MAVIC 3 - ISO TEST 4K (Low Light) - Mavic 3 vs Mavic 2 vs Inspire X5S

[canadian drone videos]

Have you seen a test somewhere with all ISO variants and all-color profiles M2 vs M3 vs X5S with attached source files?

The test results don’t even show the full low light capability of the x5s, the 15mm goes to F1.7, this test isn’t even using a Dji lens on the x5s. So if you wanted to see the difference of possible best low light performance from all 3 camera sensors this test is still inconclusive.

All I’m saying is as a former owner of the x5s I’m more curious to know how this 4/3rds M3 sensor holds up against the x5s at F2.8? The M2 isn’t in the same league, the M3 is obviously way better and can be on par with the x5s.

Using your analogy, instead of comparing a Honda Civic to a Ferrari I want to compare a Ferrari to a Lamborghini.

 

[brett8883]

The test results don’t even show the full low light capability of the x5s, the 15mm goes to F1.7, this test isn’t even using a Dji lens on the x5s. So if you wanted to see the difference of possible best low light performance from all 3 camera sensors this test is still inconclusive.

All I’m saying is as a former owner of the x5s I’m more curious to know how this 4/3rds M3 sensor holds up against the x5s at F2.8? The M2 isn’t in the same league, the M3 is obviously way better and can be on par with the x5s.

There’s a larger difference between the X5S and the M3 than there is between the M2 and M3. The sensor on the M3 is about twice the size as the M2 but the X5S lens let’s in 3-4x more light than the M3. Larger sensor or faster lens does the same thing in this context.

Using your analogy, instead of comparing a Honda Civic to a Ferrari I want to compare a Ferrari to a Lamborghini.

Then compare the X7 at f/2.8 to the X5s at f/1.7. That would be an interesting test.

 

[canadian drone videos]

There’s a larger difference between the X5S and the M3 than there is between the M2 and M3. The sensor on the M3 is about twice the size as the M2 but the X5S lens let’s in 3-4x more light than the M3. Larger sensor or faster lens does the same thing in this context.

Then compare the X7 at f/2.8 to the X5s at f/1.7. That would be an interesting test.

I don’t know if I agree with that? Or I might be misunderstanding what your saying? When I use, say, f5.6 on the M3 it’s relative to f5.6 on the x5s on all the same settings. I also know from experience 1.7 doesn’t let in 3-4 times more light then 2.8 on the 15mm x5s DJI lens. If you put the DJI 15mm lens on the x5s set it to 2.8 and set the M3 to 2.8 and use identical settings you will be hard pressed to see any difference in the two images, except a slightly wider FOV on the M3. If you compare an M2 to an M3 like we see in this test there is a huge difference between the M2 and M3. Thus, the M3 is closer to the x5s then it is to the M2.

Someone should just do the test to put this argument to rest. What I don’t need is a test between the M2-3, it tells me nothing. The only similarity between the M2-3 is the name and the fact both are folding drones. I never even considered buying a Mavic of any kind till I heard this 3 has the x5s sensor, which I’m very familiar with.

The X7 has a full frame sensor, but only f2.8 lenses. However because the sensor is twice as big it would have even better low light performance then the x5s.

 

[Netvise]

I don't shoot at night on X5S with an aperture of 2.8 I don't need such a test. If you need such a test, take two devices and do it.

I think this particular test was useful to many. Thanks for the feedback.

 

[brett8883]

I don’t know if I agree with that? Or I might be misunderstanding what your saying? When I use, say, f5.6 on the M3 it’s relative to f5.6 on the x5s on all the same settings. I also know from experience 1.7 doesn’t let in 3-4 times more light then 2.8 on the 15mm x5s DJI lens. If you put the DJI 15mm lens on the x5s set it to 2.8 and set the M3 to 2.8 and use identical settings you will be hard pressed to see any difference in the two images, except a slightly wider FOV on the M3. If you compare an M2 to an M3 like we see in this test there is a huge difference between the M2 and M3. Thus, the M3 is closer to the x5s then it is to the M2.

Someone should just do the test to put this argument to rest. What I don’t need is a test between the M2-3, it tells me nothing. The only similarity between the M2-3 is the name and the fact both are folding drones. I never even considered buying a Mavic of any kind till I heard this 3 has the x5s sensor, which I’m very familiar with.

For anything but low light they in theory should be similar. In low light, however, things change because of the faster lens on the x5S. This thread is about low light capability so that is what we have been focusing on. However, in full light you are correct that the M3 and X5S should be similar. Maybe this is the cause of confusion?

Low light capability is based on how much total light hits the sensor. The total light hitting the sensor and therefore lowlight capability comes down to two factors. The size of the sensor and amount of light that enters the lens.

The amount of light that enters the lens is determined by the aperture opening and the focal length. The larger the aperture opening and wider the lens the more light comes in. To standardize this ratio of focal length to aperture opening they came up with the f-stop system. F-stop is focal length divided by aperture opening. Using F-Stop we can compare the amount of light entering the lens regardless of focal length.

f/1, f/1.4, f/2, f/2.8, f/4, f/5.6, f/8, f/11, f/16, and f/22 each represent a doubling of the amount light allowed in by the aperture called “stops.” f/1 is 2x as much light as f/1.4, 4x as much light as f/2, 8x as much as f/2.8 and so on. f/1.7 is between f/1.4 and f/2 so it’s more than 2x the light as f/2.8 but not quite 4x. It’s 3x as much light. This is an unarguable fact.

Sensor size also contributes because a larger surface area is able to capture more light. Double the sensor size and it will be able to capture twice the light. Therefore a sensor that is twice as big acts the same as the difference between f/2.8 and f/2.

The X7 has a full frame sensor, but only f2.8 lenses. However because the sensor is twice as big it would have even better low light performance then the x5s.

The X7 is an APS-C sensor. It is about 64% larger than the X5S representing about a 2/3 stop advantage. A full frame sensor is roughly 2.3x larger than the X7. The X5S only needs to be at f/2.2 to equal the low light capability of the X7 at f/2.8. At f/1.7 the X5s has 2/3rd stop advantage on the X7 making the x5s still the King of low light. In fact you can get lenses that go down to f/1.4 which would be even more of an advantage.

 

[brett8883]

The tests show exactly what you would expect to see if you understand the mechanics of how a camera captures light.

 

[Surge]

Watching the video - how on earth is a “low light” test done at 400 ISO!? Is that a joke?
Almost setup so the Mavic 2 can excel. Crank the ISO up to 1600 or 3200 and let’s compare!

Where the M3 is obviously superior even in your fake low-light test is in dynamic range and sharpness, vs. the M2.

 

[Netvise]

Watching the video - how on earth is a “low light” test done at 400 ISO!? Is that a joke?
Almost setup so the Mavic 2 can excel. Crank the ISO up to 1600 or 3200 and let’s compare!

Where the M3 is obviously superior even in your fake low-light test is in dynamic range and sharpness, vs. the M2.

ISO 1600 or 3200 in the test are presented.

In what place is it fake? The sources are in the video description. You can study them personally.

 

[Netvise]

I have all these three devices. What's the point of me faking something? I was interested in comparing them. What I've done and shown you.

 

[Surge]

ISO 1600 or 3200 in the test are presented.

In what place is it fake? The sources are in the video description. You can study them personally.

Ok thanks. I have not finished watching. Just saw the opening “low light” at ISO400, which is a daylight ISO setting and immediately assumed the creator doesn’t know what low light is.

 

[brett8883]

Ok thanks. I have not finished watching. Just saw the opening “low light” at ISO400, which is a daylight ISO setting and immediately assumed the creator doesn’t know what low light is.

Low light just means that, taken when there is low amounts of light which is clearly the case in the video.

 

[kodiak]

Now that I have a M3C in my hands I've been able to do some very unscientific but direct testing between the X7 and the output of the L2D-20c. I've gotta say the D-Log PR 5.1K output of the M3 looks very good. Excellent detail, range and overall IQ. The LOG profiles match up well. So far I've only done extremely limited low light testing. I think the X7 will be noticeably superior in all aspects but there is no doubt the M3C PR output is very usable and could cut in well- especially in the right kind of low light situations and even more so with the right massasging in post. For sure as an auxiliary tool the M3C is going to fit right in.

 

[Deleted member 77603]

"The X7 is an APS-C sensor. It is about 64% larger than the X5S representing about a 2/3 stop advantage. A full frame sensor is roughly 2.3x larger than the X7. The X5S only needs to be at f/2.2 to equal the low light capability of the X7 at f/2.8. At f/1.7 the X5s has 2/3rd stop advantage on the X7 making the x5s still the King of low light. In fact you can get lenses that go down to f/1.4 which would be even more of an advantage."

Exactly...its why f2.8 is a "let down" on cameras like the X7, and the various faster lenses you can put on the X5s give it an advantage.

 

[ianwood]

Good discussion.

Again, I think the most interesting low-light comparison would be X5S with Oly 12mm vs. M3C, ProRes 422HQ. The Oly 12 is the same field of view, widely used and considered a very good lens. It gives you the closest like for like comparison so you can evaluate differences in the sensor, lens, and image processing of each.

Run a matrix of tests between ISO 400-1600 and f2*- f4 (*f2.8 minimum for M3C).

Look at overall image quality, color, saturation, shadow detail, highlight detail, distortion, vignetting, chromatic aberration and noise.

Note: would need to verify which filming modes are open gate without binning on each sensor and make sure to use those and scale to a common size. This would provide the truest like for like comparison of sensors.

Bonus comparison: X5S C-DNG minimally processed through Resolve or After Effects.

 

[brett8883]

Exactly...its why f2.8 is a "let down" on cameras like the X7, and the various faster lenses you can put on the X5s give it an advantage.

Yep and you could even push the X5S further by using manual lenses. There are some f0.95 and even a f/0.8 manual lenses that would likely work if you balance them with some counter weights. You’d of course have to set focus to infinity before take off and depth of field might actually come into play but it would work.

 

[kodiak]

Okay, if you look at the DJI white paper on their D-Log and D-Gamut color space it specifically lists the X5S as a 10-bit signal. So while I can record ProRes 4444XQ it does not utilize the full 12 bits. I guess similar to how it does not utilize the alpha channel. I think this is also supported by the fact that the X5S can not do PRR, which utilizes a minimum of 12 bit encoding.

 

[D_The_Producer]

Hi!

Many people are interested in how the sensor works in Mavic 3.
I compared Mavic 3 vs Mavic 2 vs Inspire 2 X5S

I shot everything on video for you.
Enjoy watching and subscribe (I'm preparing a flight time test).

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

This is such an amazing video, man! I will use it for my class, I teach a drone audiovisual course. Thank you!

 

[brett8883]

Okay, if you look at the DJI white paper on their D-Log and D-Gamut color space it specifically lists the X5S as a 10-bit signal. So while I can record ProRes 4444XQ it does not utilize the full 12 bits. I guess similar to how it does not utilize the alpha channel. I think this is also supported by the fact that the X5S can not do PRR, which utilizes a minimum of 12 bit encoding.

Can you link to the the white paper you found? The one I know of doesn’t seem to mention this. We know the X5s is capable of recording 12 bit because it records 12 bit raw. As to why the X5s doesn’t record ProRes Raw I just assumed it was a business decision rather that a physical limitation but its possible there’s more to it.

 

[kodiak]

Yes- sorry I didn't include that earlier.

White Paper on D-Log and D-Gamut of DJI Cinema Color System