First Light with new ASI585MC Pro.: Page 4

First Light with new ASI585MC Pro.

Kevin_A

w7ay that is what I thought but lots of pundits out there that tell the masses that darks are not needed regarding the 2600. I do my darks indoors when my skies are bad and just let it run. I usually do 30 of each 120, 180, 240, 300s and it takes me a whole day every 6 months. I am just now waiting again for my new Esprit to arrive ( it was $1000 off so I had to get one) and the skies to clear so I can image again. There has been zero clear nights since the eclipse. Thanks again for your expertise and data! I run a PC with a i12700 so it runs well but only 64Gb ram. Cheers!

w7ay

Kevin_A I am just now waiting again for my new Esprit to arrive

You getting the 120 with the 0.77x, right?

Eventually, you will end up with the FSQ106 anyway, so why not get it now, and enjoy it for a couple more years :-). You and I are not getting any younger :-).

After thinking that I should go back and play monochome cameras again, mounting the ASI2600MM and filter wheel (loaded with 7 heavy 3mm thick Chroma 36mm filters), the weight of my Baby-Q came to about 6.3 kg. I (5'2", and 75 years old) can handle it, but barely.

I had bought the FSQ-85 instead of the FSQ-106 just because of the weight. The barebones FSQ106 is 7 kg. Add tube bands, Losmandy dovetail, electronic focuser, and camera, and I am sure I cannot handle it.

A barebones FSQ-85 (from specs) is 3.6 kg. So, the mounting bands, Loasmandy plate, EAF, loaded 36mm EFW, and ASI2600MM must account for the 6.3-3.5kg = 2.7 kg.

Add 2.7 kg to 7 kg for the barebones FSQ106, and we have the weight of almost a 25 sack of rice, and am sure I cannot lift and mount that on an equatorial mount. I can barely load a 25 lb sack of rice into the trunk of a car :-).

I might be able to do it without the camera and filters, but I don't really want to have open optical paths in the dust outdoors while attaching a camera to a mounted OTA.

But I am still tempted by the real FSQ and its 88mm image circle (and f/3 with Tak's reducer -- that reducer costs more than Chinese OTAs :-) -- and still looking for ways (4040 aluminium extrusions is what I am playing with currently :-) to mount the RST135 at a height that I can handle (like with the dovetail clamp no more than 3 feet above ground). The RST on top of a 800mm tripier is just too tall for me. Perhaps a gurney-like platform to help lift the OTA to reach the mount :-).

First world problems :-).

Chen

w7ay

Kevin_A

OK... back to your original topic (cooled ASI585MC) :-).

First, lets look at a ASI2600MC (which both of us have, so we can be in sync) at -10ºC.

This is the RMS error (a.k.a. standard deviation, and that is the square root of the variance -- thus the "root" part of the term RMS) for the ASI2600MC that I just measured earlier today. That is the ordinate axis.

The abscissa is the gain in 10x dB, thus basically the ZWO "gain."

Each color "curve" is a different exposure time.

Notice the abrupt reduction of noise at 100 (where the HCG kicks in). So, if you see a big drop, that means the read noise (which is affected by HCG) is predominant, and if you don't see a drop, then dark current noise (which is not affected by HCG) is the predominant noise for that exposure value.

Notice that the ASI2600MC is pretty much read-noise limited. It is not until 180 second exposures that the HCG drop drops. And even at 180 seconds, it has not yet completely vanished. Dark current noise is not a problem.

Now, lets look at ZWO's cooled ASI585MC (same -10ºC):

Interestingly, HCG is effective even with dark current noise at 120 second exposures.

However, if you look carefully, dark current noise is large enough that the total noise at HCG (gain = 252) is never really lower than using a gain of 150 or so.

And, by gain of 150, the total noise is already very high. The total RMS noise for a 10 second (not typo 10 sec, not 100 sec) exposure (dashed green line) at gain 150 is already at about 40 ADU.

Now go look back and look at the ASI2600MC chart... a 60 second exposure at HCG has a total noise of just a little over 40 ADU.

The ASI585MC is definitely not a good long exposure camera. You may be able to use this camera, if you use 10 second type exposures, and let stacking reduce the strong read noise.

Now, if you kick HCG in at gain 150 instead of 252 that ZWO chose, it might actually become a usable long (120 second-ish) exposure camera. As it is (with HCG at 252), the noise is already so high as to be unusable.

Pity the people who buy this camera just because it is finally a cooled camera that they can afford. It is money down the drain (unless you are a CAA type who don't need long exposures).

I guess if you are in a Bortle 9 area, the noise would be dominated by the sky background, and the camera noise is not such a big factor.

Chen

w7ay

Kevin_A t

Its a shame since I received the M42-M54 11mm thick ring from PreciseParts (black ring on the ASI585 on the right).

Allows me to hook the camera to the M54 that I have standardize my interfaces to (maintains constant 17.5mm flange focal distance). But I have enough other ZWO cameras that uses the 11mm ring, so M54 ring will remain quite useful.

Chen

Kevin_A

w7ay it is too bad the HCG was not at 150. Maybe it was a Sony implementation but not sure why! Is it a firmware code setting or built into the electronics when HCG occurs?

w7ay

Kevin_A it is too bad the HCG was not at 150. Maybe it was a Sony implementation but not sure why! Is it a firmware code setting or built into the electronics when HCG occurs?

My impression is that HCG is a state (a hardware bit). ZWO does not support user enabling HCG, so they select some gain when they automatically turned it on. I twould not be surprised if ZWO is the one who selected 25.2 dB.

Yeah, 150 would have been a nice place to turn on HCG, bringing the total noise for a 60 second exposure at HCG down to a really nice low level.

BTW, one of the things I want to try next is to use L from mono camera, but extract RGB from a color (OSC) camera, instead of taking L, R, G, B from a mono camera. I know others have tried it, so it has to work pretty well.

If we spend T seconds on the OSC camera, and all else is equal, the red channel has about 1/4 of the photons, geeen has 1/2 of the photons and blue has 1/4. With a mono camera, you would probably spend 1/3 of T on each R, G and B. So OSC red gets (1/4)/(1/3) = 3/4 of what red gets from mono camera for the same total of T seconds. That's not too bad.

I already have an 2600M for L, and 2600C for the OSC channels. And I already have two mounts, so both L and RGB can be captured on the same night. With ASIAIR, would need two ASIAIR, but I already have 3 ASIAIR plus -- I currently have one set up exclusively for indoors use :-). With Indigo sky, both cameras can be controlled from one Raspberry Pi 5 on a single mount, and only need one guide scope.

Chen

Kevin_A

w7ay your logic behind that adds up for sure.
I am still going to see what I can get out of the 585mc pro for galaxies only as I love a challenge and I will see if I can use 60s shots with my 805mm scope at HCG. If not and it does not pan out then I still have a 585 that I will use without cooling for lucky imaging on the moon etc. Haha

Back to flats and flat darks… do you think there is any difference in them being 3-5s long vs 0.5-2s long? I seem to have good luck with longer flats but not sure if it makes any difference.

w7ay

Kevin_A do you think there is any difference in them being 3-5s long vs 0.5-2s long? I seem to have good luck with longer flats but not sure if it makes any difference.

I am not sure it makes a difference, either. It may depend on whether the light panel has fluctuations.

LED panels with poorly filtered power supplies could have 60 Hz ripples (or 50 Hz for folks across the pond), and modulate the exposure -- you may have seen bands of light when exposing at about 1/30 second, when pointed at a garage fluorescent bulb, for example. The same would happen to an LED panel whose power supply has 60 Hz ripples. In that case, using long exposures would average them out somewhat.

EL panels use 400 Hz and above, I think, and any modulation would be seen only with much shorter exposures. So, they are more immune to exposure duration.

Speaking of which, I know some people try to be cheap and use LED drawing panels as flat lights. I don't see how it turns out well, because "white" LED are made from R,G and B LEDs, and those each have moderately narrow bandpasses -- it would only excite a wideband filter at only some spots on the spectrum, and would have very little light output at the H-alpha etc narrowband filter passbands. Penny wise, pound foolish -- eventually the Amazon drawing panels go into the landfill, and you have to buy an EL panel anyway to do things properly; "upgrading" to the proper panel. Again, I believe in buying something only once.

There are "full-spectrum" LED that people like using for "plant" (heh, heh) growing and the spectrum is more of a continuum. But I don't think the cheap Amazon drawing panels use them. Even plants don't like narrow band LEDs :-).

Chen

Kevin_A

w7ay Good to know about the light panels. I use a cheap one but not illuminated. I found out too that mirrorless camera sensors do not like them either as the PDAF (focusing) pixels start banding, so I just use the panel for dimming my ambient light. Seems to work fine for me.

Kevin_A

w7ay is there a good brand of Electroluminescent light panels?

w7ay

Kevin_A is there a good brand of Electroluminescent light panels?

I use the Gerd Neumann Auroras (two sizes, one for everything, and a larger one for >5" OTA tubes). Works pretty well. DHL gets here just as fast as UPS :-). Just be sure to get the 12 V ones instead of the 220V ones.

I top it off also with the Daystar FlatCaps to further diffuse the light -- I have one for the Baby-Q and one for the FOA-60Q. Definitely not neccessary, the panel is way out of focus already, but I am OCD :-).

Chen

w7ay

Kevin_A

Kevin, I found a good combo to get f/4 from the Samyang/Rokinon.

The cheap Tiffen 39mm filter with the glass removed has a clear aperture of 33.7mm, which givesreally close to f/4 for 135mm.

https://www.amazon.com/gp/product/B0BBQ4GM9C

So, step down the 77mm filter thread to 39mm thread, and insert the glassless "filter." The aperture appears quite clean and round.

Since getting my Rokinon, I have increased my existing collection of step-down/clear-filter combos to these:

The 37mm and 39mm needed two steps of step-down, since I could not find any 77mm-39mm and 77mm-37mm step-down rings. So I chose 55mm as the intermediate; 77mm-55mm is one of the existing step-down rings.

Darn lens should be clear as a bell at f/4. (Sorry for mixed metaphores. :-)

I am always nervous about placing an iris at some place other than a nodal point of the lens. But so far have not found that it increased the vignetting. And using a glassless filter gives a very clean aperture with no diffraction spikes. I have not seen anyone else mentioned using apertures from camera filters, by the way, but just using step-down rings themselves, which do not have a clean circular aperture due to the filter thread. $10 to get rid of nuisance diffraction seems worth it.

Chen

Kevin_A

w7ay I cannot wait to see the results… you have peaked my curiosity!

w7ay

Kevin_A I cannot wait to see the results… you have peaked my curiosity!

Here you go...

halfway clear enough skies so the diffraction spikes don't get smeared by the clouds. Just out of astronomical twilight. I used Alioth (magnitude 1.75, and far from the full Moon) with an ASI6200MC set to an APS-C ROI. Gain of 10 dB (HCG), and 60 seconds, and cooled to -10ºC, with ASIAIR. like so:

I centered Alioth as well as I can, so there is no infuence by lens aberrations.

Here is a crop at 1:1 scale (1 camera pixel == 1 pixel in this image) with the lens set to T=4. Recall that my Rokinon is a cine version of the Samyang, so everything is calibrated in transmission values (wide open is T/2.2, so there is a loss of 10% through the glass elements -- should be the same with the photo version of the Samyang f/2.0). Also, to comply with cine folks' usage, the T ring does not have click stops, so I am just approximating by eye where T/4 is.

Next image is the T-stop set to wide open (T/2.2), and I added a 77-55mm step down ring, followed by a 55-39mm ring, with a (cheap) Tiffen 39mm UV filter that has the glass removed. I.e., I am just using the metal filter holder as an aperture. The clear aperture of the Tiffen's retaining ring is about 33.7mm, so that gives me the f/4 at 135mm (my Rokinon actually plate solves to 129mm focal length). Some other brand's 39mm filter may (and probably will) give a slightly different clear aperture.

Neat, eh? Practice agrees with theory. I don't know why no one has thought of cleaning up step-down ring filter threads this way.

The Tiffen UV filters are pretty cheap, and I chose it because all of the Tiffens' retaining ring that I played with seem to be loose enough to use a simple lens spanner -- there are some high end filters that really torqued down the retaining ring.

Be sure not to destroy the fragile retaining ring as you remove it (don't even try it, if you don't have a lens spanner; I assume all astrophotographers owns at least one lens spanner), since that has to be mounted back, without the glass, to form the actual circular aperture.

Chen

w7ay

Kevin_A

By the way, Kevin, since the Moon is too bright to do anything serious, I started to look at guiding the encoder version of the mount (RST-135E), and it looks like there is some improvement in guiding at 1 second than at 0.5 sec. ASIAIR's multistar centroid algorithm sucks, since with that many stars, longer exposures should not improve guiding so much.

Chen

Kevin_A

w7ay my Rokinon plate solves to 129mm too! Nice results. Can you post a photo of what it looks like without the lens hood on with the ring setup on? Thanks!
I still think I would do the setup at close to f2.8 for 2 reasons. I think it is like buying a fast car but only driving it slowly…. So I would still want it close to max speed but a bit better handling.
F4 is sharper but a bit more abberations than f2.8 and why buy a fast lens and use it 2 stops slower… 1 stop would be my setting, but that’s just me! Haha My lens spanner is toast as it was a cheap one so if you have a recommendation of a better design then let me know what you have been using.

w7ay

Kevin_A Can you post a photo of what it looks like without the lens hood on with the ring setup on?

OK, first the stepdown rings by themselves. This first image is for the case of a single stage step-down (77mm-40.5mm):

The large disc is the stepdown ring, and the protrusion is a 40.5 UV filter with the glass removed. Nice thing about camera filter sizes is the lens caps that are available; shown on the right of photo.

I could not find any 77mm-39mm stepdown ring, so the next photo shows a 55mm-39mm step down mounted on a 77mm-55mm stepdown, with a 39mm UV filter mounted on the 55-39mm ring, with its glass removed:

Now, the clear aperture of the 39mm filter is not 39mm of course. In the case of the Tiffen, it is 33.7mm (thus the f/4).

OK, finally, here is the ring mounted on a Rokinon. The camera band is from Astrodymium, and I had drilled and tapped the 3D material and also the body of a ZWO EAF to directly mount the EAF on the tube band. Notice my temperature probe hack on the EAF :-)

The lens hood works as normal (since photographers need a lens hood when they use filters),

By the way, did you notice something in the Alioth image? It completely lacks any diffraction spikes. This means there is no aperture vignetting at f/4, when done this way. If there is any vignetting, you will see two "negative diffraction spikes" around a star, as discussed here (last post there is the best one; I have a similar program for macos as the Maskulator that I had used to find the inverse filter for non-blind deconvolution; my program can actually also control the intra-focus and extra-focus positions too):

https://www.cloudynights.com/topic/547028-anyone-know-what-causes-this-artifact-on-bright-stars/

So, it is perfectly safe to do f/4 aperture control this way. I will need to test the other aperture for vignetting.

Again, here is the clear aperture (and 135mm f-number) for various step-down rings and Tiffen filters:

"Thread" is the stepped-down thread, i.e., "67mm" indicates a 77mm-67mm step-down ring. For an f/2.8, try either a 77mm-55mm step down or a 77mm-52mm step down, together with the 55mm or 52mm Tiffen UV filter. The nice thing about the cheap Tiffens is that they seem to make it for pretty much any size.

BTW, be careful when you buy from Amazon, that the ring is a step-down ring. A step-up ring (filter that is larger than camera lens) will not work.

Chen

w7ay

Kevin_A My lens spanner is toast as it was a cheap one so if you have a recommendation of a better design then let me know what you have been using.

My Amazon records say that I had bought this one in 2015 (no longer available). These tools last forever if you take care of them. Lots at Amazon with the same style, though.

https://www.amazon.com/gp/product/B00J5F6O92/

However, after mentioning lens spanner yesterday, I saw this one at Amazon:

https://www.amazon.com/gp/product/B0C4P5L9PC

Can save a lot of time with adjustment, so I could not resist, and ordered one :-). If it lasts 10 years too, it will be well worth the money.

Chen

w7ay

Kevin_A I still think I would do the setup at close to f2.8

I normally use between 1 and two stops down with camera lenses, but the Rokinon appears to be quite good one stop down from the initial images I took.

The MTF graph that you posted still shows better MTF (smaller stars) at f/4 though.

BTW, take a careful look at the two f/4 images of Alioth. Even the dimmer stars from the artifical aperture shows a tighter spot size. You cannot see the diffraction spikes directly (too small and dim), but the stars appear smaller and "rounder."

Chen

w7ay

Kevin_A

Ooops, sorry about the vignetting comment. Alioth was at the center of the frame, so of course no aperture vignetting.

So I still need to test vignetting by placing a bright star at the corner of the frame to look for those diffraction notches.

Chen

Kevin_A

w7ay they were the two spanners I was looking at as well. I noticed the lack of difraction spikes… that’s what interested me the most. At f4 mine were wicked so your stepdown design would work well with me.

« Previous Page Next Page »