New camera with very impressive IMX 482 new Sony sensor

moson

That's a very good sensor, thanks for sharing us! Our devs will working hard to provide our customers a better camera which use a better sensor in the near future.

tjn

Hi zwo team

Should the IMX482 have similar NIR capabilities to imx462, this could open up an entirely new market for sub urban astronomers.

I have done some studies within the field of NIR-RGB astronomy of galaxies as an alternative to LRGB. Until now the results are very positive.

What it requires is a NIR sensitive camera with very low noise. Until now I have used a cool modded asi224mc.

What is particular interesting about this method is that it beats light pollution, and is less effected by seeing. Stars have a tendency to “explode”, which is an effect of scope diffraction limitations. Though The diffraction limitation is to be dealt with by dither/drizzle convolution. The result is a suburban photo technique for galaxies, less affected by seeing and light pollution.

The challenge: to find a low noise camera that is sensitive in NIR, since drizzle convolution is dependent of statistical spread between adjacent pixels.

I’m writing an article on the topic, and will be very interested in doing some initial testing, if the IMX482 is indeed NIR sensitive.

Best regards

Thomas Gera Jensen

Support@ZWO

Hi Thomas，

Thanks for your feedback, I have conveyed your message to our devs and we will consider this.

tjn

Thanks

Please let me know if I can be of any assistance. I currently use a cool modded asi224mc as test camera and the zwo850nm pass filter, for NIR shot of galaxies. Even though it is 3 times weaker in NIR sens. Than I expect the imx482 to be, results are great.

Best regards

Thomas

Support@ZWO

Thank you for your support.:)

Relativist

I’m very interested in knowing if this camera is upcoming. I do 100% EAA and the sensitivity of this sensor would be good to have.

Relativist

Any update?

elpajare

Could someone explain to me the difference of measuring sensitivity on Sony chips?
In some sheets I see that they measure it in mV and in others in Digit
What is the equivalence?

lateralus35

Also looking forward for a cooled camera with this sensor. Any update would be greatly appreciated.

xs_man

Hi Sam,
Hi Catherine,
Hi all,

More than one year has passed and as the other persons of this topic,
I wonder if an ASI camera with IMX482 will be released before end of
2021 ?

As lateralus35, I would like a cooled version.

Please keep us informed.

Many thanks.
Best.

Albéric

xs_man

Hi Sam,
Hi ZWO Team,
Hi all,

I see on the ZWO Facebook that 2 new cameras are "on the fire" a ZWO ASI 485
and a ZWO ASI 482. I don't have Facebook (and I will never use it..) so I post
my own impressions here.

Great and many thanks to the team for all your efforts !!

Now I take a breathe and take a look to the datas.

So, no surprise about ASI 485, this sensor has been integrated by other
manufacturer and the datas are very close. Read Noise with minimum near 0.7 e-,
OK great.

Now the ASI 482... I wait for it for 1 year... And frankly I was expecting
better Read Noise. Near 1 e- at minimum. Here, by extrapolation at Gain = 450/500
the value is near 1.5 e- so higher than I expect. It's not a bad value but not
what I expect. If we compare the Read Noise curve of ASI 482 and ASI 485, the curve
of ASI 482 is nearly exactly 2x the curve of ASI 485. So the logical explanation
is that IMX 482 seems to behave as a simple soft binned 2x2 of IMX 485. No
"magic Quad Effect". Sad but true at first sight.

Please, could it be possible to have a comparison of all curves (Full Well, Gain, DR,
Read Noise) between ASI 482 in bin 1x1 mode and ASI 485 in bin 2x2 mode ?

Now, do you plan a cooled version of one of thoses 2 caméras ??
The best would be an ASI 482 cool as RisingCam now have a cooled camera with IMX 485.

Many thanks for your answers and... pain.

Albéric

miopic

xs_man

If we compare the Read Noise curve of ASI 482 and ASI 485, the curve
of ASI 482 is nearly exactly 2x the curve of ASI 485. So the logical explanation
is that IMX 482 seems to behave as a simple soft binned 2x2 of IMX 485. No
"magic Quad Effect". Sad but true at first sight.

The software bin2 will result in sqrt(2) = 1.4 better signal/noise ratio. Thus, a ratio of 2 might be indicative of the hardware "quad bin effect".

EDIT. I do not know what happened to me. I was wrong. Alberic was right. One pixel of 482 is composed of 4 pixels of the 485. Thus, sqrt(4) = 2.

xs_man

After the "schock" of the new arrivals and relative déception about read noise for ASI 482,
let's have a breathe and think about the main benefit to choose this sensor.

For EEA use and Deep Sky lucky Imaging (to add color layer to a luminance layer), the
image sampling is not the main obsession. So if we compare available sensors at the
same F/D ratio (for one instrument), the most interesting parameter is the ratio between
read noise (minimum) to the surface of pixel.
Considering that the sensors have nearly similar QE :

IMX 224 / IMX 485 :
Minimum read noise : 0.7 e-
Surface of one pixel : 3.75x3.75 = 14.1 µm2
R = 0.050 e-/µm2

IMX 462 :
Minimum read noise : 0.5 e-
Surface of one pixel : 2.9x2.9 = 8.4 µm2
R = 0.059 e-/µm2

IMX 294 (Quad Bayer - color version):
Minimum read noise : 1.2 e-
Surface of one super-pixel : 2.4x2.4x4 = 23.0 µm2
R = 0.052 e-/µm2

IMX 485 in bin 1x1:
Minimum read noise : 0.7 e-
Surface of one super-pixel : 2.9x2.9 = 8.4 µm2
R = 0.083 e-/µm2

IMX 482 (Quad Bayer) :
Minimum read noise :1.5 e-
Surface of one super-pixel : : 2.9x2.9x4 = 33.6 µm2
R = 0.045 e-/µm2

We have a winner : IMX 482 !!

So even with 1.5 e- of read noise, the IMX 482 is the best sensor available reported to this
specific parameter...

So if a ZWO ASI 482 COOLED is launched in next months, I will buy it !!
Please let us know.

Many thanks.

Albéric

Tech@ZWO

xs_man So if a ZWO ASI 482 COOLED is launched in next months, I will buy it !!

Hi Albéric,
Sorry to tell you there is no cooled version for the moment.

xs_man

Thanls for your answer !
That's a bad news for me and other customers for which cooled version is
useful.

Albéric

Relativist

I agree, a cooled version would be very nice for EAA. Please let us know if you will be making one.

Legeeloo

Hi team,
Do you think it will be possible to make a driver who take acount about the quad bayer matrix of the IMX482 ?
In this case, the user could choose between 1x1 binning and 2.9ųm and binning 2x2 and 5.8ųm.
This ASI482mc would be very versatile ! No ?

lettuceman

If we're talking strictly low light sensitivity(ie snr doesnt matter), would you just look at the relative QE of each sensor to determine which one is the best? I'm curious on which one would work best for live viewing.

Oh, and also, sony has shipped out their starvis 2 sensors early this year. Has there been shipping delays? I haven't seen them being implanted in any astro products anywhere.

xs_man

I finally choose a cooled ASI 533 and buy it.
Can't wait for a cooled ASI 482 !

Albéric