Please select Into the mobile phone version | Continue to access the computer ver.
Noise in video?
396 8 2020-7-23
Uploading and Loding Picture ...(0/1)
o(^-^)o
SLiWooDy
First Officer
Flight distance : 126234 ft
  • >>>
United Kingdom
Offline

Noise in your videos with MA2?
This might be helpful:

2020-7-23
Use props
DJI Stephen
Super Moderator

Offline

Hello there SLiWooDy. Thank you for reaching out and for posting this helpful video. Great find and thank you for your support.
2020-7-24
Use props
DAFlys
Captain
Flight distance : 1243005 ft
United Kingdom
Offline

Nice tips.  
2020-7-24
Use props
A J
Captain
Flight distance : 13838848 ft
  • >>>
United Kingdom
Offline

Thanks for sharing
2020-7-24
Use props
M4cr0s
Second Officer
Flight distance : 1404308 ft
Norway
Offline

As I understand it "normal" is a picture profile, a set of parameters the camera applies to the raw data captured, color, tonality, contrast, sharpening, nr and so forth. The disadvantage with this, theoretically, is that as contrast is increased and colors boosted..and data is thrown out or "drawn out" to cover a wider range than there is real information to cover. Hence, there is less room tweak the footage in post before it falls apart (posterization). D-cinelike is just the same data with less of this throw on top, which should preserve more data, or make it more malleable.  So equal or better than "normal" should be possible with careful post processing. However, might be tricky to replicate manually.

You guys that have worked a fair bit with drone footage, what do you experience?
2020-7-24
Use props
Ice_2k
Captain
Flight distance : 1132575 ft
Romania
Offline

this goes against pretty much all of the advice I saw online from seriously experienced videographers so I'm gonna pass on it.... I will do my own tests however as it kind of peaked my interest.I did some tests about a month ago but I was focused on HDR vs D-Cinelike and if it's really worth using HDR as it comes with quite some downsides.
2020-7-24
Use props
hallmark007
Captain
Flight distance : 6059744 ft
  • >>>
Ireland
Online

M4cr0s Posted at 7-24 02:16
As I understand it "normal" is a picture profile, a set of parameters the camera applies to the raw data captured, color, tonality, contrast, sharpening, nr and so forth. The disadvantage with this, theoretically, is that as contrast is increased and colors boosted, data is thrown out or "drawn out" to cover a wider range than there is real information to cover. Hence, there is less room tweak the footage in post before it falls apart (posterization). D-cinelike is just the same data with less of this throw on top, which should preserve more data, or make it more malleable.  So equal or better than "normal" should be possible with careful post processing. However, might be tricky to replicate manually.

You guys that have worked a fair bit with drone footage, what do you experience?

I think you’re right, this is a sharpness test nothing else and OP has looked at nothing except sharpness.
2020-7-24
Use props
SLiWooDy
First Officer
Flight distance : 126234 ft
  • >>>
United Kingdom
Offline

Ice_2k Posted at 7-24 07:39
this goes against pretty much all of the advice I saw online from seriously experienced videographers so I'm gonna pass on it.... I will do my own tests however as it kind of peaked my interest.I did some tests about a month ago but I was focused on HDR vs D-Cinelike and if it's really worth using HDR as it comes with quite some downsides.

I encourage everyone to run the tests for themselves and publish their findings.
(Obviously without any noise reduction software)
I think they may surprise a lot of people, especially in low/mid light.
2020-8-4
Use props
Mirek L
First Officer
Flight distance : 609724 ft
Canada
Offline

This is all very interesting.
However, OP, as he said himself, applied post-processing to D-Cinelike profile to make it look tonally similar to Normal video. Depending on editor he used, his skills and the output parameters for the resulting video, this may have contributed to much worse quality. Also, I find that YouTube encoding may really have quite dramatic effect on messing up your video.

In addition, he said that H.265 has similar quality but smaller file size. Did he really check that?
I did extensive tests on MA 2 H.264 versus H.265 footage and did not see almost any difference in file size (H.265 were marginally smaller) when I had videos of the same subjects, filmed in a same way and of the same length. This led me to conclude that DJI is simply packing more data into H.265 files without affecting file size. Having said that, I did not see much of a quality difference between the two either. I expected it, but could not see it.

Couple of months ago I did D-Cinelike versus Normal, versus HDR tests, My conclusions were as follows:
  • Do not use Normal unless you are not able to do post-processing. Normal behaves just fine in most situations but may fall apart when the lighting gest challenging. This fault is much more visible than scrutinizing small amount of noise and comparing it.
  • Use D-Cinelike instead of HDR because it gives you a bit more flexibility in post-processing. I could get as good or better results from D-Cinelike than HDR.


I would be very interested to see some other examples from other forum members proving or disproving what was said on the video.

I would just add, if you are doing your own comparisons, compare original files without any processing applied (straight from the card, no editor, no YouTube conversion). Accept that D-Cinelike will be flat and less contrasty but look for quality and details and not color and contrast.

Mirek
2020-8-4
Use props
Advanced
You need to log in before you can reply Login | Register now

Credit Rules