[leonine]

Hello everyone,

Restarting the discussion here.

I got a few night aerial images printed recently. The PP was done by someone good at it. Some images are reasonably good upto 8x10 size. Topaz DeNoise was used in the image processing work flow.

The Pentax K1 with the 77mm Limited f/1.8 gave me decent images at 6400 ISO at about 1/45 sec with K1 IBIS on, but only in situations where the experienced pilot managed to inch forward slowly while keeping the helicopter vibrations low to the extent he could. These images are sharp enough, so the IBIS of K1 seems to work.

I can't share the images as yet due to the confidentiality clause.

I believe the Fujifilm GFX 100 or the new 100S with IBIS upto 5.5 stops and coupled to Pentax 645 55mm f/2.8 or 90mm f/2.8 might produce better images. Any thoughts on this?

Will the Pentax 645 55mm f2.8 or 90mm f/2.8 resolve adequately on the 100MP sensor of the Fujifilm GFX?

How would be the high ISO performance of the Fujifilm GFX say at ISO 6400? If I can shoot at that ISO with shutter speed upto 1/100 with IBIS on and lens wide open, I think I can get pretty decent images even in the low light levels of the city in the night, going by my experience with the K1.

The thought of using the new Fujifilm GF 80mm f1.7 on GFX 100S for night-time aerial photography (because of the 1.5 stop advantage compared to the Fastest Pentax 645 lenses) is very tempting, but for the cost. To reduce the cost, but to get another extra stop of light, the Mamiya 80mm f1.9 seems another choice, but again will it stand up to the 100MP sensor of the Fujifilm GFX 100S?

Finally, what about the Sony A7S iii? Meant to be an excellent performer for low light photography, would it be better than the Fujifilm GFX 100S at ISO 6400? What about its only 12MP resolution? Ideally, I would like to print the images to 20x30 size; in the worst case, at least 16x20 size.

I would be thankful for the kind advice of the forum members.

[leonine]

Just to add, I managed to get a good copy of the Pentax 50mm f1.2 on ebay. It is reaching me in end of March. So I will get an additional stop of light compared to the 77mm Limited f1.8. Every bit of extra light gathering capacity of the lens would help to bring down the ISO or increase the shutter speed. I hope that the 50mm is as good as the 77mm wide open.

[Lowell Goudge]

Here’s one just waiting for you, at the Lockheed-Martin facility at the Palmdale Regional Airport (California). (Google maps screen shot)

I’m not sure what the list price, is though, and whether the camera is included!

If time is of the essence, the SR71 is much faster than the U2, and there is/Was one parked at the air museum/bone yard just south of the Tucson airport.

But there may be newer cameras in the current theatre tactical surveillance tool box based on both low and high altitude drones

---------- Post added 02-20-21 at 07:47 AM ----------

Just to add, I managed to get a good copy of the Pentax 50mm f1.2 on ebay. It is reaching me in end of March. So I will get an additional stop of light compared to the 77mm Limited f1.8. Every bit of extra light gathering capacity of the lens would help to bring down the ISO or increase the shutter speed. I hope that the 50mm is as good as the 77mm wide open.

The 50/1.2 is quite soft wide open., and you would be better letting the iso go up a little in my opinion that shooting with the 50/1.2 wide open

For what it’s worth, shooting at F1.8 1/100 and 6400 iso is pretty dark, in my opinion. You really risk burning out highlights, streetlights etc

[leonine]

Thanks Lowell.

There isn't much help available on a good lens-camera combination for night aerial photography without the use of gyro stabilisers.

I am pinning my hope on the IBIS in the camera, the high ISO performance and a fast lens like f/1.2.

So I am looking at options.

The Sony A7S iii is tempting being full frame at only 12MP (bigger pixel size and hence better high ISO performance), but at only 12MP, I don't know how big the print size would be.

Sony A9ii maybe next better camera, being full frame at 24 MP (and not insane 50 or 60 MP for same sensor size), with comparatively decent pixel size and maybe theoretically better high ISO performance than say the Sony A1 or the A7R iv. If I can couple it to the Sigma Art 35mm f1.2, I might have a good compromise combination.

But I need some validation before plunging into a new camera system.

Finally, if forced to choose, I would rather get an excellent small size print from the Sony A7S iii than a not as good print from the Sony A9 ii.

I need some expert advice here from guys who have delved into these situations or at least pushed the limits of low light high ISO performance.

And I have maybe a week to make up my mind.

[UncleVanya]

My thoughts... and bear in mind I do not do this kind of photography.

The a7siii is mainly useful for video. It has lower high iso performance for stills photography than a basic a7iii (which has really good low light capability). The a7Siii video is made with considerably less cropping than the other models which is why they perform better at video but not stills. Even best case low iso dynamic range lags on the a7siii vs a plain a7iii. I own and shoot a Sony A7rII which is why I have looked into the topic deeply.

I haven’t ever shot the Fuji’s, but my impression from all kinds of cross system articles and comparisons is that ibis gets less effective the larger the sensor. Dual IS type systems are the most effective overall. This would lead me to look at one of the systems that offered this. The fact that the k-1 was only good to 6400 is interesting. Was this the Mark 1? If so the accelerator chip in the Mark 2 could stretch your results without major changes.

The question of out resolving a lens comes up a lot. It’s over hyped and less meaningful than it seems. Was the shot on the k-1 made with the fa77 as sharp as a similar shot made with the D FA 85? I’m guessing it wasn’t. But was it a good image? Lenses resolve better almost always on a higher resolution sensor than they do on a lower one. The gain is higher with optically better lenses. Unless you plan on cropping that 100mp image, you won’t see worse results than a lower mp sensor would give, you are possibly going to see less gain than the larger gear would warrant. It’s more important that the general characteristics of the lens work for the photo. So using a vintage lens with bad performance outside the central portion wouldn’t be a good choice for a highly detailed edge to edge landscape shot. Obviously a super sharp lens with similar edge to edge performance with let you resolve more detail than a softer one, but if you like the output of a given lens at lower resolution, don’t be scared to try it on a sensor that may out resolve it.

Looking at both DXO and photons to photos, there’s not a huge amount to gain over the k-1, and the DXOmark folks haven’t tested the k1 MK II. Rankings - DXOMARK. They also haven’t tested the Fuji, but I’d guess it’s sensor is better than all of the listed models, but how much better is unknown. The data on photons to photos ( https://www.photonstophotos.net/Charts/PDR_Shadow.htm ) suggests the GFX 100 isn’t any better at shadow recovery at the iso you are shooting than the k1 MK II. But that data is not as easy to interpret as I’d like. Also practical application of the 645z by real shooters suggests it is better than the data suggests at shadow recovery so the measuring method for larger sensor cameras may need work.

As for the various Sony’s, the a7iii seems like a good sweet spot for what you want to do. The FE 85 1.4 should give good results. But I would have expected a k1 MK II with the d fa 85 to excel also.

One last point, shutter which got much worse as sensor resolution increased. The cause appears to be resolution related. Moving into 100mp sensor resolution may make the vibrations impact the photos more even with good ibis. Ibis wasn’t designed to reduce the impact of this kind of vibration.

[Lowell Goudge]

Thanks Lowell.

There isn't much help available on a good lens-camera combination for night aerial photography without the use of gyro stabilisers.

I am pinning my hope on the IBIS in the camera, the high ISO performance and a fast lens like f/1.2.

So I am looking at options.

The Sony A7S iii is tempting being full frame at only 12MP (bigger pixel size and hence better high ISO performance), but at only 12MP, I don't know how big the print size would be.

Sony A9ii maybe next better camera, being full frame at 24 MP (and not insane 50 or 60 MP for same sensor size), with comparatively decent pixel size and maybe theoretically better high ISO performance than say the Sony A1 or the A7R iv. If I can couple it to the Sigma Art 35mm f1.2, I might have a good compromise combination.

But I need some validation before plunging into a new camera system.

Finally, if forced to choose, I would rather get an excellent small size print from the Sony A7S iii than a not as good print from the Sony A9 ii.

I need some expert advice here from guys who have delved into these situations or at least pushed the limits of low light high ISO performance.

And I have maybe a week to make up my mind.

You need to look not just at the pixel size, between for example the A7 and A9.

My *istD is only 6 MP, but, my K5 blows it away in performance at high iso, and my K1MK2 blows away the K5 considering same pixel density,

The era of the sensor is just as important as the pixel density/size

[leonine]

My thoughts... and bear in mind I do not do this kind of photography.

The a7siii is mainly useful for video. It has lower high iso performance for stills photography than a basic a7iii (which has really good low light capability). The a7Siii video is made with considerably less cropping than the other models which is why they perform better at video but not stills. Even best case low iso dynamic range lags on the a7siii vs a plain a7iii. I own and shoot a Sony A7rII which is why I have looked into the topic deeply.

I haven’t ever shot the Fuji’s, but my impression from all kinds of cross system articles and comparisons is that ibis gets less effective the larger the sensor. Dual IS type systems are the most effective overall. This would lead me to look at one of the systems that offered this. The fact that the k-1 was only good to 6400 is interesting. Was this the Mark 1? If so the accelerator chip in the Mark 2 could stretch your results without major changes.

The question of out resolving a lens comes up a lot. It’s over hyped and less meaningful than it seems. Was the shot on the k-1 made with the fa77 as sharp as a similar shot made with the D FA 85? I’m guessing it wasn’t. But was it a good image? Lenses resolve better almost always on a higher resolution sensor than they do on a lower one. The gain is higher with optically better lenses. Unless you plan on cropping that 100mp image, you won’t see worse results than a lower mp sensor would give, you are possibly going to see less gain than the larger gear would warrant. It’s more important that the general characteristics of the lens work for the photo. So using a vintage lens with bad performance outside the central portion wouldn’t be a good choice for a highly detailed edge to edge landscape shot. Obviously a super sharp lens with similar edge to edge performance with let you resolve more detail than a softer one, but if you like the output of a given lens at lower resolution, don’t be scared to try it on a sensor that may out resolve it.

Looking at both DXO and photons to photos, there’s not a huge amount to gain over the k-1, and the DXOmark folks haven’t tested the k1 MK II. Rankings - DXOMARK. They also haven’t tested the Fuji, but I’d guess it’s sensor is better than all of the listed models, but how much better is unknown. The data on photons to photos ( Shadow Improvement of Photographic Dynamic Range versus ISO Setting ) suggests the GFX 100 isn’t any better at shadow recovery at the iso you are shooting than the k1 MK II. But that data is not as easy to interpret as I’d like. Also practical application of the 645z by real shooters suggests it is better than the data suggests at shadow recovery so the measuring method for larger sensor cameras may need work.

As for the various Sony’s, the a7iii seems like a good sweet spot for what you want to do. The FE 85 1.4 should give good results. But I would have expected a k1 MK II with the d fa 85 to excel also.

One last point, shutter which got much worse as sensor resolution increased. The cause appears to be resolution related. Moving into 100mp sensor resolution may make the vibrations impact the photos more even with good ibis. Ibis wasn’t designed to reduce the impact of this kind of vibration.

Dear UncleVanya,

Thank you for your advice.

A lot of things now make sense. For example, why the Sony A7S iii isn't spoken about much by anyone for extreme low light high ISO photography.

I have been using K1 Mark 1. I understand the Mark 2 would be better.

The IBIS of K1 seemed to work well in the vibratory environment of the helicopter. The pilot was awesome and did his best to keep the platform as steady as he could. I got gopd images at 1/90 seconds. I got a few good ones at 1/45 seconds too. So I presume a modern camera from Sony with a modern IBIS might give additional one stop more to play with allowing a lower shutter speed and hence lower ISO or a higher shutter speed for the same ISO.

Some of the images look fabulous, after certain PP and noise reduction, on the iMac . But the prints do not have even half as much pop as what you see on the iMac screen.

I surely need at least two stops of extra light.

I had not thought of the A7 iii. I will check it out.

Otherwise, I will take a leap of faith with the Sony A1 and pair it with the Sigma Art 35mm f1.2. I guess the high ISO performance of the Sony A1, being a more modern sensor, would certainly be at least one, stop better. A better and modern IBIS might give an odd stop extra. The Sigma 35mm f1.2 will give one additional stop compared to the Pentax 77mm f1.8, besides a more usable focal length and better performance wide open, based on the reviews I have read. So, considered together, I am likely to get at least two stops extra if not three with the Sony paired to the Sigma. This will allow me to shoot at ISO 3200 to 6400 range wide open at decent shutter speeds. That should hopefully suffice.

Since I also do a bit of wildlife photography, the A1 paired to a Sony 200-600 in future would be useful there. Also, I can use my Pentax 645 lenses on the Sony A1 for various purposes where the AF is not important.

The Sony A1 is going to be a huge expense, but at this age and times, I feel it would be prudent to get the best that the technology can offer, as opportunities such as these don't come very often. Well, I can even use the monster Pentax 645 600mm f5.6 on the Sony A1, which is pretty difficult to use on the 645Z for getting sharp images.

If I continue shooting cities by night in days to come, sooner than later I will get the Kenyon Gyro stabiliser unit and be done with the technical challenges for such type of photography once and for all.

---------- Post added 02-21-21 at 11:55 AM ----------

You need to look not just at the pixel size, between for example the A7 and A9.

My *istD is only 6 MP, but, my K5 blows it away in performance at high iso, and my K1MK2 blows away the K5 considering same pixel density,

The era of the sensor is just as important as the pixel density/size

Thanks Lowell for putting things in correct perspective.

I was correlating the ISO performance to the pixel size in a simplistic manner. I was wrong there. I researched over the weekend and understood that a high resolution modern sensor can possibly have much better high ISO performance than an old low resolution sensor for the same sensor size.

I think the Sony A1 might just be the best available option. I will of course also check out the Sony A7 iii before I buy into the Sony system.

[leonine]

UncleVanya,

I can't thank you enough for recommending the Sony A7 iii. Under the present circumstances, this is perhaps the best camera for night aerial photography. I hope to pair it to the Sigma 35mm f1.2 and shoot around ISO 6400. I am pretty confident this will work. I also hope that the Sony manages the big dynamic range of such images.

[TER-OR]

If you want reach, the DA300 on the K3II is a great combination. It really performs well.

[leonine]

Thank you

[UncleVanya]

WOW! Thanks for the detailed description of all cameras. I plan to choose something from the list provided, but my budget is limited. Tell me an inexpensive but high-quality device.

What is the purpose and use and constraints?

[leonine]

Hello friends,

Finally bought the Sony A7 iii with Sigma 35mm f1.2. Will be trying them out soon in the night. I will let you know if the results improve.

I waited for a long time for the K3 iii. Unfortunately, it is not yet launched in India. As and when I get it, I will pair it with say the DFA 50mm f1.4. This will give two good Focal lengths to use - 35mm and 75mm. I found that the 77mm of the FA limited is an excellent Focal Length for aerial photography from around 1000 ft height.

I also found the Topaz DeNoise an excellent tool to reduce noise.

Now my problem is: My 2009 vintage iMac and 2010 vintage macbook can't manage the Topaz software. DeNoise AI at least works on small JPEG files, but the Sharpen AI just can't be run.

What would be a good PC/ MAC to run these Topaz software including Gigapixel AI, and also for stitching Pentax 645Z RAW files (upto 8)to create panoramas and editing 4K movies from the Gopro 9 and Sony A7 iii? Will the latest iMac 27 be good enough? Or should I go in for a PC? Any thoughts on the necessary hardware configuration without going for the overkill (because unnecessary extra RAM and VRAM both cost a lot of money)? Of course, I want it to be a fast machine to speed up my workflow.

[TwoUptons]

Another advantage of the U2 is that its flight path is much more flexible than a satellite. But the OP is planning to take photos that aren't straight down, so a U2 won't do any good.

I don't have any digital aerial photos of cityscapes, but I chartered a Cessna 206 for a two hour flight over the Kluane Icefields. The cabin was pressurized (to get to 17,000 feet) so I had to shoot through glass. I used a K-30 and DA 18-135, at 18mm, wing and wing strut were intruding into my photos, even with the camera up close to the glass, but it still wasn't wide enough to get even half of Mt. Logan in the picture.



At the other end of the scale, at 135mm I was able to get some details of meltwater lakes about 3000 feet below us.

How did you pressurize a 206? That's about as leaky as they come... or did they just put you on supplemental oxygen?

Way back in the day, I worked with some folks who used a Pentax 67 bolted into the bottom of a Cessna 182 to look for people baiting fields for quail...

-Eric

[RGlasel]

How did you pressurize a 206? That's about as leaky as they come... or did they just put you on supplemental oxygen?

There were oxygen masks in case of an emergency (under the seats, if I remember correctly), but the maximum altitude was under 18,000 feet, so not a huge differential to maintain a cabin altitude of 12,500 feet. The only modification I noticed for a pressurized cabin was that all the windows were sealed and couldn't be opened; looking at a photo I took from the ground, the cockpit doors look the same.

[leonine]

Hello friends.

Update:

Did one night aerial sortie using the Pentax K1 with 77mm f1.8 Limited and Sony A7 iii with Sigma 35mm f1.2.

First impressions :

1. It was a pain using the Sigma without hard focus stop at infinity. The focus ring just goes on rotating. Just why is such design needed in a lens? I wasted some precious moments trying to manually focus the lens because the AF wasn't able to cope in focusing on the city lights from the chopper.

2. I did make a few decent images with the Sony and the Sigma. But images with the K1 and 77mm came out visibly better.

3. The K1 (old Mark 1) with the 77mm lens performed much better than the Sony with Sigma lens. That was a big surprise. I was able to make decent images on a chopper handled by an expert pilot using the K1 at at least one stop lower shutter speed than the Sony. Besides, the details in the dark areas were rendered far better by the K1 than the Sony.

While I will figure out some way to use the Sony A7 iii with Sigma 35mm f1.2 in days to come, it was a revelation that the IBIS of the newer Sony was not as good as the SR of the much older K1 in this particular application.

I am waiting for the K3 iii which I will pair with the Sigma 35mm f1.4. I am quite sure that this combination will surpass all other combinations of cameras and lenses I have used so far.

I will update you in due course of time.

Finally, how in the world do you lock the Sigma 35mm f1.2 at infinity permanently in manual mode?