Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

ToupTek ATR 585M, and more generally, what are the rules for HCG and offset?

  • Please log in to reply
6 replies to this topic

#1 Phil49

Phil49

    Mariner 2

  • -----
  • topic starter
  • Posts: 257
  • Joined: 14 Mar 2021

Posted 24 May 2025 - 03:20 AM

Hi all,

 

I've got the opportunity to buy the early bird package of the ToupTek ATR 585 + AFW-M + 7 x 1.25" filters (LRGBSHO).

 

This is my first mono camera, and after 5/6 years practicing astrophotgraphy, I'm no longer a newbie but not yet an expert...

 

Usually I let the unity gain for my OSC cameras (see my signature for current equipment) but the ATR 585 M camera is brand new and I've got the opportunity to shoot 3 targets up to now, LDN 1251, vdb 149 and currently IC1396.

 

I let the 120 gain and 50 offset, which were the default values.

 

I've came across a few posts where some people were using a gain of 100 and a higher offset, around 80/85.

 

When I asked to some how they were setting the offset value, usually I had no replies...

 

ToupTek ATR585M HCG curves.jpg

 

I can understand the notions of read noise, dynamic range, full well, but how do you optimize the offset?

 

I imagine that values are perhaps also linked to targets?

 

You wouldn't necessarily use the same values for a high-contrast galaxy like M41, a nebula like NGC 7000, a globular cluster or a dark nebula like LDN 1251, which was my first object with this mono camera!?

 

The exposure time must also be part of the equation, I imagine. I usually select 180 sec, but I've decided to test 120 sec for my last objects.

 

Thanks for your help

 

Philippe

 



#2 Tapio

Tapio

    Voyager 1

  • -----
  • Posts: 12,429
  • Joined: 24 Sep 2006
  • Loc: Tampere, Finland

Posted 24 May 2025 - 03:25 AM

I remember watching Cuiv's video of the subject.

I suggest you do the same.

https://www.youtube....h?v=mCYQoVSuEaI



#3 kevinkiller

kevinkiller

    Vostok 1

  • -----
  • Posts: 102
  • Joined: 24 Aug 2013
  • Loc: Austin TX

Posted 24 May 2025 - 05:49 AM

To determine proper offset look at the image statistics of a bias frame.

You want to have the smallest offset that produces no (or very few) pixels with 0 values.

 

The median value of the bias frame will be your offset in ADUs.

 

You can use that along with your read-noise to figure out the target median value to shoot for during light frames to "swamp the read noise" by whatever factor you like:  3x, 5x, 10x.



#4 Phil49

Phil49

    Mariner 2

  • -----
  • topic starter
  • Posts: 257
  • Joined: 14 Mar 2021

Posted 07 June 2025 - 11:41 AM

I remember watching Cuiv's video of the subject.

I suggest you do the same.

https://www.youtube....h?v=mCYQoVSuEaI

Hello,

 

Thank you very much, this is exatcly what I was looking for! smile.gif

 

That's too bad because I had already viewed this video but I had completely forgotten it because I rarely change these settings.

 

I've seen it again, and Cuiv's explanation and demonstration and illustration on a way to check-out, and more importantly adjut the offset, is great as usual!

 

I've changed my default gain that was set to 120 out of the box to 100, as well as the offset from 50 to 460 apparently to get a result similar to Cuiv's.

 

I had also left the "Low conversion gain" by default, and the offset was less, close to 200 if I remember, and way higher wth "High conversion gain" at 460.

 

Does anybody know which mode is advised between low and High conversion gain?

Does it depend on the object targeted, galaxy, nebula, or dark nebula?

 

 

Kind regards

 

Philippe



#5 NuovaApe

NuovaApe

    Ranger 4

  • -----
  • Posts: 352
  • Joined: 07 Aug 2021
  • Loc: England

Posted 08 June 2025 - 11:01 PM

Some days I miss going into a photography shop, asking for ISO100 or ISO800 camera film and knowing exactly what they were and why I need them.

 

The bigger the number, the faster the film. Simple.

 

I've yet to see one good explanation of Sony low vs high gain.

 

google "sony imx high conversion gain"

This technology amplifies electrical signals immediately after light is converted to a signal, reducing noise. It achieves this by using a floating diffusion with lower capacitance, resulting in a higher voltage and increased sensitivity, even with a small amount of charge from a low number of incident photons.

 

google "sony imx low conversion gain"

 

It's essentially the opposite of high conversion gain (HCG), which amplifies the signal to improve image quality in low light

 

ARRGGHH! 

 

But further down the LCG info

 

LCG prioritizes minimizing noise in bright conditions, which can be a concern when the signal is strong.

 

"When the signal is strong" - something you DSO dudes and dudettes create by stacking lots of subs.

I only do solar these days, and I'm forever messing with gain/exposure sliders like a church pipe organ player banging out Bachs Toccata.

 

I would advise "less is more".

High gain is great for realtime video feeds. Our eyes can't boost the signal like electronics can.

But sometimes "more is more".

 

"Offset" can help defeat what is known as "black crush" in the LCD/OLED telly world.

 

Best, Ed.



#6 prookyon

prookyon

    Explorer 1

  • -----
  • Posts: 91
  • Joined: 26 Jul 2024

Posted 10 June 2025 - 02:03 AM

If you are stacking DSO images then High Conversion Gain (HCG) is better.

Basically it was an improvement that was added to Sony sensors at some point. Astrophotography (the stacking of many images kind) also happens to benefit from it. By using LCG you are just going back to the days before such technology existed. There are (much) longer explanations elsewhere on the internet (obviously with some conflicting opinions).

 

Offset - as was already said the goal is to just use lowest setting that avoids 0 pixel values.

 

Gain - Usually there is no need to increase it from the lowest value. But HCG only works correctly starting from some value. I use SharpCap and it does the whole gain/LCG/HCG thing very differently for IMX585 so I have no idea what the correct setting for NINA is.



#7 Phil49

Phil49

    Mariner 2

  • -----
  • topic starter
  • Posts: 257
  • Joined: 14 Mar 2021

Posted 10 June 2025 - 11:04 AM

Hello,

 

I eventually asked to ToupTek support, and they kindly replied to my questions :

 

About Image Capture Modes (LCG, HCG, HDR):

LCG (Low Conversion Gain) is designed for maximum dynamic range, suitable for very bright targets or broadband filters under dark skies.
HCG (High Conversion Gain) is ideal for general deep-sky imaging, as it offers a lower read noise while preserving a good dynamic range — we recommend HCG as the default mode for LRGB and narrowband imaging.
HDR (High Dynamic Range) combines both gain curves to extend signal depth — this mode is still under active refinement and may behave differently depending on your camera’s firmware version. We recommend using HCG unless you specifically want to test HDR behavior.

 

Therefore, I consider that HCG is correct, this is the mode I've selected for my new serie of sessions.

 

Regarding gain, of course, I changed from default 120 to 100, and for the offset, I followed the explanation provided by Cuiv The Lzay Geek in his excellent Youtube video mentioned by Tapio :

 

https://www.youtube....h?v=mCYQoVSuEaI  

 

Checking with an exposure time of 0.001 sec, I checked the behaviour of the curve in NINA's Statistics, and had to set it at 420 to be in the same conditions as Cuiv.

 

I've started a serie as said above, with those settings, but visually, I already noticed on a photo taken with those settings that it looks quite different.

 

I'll be able to confirm when I finish my serie.

 

Best regards

 

Philippe


  • Elminster likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics