Avigilon doesn’t compress the image as the other manufacturers do. ACC stores all the possible dynamic range of the camera. When you display the video in ACC, the client adapts the range dynamically showing you the low lights or the highlights. You need ACC to get the best of H5A cameras.
We've tested the Avigilon H5A in ACC vs. Exacq, and found a marginal difference in terms of brightness and detail between the two. In this scene, ACC was able to provide a slightly brighter background in the overall view, and marginal detail improvement zoomed in to the plate. Images have been added to the test report.
However, I've embedded the images here in this reply as well for convenience sake.
Would like to see more detailed examples of settings like this, along with commentary. From the mfr side, it would be nice to see HONEST assessments of how much settings tweaks can actually do, coupled with documentation that adequately describes how to use the settings.
On a related note, as of right now, I see serious gaps in the test results here on IPVM of the Wisenet X series cameras. The XNV-8020R, for example, is priced competitively even with the decent Hikua models, and in the initial AI report here a couple years ago it was stated that by using Loitering mode almost all false alerts can be eliminated. After a couple IPVM staff replies where it appears that opinion has changed, I did not get confirmation of whether that initial report was essentially useless at this point in time. The initial report text is in direct conflict with an average performance rating in the recent analytics shootout.
I also don't know how the color low light or IR performance of that model ranks against the newer models. Maybe I missed it, but the Hanwha models that have been tested have different resolutions, sensor sizes, and/or maximum aperture sizes than the XNV-8020R and similar models close to it in the lineup. It might be the good enough Hikua alternative I've been looking for, but unfortunately I cannot find the test data to support the purchase.
Finally, if you work for Hanwha, let me strongly suggest that your camera models are not "partitioned" in illogical ways. The new AI cameras coming out, for example, should also include models with maximum low light color and IR performance. This means F1.6 or smaller apertures aren't going to cut it. Get F1.2 or even F1.0 on there, with 0.6 inch sensor size or so for 4 MP and 1 inch or more for 4K sensor sizing, coupled with very good to excellent analytics, and compete for best of show with Axis on the low light side and Avigilon and Bosch on the analytics side. Don't do "ExtraLUX" on models with little or no analytics capability, and don't do average to below average low light performance on the AI models either. (I For example, avoid Hik's model of small sensor and small aperture Acusense models, versus basic limited analytics on larger sensor and larger aperture ColorVu models. A nonsensical partition, but typical for Hik.)
Other than LPRs and maybe hallway cams, I wouldn't be messing with releasing 1080P models at this point either, whether "ExtraLUX" or "MegaLUX" or whatever. There are way too many good 4K options out there that simply blow away 1080P in terms of identification at greater distances that the small advantages that may exist in low light do not balance the equation. Plus it's only going to swing more to the higher resolution side in the next couple years. 4 MP should be the minimum for general purpose wide angle use for new model releases.