Keep it. Fly it. Enjoy it!! What else is there that is comparable? Unless you want to buy a Phantom 4 Pro, then keep it. Overall, it's a great drone!
Mavic Pro-2 vs Zoom and P4P - Line Skipping + Sensor Heat?
- Thread starter Cliff_622
- Start date
Keep it. Fly it. Enjoy it!! What else is there that is comparable? Unless you want to buy a Phantom 4 Pro, then keep it. Overall, it's a great drone!
Maybe, but Ambarella video processors like the H1, H2, H22 and H3 are dedicated processors that are fully designed to handle raw sensor readout, deBayering, noise reduction, color processing, h,264 and h.265 compression .without affecting other non-video functions of the drone.
CT
Yes, the IMX183 has two native recording modes for full FOV, this is 5.5K (10bit max 25fps) and 2.7K (10bit max 60fps). 4K looks like 2.7K as they're indeed upscaling 2.7K FOV 30fps 10bit to 4K30 FOV 30fps 10bit.
My theory is different, to me it is obvious the current M2P FW is far from being finalised, and DJI Marketing wanted to launch in August 23th the M2P highlighting 10bit in all 4K modes/framerates. Let's not forget the M2P shipped with the 00.60 FW that didn't even have 10bit mode in 4K FOV.
So clearly they had to add the 10bit in 4K FOV -among other things- as a last minute addition (FW 01.00 launched on August 28th) they figured out the easiest and fastest way to provide a 4K 10bit FOV across 24, 35 and 30fps was to simply upscale 2.7K 10bit 24/25/30 to this. I bet some beers the logic for all 4K FOV modes is the same (2.7K upscale at given fps) and then simply encoded in target codec user selected with H264 set to 8bit and H265/HEVC set to 10bit.
So I believe there is room for improvement and basically do what the P4P does in 4K FOV 8bit at 24/25/30fps which is 5.5K downsample to 4K and then in 5.5K>4K FOV 10bit only at 24 and 25 fps. 4K FOV 30fps 10bit is simply not doable by the sensor most likely due to bandwidth limitations with the SOC.
I don't think there are any thermal limitations in the 1" sensor as precisely the M2P camera body is constructed in aluminum/magnesium allow (good heat transmission) moreover the rear part of the camera is shaped as a heat sink (even better disipation) and this is a drone so plently of airflow in the camera heat sink. This as opposed to the P4P which is inside a plastic body with no heat sink. If anything I would say the M2P sensor has to be cooler than the P4P sensor.
The Sony IMX183CQJ, if used, CAN deliver 12bit in full pixel read out in 5.5K up to 16:9 up to 26p, so enough for 24/25p up to 10bit since you should (!) use more bitdepth when gamma processing (HLG/D-LOG M) is used before compression/saving.
IMX183CLK-J/CQJ-J | Sony Semiconductor Solutions
DJI writes in the manual on p. 38 and in the FAQs that the 5.5K sensor (5.7K???) is downsampled (!!) to 4K „resolution“ (!!)
First there is no „downsampling“ as with the P4pro, there is only binning or skipping since even near neighbour downscale would result in better resolution.
Next we get no advertised 4K „resolution“. We get a picture with 4K „size“ but only with about 2.7K „resolution“ and that‘s a big difference!
The P4pro shows nearly 4K (luma-) resolution but not the M2P, not in HQ neither in FullFOV.
What is DJI doing here, promising us downsampling and 4K resolution and delivering much less!?
Maybe some here are right and DJI is working on this.
Maybe we should „ask“, better demand the DJI support „when“ we get 4K „resolution“ as described also in FullFOV instead of 4K picturesize with about 2.7K resolution.
The FW release notes show actually nothing about current limitations in this regard as far as I know.
No demand, no deliver.
IMX183CLK-J/CQJ-J | Sony Semiconductor Solutions
DJI writes in the manual on p. 38 and in the FAQs that the 5.5K sensor (5.7K???) is downsampled (!!) to 4K „resolution“ (!!)
First there is no „downsampling“ as with the P4pro, there is only binning or skipping since even near neighbour downscale would result in better resolution.
Next we get no advertised 4K „resolution“. We get a picture with 4K „size“ but only with about 2.7K „resolution“ and that‘s a big difference!
The P4pro shows nearly 4K (luma-) resolution but not the M2P, not in HQ neither in FullFOV.
What is DJI doing here, promising us downsampling and 4K resolution and delivering much less!?
Maybe some here are right and DJI is working on this.
Maybe we should „ask“, better demand the DJI support „when“ we get 4K „resolution“ as described also in FullFOV instead of 4K picturesize with about 2.7K resolution.
The FW release notes show actually nothing about current limitations in this regard as far as I know.
No demand, no deliver.
The reason I think it's not just 2.7K upscaled is that there does appear to be a resolution difference between "Full FOV 4K" and 2.7K in Cliff's video. It's very small, but it's there - the Full FOV 4K has slightly more resolution. That would rule out a 2.7K upscale: you can't create more detail via upscaling.
Has anyone even confirmed that the sensor being used is in the IMX183 family? Everyone keeps saying it's this sensor, but as far as I can tell it's just an educated guess.
Even if it is the IMX183CQJ-J, the publicly available specs for that sensor come from a small product marketing sheet. I work in the semiconductor industry and it would not surprise me at all if there are other variants of the sensor that aren't publicly listed, or other unlisted readout modes on the existing SKUs - stuff like that is very common.
My point is, just because there is not a line-skipping or binning readout mode listed on that page, does not mean it does not exist. It may not be listed or DJI may be using a different variant of the sensor entirely.
The point is that DJI makes us believe that the resulting FullFOV image is „down sampled“ from the 5.5K sensor and that the resulting image has 4K „resolution“ (!) that it has not!
It has 4K image size but with lower resolution as we learned!
to be found here:
Mavic 2 - Specifications, FAQs, Videos, Tutorials, Manuals - DJI
as also in the manual on p.38.
It has 4K image size but with lower resolution as we learned!
to be found here:
Mavic 2 - Specifications, FAQs, Videos, Tutorials, Manuals - DJI
as also in the manual on p.38.
Completely agree with Chunky. Really really need to press for some kind of formal response fron DJI. They should at least entertain these concerns and let us know if these findings are temporary, or if we are stuck with misleading or even fake marketing.
I gave them a week to respond to me privately. They watched it 70+ times and declined to comment to me. I love these guys too,...I swear I do!!, but I dont think they will. They never wanted this to come out like this and I dont think they really want to deal with it. They have a "move along,....move along,...nothing to see here" mentality. They will CERTAINLY never tell us whey they killed the original "See the Bigger Picture" even in such an emergency manner.
Roamer105
Well-Known Member
You are correct in that they control the company. But just for a moment: why would you go and buy something that is neither a threat nor competition AND no value? Because that would be the case if they destroy the only thing - the reputation of Hasselblad. Of course nobody expects a $1400 flying camera to be in the same class as the 10K Hasselblad cameras. However nobody should for any reason expect that this camera should be worse than any other cheap camera.
I am quite sure that to start with the optics are far superior from what we have seen so far. Anything else I expect them to fix by firmware in the near future. Just remember the early P4P or Mavic video issues - we are still on the ground floor with this but have a lot more “material” to work with for improvements.
I for one am very optimistic and positive about this aside from being quit happy with the current (already an improvement over my MP) image and video quality.
I'm very cynical, and believe the issues will be fixed and released as a Mavic 2 Pro Platinum edition. They will force us to upgrade to fix these deficiencies....
Roamer105
Well-Known Member
This is also what I fear but that would not be right and loyal since they described and promised the quality now for the M2P, not for a future product!
I think there must be put more pressure on DJI, maybe more YT videos.
Ask DJI support when we will see the improvement and make them clear that they are wrong with their behavior.
I already did and I was told „maybe“ something is coming, we are still on 1.00.
„I“ want „we will solve this soon“, everything else is hardly acceptable.
I also saved the original manual and made screenshots of the FAQs. I had a case with Panasonic where they removed described functionality from the LX100 and simply changed the description on their website later, that seems to be common style.
If a lawyer really wanted to play, they could just sue DJI if they changed the specs, on account of influencing purchase decision on misinformation.
DJI actually tells us that we get 4K „resolution“ - even in FullFOV.
Maybe many of us know that real 4K „resolution“ is not achieveable with very most 4K cameras.
This is only close possible with 4K 3 chip cameras or oversampling cameras like Samsung NX1, Sony a6300/6500 or a7M3 (all „6K“ to 4K) and for example RED 6/8K/Sony Venice cameras after downscale due to debayering effects, there even only for luma signals, not for chroma and at least not for red/blue detail because of the lower pixel count and color subsampling.
The Phantom 4 pro also does oversampling from about 5.5K and was well known for it‘s power if detail.
I think nobody would claim if we would get nearly the same resolving power of the P4pro also with the M2P and actually nobody really understands why we currently only get about 2.7K where it shouldn‘t be a limitation of computing power or something similar.
P4pro is about 2 years old now and the M2P is nearly the same price range.
DJI markets them both as consumer drones, so no „pro premium“ vs. „pro standard“ visible for the customer.
Maybe many of us know that real 4K „resolution“ is not achieveable with very most 4K cameras.
This is only close possible with 4K 3 chip cameras or oversampling cameras like Samsung NX1, Sony a6300/6500 or a7M3 (all „6K“ to 4K) and for example RED 6/8K/Sony Venice cameras after downscale due to debayering effects, there even only for luma signals, not for chroma and at least not for red/blue detail because of the lower pixel count and color subsampling.
The Phantom 4 pro also does oversampling from about 5.5K and was well known for it‘s power if detail.
I think nobody would claim if we would get nearly the same resolving power of the P4pro also with the M2P and actually nobody really understands why we currently only get about 2.7K where it shouldn‘t be a limitation of computing power or something similar.
P4pro is about 2 years old now and the M2P is nearly the same price range.
DJI markets them both as consumer drones, so no „pro premium“ vs. „pro standard“ visible for the customer.
Bazzam
Well-Known Member
- Joined
- Apr 12, 2017
- Messages
- 132
- Reactions
- 123
- Age
- 49
Or follow the advice of unknown,subjective,unverified random people on Mavic pilots forum [emoji39] I'm off to submit my footage to NatGeo [emoji39]
P4P V1 range with clear LOS set to Auto 5.8 Ghz will easily deliver up to 5 miles. Don't sell it short!
Not my writing, but a response to a thread. This comment makes more sense than this whole thread. I will just leave it here.
From someone who works in TV for the past 40 years... What this guy is claiming is totally bogus. 1080 and 4K are exactly the same resolution. 4K is just 4x 1080 pictures joined together to give you more real estate. The only real benefit of 4K is in editing you have a lot more latitude to digitally crop a piece of the picture and still have 1080 resolution. That's it.
Where this guy is incorrect is that the resolution changes depending on what mode you're in. In fact is, in either mode you still have either 4K (3840x2160) or HD (1920x1080) resolution being recorded. Because they're all fixed focal length lenses, what changes is the processing that zooms them in to change the field of view. It does not change the recorded resolution but like any digital zooming it will slightly degrade the actual picture quality.
Resolution in television was originally measured in scan lines back in the days of SD but now it's measured in pixel lines in digital. So no matter what mode you're in, you're still getting 2160 (4k) or 1080 (HD) lines of resolution.
When he talks about "pixel skipping," what he's really referring to is compression, not changes in resolution. Video is compressed to save data in the recording. More compression = poorer picture.
As for his claim of "line skipping," I'd like to see some proof of his claim that they're skipping scan lines on the sensor. If this were true, it would mean that you're not getting a true HD image out of the sensor and that makes absolutely no sense. If you eliminate half the lines of resolution you're reducing the image down to slightly higher than SD resolution. That would be extremely noticeable and most people would find it unacceptable in any conditions.
As for his tests, If this guy was doing a legitimate test of the images he would have had a serious set of SMPTE camera charts to show both resolution and coloremetry. The fact that he printed out a chart and it's not correctly scaled tells me he has no clue what he's talking about. Firstly what he used was not a legitimate SMPTE resolution chart. The link below shows what legitimate SMPTE (Society of Motion Picture and Television Engineers) resolution charts look like. They start at about $2 grand each. Clearly not something you print off of the internet. Besides which, there is no home printer capable of printing the kind of resolution required for a real resolution chart to actually work. http://dsclabs.com/test-charts/resolution-focus-test-charts/
The last thing that blew all credibility for me of his claims is that he was using some kind of pot with a piece of the field turf from the super bowl in it as a reference for focus. What a joke. This is no legitimate test. Again there are proper, calibrated focusing charts available to do this properly.
So in short, I think you can distill the changes in the images down to two things... video compression and processing and likely quality of the glass in front of the sensor. Let's face it, folks, you're not getting the same quality glass in a $1200 drone as you would in a $150k UHD television camera lens.
FYI, this is what a $150k television camera lens looks like... and there's a reason why they cost $150k: https://www.usa.canon.com/.../4k-uhd.../uhd-digisuper-122
The only one thing he said about the drones that made any sense is that they all look pretty good.
From someone who works in TV for the past 40 years... What this guy is claiming is totally bogus. 1080 and 4K are exactly the same resolution. 4K is just 4x 1080 pictures joined together to give you more real estate. The only real benefit of 4K is in editing you have a lot more latitude to digitally crop a piece of the picture and still have 1080 resolution. That's it.
Where this guy is incorrect is that the resolution changes depending on what mode you're in. In fact is, in either mode you still have either 4K (3840x2160) or HD (1920x1080) resolution being recorded. Because they're all fixed focal length lenses, what changes is the processing that zooms them in to change the field of view. It does not change the recorded resolution but like any digital zooming it will slightly degrade the actual picture quality.
Resolution in television was originally measured in scan lines back in the days of SD but now it's measured in pixel lines in digital. So no matter what mode you're in, you're still getting 2160 (4k) or 1080 (HD) lines of resolution.
When he talks about "pixel skipping," what he's really referring to is compression, not changes in resolution. Video is compressed to save data in the recording. More compression = poorer picture.
As for his claim of "line skipping," I'd like to see some proof of his claim that they're skipping scan lines on the sensor. If this were true, it would mean that you're not getting a true HD image out of the sensor and that makes absolutely no sense. If you eliminate half the lines of resolution you're reducing the image down to slightly higher than SD resolution. That would be extremely noticeable and most people would find it unacceptable in any conditions.
As for his tests, If this guy was doing a legitimate test of the images he would have had a serious set of SMPTE camera charts to show both resolution and coloremetry. The fact that he printed out a chart and it's not correctly scaled tells me he has no clue what he's talking about. Firstly what he used was not a legitimate SMPTE resolution chart. The link below shows what legitimate SMPTE (Society of Motion Picture and Television Engineers) resolution charts look like. They start at about $2 grand each. Clearly not something you print off of the internet. Besides which, there is no home printer capable of printing the kind of resolution required for a real resolution chart to actually work. http://dsclabs.com/test-charts/resolution-focus-test-charts/
The last thing that blew all credibility for me of his claims is that he was using some kind of pot with a piece of the field turf from the super bowl in it as a reference for focus. What a joke. This is no legitimate test. Again there are proper, calibrated focusing charts available to do this properly.
So in short, I think you can distill the changes in the images down to two things... video compression and processing and likely quality of the glass in front of the sensor. Let's face it, folks, you're not getting the same quality glass in a $1200 drone as you would in a $150k UHD television camera lens.
FYI, this is what a $150k television camera lens looks like... and there's a reason why they cost $150k: https://www.usa.canon.com/.../4k-uhd.../uhd-digisuper-122
The only one thing he said about the drones that made any sense is that they all look pretty good.
Thanks for your comparision video, despite the root cause of the crappy 4K FOV is not yet known (and probably never will as DJI is known to not acknowledge issues but rather silently fix them), with your video the writting is in the wall, M2P delivers less detail than M2Z in all video modes and in particular the FOV mode delivers similar detail as 2.7K FOV.
This certainly triggered internal discussions at DJI plus it's getting traction across different RC forums/YouTube, so that will force them address the issue. With the superior H3 SOC onboard, I still believe there is no reason for them to not downsample from 5.5K as good as the P4P does.
I own both the P4pro and the M2pro and easily SEE the difference in resolution on my 4K TV!
I also did the first resolution comparisons here with the colored ISO charts.
I also wouldn’t see NatGeo as a reference in IQ when I compare footage to my Sony a7M3 ;-)!
I am not quite sure if that is correct. Resolution can be defined as how close lines can be to each other, yet still be discerned as two separate lines. By nature, 4k will have a lot more pixels recorded vs 1080p, and hence will allow for more resolution (if you stay in the same spot relative to the screen), especially at larger screen sizes and the closer you stay at the screen.
Similar threads
