SV205 expectations?

[fommil]

Hi all,

I am just getting started with astrophotography so I got the cheapest thing around, an svbony sv205 (IMX415 over the webcam protocol). I do not currently have a Windows or x86 laptop to be able to use sharpcap so I'm using python code to import frames over the standard webcam / V4L to get myself into the whole thing.

I have an entry level 130mm (5") Newtonian at f/5; a celestron starmaster. I've got bortle 4 skies and a clear view to polaris.

I'm able to set exposure time, width, height and gain (I upgraded the camera's firmware to unlock this option) but even with 1 second exposures I am not getting anything coming through. Total darkness. If I shine my red LED down the tube, it does register it, but I'm not seeing any stars. I was red dot pointing at Polaris, but nothing, then I decided to go point at a very bright Jupiter, and still nothing. I moved the focuser all the way from both extremes, very slowly, and nothing changed. I basically can't even get anything to focus with, nevermind anything else. In hindsight, I should have put my eyepieces on to really center it with a 9mm, just to be absolutely certain.

Does anybody with this camera have a frame of what I should reasonably expect with a 1 second exposure? Ideally at a few different gain settings. And if you don't have that level of data around, do you know what gain you're typically using?

The following is the calculation of my FOV through my eyepieces and the camera, so another possibility is that I just wasn't pointing directly enough at the objects to see anything. But I'm getting absolutely nothing through at all, I'd really be expecting to pick up faint stars or something at the very least with a 1 second exposure. Even my eyes can see faint objects with this level of zoom, and they are much less sensitive than a digital camera with 1 second exposure. Or is my intuition all wrong here? (urls are kinda janky, pasting the raw form)

Code: Select all

http://astronomy.tools/calculators/field_of_view/?fov[]=88|274|||1||&fov[]=88||13745||1|1|0&fov[]=88|232|||1||&messier=31

[admin]

Hi,

I think the best approach for you would be to try to focus on something in daylight - ideally a distant object (tree/roof/etc) on the skyline - this gives you the advantage of not having to worry about the things you are viewing being really dim at the same time as you are trying to focus.

Hopefully you will be able to get focus on the distant target - then focus for the sky will be a little further in. I do recall that with some cameras like DSLRs, it may be impossible to get focus with a typical 130mm f/5 Newtonian as it is impossible to move the focuser far enough inwards to get the sensor to the focal plane of the telescope. Hopefully that won't be the case with the SV205, but at least in daylight you will be able to see what is going on.

Once you have daylight focus, the next target should be the moon as it is bright enough to find even if somewhat out of view due to the sky glow around it.

Even given all of that, you will be limited by the maximum 1s exposure and the 8 bit depth available from the SV205 - along with the small pixels, you will be limited in seeing stars (the camera is more suitable for the moon and planets).

Hope this helps,

Robin

[fommil]

Thanks Robin, yeah focussing in the daylights sounds like a good idea. I will try that out. It should also let me gather some bias data to see what's going on, although I expect it will saturate fairly early on and I'll have to dramatically reduce the exposure length. It should let me gather some ballpark figures for the bias to noise chart, but I will plan on creating a more reliable chart one day for 1 second exposures in low light.

Once I can figure out bias and focussing, my next question was going to be about the bit depth: the chip is 10 bit and the svbony marketing material says 10 bits "ADR" but they don't specify anywhere what that means. Are they letting the camera saturate early (dropping the top 2 bits) or are they throwing away the 2 bits of sensitive information? Or are they doing some crazy thing where they drop the bottom bits as they get near the top bits, which would cause all kinds of havoc if I get anywhere near the top 3 bits. I've also heard people say that there's some colour correction happening in the conversion to YUYV format but I don't really understand what they mean by that... I'd like to understand what that means in terms of starting with a raw RGB 10 bit value and converting it into the RGB channels I get from V4L (and I've still to confirm what the bit depth is there, it is claiming it supports HDR).

To address the small pixels, which means small field of view, I have a cheap 0.5x reducer on order. There are no technical specs about what that does in terms of aberration or transmission loss by frequency, but I'd be interested to know of any ballpark figures. These things can get very expensive so I would have expected there to be some standardised test that they can take. There doesn't appear to be much of a market for anything besides a 0.5x reducer, I'd be interested to know if there's any cheap 0.33x products out there.

[admin]

Hi,

given that the camera is working as a webcam (making it impossible to have specialist options to turn on/off features), I'd expect that the image processing is working something like

10 Bit ADC => Discard bottom 2 bits to get 8 bit raw pixel values => Debayer to RGB image => Convert RGB to MJPG or YUY2 => Image Data flows to computer

On the computer side, SharpCap turns the YUY2 or MJPG back into RGB for processing, but of course the original pixel values of the raw image are long gone at that point - first the debayering and then the compression (of either sort) will blend pixel values together and affect them via the lossy nature of the compression.

On the reducer side, I think that only 0.5x are available in the simple '1.25 filter screw thread' form. There used to be 0.33x reducers available for SCT telescopes, but I think they introduced a lot of distortion, so you don't see them much.

cheers,

Robin

[fommil]

It prompted me to go read the wikipedia page on this https://en.wikipedia.org/wiki/Y%E2%80%B2UV

So now I understand why the colour conversion is technically lossy. But my understanding is that there's no lossy compression with the YUYV format, that's only in MJPG mode.

Exactly which two bits are dropped is still to be confirmed. It's annoying that they don't document this. It's just really silly if they are dropping the sensitive end, given that it's marketed to be a 15 FPS planetary camera.

[fommil]

gain definitely impacting the luminosity, here's with gain set to 128 (max) and 0 (min) on the same scene (a neighbour's tree wobbling in the wind), with lowest resolution and 0.01 second exposures. This definitely let me find the rough focus sweet spot, so now I just need to repeat this at night when pointed at any kind of object to try and get some data. Thanks!

[admin]

Hi,

glad you are on the right track now with the focus and other settings

There is no need to worry about the 8 bit/10 bit thing for planetary imaging - planetary imaging is conducted with short exposures (30ms or less) and high gain, leading to noisy individual frames. The bottom few bits of the data in 8 bits will be noise anyway, so if you had 10 bits you would just be sampling the noise in more detail to no actual benefit (in fact a net loss as the size of saved files would have to increase).

YUY2 is "less lossy" than MJPG, but it still sub-samples the colour information (ie colour is represented with less resolution than brightness). This isn't usually a significant issue, but worth being aware of.

cheers,

Robin