ASI1600MM-Cool, odd exposure behavior around 1 second

han.k

At the Cloudynights forum an odd behavior is described. For exposures above one second the ASI1600MM camera is still recording light. Has something to do with the cutover point for camera-controlled vs. driver-controlled exposure length.

https://www.cloudynights.com/topic/547440-asi1600mm-cool-odd-exposure-behavior/

The current ASCOM driver V1.0.2.25 still show the cutover point at 1 second.

I still have an old driver V1.02.16 with a cutover point at 5 second but that one produces sometimes weird temperature control problem.

The make flats I would like to have a cutover point at 5 or better 10 seconds. Are there drivers available with an other cutover point?

Is there a better solution available?

Han

Sam@ZWO

as the camera can transfer data very fast when connect USB3.0

so we adjust the time point to 1s when connect USB3.0

it's still 5s when connect usb2.0

now we already set it to 2s and will release a new version of ASCOM soon

please try this beta one

ASICamera ASCOM Setup V1.0.3.1.zip

han.k

Sam,

Thanks for providing this driver.

I'm currently testing the driver. Yes it works well. I see clearly the behavior changing from 1.99 to 2 seconds exposure. The 1.99 sec is just enough to take some h-alpha flats but with low signal levels. As soon the exposure is 2.0 seconds the gradient comes in.

However using USB2.0 instead of 3.0, I SEE ALWAYS A GRADIENT !!, The gradient with USB 2.0 is always there for any exposure, 0.9 or 1.99 or 4.99 seconds!! It would be nice if with USB 2.0 the switch over point would be at 5 seconds but I don't see it. The same problem with drivers 1.0.2.23 and 1.02.25. The old version 1.01.16 is too buggy to use.

The driver is reported in the ASCOM setup as 6.0.3.1 instead of 1.0.3.1 as in Windows. Same for the old drivers.

Yes this driver is good, but I still would like to have a special version with the switch-over at 5 seconds, or adjustable by the user, or 5 seconds at USB 2.0. Any amp glow can be taken away with flat-darks.

Regards, Han

Sam@ZWO

han.k wrote:
Sam,

Thanks for providing this driver.

I'm currently testing the driver. Yes it works well. I see clearly the behavior changing from 1.99 to 2 seconds exposure. The 1.99 sec is just enough to take some h-alpha flats but with low signal levels. As soon the exposure is 2.0 seconds the gradient comes in.

However using USB2.0 instead of 3.0, I SEE ALWAYS A GRADIENT !!, The gradient with USB 2.0 is always there for any exposure, 0.9 or 1.99 or 4.99 seconds!! It would be nice if with USB 2.0 the switch over point would be at 5 seconds but I don't see it. The same problem with drivers 1.0.2.23 and 1.02.25. The old version 1.01.16 is too buggy to use.

The driver is reported in the ASCOM setup as 6.0.3.1 instead of 1.0.3.1 as in Windows. Same for the old drivers.

Yes this driver is good, but I still would like to have a special version with the switch-over at 5 seconds, or adjustable by the user, or 5 seconds at USB 2.0. Any amp glow can be taken away with flat-darks.

Regards, Han

the switch point should be 5s with USB2.0, we will check

chad

Hi hank，

I have tried to take the image which you mentioned here。

https://www.cloudynights.com/topic/547440-asi1600mm-cool-odd-exposure-behavior/

I noticed that the pixel value in your image is small.

But in my image, the pixel value is still much larger than yours, even if I used the lowest setting.

So I want to know your setting and how you shot it.

Thanks

chad

han.k

Hello chad,

I guess you referring to the images of user mikefulb. They are not mine. Yes the values are very low.

For my ASI1600MM-cool, I measure a bias value of around 275 in a scale of 0-64k using binx2. After 200 seconds exposure at -15 degrees Celsius it is maybe 330. I only use unity gain settings and binx2.

For flats I try to reach a value of 20k to 30k at binx2. With a h-alpha filter in place, I need about 8 seconds exposure to reach these values using my EL-panel as light source.

Regards, Han

chad

Hello Han，

I tried it, but I did not get a gradient .

So can you send me your problematic photo？

Thanks

chad

han.k

Hello Chad,

I will make tomorrow new exposures and send them to you.

To see the effect, you need a flat light source in front of the camera. At 1.99 seconds the exposure readout is flat. All pixels have more or less the same value because they are equally exposed by light.

At 2 seconds a different read-out mode is used. The last pixels read-out will have an longer exposure then the pixels of the first read out. At 2 second or longer the camera can't stop pixels recording light photons. So at 2 seconds exposure, one side of the sensor will be exposed at 2 seconds of light and the other side maybe at 2.5 seconds. The effect become less visual at longer exposures since the different will be smaller, eg. 100 second and 100.5 seconds.

It would be ideal if this switch point of 2 seconds can be set by the user in the driver. I think also the default setting of the gain and bias should be at unity gain but that is an other discussion.

The switch point using a USB2.0 connection was selected to be 5 seconds since it takes more time to readout the camera so the effect should be stronger, but I don't see it in the latest drivers.

Regards, Han

chad

Hi Han.k，

Waiting for your photos。

Based on the sensor's datasheet, it supports the rolling shutter mode and the global shutter mode. At first, I wanted to use the rolling shutter in long exposure mode. However, it's not stable, sometimes there will be a wrong picture. So I changed it to the gloabl shutter in long exposure mode.

In the global shutter mode, a gradient is possible。As we know, the severity of this problem is related to the readout speed and exposure time. So I changed the switch time from 2s to 5s in usb2.0, and keep it at 2s in usb3.0. we test it and it works well.

As far, it is not easy for us to add a API for user to control the switch time. Because some kinds of sensor have limits for the switch time. A wrong switch time will cause a lot of unknown problems.

Regards

chad

han.k

Hello Chad,

Attached an FITS file with exposure of 2.00 seconds at binx4, -15 degrees C.

Below also a JPG image. The difference between top and bottom 2 sec* (23500-21800)/21800=> 0.15 seconds is clearly visible. At next post I will attach an 1.999 seconds exposure.

Is it possible to have simple checkbox in the driver to force the switch point from 2 seconds to 5 seconds? A 5 seconds exposure will allow many users in making good flats.

Regards, Han

UV 2.00 sec binx4.jpg

han.k

And here at FITS attached at 1.999 seconds exposure under the same conditions. Flat is normal and symmetrical as it should be.

UV 1.999 sec binx4.jpg

Han

chad

Hi Han.k,

Is the image in the attachment taken with USB3.0? or USB2.0?

Regards

chad

han.k

Both images are taken with USB3.0.

I see always gradient with USB2.0. Do you need a USB2.0 image?

Han

chad

In USB3.0, the switch time now is 2s. So the difference between 1.99s and 2s is normal。Do you think this switch time 2s is enough? Should I change it to 5s or longer?

In USB2.0, the switch time is 5s, there should be no difference between 1.99s and 2s, except 4.99s and 5s. Is it better if I extend the switch time longer than 5s?

Regards

chad

han.k

Hello Chad,

I did an other test and I confirm the switch point for USB2.0 is at 5 seconds and working! Below a screenshot of the gradient at 5 seconds using USB2.0. It is severe, most likely while USB2.0 is not working at optimum speed. With this gradient imaging is not workable. Some pixels are 1.6 seconds longer exposed!

Assuming a gradient of 1% is acceptable, the switch point for USB2.0 should be at 106 seconds. The amp glow is not a problem since dark subtraction will solve it. At 4.99 seconds the gradient by amp glow is at value 378 against an average pixel value of 308, so still minor.

The switch point will be always debatable and depending on the setup of the user. So I think you should allow the user to select one of a number of fixed values E.g. 1, 2, 5, 10, 15, 20, 30 (50, 100?) seconds. Default at 2 seconds. See below a proposed modified menu:

Regards, Han

Menu proposal:

menu proposal.png

USB2.0 gradient

cls 5.0 sec, binx4.jpg

chad

Hi han.k,

Great ideas!

But it is a great changing for us. So we will have a discussion about this issue.

I will keep you update if any further information。

Best Regards

chad

Hi Han.k，

I have good news and bads news , the good news is that I solved the problem.

While the bad news is it only works for the latest product---ASI1600MM Pro, there will be no more gradient issue， for the previsous 1600 cooled cameras,Due to the limitated design of software and hardware design, it does make worse image quality with this method, we decide to only make improvements in our next generation of our driver insted of current version, pelse keep an eye on our latest products and new drivers.

Thanks

chad

Hi han，

We decided not to use your method, not because it was not good. It's our current code framework that determines that we can't use this method.

Because this change affects all of our products and causes problems for other products. Also, this change is not a perfect solution, as I said before, the hardware must be changed to solve the problem.

If you really need it, I suggest you contact our customer service after our 1600MC/MM-Pro products come out.

Thanks

chad