As an Amazon Associate I earn from qualifying purchases from amazon.com

Sony A1 Lab Check – Rolling Shutter, Dynamic Vary and Latitude


Sony A1 Lab Test - Rolling Shutter, Dynamic Range and Latitude

We not too long ago put Sony’s new flagship of the Alpha collection, the A1 by means of our lab take a look at procedures – curious to listen to how good it truly is? Then learn on …

After testing the Sony a7S III a while in the past, it was good to carry a Sony Alpha digital camera in my fingers once more. I actually do just like the relatively small bodily measurement and weight compared to different full-frame cameras just like the Panasonic S1 or S1H for instance.

Underneath the hood, the Sony A1 presents spectacular options (learn extra about these here), like internally recorded full-frame 8K as much as 30 fps or 4K 120fps, all in 10bit with correct LOG encoding in S-Log2 and S-Log3, and an extra ProRes RAW recording chance with the exterior ATOMOS NINJA 5 recorder (see our article right here). Additionally, my pricey buddy and colleague Johnnie made a evaluate of the digital camera together with one in every of his particular “Johnnie signature” mini-documentaries here.

Utilizing the NVIDIA GeForce RTX3090 card with the Sony 8K H265 information

Because of NVIDIA we’ve got a loaner RTX3090 card at our workplace, and imagine it or not, utilizing the 8K H265 information on an 8K (7680×4320) DaVinci Resolve (17.2) timeline was easy and relatively fluid. GPU reminiscence utilization was round 10GB, with the GPU working at round 30% load. Including noise discount (3 frames temporal and spatial) after all causes an enormous drop of playback, however you continue to get about 8 fps with the settings depicted additional down on this article (5 stops below with noise discount) – after all, GPU load is then hovering round 90% and 16GB reminiscence is used. That is actually astonishing, as most graphic playing cards are stumbling already to playback 4K H265 information on a FullHD (1920×1080) timeline. In 8K, we’re speaking 33 megapixels of decision per body (UHD = 8.3 megapixels)! Spectacular! However now, let’s leap into the lab outcomes!

Rolling Shutter of the Sony A1

Capturing at 8K full-frame 25 frames per second, an excellent rolling shutter of 16.6ms is obtained (decrease is best). That’s on par with the Canon R5 for instance (which confirmed 15.5 ms in 8K DCI (17:9, which is 7% much less image peak).

Rolling shutter of the Sony A1 in 8K at 25fps.

Taking a look at 4K full-frame recording at 25 frames per second, the rolling shutter drops considerably to eight.1ms. Clearly, some type of pixel binning is occurring right here – so it is going to be fascinating to see if the dynamic vary will profit from taking pictures within the downsampled 4K full-frame mode vs. 8K full body. In APS-C crop mode at 25fps, the rolling shutter is 11.1ms.

These are very good rolling shutter values for an 8K full-frame digital camera! For comparability, the Sony FX6 (in addition to the Sony a7S III) reveals an 8.7ms rolling shutter for full-frame, 25fps UHD recording, the ARRI ALEXA Mini LF reveals 7.4ms for UHD 25fps full-frame recording.

Dynamic Vary of the Sony A1 at ISO800 utilizing inside recording

Please head over here to see how we take a look at dynamic vary. Capturing our Xyla21 chart at ISO800 utilizing the SGamut3.Cine / S-Log3 mode in 8K 25fps inside (H265) recording reveals the next waveform chart (on an 8K timeline):

Waveform plot for the Sony A1 at ISO800 in inside 8K SG3.C/SLOG3 mode. Picture Credit score: CineD

12 strong stops will be recognized above the noise ground, and a relatively strong thirteenth cease, in addition to a faint 14th cease, are seen contained in the noise ground.

That’s in step with the IMATEST consequence at 8K decision (full-frame), exhibiting 11.9 stops of dynamic vary at a sign to noise ratio of two (SNR), and 13.2 stops at SNR = 1.

IMATEST inside 8K dynamic vary consequence for the Sony A1 taking pictures SG3.C/SLOG3 at ISO800. Picture Credit score: CineD

Additionally, within the center graph of the three diagrams above the SNR = 1 line (blue) about 1 – 1.5 extra stops are recognized by IMATEST.

Downsampled inside 4K recording

Now, let’s take a look on the downsampled 4K (UHD) full body picture of the Sony A1 taking pictures inside UHD XAVC HS:

IMATEST inside 4K (full body) dynamic vary consequence for the Sony A1 taking pictures SG3.C/SLOG3 at ISO800. Picture Credit score: CineD

As you may see, taking pictures in UHD full body on the Sony A1 there appears to be some type of downsampling occurring internally, which mixes pixels and due to this fact reduces the noise, main to higher dynamic vary outcomes: 12.7 stops for SNR = 2 and 13.8 stops for SNR = 1.

Exterior 4.3K ProRes RAW recording at ISO800

Switching the Sony A1 to exterior RAW recording over HDMI, utilizing the Atomos Ninja V we’re getting a extra noisy picture, however the dynamic vary lead to IMATEST (in comparison with inside 8K recording) will increase to 12.3 stops at SNR = 2 (13.3 stops at SNR = 1), see beneath. Clearly, for exterior ProRes RAW recording once more some type of downsampling of the total sensor decision to 4328×2446 ProRes RAW is occurring, thereby growing the dynamic vary outcomes on the SNR = 2 and SNR = 1 thresholds.

Sidenote: for some motive, we couldn’t get the exterior ProRes RAW recording to work with the Sony A1 – information have been exhibiting positive on the Ninja V, however corrupted when performed again on the Ninja or in Premiere Professional. Earlier this week a brand new firmware for the Ninja V was launched, which solved our points. However sadly, that meant there was no time to redo the latitude take a look at in ProRes RAW. However I’m glad at the very least with the assistance of my colleague Florian we have been capable of repeat the dynamic vary take a look at.

The waveform plot once more reveals strong 12 stops, additionally a relatively strong thirteenth cease will be seen in addition to a faint 14th cease:

Waveform plot for the Sony A1 at ISO800 in exterior 4.3K ProRes RAW mode (developed to SG3.C/Slog3). Picture Credit score: CineD
IMATEST 4.3K (full body) exterior ProRes RAW dynamic vary consequence for the Sony A1 at ISO800. Picture Credit score: CineD

Additionally, the IMATEST ProRes RAW outcomes present about 2 extra stops buried within the noise ground above the 13.3 stops SNR = 1 threshold (blue line within the center graph above).

These are actually good outcomes, representing the cutting-edge of what’s potential with client 12bit A/D digital camera sensors – truly surpassing the Panasonic S1H, S1, S5 for instance. The Canon R6 falls means behind in that comparability (see our lab article right here), and the Canon R5 reaches the same stage (article here) solely when taking pictures inside RAW – and this can be a very distinctive function of the R5 that no different SLR model, full-frame digital camera on this worth bracket presents but. And sure, there’s after all the beautiful Z CAM E2-F6 that shoots inside ZRAW – however ZRAW is one other story by itself, see our article here.

To me, the picture of the Sony A1 seems to be extra pure, much less processed than the one of many Sony a7S III (which is hampered by an excessive amount of inside noise discount that can not be turned off).

Latitude assessments with the Sony A1

Latitude is the potential of a digital camera to retain colours and element when over- or underexposed and introduced again to reference publicity. It is rather revealing, particularly when underexposing, how properly the codecs maintain up and particularly, what the noise construction seems to be like.

As talked about earlier, attributable to our points with the Atomos Ninja V we weren’t capable of do the latitude take a look at in ProRes RAW, so we used inside 8K H265 recording as a substitute.

As traditional, for our underexposure take a look at, we organized the scene by adjusting the studio gentle in order that my colleague Johnnie’s face stayed a tad beneath clipping (the purple channel) – at F1.4 and 360° shutter angle (utilizing an 85mm T1.5 Zeiss Compact Prime). Stopping down from there in 1 cease increments we finally reached a luma worth of round 60% – that is our base publicity (for the data, it isn’t all the time precisely the 60% worth because it relies upon a bit on how code values are distributed with the respective LOG curve).

From there, we proceed to underexpose by stopping down the iris of the lens till F8 is reached, then shutter angles are decreased to permit additional lowering the publicity. For all pictures, 8K inside SGamut3.Cine/S-Log3 at ISO800 settings have been used on the Sony A1.

Now let’s take a look on the first shot. As anticipated, the 4 stops above base publicity picture clip Johnnie’s brow (and some patches of the DSC Shade checker on the left, as do all Sony cameras with S-Log3), therefore the primary usable picture is at 3 stops above our (arbitrarily chosen) base publicity studio scene:

Sony A1 3 stops overexposed, introduced again to base publicity. Picture Credit score: CineD

Not a lot to see from 2 stops over, 1 over, and so forth… so lets leap forward to the three stops underexposed scene, pushed again to base publicity:

Sony A1 3 stops underexposed, introduced again to base publicity. Picture Credit score: CineD

A positive noise begins to seem, however all seems to be good – no shade shifts, no stripes throughout the picture, nothing. Superb, let’s transfer on to 4 stops below, pushed again:

Sony A1 4 stops underexposed, introduced again to base publicity. Picture Credit score: CineD

Once more, seems to be superb! So let’s see what the 5 stops below, pushed again picture reveals:

Sony A1 5 stops underexposed, introduced again to base publicity. Picture Credit score: CineD

Luma and chroma noise turns into very evident within the picture, particularly within the darker elements of the picture – on the darkish facet of Johnnie’s face for instance, or take a look at his T-shirt.

A little bit of noise discount in DaVinci Resolve (see settings beneath) does an excellent job on eradicating extreme noise and I discover the 5 stops below completely positive:

Sony A1 5 stops underexposed, introduced again to base publicity utilizing noise discount. Picture Credit score: CineD
Noise discount settings in DaVinci Resolve 17.2 for the 5 stops underexposed, pushed again picture. Picture credit score: CineD

Now, the 6 stops underexposed picture might be fascinating, as we’ve got already reached the noise ground within the shadows – Johnnie’s face is at round 60% luma at base publicity, 5 stops beneath which can be the darkish elements within the studio scene (at 5 stops below we went from 3 above base to five below = 8, plus 5 stops for the darkish elements within the picture = 13).

And right here, at 6 stops underexposure, pushed again the picture can’t be recovered anymore – massive blotches of chroma noise together with heavy luma noise corrupt the picture to an extent, that even when utilizing heavy noise discount the colours can’t be re-established:

Sony A1 6 stops underexposed, introduced again to base publicity. Picture Credit score: CineD

Utilizing noise discount yields the next consequence:

Sony A1 6 stops underexposed, introduced again to base publicity utilizing noise discount. Picture Credit score: CineD
Noise discount settings in DaVinci Resolve 17.2 for the 6 stops underexposed, pushed again picture. Picture credit score: CineD

As will be seen above, shadows turn out to be pink, the picture reveals heavy banding and can’t be absolutely recovered – recreation over.

Nonetheless, we’ve got to state that 8 stops of publicity latitude (3 above to five beneath) is a cutting-edge consequence for a client sensor, once more in keeping with loads of different cameras on this worth bracket and beneath, just like the Panasonic S1H, S1 and S5. A cinema digital camera with a extra subtle sensor design just like the ARRI ALEXA Mini LF (examined here) for instance reveals 10 stops of publicity latitude in our studio scene.

Abstract

As anticipated, the Sony A1 reveals state-of-the-art client digital camera leads to the lab take a look at in all disciplines: rolling shutter, dynamic vary, and latitude, utilizing the inner codecs. This digital camera clearly reveals that previous paradigms like the necessity for big pixels to realize good dynamic vary usually are not true anymore. In spite of everything, we’re speaking a couple of 50 megapixel sensor right here!

To me it appears relatively clear that so long as these cameras use the present 12bit sensor learn out structure for video we can’t count on way more than these outcomes. Simply on your reference, the benchmark to date is the ARRI ALEA Mini LF which reveals about 2 stops extra within the dynamic vary and latitude self-discipline.

Have you ever already used the Sony A1? How do you just like the picture? Tell us within the feedback beneath!



We will be happy to hear your thoughts

Leave a reply

Scatteredbuys
Logo
Enable registration in settings - general
Compare items
  • Total (0)
Compare
0
Shopping cart