AstroBin is the home of astrophotography - Share full resolution images, preserve acquisition details, discover equipment, discuss techniques, and learn from the community.Discover AstroBin

IMX455 vs IMX571 vs IMX533: Do they all produce the same resolution and same level of detail?

CMOS Acquisition Optics
5 replies•211 views
Tommy Mastro avatar
Tommy Mastro avatar

My question is regarding the three Sony sensors (IMX455 vs IMX571 vs IMX533), all of which have the same 3.76 micron pixels. Assuming all the other specs were equal (which I understand they are not), do all three sensors produce the same resolution and same level of detail.

Hypothetical: Let’s assume I took two images from two identical AP rigs (same telescopes, focal lengths, etc), except one had a camera with the IMX455 sensor and the other had the IMX533 sensor. If I cropped the 455 image to the exact dimensions of the 533 FOV, would the two resulting images still have the same resolution and detail?


In other words, once I heavily crop the 455 image, I would then need to zoom in to make it the same size on my PC monitor as the 533 image. By zooming into the cropped 455 image, will it begin to pixelate sooner than the 533 image?

Does this make sense? Hopefully I articulated my question clearly enough.

Thanks,

Tommy

Well written Respectful Engaging
SemiPro avatar
SemiPro avatar

Assuming everything is the same? Then yes the level of detail would be the same.

Maybe things change once we start bringing in the specific characteristics of each sensor (QE curves, bit depth, etc etc), but the image scale will remain the same regardless.

The only thing cropping will do is change the FoV of the final image. Unrelated to your question, but I have found this can be a confusing concept to grasp for people coming from a terrestrial photography background.

Well written Helpful Concise Engaging
Tom Zepf avatar
Tom Zepf avatar

Yes, you are largely correct. Given the same imaging train, if you crop to the size of the 533, they should behave nearly identically. The major difference is the 14bit ADC on the 533 and the 16bit ADC on the 455 and 571. Though they all have the same ~50kev full well, you are losing those two bits of ADC so while “geometry” is the same if cropped, you won’t have the same ability to detect differences in light between pixels. I’m not sure it would make a huge difference in most cases, but something to think about. Personally, if I could afford the APS-C sensor (and filters, if needed) I would go for it. It will be more flexible in the end, and you would have a slight advantage with the 16bit ADC.

BTW, feeding “sony IMX455 vs IMX571 vs IMX533 sensor comparison including full well” into google search gave back some reasonable AI results

Well written Helpful Respectful Engaging
Tony Gondola avatar
Tony Gondola avatar

The basic answer to your question is yes. The FOV will change depending on how many pixels each sensor has but when you view each at 1:1 or 100%, they will be pretty much the same.

As for the differences between 14 bit and 16 bit it’s important to remember that the final product is going to be viewed in most cases on an 8 bit screen. I’m sure there might be some really nuanced differences between the two but it’s not much. I would guess you might see it if you were stretching an image that was really down in the mud and possibly up in highlights that are very close to pure white. Yeah, nah….

Well written Helpful Concise Engaging
Alex Nicholas avatar
Alex Nicholas avatar

Having owned both the 533 and 571, and one of the fabled ASI1600MM’s (3.75um), detail wise, yes - 3.76um pixels at ‘x’ focal length will produce ‘y’ arcsec/pixel resolution.

And while Tony's not ‘wrong’ about 16bit vs 14bit, I would say it’s REALLY dependent on what you shoot, what pixel scale you’re achieving, and how much integration you’re willing to give a target.

At 10h each on a target with reasonable to high brightness, you’ll likely never see the difference… at 5h you probably will, and at 20h on a dark nebula or extended reflection nebula - you certainly will too… I also have an IMX366, which suffers in the resolution department pretty heavily with 4.4um pixels vs 3.76um, but the difference for example, with my redcat51, is 3.1”/px vs 3.6”/px. This SHOULD, in theory, result in a higher SNR, but it doesn’t… I currently have over 14h of Ha/OIII dual narrowband data on the M8-M20 region with the IMX366 (14bit) and it still needs considerably more integration. 14h with the IMX571 mono (spread as 7h Ha, 7h OIII) would be SIGNIFICANTLY in terms of background smoothness and the ability to finesse structure out of the dimmest areas of nebulosity… I can see that detail present in the IMX366 data - but I can’t massage it out like I could with the IMX571.

Yes, most people output their images for display on 8bit screens, but the varying levels of brightness afforded by a 16bit sensor make processing that data considerably easier.

Well written Helpful Insightful Engaging
Tommy Mastro avatar
Tommy Mastro avatar

This is all great information. I do own all three sensors. Since I would only need two at any given time, this gives me the option of selling one of them. And since the smallest sensor is the 533, it makes sense to choose this one to sell.

I had originally bought the 533 to use with my Hyperstar setup, which I have since sold. The smaller, square sensor was convenient for the Hyperstar.

I now have two 10” RC scopes. My plan now is to frame a target using the 571, then match that framing with the 455. I’ll use one for SHO and the other for RGB stars, then crop the 455 to match the 571. Obviously, I’ll need far less integration time for RGB stars, so after a few hours, I would switch it to SHO as well.

This is useful since I can’t afford to own two 455s (pricey little guys!).

Thanks everyone.

Tommy

Well written Respectful