ToupTek SkyEye62AM - Sony IMX455 Sensor v4 - settings

[Peter_S]

Hi Robin, hi forum,
for my Touptek SkyEye62AM (Noctutec MILAN) with the sensor v4 am trying to get data for best settings (deep sky).
As already discussed in earlier threads the ToupTek has 2 settings (LCG / HCG) which I´d like to test separately.
When choosing the ASCOM driver SharpCap crashes. Taking the native driver one gets mixed data as you described (auto switch).
ToupTek has published sensor data (of the previous version of the sensor) which show the difference in read noise:

Skyeye62AM_curves HCG.jpg (123.67 KiB) Viewed 5116 times

Skyeye62AM_HCG.jpg (67.33 KiB) Viewed 5116 times

Skyeye62AM_curves LCG02.jpg (127.9 KiB) Viewed 5116 times

Skyeye62AM_curves LCG.jpg (67.4 KiB) Viewed 5116 times

There are mainly two situations, as Stathis wrote in another thread:

- aiming for high dynamic range
- aiming for low noise (faint signals)


I image with APT and have to set either LCG or HCG (no auto switch); my exposures are 240 - 300 sec; 3 nm NB filters.
My data attached, made with a L filter.

SkyEye62AM~MONO16 (Ultra Low Noise).json

(15.26 KiB) Downloaded 93 times

2024 03 24 SkyEye62Am sensor data v2.txt

(1.68 KiB) Downloaded 547 times

Until now I couldn´t test the sky brightness due to the bad weather.

Interpreting Touptek´s graphs I´d take LCG mode in both situations due to the almost equal dynamic range but the better read noise (?)
Do you have any suggestions concerning the settings?


thanks, Peter

[admin]

Hi Peter,

the thing that I think is missing in your analysis of whether to use LCG or HCG is a consideration of the effect that the different read noise values on sub-exposure lengths...

If you pick gain 100 in LCG and compare to gain 100 in HCG (both with low noise option), the read noise values are 2.72 and 1.95 respectively. The sub exposure length needed to properly expose for deep sky is proportional to the square of the read noise, so using LCG you will need to use sub exposures twice as long as for HCG. As an example, HCG might need 60s exposures and LCG might need 120s.

That means that in an hour of imaging you would get 60 HCG subs and only 30 LCG subs. While the individual subs in LCG have more dynamic range, the fact that you have twice as many HCG subs out of the same amount of imaging time will - I think - at least compensate for this and quite likely leave HCG slightly ahead in terms of dynamic range of the final stack (the SharpCap smart histogram tool will work this out exactly). The other advantages to the shorter subs in HCG is less pressure on the guiding/tracking and less loss if an individual sub is spoiled by satellite trails, etc.

As the read noise difference between HCG and LCG is not as dramatic in this case as for some sensors, the choice of HCG is less clear cut, but still an option very much worth considering.

cheers,

Robin

[Peter_S]

Hi Robin,
thank you very much for your answer (on sunday) - now it´s clear...
as soon as I can measure the sky brightness I´ll post an update.
Peter

[Joerg@skyRefl]

Hi,

the current Touptek SDK was recently implemented in sharpcap.
There nevertheless are only 2 drivers for 16bit currently. These do not cause any sensible results at the sensor analysis.

SkyEye62AM~MONO16
SkyEye62AM~MONO16 (Ultra Low Noise)

Are there no correct driver configurations for the camera at the moment currently?
Other software providers have already converted this.

HCG Mode
HCG Mode (ultra low noise)
LCG Mode
LCG Mode (ultra low noise)

Joerg

[admin]

Hi Joerg,

expected behaviour for this camera is that the LCG/HCG switch is merged into the gain control, meaning that a sensor analysis will produce a graph with a cliff edge drop in it at gain 300 (similar to the ones you see on the product page for the ZWO 6200 camera - that model also merges the LCG/HCG into the gain control).

You may be able to force the camera into either LCG or HCG mode and analyse that by use of the appropriate ASCOM driver from the camera manufacturer.

cheers,

Robin

[Joerg@skyRefl]

Hi Robin,

in principle, I only would like to work with native camera drivers.
The separation to HGC and LGC is carried out for example in the software N.I.N.A. with this drivers (toupcam.dll).

With the both Sharcap 16bit drivers no-one becomes produces a cliff edge.
The ReadoutNoise in the sensor analysis is not a Low Noise in the two cases.

MONO16 (Ultra Low Noise) 4.29e for GAIN 100 and FW 106k to 2.47e for GAIN 10000 and FW 1.6k
MONO16 5.52e for GAIN 100 and FW 107k to 2.73e for GAIN 10000 and FW 1.6k

Joerg

[admin]

Hi,

that's interesting - those do look like HCG figures for the analysis.

Are you using the very latest version of SharpCap 4.1 from our downloads page? Worth checking in case you have an older version and this issue is already resolved in the latest...

If you are on the latest version you should see one of two effects depending on whether SharpCap recognizes your camera sensor as being an IMX455 or not

1) Sensor recognized via the camera model name - the gain control will switch LCG/HCG as I described above. The gain control will have a larger range than in other applications like NINA because of this

2) Sensor not recognized - there should be a 'Read Mode' control among the controls on the right that lets you select LCG or HCG manually and the gain would have the normal range seen in NINA

It sounds like you are getting something that isn't quite either of these options though - almost as though SharpCap is not detecting the camera having the HCG/LCG option at all. If this *is* the case, then we can confirm by doing the following

* ) Tick the option to log extra info about camera hardware communications in the SharpCap settings logging page
* ) Open the camera
* ) Close it again pretty much straight away (otherwise the log will be huge)
* ) Send me the full log, which I can check. A quick search in the log text for 'FLAG_CG' will reveal if the camera is reporting a conversion gain option or not

thanks,

Robin