ZWO/Player One IMX 585 Sensor Image Showcase

[Chris]

Here is the AVI of Saturn - Evostar 120 achromat, 2.5x Barlow and ASI 585mc working at 2500mm effective focal length:

Obviously Saturn is going to be dimmer compared to Jupiter and the Moon so you need to up the exposure which effects the frame rate.

 Will also test frame rates with the Moon and Jupiter and make a YT vid about this. I'm currently working on a Vid for the HCG comparison between Gain 150 and 252 so maybe after that

 

 

 

2022-08-20-2219_5-CapObj_pipp.avi

Edited September 25, 2022 by Chris

[Adam J]

1 hour ago, Chris said:

I was surprised also, however this was one night so far so I wont set this in stone until I've done some further testing

Need to also hear from others with the ASI 585 as to their experience. 

 

The issue is that:

640x480 = 307k 

3856 X 2180 = 8.4million

So one 27th of the data rate cropped than at full resolution. 

So if it was getting 10Hz cropped it would take about 3seconds for the image to download at full resolution. That's CCD territory, so something is wrong. 

Adam 

[bottletopburly]

3 hours ago, Chris said:

I was surprised also, however this was one night so far so I wont set this in stone until I've done some further testing

Need to also hear from others with the ASI 585 as to their experience. 

 

I did an initial test on mine the other week before sharcap crashed and fps was good mind all my system is usb3 and ssd  drive since updated ascom driver to latest ascom driver so hopefully next outing will be good to test 

[Dark Raven]

12 hours ago, Chris said:

I'm currently working on a Vid for the HCG comparison between Gain 150 and 252

Oh look, some bloke made a video about the same exact thing. What are the odds? 😂

Joke aside, great video Chris. 👍

[CraigT82]

Wouldn’t you need to test one of the early cameras with HGC at 150 against one of the newer cameras with the HGC at 252? 
 

Or is it the case that the cameras themselves haven’t changed it was just the specs on paper that were wrong initially? 
 

ZWOs answer about the change is vague doesn’t really clarify anything, which makes me think they really messed up! 

[Ratlet]

2 hours ago, Dark Raven said:

Oh look, some bloke made a video about the same exact thing. What are the odds? 😂

Joke aside, great video Chris. 👍

Like the idea for the background.  Looks really good.

[PeterC65]

10 hours ago, CraigT82 said:

Wouldn’t you need to test one of the early cameras with HGC at 150 against one of the newer cameras with the HGC at 252? 
 

Or is it the case that the cameras themselves haven’t changed it was just the specs on paper that were wrong initially? 
 

ZWOs answer about the change is vague doesn’t really clarify anything, which makes me think they really messed up! 

As I understand it, the LCG / HCG switching point is not set by Sony or by the sensor, but buy the camera designer. You can see this with the IMX585 where the switching point is different for the ASI585MC and the Uranus-C. It is very unlikely that ZWO would design a camera with a 252 switching point and then publish a spec curve that shows at at 150. Much more likely that the camera design has changed.

@Chris, have you carried out a SharpCap sensor analysis for your ASI585MC and if so, what does it show? It would be interesting to know what the switching point (from SharpCap) is for someone with an early release of the camera.

While the M31 and M33 image comparison is interesting I'm not sure that it proves anything. The higher gain would be expected to produce a brighter image whatever the LCG / HCG switching point.

What I don't understand is the point of LCG mode at all. As @Chris has mentioned, he will now always use the camera at 252, so doesn't this mean that all gain settings below the LCG / HCG switching point are of no use? If so, then camera designer should set the switching point to the lowest gain possible. I can see no reason not to operate an astro camera in HCG mode at all gain settings. The switch makes no difference to full well depth or to e/ADC gain, and it increases dynamic range and reduces read noise. I have posed this question here and after much technical discussion it still isn't clear to me why astro camera use LCG mode. The best explanation seems to be that some read noise is needed to mask quantisation noise at high gains, but this rationale would result in much higher gain LCG / HCG switching points than we see.

While I've been typing UPS have finally delivered my new Uranus-C camera. It's looking cloudy here for the next few days (what a surprise!), but I will post the SharpCap sensor analysis later.

[PeterC65]

Here is the SharpCap sensor analysis for the Uranus-C (RAW16).

As expected, the LCG / HCG switching point is at 180 (18dB). The maximum gain that it is possible to set in SharpCap is 800 (80dB) so most of this range is in HCG mode. The minimum read noise is mostly 0.8e which is the same as for the ASI585MC, but it does drop below 0.7e at maximum gain.

[PeterC65]

I've been trying the new Uranus-C camera in daylight. The image is looking very promising compared with my GPCAM2 327C, a wider field of view but also clearer and with better control over the camera.

There has been some discussion about frame rate with the ASI585MC. With the Uranus-C at maximum resolution (3856 x 2180) I'm getting 2.6fps. At 1024 x 768 I'm getting 28.2fps. At 640 x 480 I'm getting 63.7fps. That's with a brand new Dell i7 laptop and USB3.0. I do have a USB3.0 router in the way which makes no difference except at 640 x 480 when I'm getting 72.2fps with the camera connected directly to the laptop.

[Magnum]

14 minutes ago, PeterC65 said:

I've been trying the new Uranus-C camera in daylight. The image is looking very promising compared with my GPCAM2 327C, a wider field of view but also clearer and with better control over the camera.

There has been some discussion about frame rate with the ASI585MC. With the Uranus-C at maximum resolution (3856 x 2180) I'm getting 2.6fps. At 1024 x 768 I'm getting 28.2fps. At 640 x 480 I'm getting 63.7fps. That's with a brand new Dell i7 laptop and USB3.0. I do have a USB3.0 router in the way which makes no difference except at 640 x 480 when I'm getting 72.2fps with the camera connected directly to the laptop.

try adjusting the usb traffic slider, that can make a huge difference. default on my QHY462C is 40 but if I reduce it to 2 then I get full speed as per the specs, if I go to 0 then it goes slower again, so sweet spot is 2 with my cam and computer

 

[CraigT82]

5 hours ago, PeterC65 said:

As I understand it, the LCG / HCG switching point is not set by Sony or by the sensor, but buy the camera designer. You can see this with the IMX585 where the switching point is different for the ASI585MC and the Uranus-C. It is very unlikely that ZWO would design a camera with a 252 switching point and then publish a spec curve that shows at at 150. Much more likely that the camera design has changed.

Well there is two possibilities the way I see it:

A) ZWO released the camera with the old specs however those charts were wrong (maybe this “incorrect testing” they mentioned?) and they have quietly corrected and updated the specs and the camera hasn’t changed at all and always switched to HGC at 252. 

B ) ZWO released the camera with the HCG set at 152, then decided at some point to change this on the cameras to the new HCG point and released those into the wild, again quietly.
 

Either way what I find fishy is how quietly they changed the specs… not even FLO knew about it! They didn’t tell their customers or dealers?! You’d think that if it was scenario B and this change made the camera better in some way they’d be making a song a dance about how it’s better than their competitors models, but I haven’t seen anything like that.
 

This makes me think it might be scenario A and the lack of announcements are because they are keen to avoid a rush of complaints/returns from people who are miffed that their camera didn’t work as it was advertised. 
 

I will accept they ZWO may have made announcements/explanations as to the change of specs somewhere and I just haven’t come across them. I’m all ears if so. 

[Chris]

11 hours ago, PeterC65 said:

have you carried out a SharpCap sensor analysis for your ASI585MC and if so, what does it show? It would be interesting to know what the switching point (from SharpCap) is for someone with an early release of the camera.

I ended up using ASICap for the one session of planetary imaging so far because SharpCap wont communicate with my 585 despite updating the drivers. I need more time to problem solve this so I can get the graph with the switching point. However, I can probably show the switching point by taking an exposure at say gain 249 then 252, if there's a switching point at 252 the noise should drop even though I'm increasing the gain by only 3. If this isn't noticeable in the real world on an actual image to our eyes then perhaps we're worrying about nothing?   

11 hours ago, PeterC65 said:

While the M31 and M33 image comparison is interesting I'm not sure that it proves anything. The higher gain would be expected to produce a brighter image whatever the LCG / HCG switching point.

For example, if there wasn't a switching point you would expect noise to increase and dynamic range to decrease with the increased brightness of the image right? to my eye the noise almost looks less at 252 compared to 150, what do you think? 

The gain 252 images didn't just look uniformly brighter, the objects seems to pop more from the background. The switching point also boosts the dynamic range back up to 11 stops according to the graph. although this is more do with the colour depth (2 to the power of 11 shades) so not sure if this is why the gain 252 looks more contrasty? 

The test was simply to show why ZWO changed their written specs, from one specific switching point to another, and I can see why because the gain 252 images look noticeably better to my eye. 

I think the camera has always had the switching point at 252 and I'm speculating here, but perhaps when they first published the specs of 4 new cameras they got a graph mixed up perhaps? e.g. their new ASI 662 has a switching point of 150 maybe that one got mixed up with the 585?  

 

 

 

Edited September 26, 2022 by Chris

[Chris]

23 hours ago, CraigT82 said:

Wouldn’t you need to test one of the early cameras with HGC at 150 against one of the newer cameras with the HGC at 252? 
 

Or is it the case that the cameras themselves haven’t changed it was just the specs on paper that were wrong initially? 
 

ZWOs answer about the change is vague doesn’t really clarify anything, which makes me think they really messed up! 

I don't believe its the cameras that have changed, rather they made a mistake with the written specs. 

 

 

[Magnum]

I just noticed the UK dealers have increased all the ZWO prices because of the pound crashing, well accept Rother Valley, not sure if they will honour the old prices though. I bet the dealers dont reduce the prices when the pound recovers though 😛 

 

Edited September 27, 2022 by Magnum

[PeterC65]

13 hours ago, Chris said:

However, I can probably show the switching point by taking an exposure at say gain 249 then 252, if there's a switching point at 252 the noise should drop even though I'm increasing the gain by only 3. If this isn't noticeable in the real world on an actual image to our eyes then perhaps we're worrying about nothing?

I think that's a good idea. It would show the effect of LCG / HCG switching independently of gain. Perhaps you could take images under the same conditions at 150, 249 and 252. That would show the effect of gain alone (comparing 150 and 249) and te effect of switching alone (comparing 249 and 252).

When I get a clear night I will do something similar with my Uranus-C (I'll take images at 150, 178, 180 and 252).

13 hours ago, Chris said:

For example, if there wasn't a switching point you would expect noise to increase and dynamic range to decrease with the increased brightness of the image right? to my eye the noise almost looks less at 252 compared to 150, what do you think?

The gain 252 images didn't just look uniformly brighter, the objects seems to pop more from the background. The switching point also boosts the dynamic range back up to 11 stops according to the graph. although this is more do with the colour depth (2 to the power of 11 shades) so not sure if this is why the gain 252 looks more contrasty? 

Read noise always decreases as gain increases, even without the LCG / HCG switch, but there may be other sources of noise that do the opposite. Dynamic range decreases with increasing gain. The images at 252 certainly look better to my eye. I'm not sure I'd say there was less noise, but certainly more dynamic range (contrast). I think noise is a hard thing to spot by eye / ear in the presence of a wanted signal, unless it is extreme.

What is the maximum gain you can set for the ASI585MC?

For me the issue with the position of the LCG / HCG switching point is how much of the gain range does it make unusable. For the Uranus-C it is at 180 / 800, so just 22.5% of the gain range is in LCG mode and therefore unusable. I'm assuming here that most people set the gain to just above the LCG / HCG switching point which I believe is the standard advice.

14 hours ago, Chris said:

I think the camera has always had the switching point at 252 and I'm speculating here, but perhaps when they first published the specs of 4 new cameras they got a graph mixed up perhaps? e.g. their new ASI 662 has a switching point of 150 maybe that one got mixed up with the 585?

Unless ZWO enlighten us, I don't think we will know why the spec / camera changed. What bothers me most is that different camera manufacturers using the same sensor have made quite different design decisions. Why that is is the question that I'd like answering.

[Chris]

7 hours ago, PeterC65 said:

What is the maximum gain you can set for the ASI585MC?

I shall check this next time I'm out imaging

7 hours ago, PeterC65 said:

Unless ZWO enlighten us, I don't think we will know why the spec / camera changed. What bothers me most is that different camera manufacturers using the same sensor have made quite different design decisions. Why that is is the question that I'd like answering.

ZWO replied saying they made a mistake with their testing which implies that the HCG was always 252 but they tested this incorrectly? or mixed something up perhaps? Anyway they used the word mistake.  I have to remember that English isn't their first language so something could be lost in translation. 

I think the important thing to establish is the best setting for the camera you're using, you'll soon establish this when you do some test shots as you've outlined

 

Edited September 27, 2022 by Chris

[PeterC65]

I managed a session with the new Player One Uranus-C camera last night. Not great conditions and with hindsight the focus may have been off a little. I aways find the first session with new kit to be a little frustrating so more testing is needed, but I took some images of M33 at different gain settings and came to some initial conclusions.

Bear in mind that I do EEVA (rather than AP) and with an AZ GOTO mount, so long exposure images are not really an option. Instead, I captured the output from live stacking 20 frames, auto stretched, at 4s exposure. I tried gains of 175, 180, 250, 350, 450, 550, 650 and 750.

There was noticeable banding at 175 which went away at 180. The noise seemed to reduce significantly between 180 and 250 then reduced further but more slowly to 450 after which it rose again. I'm thinking this was the read noise diminishing and then being overwhelmed by some other source of noise. The level of detail gradually increased with gain, particularly from 450 upwards, but overall the best gain was at 450 I think.

Here is the (horrible) capture with gain at 175:

And here is all that banding gone by a gain of 180:

The best capture I think is at 450:

Then it gets very noisy as you approach 750:

Next time I get a properly clear night I will make sure the focus is spot on and carry out a more systematic test for varying gain and exposure.

These were all taken with a 72mm / 432mm APO by the way, with no Reducer / Barlow and just a Baader Skyglow filter.

 

Edited September 29, 2022 by PeterC65

[Chris]

On 29/09/2022 at 21:44, PeterC65 said:

I managed a session with the new Player One Uranus-C camera last night. Not great conditions and with hindsight the focus may have been off a little. I aways find the first session with new kit to be a little frustrating so more testing is needed, but I took some images of M33 at different gain settings and came to some initial conclusions.

Bear in mind that I do EEVA (rather than AP) and with an AZ GOTO mount, so long exposure images are not really an option. Instead, I captured the output from live stacking 20 frames, auto stretched, at 4s exposure. I tried gains of 175, 180, 250, 350, 450, 550, 650 and 750.

There was noticeable banding at 175 which went away at 180. The noise seemed to reduce significantly between 180 and 250 then reduced further but more slowly to 450 after which it rose again. I'm thinking this was the read noise diminishing and then being overwhelmed by some other source of noise. The level of detail gradually increased with gain, particularly from 450 upwards, but overall the best gain was at 450 I think.

Here is the (horrible) capture with gain at 175:

And here is all that banding gone by a gain of 180:

The best capture I think is at 450:

Then it gets very noisy as you approach 750:

Next time I get a properly clear night I will make sure the focus is spot on and carry out a more systematic test for varying gain and exposure.

These were all taken with a 72mm / 432mm APO by the way, with no Reducer / Barlow and just a Baader Skyglow filter.

 

Very interesting Pete, it shows how worth while it is taking test shots at different increments of gain regardless of the published HCG switching. I look forward to taking finer increments with the ZWO and seeing what happens. Player one state 180 but your 450 results are a big step up. I don't know if the live stacking effected anything as you would expect the published gain 180 HCG to be correct. 

[bottletopburly]

I tried out my camera the other night but fps were shocking but noted camera was showing connection via Usb2 so I tried my back up  active cable and now showing as usb3 and fps 192fps at 640 x480 so that’s a lot better , hopefully next outing should be a lot better 

[PeterC65]

Last night I did some more testing with the Uranus-C, checking gain / exposure / binning combinations, mostly with M33 and also with M34 and the Double Cluster (as a brighter target). Bearing in mind that I'm doing EEVA and so need short exposure times and am not looking for photographic quality, the conclusions were:

• Gains above 600 introduce unacceptable amounts of noise. The 800 gain was too noisy for live stacking to work. This explains why the spec for the camera only shows gain figures up to 500 (SharpCap can set the gain up to 800).
• Below 180 (in LCG mode) there is always horizontal banding in the live stack, less so in individual frames. This disappears completely at 180 gain, suggesting it is an artifact of LCG mode.
• The background noise falls noticeably as the gain increases from 180 to 250 while the target clarity remains the same. Above 250 the background noise starts to increase but the target clarity increases more quickly up to 600 gain. The best compromise between background noise and target clarity is around 400 / 450 gain.
• There is no discernable amp glow under any conditions!
• 2x2 binning is like increasing the exposure time, but comparing several different targets, the 1x1 image is slightly clearer with the same gain and exposure settings. 4x4 binning just loses much of the detail and I won't be using it. 3x3 binning caused red, green and blue artifacts to appear all over the image making it unusable (it's not clear whether 3x3 binning is supported by the camera but SharpCap allows it).
• When ROI or binning is enabled, I was seeing horizontal banding along the top edge of the image, gradually reducing to about 5% of the way down. I've asked Starfield Optics about this, and the 3x3 binning.

For a couple of the targets, I ended up rotating the camera through 90 degrees to get better framing and it occurred to me that by doing this I was getting the same field of view as you get with IMX533 based cameras but more resolution in both the horizontal and vertical (by rotating the camera).

Since this is the IMX585 image showcase thread, here is a nice live stack of the Double Cluster, 23 frames at 4s exposure and 180 gain, from the Uranus-C paired with a Photoline 72mm/432mm APO refractor.

I just love that I can see this much detail on the night!

 

[Ratlet]

Looks like my favourite cut price astronomy maker is bringing out a camera of their own with this sensor:

https://www.svbony.com/blog/one-of-the-new-products--sv705cc-eaa-/

Itll be interesting to see how it compares.

[Pixies]

I've just put in an order with FLO for the ASI version. I bit steeper, I know, but I have an idea I might end up with an ASIAir before long.

[nfotis]

7 hours ago, Ratlet said:

Looks like my favourite cut price astronomy maker is bringing out a camera of their own with this sensor:

https://www.svbony.com/blog/one-of-the-new-products--sv705cc-eaa-/

Itll be interesting to see how it compares.

Quite intriguing, indeed. I wonder  what's the expected pricing, too.

And I see that they're going to introduce a mono sensor, too.

N.F.

 

[PeterC65]

On 26/09/2022 at 14:51, PeterC65 said:

There has been some discussion about frame rate with the ASI585MC. With the Uranus-C at maximum resolution (3856 x 2180) I'm getting 2.6fps. At 1024 x 768 I'm getting 28.2fps. At 640 x 480 I'm getting 63.7fps. That's with a brand new Dell i7 laptop and USB3.0. I do have a USB3.0 router in the way which makes no difference except at 640 x 480 when I'm getting 72.2fps with the camera connected directly to the laptop.

CORRECTION

I realised today that I've been using a USB 2.0 cable between by USB hub and the Uranus-C camera, hence the reduced frame rate.

Now with a USB 3.0 cable directly connecting the Uranus-C camera to the laptop I'm getting 46.3fps at maximum resolution (3856 x 2180).

That's using the 2m cable that came with the camera. If I use a 5m USB 3.0 cable the camera will not reliably connect. If I use the 5m USB 3.0 cable between the laptop and a powered USB 3.0 hub, and then the 2m USB 3.0 cable between the hub and the camera, then the camera connects but the frame rate varies between 5.2fps and 0.4fps. With a 2m USB 2.0 cable between the hub and the camera (the setup I've been using) the frame rate is a reliable 5.2fps!

So to use a fast frame rate you need a short USB 3.0 cable, probably without a hub. With a hub or with longer cables only USB 2.0 is reliable.

[MichaelBibby]

On 07/10/2022 at 06:28, PeterC65 said:

Last night I did some more testing with the Uranus-C, checking gain / exposure / binning combinations, mostly with M33 and also with M34 and the Double Cluster (as a brighter target). Bearing in mind that I'm doing EEVA and so need short exposure times and am not looking for photographic quality, the conclusions were:

• Gains above 600 introduce unacceptable amounts of noise. The 800 gain was too noisy for live stacking to work. This explains why the spec for the camera only shows gain figures up to 500 (SharpCap can set the gain up to 800).
• Below 180 (in LCG mode) there is always horizontal banding in the live stack, less so in individual frames. This disappears completely at 180 gain, suggesting it is an artifact of LCG mode.
• The background noise falls noticeably as the gain increases from 180 to 250 while the target clarity remains the same. Above 250 the background noise starts to increase but the target clarity increases more quickly up to 600 gain. The best compromise between background noise and target clarity is around 400 / 450 gain.
• There is no discernable amp glow under any conditions!
• 2x2 binning is like increasing the exposure time, but comparing several different targets, the 1x1 image is slightly clearer with the same gain and exposure settings. 4x4 binning just loses much of the detail and I won't be using it. 3x3 binning caused red, green and blue artifacts to appear all over the image making it unusable (it's not clear whether 3x3 binning is supported by the camera but SharpCap allows it).
• When ROI or binning is enabled, I was seeing horizontal banding along the top edge of the image, gradually reducing to about 5% of the way down. I've asked Starfield Optics about this, and the 3x3 binning.

For a couple of the targets, I ended up rotating the camera through 90 degrees to get better framing and it occurred to me that by doing this I was getting the same field of view as you get with IMX533 based cameras but more resolution in both the horizontal and vertical (by rotating the camera).

Since this is the IMX585 image showcase thread, here is a nice live stack of the Double Cluster, 23 frames at 4s exposure and 180 gain, from the Uranus-C paired with a Photoline 72mm/432mm APO refractor.

I just love that I can see this much detail on the night!

 

Hi Peter, you mentioned seeing horizontal banding at the top of the image extending about 5% of the way down when ROI or binning is enabled, I've been doing some test shots on terrestrial objects (clouds as far as the eye can see on the East coast of Australia, extending far into the future...) and noticed the same thing. Did you hear back from Starfield Optics about that (and whether 3x3 binning is supported)?

Thanks, Michael.