Hello everyone,

I'm new here and as the Thred suggests I'm having troubles with the D-cinelike profile on my new Osmo Action. It seems like all my recordings get realy noisy, especially in saturated areas and after color grading. It should be mentioned I don't have any experience whatsoever with actioncams in general or grading their clips. I'll try to add a short Clip to  ilustrate the issue.
Here is the clip: https://youtu.be/VZ7Z7KC8yJ8
It's the raw footage first and then after applying some contrast and color. After that I did add some sharpness but that makes the image pretty much useless.
As mentioned this is my first actioncam but I don't feel like 4k 30fps footage should look like this. Maybe I did something wrong when grading (but I find it pretty noisy as it was out of the camera)? Or maybe I messed up with the in-camera-settings (this was shot at 4k30fps, rs, d-cinelike, maxISO 800, wb 5500k)? Oh yes, I use a SanDisk Extreme 128GB (formatted twice already) wich I believe is the recomended SD Card and I grade in DaVinci Resolve.
Hopefully someone with a bit more experience can help me out here ;)
Thanks in advance.

Greetings from south germany

Chris

Do you remember what the camera used for ISO and shutterspeed?
It looks like a lower shutterspeed (which gives some motion blur) and a higher ISO value.
Maybe make some test shots without moving at a lower ISO (100-200 max)

looks good to me if asked on that overcast day.

Thanks for your replys. It's very much appreciated.
ISO and shutter was on automatic and I didn't check it since the camera was mounted on my chest mount.
I'll try to get more test footage during the daytime hours without much movement and an eye on the settings.
Would you recomend adjusting shutter and ISO manually when shooting in d-cinelike? Maybe lower even lower fps?

The result is the same when its sunny and not overcast. If you look at the gras on the left it's pretty servere.

Montfrooij Posted at 12-15 12:21
Do you remember what the camera used for ISO and shutterspeed?
It looks like a lower shutterspeed (which gives some motion blur) and a higher ISO value.
Maybe make some test shots without moving at a lower ISO (100-200 max)

So I tried with ISO100 and the auto setting put the shutter at around 1/1000. The Image was still noisy.
When I put the shutter at 1/60 and keep ISO100 the image gets waaay overexposed an it clips all the highlights. So no joy there as well. Today was pretty overcast again, so maybe the smal sensor doesn't like that very much?
Resolution and frame rate did not have any impact.

Here's a short clip (https://youtu.be/2r7a4kVgKgA)
You can see what I mean though, right?! If it's just my screen beeing sketchy that would be a bummer. ;)

Chris_rebon Posted at 12-16 12:12
So I tried with ISO100 and the auto setting put the shutter at around 1/1000. The Image was still noisy.
When I put the shutter at 1/60 and keep ISO100 the image gets waaay overexposed an it clips all the highlights. So no joy there as well. Today was pretty overcast again, so maybe the smal sensor doesn't like that very much?
Resolution and frame rate did not have any impact.

I don't see a clip, but I would not expect super sharp images on an overcast day. Especially not when moving fast.
Better try again with a brighter day

I don't see any noise in your first footage, but I see lack of sharpness.
Your second clip is way under-exposed.
Since you use Davinci Resolve, what value do you put your sharpness on? I have mostly used 47, but I think 44 will be a better choice (need to test it out a bit more).
Also when exporting for youtube in 4K, use atleast 100Mbit/s quality (100.000 kbit/s) in Resolve export and mp4.

Remember also that when you are shooting 4K @ 30 fps, your shutter should not go above 60 fps, always use double framerate for video. 4K @ 60 fps will give you 120 fps shutterspeed.
And use ND-filter to get the right exposure. I have my camera set at max ISO 400 in summertime for less noise, and switch between ND16-32 (sunny) and ND4-8 (overcast).

If you look at this video, I used ND16 on an overcast day and recorded in Cinelike (yes a little too less colors in this video):


I have created LUT's for my OA with ND-filters/CPL-filters and without filters.

johansenfoto Posted at 12-17 01:22
I don't see any noise in your first footage, but I see lack of sharpness.
Your second clip is way under-exposed.
Since you use Davinci Resolve, what value do you put your sharpness on? I have mostly used 47, but I think 44 will be a better choice (need to test it out a bit more).

Because of the noise I stayed away from adding sharpness. It made the image just grainy but I fount the clips I uploaded to be not too noisy on youtube. Might be an issue with my computer or the HDMI cable.
I put sharpness on 48 most of the time and add some midtone detail to make up for it.

For the shutterspeeds you recomend it is definitely required to use ND filters I guess?! Do you need to change those often when recording in different conditions during the same day i.e. shadows, direct sun. How versatile are you when using those? I record out on my bike most of the time and don't want to worry about filters an setting shutter speeds all of the time....

Chris_rebon Posted at 12-17 05:05
Because of the noise I stayed away from adding sharpness. It made the image just grainy but I fount the clips I uploaded to be not too noisy on youtube. Might be an issue with my computer or the HDMI cable.
I put sharpness on 48 most of the time and add some midtone detail to make up for it.

I follow the same path everytime I'm going out recording, if sunny I use mostly ND16 during summer and let Auto ISO handle the darker areas. But there is some cases where I should used ND32 when the sun and sky is really bright. Check before you record that the sky look descent, then you are usually good. Because with video it's easier to bring up shadow in post, than bring down highlights.

I forgot to mention that I always record in 60 fps (120 fps shutter) so if you always record in 30 fps, then you need double ND-strength (ND32 instead of ND16) to give the same exposure.

johansenfoto Posted at 12-17 05:36
I follow the same path everytime I'm going out recording, if sunny I use mostly ND16 during summer and let Auto ISO handle the darker areas. But there is some cases where I should used ND32 when the sun and sky is really bright. Check before you record that the sky look descent, then you are usually good. Because with video it's easier to bring up shadow in post, than bring down highlights.

I forgot to mention that I always record in 60 fps (120 fps shutter) so if you always record in 30 fps, then you need double ND-strength (ND32 instead of ND16) to give the same exposure.

Hello,

I am new in the world of action cams and try to make my way to using it properly. Luckily I was able to receive some ND filters from a friend who doesn't need them right now.
Coming back to the comment: Surely I know the shutter rule and hence 1/120 is recommended when recording 60fps. However, in another thread it was mentioned that the EIS capabilities decrease as the shutter speed slows down...a boundary of 1/240 was mentioned there (Thread).

Is it right then that the use of ND filters goes on expense of Rocksteady stabilization due to the slower shutter speed? Made a quick test outside (running) and I have to admit that the 1/120 shutter ND-video seemed way more shaking than the one with standard lense and auto shutter speed.

ActionFlyer Posted at 12-20 03:12
Hello,

I am new in the world of action cams and try to make my way to using it properly. Luckily I was able to receive some ND filters from a friend who doesn't need them right now.

Not sure what to say about this, because when I first time read about ND-filters are making rocksteady worse, my experience didn't show any difference that I could notice when doing test myself.

It could be that rocksteady works better with slower shutterspeed, or it could be the effect of just blur that make it look more steady.

I also remember when OA came out and saw a video where that person complained about ND-filters made rocksteady not work properly (it did show in the video), but I didn't have any problem at the same settings back then.
A little notice, I use original ND-filters sold by DJI shop here in Norway, but not sure if 3rd party filters will or will not affect rocksteady.

Rocksteady relies on high shutter speeds to function, so you cannot use the 180 degree rule (shutter speed at double the frame rate). The loss of stabilization brought by slow shutter speeds might be acceptable, or even unnoticeable depending on how much and what type of correction is needed; but in general  cinematic shutter speeds and software image stabilizations like Rocksteady are mutually exclusive.

Sounds like higher frame rates would also be beneficial for getting a more stable image?!
I heard the RockSteady algorithm does prefer 30 and 60 fps anyway....

I agree with the point made by johanasfoto, it is the lack of sharpness / resolution you are seeing in the grass.

I found the same thing with the action.      

Rocksteady and dewarp also eat up some resolution. If you switch these options off you will find you gain some resolution.

Grass seems to highlight the lack of resolution , I guess it’s the fine detail of the blades you notice missing.

I’ve tested the same foliage scene against an osmo pocket and gopro8,  the action has a noticeable lack of sharpness. ( you can tweak sharpness in post, but the info just isn’t there ) . However overall picture quality on the action is pleasing, it just misses detail.

As mentioned above overcast days never look great, but my particular comparisons were done in great summer light.

CemAygun Posted at 12-21 21:37
Rocksteady relies on high shutter speeds to function, so you cannot use the 180 degree rule (shutter speed at double the frame rate). The loss of stabilization brought by slow shutter speeds might be acceptable, or even unnoticeable depending on how much and what type of correction is needed; but in general  cinematic shutter speeds and software image stabilizations like Rocksteady are mutually exclusive.

Thank you for this clarification. I am still quite new and want to dive deeper in the possibilities of settings I have

ActionFlyer Posted at 12-30 06:28
Thank you for this clarification. I am still quite new and want to dive deeper in the possibilities of settings I have

You're welcome

MKosmo Posted at 12-23 05:48
I agree with the point made by johanasfoto, it is the lack of sharpness / resolution you are seeing in the grass.

I found the same thing with the action.      

The total resolution loss of Rocksteady + Dewarp is insane. You are barely shooting 2.7K with those settings (it is then upscaled to 4K if you choose 4K).

Apart from a select few, most action cameras suffer the same fate as software image stabilization crops too much into sensors that need almost a full, untouched readout to record 4K.