Zwo asi 294 is now 8340*5644 Pixels

[Andycayman]

Hello

Anyone seen this before. My almost new asi294 has gone from 11mega pixels to 43mega pixels. obviously the onboard binning has failed. Guess i am wondering if i should just Binn in spg. Is there any difference to on camera or software binning.  Still under Guarantee  and is only two or three months old bought from FLO .

any advice would be appreciated

Andy

[vlaiv]

I would say that preferred way of binning cmos is in software rather than in drivers / on chip, so you won't loose anything by binning in software after calibration and debayering.

In fact, because this is OSC camera - first debayer without interpolation (super pixel mode) and this will give you already twice reduced sampling rate and then decide if you want to bin additionally.

Binning interpolated pixels actually does nothing as both binning and interpolation are sort of average value - and averaging the averaged value won't change anything.

[Andycayman]

This is the mono 294 does that make any difference

 

Edited December 1, 2020 by Andycayman

[vlaiv]

6 minutes ago, Andycayman said:

This is the mono 294 does that make any difference

 

Oh it does. You can bin it in software without any issues - bin by x2, x3, ... what ever you like.

Only problem is with debayering and binning as I described briefly above - it is only important for OSC sensors (and I thought this is OSC camera).

Binning in software still stands as better option.

 

[Andycayman]

this camera is supposed to be very quantum efficient and very low nose after 120 gain any idea how that would be effected by the lack of on board binning

 

 

Andy 

[vlaiv]

1 minute ago, Andycayman said:

this camera is supposed to be very quantum efficient and very low nose after 120 gain any idea how that would be effected by the lack of on board binning

 

 

Andy 

I think it will make absolutely no difference in normal operation mode. It could be that this is just glitch in drivers as drivers are expecting OSC version and getting Mono version or something.

Maybe new version of drivers will sort out this issue.

It could also be that firmware for camera has binning for OSC mode somehow although sensor is mono.

In any case - can you shoot normally at regular resolution and get proper image (no binning - just regular shooting)? If yes, then I suspect you don't need to worry about any lack of internal binning.

[Andycayman]

Cheers for your input

[tooth_dr]

51 minutes ago, Andycayman said:

Hello

Anyone seen this before. My almost new asi294 has gone from 11mega pixels to 43mega pixels. obviously the onboard binning has failed. Guess i am wondering if i should just Binn in spg. Is there any difference to on camera or software binning.  Still under Guarantee  and is only two or three months old bought from FLO .

any advice would be appreciated

Andy

I have been browsing the QHY website and their 294M model has a feature called called extended pixel mode:

”The original 294M chip, the SONY IMX492, is an 11.7-megapixel chip.When the extended pixel mode is enabled, the original single pixel can be split into four pixels to about 46.8 million pixels.The process is similar to a reverse bin”

Have you enabled this?

[vlaiv]

5 minutes ago, tooth_dr said:

I have been browsing the QHY website and their 294M model has a feature called called extended pixel mode:

”The original 294M chip, the SONY IMX492, is an 11.7-megapixel chip.When the extended pixel mode is enabled, the original single pixel can be split into four pixels to about 46.8 million pixels.The process is similar to a reverse bin”

Have you enabled this?

I'm not even going to ask why would one want to do this

[Taman]

1 hour ago, Andycayman said:

Hello

Anyone seen this before. My almost new asi294 has gone from 11mega pixels to 43mega pixels. obviously the onboard binning has failed. Guess i am wondering if i should just Binn in spg. Is there any difference to on camera or software binning.  Still under Guarantee  and is only two or three months old bought from FLO .

any advice would be appreciated

Andy

I spotted this yesterday, ZWO have just released a new driver update that unlocks the binning on this camera.

Have you just downloaded the latest update? If so, you should be able to now choose between the two modes. I have the same camera, but not updated it yet.

Tony.

Edited December 1, 2020 by Taman

[matt_baker]

1 minute ago, Taman said:

I spotted this yesterday, ZWO have just released a new driver update that unlocks the binning on this camera.

Have you just downloaded the latest update? If so, you should be able to now choose between the two modes.

Tony.

How does that work? Was the camera 43MP to begin with but was binned down to 11 by ZWO?

[Taman]

Just now, matt_baker said:

How does that work? Was the camera 43MP to begin with but was binned down to 11 by ZWO?

The sensor is 43MP and ZWO binned it to get 11MP. The new update unlocks the full resolution of the camera. There should be an option to switch between the two.

[matt_baker]

3 minutes ago, Taman said:

The sensor is 43MP and ZWO binned it to get 11MP. The new update unlocks the full resolution of the camera. There should be an option to switch between the two.

Why would they bin it in the first place?

[knobby]

Improved sensitivity ?

[vlaiv]

3 minutes ago, matt_baker said:

Why would they bin it in the first place?

Because they used ASI294 name for it for some strange reason

Color camera uses IMX294 color sensor.

Mono camera uses different sensor all together - IMX492.

It looks like it all started with IMX294. This sensor has very strange structure - it has double bayer matrix and this is used in special mode:

It actually has four times more pixels - that you can't access as individual pixels. Sensor either does normal mode where it sums 2x2 groups of pixels or does HDR mode where there are pixels that do short integration and pixels that do long integration - and they are summed on camera to produce HDR image.

It looks like IMX492 kept pixel sizes but skipped on that HDR thing and you can access them in single and "double" mode (one that we would call normal mode).

I'm not sure what is better thing to do. I would say go for single pixels, but that will create massive files for storage and processing and I'm not sure if you'll get any benefit as pixels will be tiny at ~2.3µm.

 

[DaveS]

Might be worth doing for a short focus 'scope or camera lens.

[vlaiv]

28 minutes ago, DaveS said:

Might be worth doing for a short focus 'scope or camera lens.

I think that 2.3µm pixel size is going to be useful only in these two cases:

- planetary imaging with ROI - single pixel should have low read noise - about half of that of "whole" pixel (as whole pixel is just summed 4 small single pixels)

- possibility to do x1.5 bin in relation to regular pixel size without pixel to pixel correlation. When you download 2.3µm image - you can do 3x3 bin and that will effectively give you ~7µm pixel size - not something you can get from 4.63µm with integer binning

With short focus scope or camera lens - that pixel size will over sample a lot.

Let's say you want to do higher resolution image at 1.5"/px. With 2.3µm pixel size, you need 316mm of focal length.

Imagine that you have rather fast small telescope at F/4.5 and it has 316mm of focal length - how large aperture is that? That is 70mm of aperture.

Calculated size of airy disk for 70mm of aperture is 3.67", so you see, it does not really make sense to go with 1.5"/px as airy disk size itself is guarantee that you won't resolve enough to go with 1.5"/px

You'll then say - ok, let's go with more reasonable 2"/px, but then, needed focal length will be ~237mm, and even if you have F/4 scope - you'll have small aperture at about 60mm or less to get that focal length. Guess what? Airy disk size now jumps to 4.28" - again too large for selected sampling rate ...

In any case, you are over sampling. With lenses, things are just worse than with diffraction limited scopes as lenses are not diffraction limited and star blur is larger than airy disk because of that (often much larger) - lens often require pixel sizes to be around 10µm or larger in order to get properly sampled sharp images.

 

[Alien 13]

This camera sounds perfect for camera lenses especially when you get down to the short f/l Samyangs.

Alan

[DaveS]

This is what I get if I put my TS 80mm f/4.4 and 2.3 micron pixels into Astronomy Tools

[vlaiv]

2 minutes ago, DaveS said:

This is what I get if I put my TS 80mm f/4.4 and 2.3 micron pixels into Astronomy Tools

I know, however, that calculator is simply flawed.

You can check this yourself, since you have 80mm F/4.4 scope. Take any of the subs you taken with that scope and measure FWHM in arc seconds. You will find that most of them are above 3" FWHM.

Above calculator assumes that seeing is all there is to star FWHM - but in reality it is only one of couple of components that impact resulting FWHM of stars. That is first problem with said calculator.

Second problem is conversion from FWHM to sampling rate. There is rather simple conversion between the two - factor of x1.6. Divide FWHM of your star with x1.6 and that gives you ideal sampling rate for that level of blur.

Let's say that you really have 2" seeing, and you have excellent mount that is capable of 0.3" total RMS error, with 80mm scope your resulting FWHM will be 2.54" and hence ideal sampling rate will be ~ 1.58"/px.

Even in best case scenario (mind you, 0.3" total RMS is premium mount territory - not something you can do with HEQ5 or EQ6 and more expensive mounts will often struggle), optimum sampling rate is far away from lower bound of 0.67"/px suggested by above tool.

More realistic scenario of 2" seeing, 1" total RMS guide error and 80mm scope will give you ~3.4" FWHM or 2.125"/px sampling rate.

[DaveS]

I'll have to have a dig through some old images as I haven't used the 80mm for a while now. I may have something with the ASI 183. The combination would have been riding piggyback on the 130mm apo, on the DDM 60, certainly capable of better than 0.35", on occasion below 0.2"

[vlaiv]

4 minutes ago, DaveS said:

I'll have to have a dig through some old images as I haven't used the 80mm for a while now. I may have something with the ASI 183. The combination would have been riding piggyback on the 130mm apo, on the DDM 60, certainly capable of better than 0.35", on occasion below 0.2"

I know that your mount is certainly capable of ~0.2-0.3" RMS values - but let's face it, only couple of mounts available to amateur community can do that (Mesu, 10micron, etc ...) and even that requires very good seeing with 80mm scope to be able to use resolutions of ~1.6"/px.

It is far easier to achieve those resolutions with 6" scope as Airy disk size has much smaller impact on total FWHM with larger scope.

[Adam J]

3 hours ago, Andycayman said:

Hello

Anyone seen this before. My almost new asi294 has gone from 11mega pixels to 43mega pixels. obviously the onboard binning has failed. Guess i am wondering if i should just Binn in spg. Is there any difference to on camera or software binning.  Still under Guarantee  and is only two or three months old bought from FLO .

any advice would be appreciated

Andy

You don't say but this means something different depending on if it's the mono or OSC version. The original OSC does operate in a different HDR mode at low gain. What gain are you using? 

[wimvb]

10 hours ago, vlaiv said:

It could be that this is just glitch in drivers as drivers are expecting OSC version and getting Mono version or something.

It’s not a glitch, it’s a feature.

I would avoid shooting at full resolution for dso. My latest image used 344 90 s exposures, for L. At 22Mbyte each, that pit a sizeable dent in my disk storage space. The full res image would have abt 90 Mbyte per image, close to 30 GB for those 344 subs, at no gain.

10 hours ago, vlaiv said:

looks like it all started with IMX294. This sensor has very strange structure - it has double bayer matrix and this is used in special mode:

It actually has four times more pixels - that you can't access as individual pixels. Sensor either does normal mode where it sums 2x2 groups of pixels or does HDR mode where there are pixels that do short integration and pixels that do long integration - and they are summed on camera to produce HDR image.

This can possibly be very bad news for us, since it means that the sensor likely does on chip image manipulation. And that can interfere with image calibration. Imo, smart sensors are NOT good news for astrophotography.

[michael.h.f.wilkinson]

The small pixels might find their use in RASAs