Insta360 Luna Ultra Lab Test: Rolling Shutter, Dynamic Range and Exposure Latitude

Lab Test

by Gunther MachuJuly 30^th, 2026

Subscribe to Lab Tests

Pretty much around the same time that the DJI Osmo Pocket 4 Pro had arrived at our CineD headquarters, we also got the new Insta360 Luna Ultra gimbal camera. As we worked hard to make our standard CineD lab test work with the DJI cam, we were also in a good position to use this setup for the new Insta360 Luna Ultra camera. Hence, we can directly compare the two! Interested to hear more? Then read on …

The Insta360 Luna Ultra gimbal camera is now the second most serious camera in a new market segment that DJI developed and has so far exploited on their own – the small, really pocketable gimbal camera market segment.

Please read our article here about the launch and specs of the Insta360 Luna Ultra.

We have also just published our Lab Test of the Osmo Pocket 4P. In case you missed it, check it out here.

It is a really interesting value proposition: you get a very small, very capable camera with a gimbal that is very affordable and as easy as it gets to shoot with. Previously, action cameras like the GoPro Hero, then the DJI Osmo Action, and also Insta360 action cameras occupied this point-and-shoot segment, but there is a very important difference: the gimbal. For one, you have a much better handgrip to hold the camera; secondly, the gimbal makes your movements super smooth, and the horizon stays level; and thirdly, you can comfortably use ND filters to create motion blur without the motion blur-induced stabilization artifacts of action cameras’ electronic stabilization when the shutter speeds are double the frame rates.

The Luna Ultra pairs a Leica Summicron main lens (1-inch sensor, F1.8 aperture, 20mm full-frame equivalent, 9cm minimum focus) with a secondary telephoto module built on a 1/1.3-inch sensor (F2.0, 60mm equivalent, 15cm minimum focus).

Now, we had a lot of issues trying to run these fixed-lens devices through our standard lab test; more about that in the DJI Omso Pocket 4 Pro lab test and the video above.

Again, a big, big thank you to Florian Milz, who made everything work at the end and helped to shoot and analyze the test!!

The Luna Ultra was especially challenging because there was no way to properly judge the exposure in I-Log. Hence, Florian was running back and forth a thousand times between the studio and the laptop to find the spot where things started to clip …

The DJI Osmo Pocket 4P was so much better, as you could hook it up in webcam mode with a Blackmagic dock to connect an external monitor with proper exposure tools.

Also, it is worth mentioning that we used the latest firmware v1.0.283 from the 18th of July, the “high” bitrate setting, and, of course, 10-bit modes. Internal noise reduction cannot be modified / or turned off anywhere.

Rolling Shutter of the Insta360 Luna Ultra

As usual, we use our 300Hz strobe light, which produces a sequence of black-and-white bars, a characteristic of the readout nature of rolling shutter sensors.

For the wide camera in 8K, we get the following result:

All fine and good, 10.9ms (less is better) is a good but not exceptional result. We get the same for 4K25 and 4K60 modes. In 4K120, the read-out mode changes, yielding 7.4ms. However, the tele camera had a surprise for us:

Ouch – in 8K mode, 25.3ms are obtained. That is a rather bad result!! In tele mode for 4K25, however, the readout speed is again at 10.9ms.

Dynamic Range of the Insta360 Luna Ultra

Now there is not much information available about the “native” ISO settings for both the wide and tele cameras; hence we reached out to our contacts at Insta360, and they confirmed that for ISO400 we should get the best dynamic range results using I-Log.

If you are not aware of how we test dynamic range, please head over here first.

Let’s start with a waveform of 8K25, I-Log at ISO400:

Oh – the noise floor looks super clean. That is not a good sign of things to come …

IMATEST exhibits the following results:

Unfortunately, this IMATEST result reveals quite a few things:

• The values for Slope-based DR, single-to-noise ratios (SNR) of 1 and 2 are all the same: 13 stops. A clear sign of very aggressive in-camera noise reduction. IMATEST can no longer discern the different patches /stops on the Xyla21 chart anymore, thereby invalidating the result.
• The Noise spectrum graph on the lower right-hand side shows rapidly declining amplitudes towards higher frequencies (= resolution), dropping to below 0.2 even at low resolutions. Hence, aggressive noise reduction is reducing details and renders the image mushy. This is not an 8K image!

Now, we also looked at 4K25 or 4K60 modes for the wide camera, all the same:

Overall, just judging from the waveform plot, we can see that the wide camera on the Insta360 Luna Ultra has about 5 stops less dynamic range than the DJI Osmo Pocket 4 Pro’s wide LOFIC sensor. And this result will be confirmed later on in the Exposure Latitude section below. Sorry for the spoiler.

Now, let’s have a look at the second tele camera. The 8K25 I-Log ISO400 waveform plot looks like this:

Interesting – here, we don’t see such an aggressive noise reduction at work. IMATEST confirms this:

We get 9.7 / 11.1 stops at SNR = 2 / 1. That is a proper result and about 2 stops less than the DJI Osmo Pocket 4 Pro tele camera. This image looks absolutely fine!

Strangely though, if we look at the tele 4K25 image, it is similar to all the wide modes in I-Log:

Again, we see the ultra-aggressive noise reduction in place, and the same 12.6-stop values across SNR = 2, 1, and slope-based DR.

I have no idea why the tele 8K25 ISO400 I-Log mode is the notable exception here.

Except for the tele 8K25p result, we will not add these IMATEST scores to our database, as the evaluation basically failed.

Exposure Latitude for the Insta360 Luna Ultra

As stated in previous articles, the exposure latitude is a camera’s ability to retain detail and color when over- or underexposed and then pushed back to a base exposure. This test is very revealing, as it pushes the complete image pipeline of any camera to its absolute limits, not just in the highlights but mainly in the shadows. And typically, it’s in the shadows where the image pipeline and codecs reveal their flaws.

Well, after having seen the dynamic range results, my expectations were low. But have a look yourself. Again, we used 8K25 10bit I-Log ISO400 on the wide and tele camera.

Also, we wanted to use the official I-Log to Rec 709 LUTs we downloaded from the Insta360 homepage, but they had a very pronounced pink tint with totally skewed colors. Hence, we used an alternative LUT that our friend Philip Bloom shared with us.

Our studio base exposure is (arbitrarily) set to an (ungraded) luma value of 60% on the forehead of our subject on the waveform monitor. In this case, my colleague Nino:

Like in the DJI Osmo Pocket 4 Pro lab test, the image is quite distorted because we had to move the wide camera very close to Nino.

We already see quite a bit of noise in this base exposure image, especially at the edges. And posterization is also visible (concentric bands of different colors)…

From here we can overexpose 4 stops:

This is a good-looking image – overexposure hides most of the image flaws!

From base exposure we go one stop down, to 5 stops of exposure latitude:

And … we are already reaching the limit. Especially the moving image in our video above reveals the blotchy, hovering chunks of pink, nasty-looking chroma noise. Just look at the edges of the image, or the dark spot between Nino and the colorcheckr. To me, this looks like too much temporal noise reduction, as you can see a pumping in the image (it starts OKish, then gets worse with each successive frame) – it could also be partially related to the implementation of the H265 long GOP codec which has a rather low bitrate of ~120Mbits/s – clearly not enough for an 8K image! Although the shadow side of Nino’s face looks still good (which is normally our criterion), I find it hard to tolerate this.

Two stops underexposed, 6 stops of exposure latitude looks like this:

Now things get really ugly; we have pushed beyond the limits (look at the video footage).

Just for reference, here is the 3 stops under, pushed back image (7 stops of exposure latitude):

Fine. The wide camera in 8K25 I-Log ISO400 is capable of 4 to a maximum of 5 stops of exposure latitude. Similar to our dynamic range findings above, this is 5-6 stops less than the DJI Osmo Pocket 4 Pro.

The tele camera in 10bit, 8K25 I-Log ISO400 looks much better, and remember that was the only mode not hampered by the aggressive internal noise reduction:

The noise in the 1-stop-under image (5 stops of exposure latitude) definitely looks more pleasing to the eye.

Let’s push to 6 stops of exposure latitude, 2 stops under and pushed back:

Noise is very pronounced, and unfortunately something else pops up: horizontal lines. Post noise reduction cannot mitigate this:

Throughout the image, horizontal lines appear at random locations. Game over again, although this looks way better than the wide camera.

For the tele camera, we can deduce that it has 3 stops less exposure latitude than the DJI Osmo Pocket 4 Pro (which scored 8 stops).

Hence, in summary we have 4 – 5 stops of exposure latitude for the wide and 5 stops for the tele camera.

Summary

For our lab test, log mode is a prerequisite, and I-Log on the Insta360 Luna Ultra exhibits serious deficiencies in the wide camera across all resolutions (tested at ISO400). The exception is the tele camera in 8K25 I-Log mode at ISO400, which shows normal behavior without the super aggressive noise reduction that is hampering the image in all other modes.

This finding is confirmed for the exposure latitude test. 4, respectively 5 stops, is the worst result of all cameras tested so far.

Rolling shutter values are OK for the wide camera, with again the (this time bad) exception of the tele 8K25 mode, which shows a very high 25.3ms of rolling shutter. The gimbal helps on this end of course, and most people probably won’t notice.

We have not looked at any other standard picture profiles, and they may be the better choice for now until Insta360 fixes the I-Log mode.

Have you shot in I-Log on the Insta360 Luna Ultra? What are your experiences? Let us know in the comments section below.