Oh wow, this is probably one of the best comparisons I've ever seen of the two of them!! Weird that the action 4 seems to be less robust in the horizon balancing, but it wins in every other area so....
Yes it does. But only in low light. Any kind of digital stabilization is effected by a slow shutter speed in low light. It has to do with the direction of motion blur in the image. The motion blur direction can feel disjointed from the movement of the stabilization as it can often oppose each other causing unnatural motions in the footage. It's best to turn it off and stabilize the camera mechanically. That is if you're trying to get the best image possible.
Very interesting review, and very clear comparison. Just received my OA4 and I start comparing it side by side with my 0A3 and OA2. I find the dynamic range much better and I have less burned skies when contrats is too high between ground & sky. Very happy with it. Only frustration is not being able to adapt my set of nd filters when needed. Great job from you, thanks
My experience shows that ND filters are good when camera is almost static, because if we film with a moving camera at 25fps shutter 50, it ruins the stabilisation or we need to switch the stab off, which is the same result. I use my OA a lot on board of my motorbike in a jittery environment, so I need Rock steady; the main problem is that at the 1000 to 2000 usual shutter, I see every fly and insect passing by very clearly. So I’m still trying to find the good fix shutter (400-800 range) with free ISO which will assure good stabilisation but enough motion blur not to see the insects passing ! Not an easy process (especially because I don’t always find volunteer flies when I experiment on the road..;it seems that they show up only to spoil my footage when I’m on my trips 😁) . For me, ND 2 (in fact a PL filter) or ND4 are a good daily base to keep a good behavior of Rocksteady and lower a bit the shutter speed (to a given ISO)
@@philsgarage64 yeah that's tough since motion blur and digital stabilization are mortal enemies. I suspect you just need to find a happy medium that you are willing to live with. Unless you go with a bulky gimbal and disable rocksteady.
@@freelyroaming you’re right : a gimbal could be the solution, but the problem is that usually the vertical axis is not stabilized. I have a Pocket 2 which works well on board as long as it’s damped on the vertical,axis (in fact, strapped on my jacket) other problem of gimbal is that it usually has a lag when the motorcycle is turning. The gimbal continues to look straight and not in the turn, so not perfect either. I ran some tests this afternoon on a good road (but not a perfect one) : shutter has to be at least at 800 to keep a good stabilisation. Every shutter below will increase the risk to have jitter. Having « motion blur » with a stabilized image on a public actioncam on a vehicle on standard road is just not possible. There’s a choice to do : stable OR blur. As you perfectly said : they are mortal enemies !(at least with electronic stabilisation process)
I think they really need to work on the horizon lock though. It works pretty well but a gopro with the max lens mod can go in a full circle without moving. Definitely seems like something they could improve for a software update and not have to wait for the Action 5
@@jcody_parker if you switch to 2.7k you can also enable total 360 degree horizon lock with both of these cameras. It's less resolution but you don't need to buy any accessories.
Both versions are shown in this video. Graded and ungraded. And the graded ones didn't really get a grade. Just a rec.709 lut directly from DJIs website. But sure, I get it. Can't please everyone.
@@freelyroaming yeah you cant lol one thing i am interested in the profiles as my dji drone is the same as action 3 i have action 1 but deciding to buy the dji action 4 or 3 comes down to the profile flats
@@FrequencyOfThought … I suspect the reason why is because many of the early OA3 cameras suffered from focusing issues due to a factory mechanical calibration issue which DJI never formally acknowledged. Like many others on YT and other Social Media platforms, I had focus issues and received 3 replacement cameras from DJI. In the end I got a full refund. So I’m very interested to see if the OA4 is better in this regard.
@@ecky_thump I'll let you know, I ended up getting one. My first real action camera besides this crappy AKASO I have which was just for recording things in the shop that I can't see. engine bay while driving etc. I bought a ZV-1II which is sony's new camera so I'll do some cool shots of night time and such to see the difference. Only thing really different on the two cameras is optical zoom.
The oa3 footage was ruined by the horrible grade, I dont know what you were doing but you crushed the blacks waay to far. It also affected the low light test to make it look pure black for the oa3.
@@SwordTomatoThere is no grade. They are exactly as it was straight out of camera. That's the whole point of the comparison. I find it rude and unnecessary for you to assume there is one and chose to call it horrible without any knowledge. For someone with no videos shot on an actual camera on your channel I can only assume you know nothing about color grading besides what you've seen from videos on RUclips. You can try to be better.
From what I see at 1:24 min the Osmo A3 has less frames per second with 8-bit mode than with 10. There are few differences between the two devices, but despite having more contrast and less light capture in a nighttime environment, I ended up liking more the DJI Osmo Action ... OA3. With the exception of the sensor, which in the OA4 is better for nighttime scenarios and color is slightly more realistic and the lower frame rate in OA3, in 8-bit mode what made me prefer the OA3 was its enormous moving image focus stability in 10-bit mode, which is much better than the OA4. Thank you both for the physical effort spent on this comparison. I hope you don't make the mistake on showing us comparisons of sports shoes for marathons. :)
WARNING: IMPORTANT ! ! ! I saw that you have a brown line on one of your fingernail. Maybe in this video, maybe in another one. This is frequently the sign of a melanoma (agressive skin cancer) under the fingernail. Have it checked by an MD ASAP. And thank for the OA4 v 3 review !
Thats up to you. My opinion is you are not watching close enough. If you care, watch it again. Look at the dark shadows and the overblown highlight areas. If none of that matters to you, then you probably don't need to upgrade.
@@freelyroaming It's maybe a 10% better image in the shadows but certainly nothing to write home about. I think DJI is too nice and updated the 3 with firmware in the last few months and that made them too close together.. But good for them keeping their old products viable. Apple should take their sensors and put them in a super durable box and start selling competing products. I think action cameras have a long way to go period.
@@FrequencyOfThought smartphone cameras like iphones do not have better sensors. It is the software in them that make the difference. iPhones have a lot more processing power which is why they can have better perceived image quality because of computational photography. Just by putting a smartphone sensor in an action cam will not look better than this. Improving the processing power and software in action cameras is what will help. But that also will increase the cost of them which I am certain most people will not be willing to pay. Honestly, I don't think smartphone videos look much better than this. Especially for how much they charge. In fact, this camera has a sensor that is double the sensor size on the latest Phone. 10% is actually a huge improvement for a single generation.
@@freelyroaming That's why these cameras look like crap is because they are simulating frames and compressing it and the DJI is the most noticeable blocky chunky artifacts that look horrible. But it is however above and beyond in other areas.
@@FrequencyOfThought You are looking at the RUclips compression. That is unavoidable when you upload to the platform and if you know what these camera videos look like you wouldn't assume that this is from the original footage. I don't know what you mean when you say simulating frames.
Best practical comparison available on entire RUclips .... Love from India ❤❤
Thank you!
Oh wow, this is probably one of the best comparisons I've ever seen of the two of them!! Weird that the action 4 seems to be less robust in the horizon balancing, but it wins in every other area so....
Yes I was surprised by that as well. Maybe the new sensor created some limitations for that ability.
Do you bother with shutter speed when you use rock steady?
I believe it affects your footage, especially in low light, but I never fully tested.
Yes it does. But only in low light. Any kind of digital stabilization is effected by a slow shutter speed in low light. It has to do with the direction of motion blur in the image. The motion blur direction can feel disjointed from the movement of the stabilization as it can often oppose each other causing unnatural motions in the footage. It's best to turn it off and stabilize the camera mechanically. That is if you're trying to get the best image possible.
Hey guys, any chance get to see more low light test & comparison ?
Yes I plan on making a whole video on varying low light conditions with the OA4. It is the biggest benefit of this new model.
@@freelyroaming look forward 💪
Very interesting review, and very clear comparison. Just received my OA4 and I start comparing it side by side with my 0A3 and OA2. I find the dynamic range much better and I have less burned skies when contrats is too high between ground & sky. Very happy with it. Only frustration is not being able to adapt my set of nd filters when needed. Great job from you, thanks
I agree. It's also too soon for a lot of aftermarket filters to be available
My experience shows that ND filters are good when camera is almost static, because if we film with a moving camera at 25fps shutter 50, it ruins the stabilisation or we need to switch the stab off, which is the same result. I use my OA a lot on board of my motorbike in a jittery environment, so I need Rock steady; the main problem is that at the 1000 to 2000 usual shutter, I see every fly and insect passing by very clearly. So I’m still trying to find the good fix shutter (400-800 range) with free ISO which will assure good stabilisation but enough motion blur not to see the insects passing ! Not an easy process (especially because I don’t always find volunteer flies when I experiment on the road..;it seems that they show up only to spoil my footage when I’m on my trips 😁) . For me, ND 2 (in fact a PL filter) or ND4 are a good daily base to keep a good behavior of Rocksteady and lower a bit the shutter speed (to a given ISO)
@@philsgarage64 yeah that's tough since motion blur and digital stabilization are mortal enemies. I suspect you just need to find a happy medium that you are willing to live with. Unless you go with a bulky gimbal and disable rocksteady.
@@freelyroaming you’re right : a gimbal could be the solution, but the problem is that usually the vertical axis is not stabilized. I have a Pocket 2 which works well on board as long as it’s damped on the vertical,axis (in fact, strapped on my jacket) other problem of gimbal is that it usually has a lag when the motorcycle is turning. The gimbal continues to look straight and not in the turn, so not perfect either. I ran some tests this afternoon on a good road (but not a perfect one) : shutter has to be at least at 800 to keep a good stabilisation. Every shutter below will increase the risk to have jitter. Having « motion blur » with a stabilized image on a public actioncam on a vehicle on standard road is just not possible. There’s a choice to do : stable OR blur. As you perfectly said : they are mortal enemies !(at least with electronic stabilisation process)
What about an overheating test?
The rocksteady+ is damn near gimbal quality
Technology is crazy
I think they really need to work on the horizon lock though. It works pretty well but a gopro with the max lens mod can go in a full circle without moving.
Definitely seems like something they could improve for a software update and not have to wait for the Action 5
Yes seriously amazing stabilization.
@@jcody_parker if you switch to 2.7k you can also enable total 360 degree horizon lock with both of these cameras. It's less resolution but you don't need to buy any accessories.
@@freelyroaming ah cool, good to know thanks
would you prefer oa3 adventure combo or oa4 standart combo package? Near price ..
OA4 for sure
thanks for answer it is gonna be my first action cam and what do you think about go pro 12 vs oa4?
@@freelyroaming
@@MaceraMuhendisi GoPro 12 is honestly a disappointing update. OA4 is better for what I use it for.
I understand what you mean. I will use for cycling, hiking and running...@@freelyroaming
But oa3 price is going down now in my country, indonesia. Difficult choice.hahah
what the difference, except the money?
Wish you hadn’t graded the dlog and action cine lite
Both versions are shown in this video. Graded and ungraded. And the graded ones didn't really get a grade. Just a rec.709 lut directly from DJIs website. But sure, I get it. Can't please everyone.
@@freelyroaming yeah you cant lol one thing i am interested in the profiles as my dji drone is the same as action 3 i have action 1 but deciding to buy the dji action 4 or 3 comes down to the profile flats
Did plz tell me what is the manufacturing date your action 3 camera
Not sure how to find that info but I just bought it a month ago. Although it was bought as a refurbished unit.
@@freelyroaming check your action 3 box
@@aashiktrading6743 Why do you ask this
@@FrequencyOfThought … I suspect the reason why is because many of the early OA3 cameras suffered from focusing issues due to a factory mechanical calibration issue which DJI never formally acknowledged. Like many others on YT and other Social Media platforms, I had focus issues and received 3 replacement cameras from DJI. In the end I got a full refund. So I’m very interested to see if the OA4 is better in this regard.
@@ecky_thump I'll let you know, I ended up getting one. My first real action camera besides this crappy AKASO I have which was just for recording things in the shop that I can't see. engine bay while driving etc. I bought a ZV-1II which is sony's new camera so I'll do some cool shots of night time and such to see the difference. Only thing really different on the two cameras is optical zoom.
The oa3 footage was ruined by the horrible grade, I dont know what you were doing but you crushed the blacks waay to far. It also affected the low light test to make it look pure black for the oa3.
@@SwordTomatoThere is no grade. They are exactly as it was straight out of camera. That's the whole point of the comparison. I find it rude and unnecessary for you to assume there is one and chose to call it horrible without any knowledge. For someone with no videos shot on an actual camera on your channel I can only assume you know nothing about color grading besides what you've seen from videos on RUclips. You can try to be better.
From what I see at 1:24 min the Osmo A3 has less frames per second with 8-bit mode than with 10.
There are few differences between the two devices, but despite having more contrast and less light capture in a nighttime environment, I ended up liking more the DJI Osmo Action ... OA3.
With the exception of the sensor, which in the OA4 is better for nighttime scenarios and color is slightly more realistic and the lower frame rate in OA3, in 8-bit mode what made me prefer the OA3 was its enormous moving image focus stability in 10-bit mode, which is much better than the OA4.
Thank you both for the physical effort spent on this comparison.
I hope you don't make the mistake on showing us comparisons of sports shoes for marathons.
:)
WARNING: IMPORTANT ! ! ! I saw that you have a brown line on one of your fingernail. Maybe in this video, maybe in another one. This is frequently the sign of a melanoma (agressive skin cancer) under the fingernail. Have it checked by an MD ASAP. And thank for the OA4 v 3 review !
Thanks. I've had it checked 2 years ago.
Am I the only one who barely notices a difference. I mean is it really enough to upgrade?
Thats up to you. My opinion is you are not watching close enough. If you care, watch it again. Look at the dark shadows and the overblown highlight areas. If none of that matters to you, then you probably don't need to upgrade.
@@freelyroaming It's maybe a 10% better image in the shadows but certainly nothing to write home about. I think DJI is too nice and updated the 3 with firmware in the last few months and that made them too close together.. But good for them keeping their old products viable. Apple should take their sensors and put them in a super durable box and start selling competing products. I think action cameras have a long way to go period.
@@FrequencyOfThought smartphone cameras like iphones do not have better sensors. It is the software in them that make the difference. iPhones have a lot more processing power which is why they can have better perceived image quality because of computational photography. Just by putting a smartphone sensor in an action cam will not look better than this. Improving the processing power and software in action cameras is what will help. But that also will increase the cost of them which I am certain most people will not be willing to pay. Honestly, I don't think smartphone videos look much better than this. Especially for how much they charge. In fact, this camera has a sensor that is double the sensor size on the latest Phone. 10% is actually a huge improvement for a single generation.
@@freelyroaming That's why these cameras look like crap is because they are simulating frames and compressing it and the DJI is the most noticeable blocky chunky artifacts that look horrible. But it is however above and beyond in other areas.
@@FrequencyOfThought You are looking at the RUclips compression. That is unavoidable when you upload to the platform and if you know what these camera videos look like you wouldn't assume that this is from the original footage. I don't know what you mean when you say simulating frames.