Air 3 vs DSLR Experiment - Help needed again... Thank you!! (First result in...)

[Wreckhunter]

Hi there!

I have made a bold claim that using the 70mm Air 3 lens and stitching a grid of photos to match a 24mm image you can create a 65MP or larger file that would match or exceed a DSLR camera 24mm image.

This is so far theory as I don't own an Air 3 to test yet but if DJI implement one-button-True-High-Res-mode using the 70mm I certainly will as the Air 3 would become a truly remarkable photo platform in the sky.

So... I was hoping someone could do a test for me which involves taking 3 photos, 2 with the Air 3 and one with a DSLR or Mirrorless camera. Thinking is simple, line up the drone at head height (unless you can fly your DSLR around ) take a picture with the 24mm wide lens then from the same spot pointing to he centre of the scene but not matching the range of view, capture a single shot with the 70mm then grab your DSLR and capture the same scene at 24mm then basically pixel peep! Can someone kindly undertake this test??

Details...

1. Find a spot to take the photos, could be your backyard does not matter, but just make sure the scene has plenty of details with lots of light.
2. Crank up the Air 3 and take it to about head height.
3. Using manual or auto exposure (as long as exposed correctly) capture a scene with the 24mm wide lens (RAW best)
4. Now from the exact same spot switch to 70mm and take a shot of the centre of the scene not moving the drone. (RAW best)
5. Now grab your DSLR/Mirrorless camera and attach a 24mm equivalent lens to cover the exact same field of view as the 24mm drone shot. So...
   • 24mm lens if Fullframe system
   • 15mm lens if Crop Sensor APS-C system
   • 12mm lens if Micro 4/3 system
6. Place the camera in the same spot the drone was located and using the 24mm equivalent capture a shot matching as close as you can the view if the 24mm wide on the Air 3 (RAW best)
7. Then if you could post a link here to the RAW files would be awesome.

Thanks so much and hopefully someone will be able to oblige.

Goal? To see if in fact the resolution of the Air 70mm Tele (if used in a 16 shot stitch - not that we are taking 16 images in this test) equals or surpasses that of a DSLR/Mirrorless Image


Many thanks

Andrew Busst

P.S. If you haven't already and you think it is a good idea to have an automated 16 stitch stills option using the 70mm on the Air 3 please vote here so we let DJI know we would find this a valuable addition. Thank you!
DJI The case for TRUE High Res on the Air 3 – Please Vote

A claim is all it will ever be, and also if that is your first and last name I would never put it anywhere on here,or anywhere else.
Not a great idea.

 

[Wreckhunter]

Hi there!

I have made a bold claim that using the 70mm Air 3 lens and stitching a grid of photos to match a 24mm image you can create a 65MP or larger file that would match or exceed a DSLR camera 24mm image.

This is so far theory as I don't own an Air 3 to test yet but if DJI implement one-button-True-High-Res-mode using the 70mm I certainly will as the Air 3 would become a truly remarkable photo platform in the sky.

So... I was hoping someone could do a test for me which involves taking 3 photos, 2 with the Air 3 and one with a DSLR or Mirrorless camera. Thinking is simple, line up the drone at head height (unless you can fly your DSLR around ) take a picture with the 24mm wide lens then from the same spot pointing to he centre of the scene but not matching the range of view, capture a single shot with the 70mm then grab your DSLR and capture the same scene at 24mm then basically pixel peep! Can someone kindly undertake this test??

Details...

1. Find a spot to take the photos, could be your backyard does not matter, but just make sure the scene has plenty of details with lots of light.
2. Crank up the Air 3 and take it to about head height.
3. Using manual or auto exposure (as long as exposed correctly) capture a scene with the 24mm wide lens (RAW best)
4. Now from the exact same spot switch to 70mm and take a shot of the centre of the scene not moving the drone. (RAW best)
5. Now grab your DSLR/Mirrorless camera and attach a 24mm equivalent lens to cover the exact same field of view as the 24mm drone shot. So...
   • 24mm lens if Fullframe system
   • 15mm lens if Crop Sensor APS-C system
   • 12mm lens if Micro 4/3 system
6. Place the camera in the same spot the drone was located and using the 24mm equivalent capture a shot matching as close as you can the view if the 24mm wide on the Air 3 (RAW best)
7. Then if you could post a link here to the RAW files would be awesome.

Thanks so much and hopefully someone will be able to oblige.

Goal? To see if in fact the resolution of the Air 70mm Tele (if used in a 16 shot stitch - not that we are taking 16 images in this test) equals or surpasses that of a DSLR/Mirrorless Image


Many thanks

Andrew Busst

P.S. If you haven't already and you think it is a good idea to have an automated 16 stitch stills option using the 70mm on the Air 3 please vote here so we let DJI know we would find this a valuable addition. Thank you!
DJI The case for TRUE High Res on the Air 3 – Please Vote

First sentence says you can create a 65mpFILE OR LARGER,So in effect this is just eating up more memory card or hard drive space.Post # 1 ,not so cool.As some people do not have that kind of extra storage space.

 

[Droning on and on...]

Different lenses with different focal lengths show a different perspective.
The case you are talking about isn't "actually removing some of the distortion that existed in the original 24mm view"
It's replacing the perspective of one lens with the perspective of a different lens.

In addition to this, there are temporal issues, just like any stitched image, if anything perceptibly moves from shot to shot.

Still, it's an intriguing way to substantially increase detail if the few issues discussed are either a non-issue for the shot, or what errors occur are acceptable for the application.

 

[Droning on and on...]

First sentence says you can create a 65mpFILE OR LARGER,So in effect this is just eating up more memory card or hard drive space.Post # 1 ,not so cool.As some people do not have that kind of extra storage space.

Not getting why you are being so nitpicky critical of this guy's idea and effort. File size these days, when we're talking tens of megabytes, even hundreds, is utterly irrelevant with the size of a small SD card these days. Why are you beating him up about it ("not so cool")?

It's an interesting idea. First pass experimentation seems to have borne good results, the intended improved detail being realized.

Stitching problems, color correction, lens distortion, etc., are all things that can be worked on to correct. Possibly eliminated.

While some have tried to be helpful, take the idea and try to offer suggestions to make it work, there's also a lot of antibody commentary that helps not at all.

 

[rddl]

You can also try AI websites to increase size of your shots. Give it a try, you will be surprised by the quality you get !
Example
It is probably the fastest way as time is money !

 

[Bussty]

Not getting why you are being so nitpicky critical of this guy's idea and effort. File size these days, when we're talking tens of megabytes, even hundreds, is utterly irrelevant with the size of a small SD card these days. Why are you beating him up about it ("not so cool")?

It's an interesting idea. First pass experimentation seems to have borne good results, the intended improved detail being realized.

Stitching problems, color correction, lens distortion, etc., are all things that can be worked on to correct. Possibly eliminated.

While some have tried to be helpful, take the idea and try to offer suggestions to make it work, there's also a lot of antibody commentary that helps not at all.

Thank you so much. I have been completely baffled about all this negativity and I really need to ignore it and just show the data I receive. To me while stitching has been around a long time on drones to be able to stitch a tele lens on an Air model to surpass the quality of a single frame will improve my stills taken from the air enormously. Maybe some just don't stitch or aren't familiar with it but I use it not just with drone images but also my Mirrorless system and is very familiar to me.

If you want a way to keep drones small but give them the ability to capture scenes in true high res, this is it. Why would you not want to explore that possibility especially when evidence supplied clearly shows a marked improvement.

I appreciate your viewpoint on this.

 

[Droning on and on...]

Dude, you have an interesting idea. It's not loony or impractical. You've shown some rough attempts with errors, but that also clearly demonstrate the viability of the concept.

Can it be refined and tuned to the point it's more than sufficient for some applications? Looks hopeful to me. Why people need to tear you down because it's not perfect at first attempt, rather than try and help improve things and make it work is puzzling.

Sold my A2S in anticipation of upgrading to an A3. Didn't pull the trigger yet, then the Mini4P hit the streets. I've been vacillating between the two, and your concept is another strike in favor of the A3, if it can work, regardless if DJI adds automation for it.

 

[Wreckhunter]

Not getting why you are being so nitpicky critical of this guy's idea and effort. File size these days, when we're talking tens of megabytes, even hundreds, is utterly irrelevant with the size of a small SD card these days. Why are you beating him up about it ("not so cool")?

It's an interesting idea. First pass experimentation seems to have borne good results, the intended improved detail being realized.

Stitching problems, color correction, lens distortion, etc., are all things that can be worked on to correct. Possibly eliminated.

While some have tried to be helpful, take the idea and try to offer suggestions to make it work, there's also a lot of antibody commentary that helps not at all.

I am not the only one that is a non believer,I voiced my opinion about this matter and that is the way it is.
In a perfect world he thinks that everyone would be on board with this IDEA.I am not picking on the individual
himself,not by any means.There is just not enough proof that this method will in fact out perform a DSLR camera.
So this will in fact be my last response on this matter,as I have other stuff going on besides living on here

 

[Bussty]

You can also try AI websites to increase size of your shots. Give it a try, you will be surprised by the quality you get !
Example
It is probably the fastest way as time is money !

Thanks rddl I use Topaz Photo AI for pretty much all my 12MP drone images, upscale 2x better than the 48MP mode from my experience. I also restitched that 16x 70mm scene the other night taking a bit more time and it stitched it perfectly I can post if you like only 36MB for the jpeg, TIFF 562MB!!!

 

[Bussty]

Thanks again to noob1957 on the DJI forum for going out of their way to test the Air 3 against their Sony FF 24MP camera.

So to recap noob1957 has found a couple of scenes in midday sun and taken wide 24mm views with both the Air 3 and a Sony FF DSLR. Here are those views…







The whole feature of this technique is to then stitch a 16x 70mm shot grid taken from the same spot as the 24mm images to effective cover the same area as the 24mm lens. We aren’t doing that in this test as we have tested the ability to stitch those images here… https://forum.dji.com/forum.php? ... D922%26typeid%3D922 and it works really well as can be seen (remember these images are just jpegs so properly converted RAWS could be even better)

So these next two comparisons are just a zoom of a single central image from the Air 3 Tele lens compared to a single frame of the Sony FF 24MP camera taken on a bright day. The Sony image is zoomed in to match the 70mm Tele of the Air 3. What we are really testing here is if we could send our Sony FF 24MP DSLR flying into the sky are we better capturing a single 24mm Wide Angle shot with the Sony DSLR or 16x Air 3 images using the 70mm lens then stitching to provide the same 24mm view as the Sony. You will have to agree that in both cases the Air 3 70mm result is significantly better…






And that result means with a little work in post processing to stitch 16x 70mm images (honestly takes less than 2 mins) you will exceed the quality of most DSLR cameras and any Micro 4/3 camera found on the Mavic 3 Pro or Mavic 3. And that fact is revolutionary for stills photographers especially if DJI make this automated 16x capture a quick and seamless process which they absolutely can do if they want to. Of course this won’t work if your image is moving but for the majority of still photographers that is not the case and with stitching software today and the ability to mask you can easily get around this for small areas of motion in your image.
You can download the source material for this latest test and see for yourself https://drive.google.com/drive/f ... bilY?usp=share_link

Thanks again for noob1957 taking the time to conduct the test and if this technique interests you please let DJI know including voting here if you haven’t already…

https://forum.dji.com/forum.php? ... D922%26typeid%3D922

Currently 44 of you think this is a great idea

 

[Meta4]

Currently 44 of you think this is a great idea

Now ask them if they think it's a great idea to wait to shoot 16 images every time they want to take a photo, and then stitch the 16 images to get a single finished mega image?
How about scenes or subjects with movement?

 

[Bussty]

Now ask them if they think it's a great idea to wait to shoot 16 images every time they want to take a photo, and then stitch the 16 images to get a single finished mega image?
How about scenes or subjects with movement?

Hi there Meta 4

In my experience I just haven't seen those two things as being issues.

Here's a heap of 24mm stitched images (much harder to stitch than 70mm) 360 Panoramas of New Zealand Landscapes by Andrew Busst These panoramas are also stitched with 35 images (way more than 16) some images need a little coaxing but these are 360 panos needing everything to align perfectly. A 16 grid is way way easier than these and way quicker.

Here is a shot with a lot of movement Lake Pukaki Outlet - New Zealand 360 Panorama | 360Cities

...and another Peregrine Loop - Kawarau River - Queenstown 360 Panorama | 360Cities

...and another Roaring Meg Power Station, Kawarau Gorge, Queenstown, New Zealand 360 Panorama | 360Cities

..so movement has never really been an issue and if it was you can easily mask in the stitching software. Having said that some images (I'm thinking a long wave for example) may cause issues but the majority would be fine. In the wave example just use the wide angle single shot and live with a smaller print.

I think given the small amount of time to shoot 16 images (I'm guessing 30-40 secs if done automatically) and the 2mins to stitch back home (this shouldn't be done in the drone), those two minor inconveniences are certainly tolerable if it enables you to make bigger prints than what even your DSLR can do (unless your are running a 65MP DSLR but when have ever been able to compare a drone image to a 65MP DSLR in the past?). Most photographers go to a place to capture a specific image so this amount of time in the scheme of things would seem very small in the whole process.

Cheers

Bussty

 

[Meta4]

..so movement has never really been an issue and if it was you can easily mask in the stitching software.

We don't all photograph static landscapes.
I was thinking of a shot where the movement is an important part of the shot, not something that you want to mask out.

 

[Wreckhunter]

Hi there Meta 4

In my experience I just haven't seen those two things as being issues.

Here's a heap of 24mm stitched images (much harder to stitch than 70mm) 360 Panoramas of New Zealand Landscapes by Andrew Busst These panoramas are also stitched with 35 images (way more than 16) some images need a little coaxing but these are 360 panos needing everything to align perfectly. A 16 grid is way way easier than these and way quicker.

Here is a shot with a lot of movement Lake Pukaki Outlet - New Zealand 360 Panorama | 360Cities

...and another Peregrine Loop - Kawarau River - Queenstown 360 Panorama | 360Cities

...and another Roaring Meg Power Station, Kawarau Gorge, Queenstown, New Zealand 360 Panorama | 360Cities

..so movement has never really been an issue and if it was you can easily mask in the stitching software. Having said that some images (I'm thinking a long wave for example) may cause issues but the majority would be fine. In the wave example just use the wide angle single shot and live with a smaller print.

I think given the small amount of time to shoot 16 images (I'm guessing 30-40 secs if done automatically) and the 2mins to stitch back home (this shouldn't be done in the drone), those two minor inconveniences are certainly tolerable if it enables you to make bigger prints than what even your DSLR can do (unless your are running a 65MP DSLR but when have ever been able to compare a drone image to a 65MP DSLR in the past?). Most photographers go to a place to capture a specific image so this amount of time in the scheme of things would seem very small in the whole process.

Cheers

Bussty

Sorry about giving you a hard time on here about this subject.It looks like you may be on to something really good here.
I am looking at adding the Air 3 to my other drones next year.My issue is I have no experience with stitching photos .lol
I fly over water most of the time, and as a hobby am a shipwreck photographer,along with coastline photos.If I could get
better yet photos out of the shipwrecks and lake coastline ,this would be fabulous
Once again I apologize to you for giving you a hard time on here about this.
Enjoy your day

 

[Bussty]

Sorry about giving you a hard time on here about this subject.It looks like you may be on to something really good here.
I am looking at adding the Air 3 to my other drones next year.My issue is I have no experience with stitching photos .lol
I fly over water most of the time, and as a hobby am a shipwreck photographer,along with coastline photos.If I could get
better yet photos out of the shipwrecks and lake coastline ,this would be fabulous
Once again I apologize to you for giving you a hard time on here about this.
Enjoy your day

No worries Wreckhunter and thank you

Can you imagine a half wall size print of a wreck with all that detail??? It would be amazing. You will have a bit of motion to contend with but I believe for the majority of shots you can mitigate that. I used to always use PTGUI and still do occasionally but I have found for smooth blends where lighting might be changing Panorama Studio 3 is working out better just check out this one I did last week with the Mini 3 Pro wide lens... The Iconic Judge and Jury - Queenstown - New Zealand 360 Panorama | 360Cities but you have to remember these are 35 image 360 pano stitches way way way harder than a simple 16 shot stitch. Tele shots are much easier to stitch as it's a flatter view with less perspective distortion. Honestly all I have done with my first sample is open up the images into the stitcher and hit stitch! Job done. There's a little cropping and maybe working out which projection is best but once you know that it is nothing plus there is a heap of advice in Youtube and in these forums (also over on DJI Forum).

I have definitely seen enough to buy the Air 3 and also will probably be doing this next year and am truly looking forward to effectively having a high resolution DSLR in the air!!

My real wish though is for everybody that is interested to let DJI know, this is a SIMPLE change on their behalf, Yes you can do this technique manually taking the photos but it will be far more accurate and hassle free if they give you a single on screen button that automates it even better if the do a portrait mode where you guess where the top third of the image is using the 24mm lens then the drone switches to 70mm and does the rest!! Cheers Bussty

 

[rddl]

I have definitely seen enough to buy the Air 3 and also will probably be doing this next year and am truly looking forward to effectively having a high resolution DSLR in the air!!

Imagine your high resolution DSLR in the air with the button you want : you will get an extra ++ large picture, far better than what the Air3 could do with your button !

 

[Bussty]

Imagine your high resolution DSLR in the air with the button you want : you will get an extra ++ large picture, far better than what the Air3 could do with your button !

So true however I don't think the flight regulators will be letting us do that anytime soon Hence why the Air 3 is so special (or a Mavic 3 Pro as has the same tele lens, would also love to see a 24mm view stitched with the 7x Mavic 3 Pro lens) Just feels for the size/weight/MP/Cost value the Air 3 hits a sweet spot.

 

[Bussty]

We don't all photograph static landscapes.
I was thinking of a shot where the movement is an important part of the shot, not something that you want to mask out.

True, I'm guessing using this technique to capture a moving cruise ship or boat for example wouldn't work if the moving object exceeded two frames of the 3x lens. Any movement contained in a single frame is usually fine. Even a highway of cars can be made to work through masking.

 

[BobaFut]

Hi there!

I have made a bold claim that using the 70mm Air 3 lens and stitching a grid of photos to match a 24mm image you can create a 65MP or larger file that would match or exceed a DSLR camera 24mm image.

Why is this a bold claim? It seems pretty self-evident to me and it's pretty common for photographers to stitch longer focal length shots into a wider angle shot to increase the detail. I mean, MOST of the "panoramas" I take (both drone and regular camera) are really just stitches of a photo I could easily get with a wider lens, but storage is cheap and processing is fun, so why not get the biggest file possible?

Where this breaks down is that that a relatively new FF camera will provide you with much greater dynamic range, reduced noise, and wider latitude for post-processing than the tiny sensor in an Air3, so it's not really an apples-to-apples comparison and raw megapixels don't tell the whole story. Also, at some point, you just aren't going to see those extra megapixels unless you are peeping or making huge prints.

 

[Bussty]

Why is this a bold claim? It seems pretty self-evident to me and it's pretty common for photographers to stitch longer focal length shots into a wider angle shot to increase the detail. I mean, MOST of the "panoramas" I take (both drone and regular camera) are really just stitches of a photo I could easily get with a wider lens, but storage is cheap and processing is fun, so why not get the biggest file possible?

Where this breaks down is that that a relatively new FF camera will provide you with much greater dynamic range, reduced noise, and wider latitude for post-processing than the tiny sensor in an Air3, so it's not really an apples-to-apples comparison and raw megapixels don't tell the whole story. Also, at some point, you just aren't going to see those extra megapixels unless you are peeping or making huge prints.

Clearly I am talking to the converted here! The thing with the Tele on the Air 3 is that down at the mid end (excluding the Mavic 3 Pro and Mavic 3 as they have tele lenes) it's the first drone to be able to stitch "within a normal 24mm view" and give a high resolution. When you want just that "standard" view but in high res there hasn't really be many options. I have done a lot of wide angle stitches and at a row of three or four perspective still looks fine but going beyond that things can start to look a little "wide" (is that the word I'm looking for?) kind of like this Gimmerburn Building | CENTRAL OTAGO Also we have never had the ability to match the current view with the 24mm with a precise automated 16 shot stitch which would be so easy for DJI to do allow you to precisely compose your shot whereas you can do it manually but would be way more convenient and faster.

I hear you about the dynamic range and that will be an issue for some shots but for a lot it will be more than fine but I agree is one of the shortcomings.

Re dynamic range one thing I always used to do was expose my 360 panos with set manual exposure but a friend told me to use auto exposure with a pretested exposure compensation and that has been ground breaking for me in getting overall good exposure everywhere The Mighty Nevis Bluff Rapids - Queenstown - New Zealand 360 Panorama | 360Cities I'm wondering if using this technique with a 16 shot stitch will kind of "fight" the need for huge dynamic range?

I am a self confessed pixel peeper but I find with drone shots people want to see the details as it's such a unique view for many so the more resolution the better.

Thanks for your post.

Cheers

Bussty