better margin when shooting in dewarp, sign that there is a possible usage of the full sensor but at this point I'm asking myself why in good light conditions the full sensor couldn't already be used to mitigate the dewarp interpolation.
Or maybe I'm completely wrong, it's difficult to pixel peep without any technical information.
And it's unlikely DJI would give us info about how they manage the sensor.

Umberto Uderzo Posted at 10-26 01:48
I guess a digital zoom based on a quad Bayer sensor is heavily dependent on light conditions. Tried doing 2.0 in low light and it's awful because the sensor works in low resolution trying to gather the most possible light information. Working in well lit environments the digital zoom has a marginal advantage over Resolve zoom, but I don't know if the detail level is comparable to what can be achieved by the Ace Pro. I noticed that there is a slight better margin when shooting in dewarp, sign that there is a possible usage of the full sensor but at this point I'm asking myself why in good light conditions the full sensor couldn't already be used to mitigate the dewarp interpolation.
Or maybe I'm completely wrong, it's difficult to pixel peep without any technical information.
And it's unlikely DJI would give us info about how they manage the sensor.

Given its a sony sensor and if they do something like clearzoon algo (what sony calls it) it should be able to get the 4k pixels it needs at about 1.8 zoom max but that is less because of stab, bayer etc etc. Still i agree its time to reconsider 'never zoom' if dji works a little more on this and make makes clear what they are doing from oversampling to pixel binning and when its 1-1 pixel....

Daniel.

Personally i don't have an Ace Pro, but Jan ( fellow member of this forum ) posted some footage and the 2x zoom looks great even on a 55" TV , so it would be great if DJI could achieve that .

O.T. I was watching some reviews of ACE Pro 2 and  some issues it have are similar to the OA5 , especially regarding the colors and the sky that look dull and cartoonish ( in one of the reviews the Go Pro 13 image in daylight is great , detailed and clear , compared to both AP2 and OA5 ) , so i wonder if these devices are both using the same image processor and sensor .

Balistic Posted at 10-26 02:28
Personally i don't have an Ace Pro, but Jan ( fellow member of this forum ) posted some footage and the 2x zoom looks great even on a 55" TV , so it would be great if DJI could achieve that .

O.T. I was watching some reviews of ACE Pro 2 and  some issues it have are similar to the OA5 , especially regarding the colors and the sky that look dull and cartoonish ( in one of the reviews the Go Pro 13 image in daylight is great , detailed and clear , compared to both AP2 and OA5 ) , so i wonder if these devices are both using the same image processor and sensor .

I fear this is a "modern" trend to apply inusitate violence to the sensor output through tone mapping because "it's what market wants". And this is cross brand behaviour. Hope producers wake up and consider also "non spartphoneish" guys.

It does look a tad better on the A5 itself, but you do lose resolving power because you use less of the lens.

Montfrooij Posted at 11-5 23:49
It does look a tad better on the A5 itself, but you do lose resolving power because you use less of the lens.

Maybe it has some advantage shooting something in the sweet focus spot, otherwise there is also the defocusing to be kept into consideration.

Umberto Uderzo Posted at 11-6 00:17
Maybe it has some advantage shooting something in the sweet focus spot, otherwise there is also the defocusing to be kept into consideration.

That too for sure.
But usually if you use a smaller portion of the lens, you also lose lens resolving power (which is also resolution, but not in megapixels, but in details)

Montfrooij Posted at 11-6 00:29
That too for sure.
But usually if you use a smaller portion of the lens, you also lose lens resolving power (which is also resolution, but not in megapixels, but in details)

You're right. But i hope that the lens is designed to allow for the sensor to resolve each single pixel. Ok, maybe i'm daydreaming.

Umberto Uderzo Posted at 11-6 00:48
You're right. But i hope that the lens is designed to allow for the sensor to resolve each single pixel. Ok, maybe i'm daydreaming.

If you use less 'lens surface', you will loose resolving power. There is no doubt about that.
And judging by the footage, the resolution of the sensor is higher than the resolving power of the lens, so you can only gain so much by zooming on the A5, but it is still better than pure digital zoom.
Not the A5's fault btw, I have quite a few cameras that have this feature and all lose detail.

Montfrooij Posted at 11-6 00:51
If you use less 'lens surface', you will loose resolving power. There is no doubt about that.
And judging by the footage, the resolution of the sensor is higher than the resolving power of the lens, so you can only gain so much by zooming on the A5, but it is still better than pure digital zoom.
Not the A5's fault btw, I have quite a few cameras that have this feature and all lose detail.

At this point i'm puzzled about why DJI placed a 40mp sensor behind such a lens. Probably it's because the sensor is intended to be used in 10 mp mode and the 40 mp quad Bayer structure only to be used for ISO noise reduction.
Anyway you're right, and i'm puzzled about Ace Pro too. Is its zoom more effective than Action 5 or is it simply the result of heavy processing? Looking at some reviews i guess the latter.

Umberto Uderzo Posted at 11-6 00:55
At this point i'm puzzled about why DJI placed a 40mp sensor behind such a lens. Probably it's because the sensor is intended to be used in 10 mp mode and the 40 mp quad Bayer structure only to be used for ISO noise reduction.
Anyway you're right, and i'm puzzled about Ace Pro too. Is its zoom more effective than Action 5 or is it simply the result of heavy processing? Looking at some reviews i guess the latter.

Yeah, I have to say I never dove into the details of those numbers and what they mean.
My D750 has 24MP, but the lens and human behind the knobs are usually the limiting factor.