Tiny chip, tiny lens?

[Ags]

Just wondering, can you pair a smaller chip like a ASI 178 (9mm diagonal, 2.4 micrometer pixels) with smaller camera lenses in the range of 14mm to 135mm?

[Davey-T]

Dunno but got both ingredients here so if we ever get a clear sky again I may try it, I find that trial and error is the best way with this sort of thing 

 

Dave

[vlaiv]

Of course you can and it will do a better job than larger sensor because it won't capture edge of the field of larger sensor where most lenses show some sort of aberrations.

You just need suitable adapter (one for mounting lens on T2).

Smaller pixels will better show any lens aberration present across whole field, focusing will be more of a challenge, but otherwise it will work well (and certainly with fast lens like Samyang F/2 135mm).

[Ags]

I was wondering if using a small chip with small pixels and a low resolution small lens might lead to soft images. But then I thought about the tiny sensors and tiny lenses on mobile phones, so I guess that concern is not very real.

I've been meaning to mod my 1100D for years, but always been put off by the cost and the fact it would be a bit less suited for daytime photos (although it can be made to work I know). But then I thought I could just put the money (and a bit extra) into one of the smaller astro cameras, and just use smaller lenses. ZWO have an EF adapter, and the ASIair would keep everything light and small...

Gear Acquisition Syndrome is a terrible thing...

Edited June 21, 2019 by Ags

[Ags]

Does anyone know if it's possible to use 1.25 filters in the ZWO lens adapter? And any tips on attaching small lens, ZWO adapter and ASI 224MC? Teleskop Service used to do a camera lens adapter with a mounting shoe, but that doesn't seem to be available any more.

[wimvb]

I guess you mean this?

https://agenaastro.com/zwo-canon-eos-lens-short-adapter-asi1600-efw.html

[Mark1489]

On 20/06/2019 at 13:39, vlaiv said:

Of course you can and it will do a better job than larger sensor because it won't capture edge of the field of larger sensor where most lenses show some sort of aberrations.

You just need suitable adapter (one for mounting lens on T2).

Smaller pixels will better show any lens aberration present across whole field, focusing will be more of a challenge, but otherwise it will work well (and certainly with fast lens like Samyang F/2 135mm).

I use the Samyang and was thinking of a QHY183C as an upgrade to my D5300 in the near future, a good match possibly? I was looking at the QHY247C with the same sensor as the Nikon but it’s quite a step up in price! 

[vlaiv]

30 minutes ago, Mark1489 said:

I use the Samyang and was thinking of a QHY183C as an upgrade to my D5300 in the near future, a good match possibly? I was looking at the QHY247C with the same sensor as the Nikon but it’s quite a step up in price! 

I think it would indeed be a good match.

At one point, I was considering following setup - ASI1600mm + TS80 reduced for mono data and Samyang + 178mc for color data. Very similar FOV, but 178mc is 3000x2000px camera and has a bit lower resolution than mono part of the setup.

QHY183C has same pixel size as ASI178, but is quite a bit larger chip, pairing that with samyang lens will get you really fast rather well corrected wide field imaging setup.

4 hours ago, Ags said:

Does anyone know if it's possible to use 1.25 filters in the ZWO lens adapter? And any tips on attaching small lens, ZWO adapter and ASI 224MC? Teleskop Service used to do a camera lens adapter with a mounting shoe, but that doesn't seem to be available any more.

Do you mean this one?

https://www.teleskop-express.de/shop/product_info.php/info/p7835_TS-Optics-Optics-Adapter-for-Canon-EOS-Lenses-to-T2-for-CCD-cameras---with-1-4--photo-thread.html

It's still available as far as I can tell.

There is one from Geoptik as well - a bit higher price:

https://www.teleskop-express.de/shop/product_info.php/info/p2836_Geoptik-Adaptor-for-Canon-EOS-Lenses-to-T2-for-CCD-cameras---with-1-4--Phototripod.html

Both of them come with 1.25" filter adapter thingy ...

[Ags]

Is 1.2 megapixels too low to be useful for DSOs? Looking at the ASI120MC-S or ASI290MC.... The 120 would be good for planets and later as a guide cam?

Edited June 23, 2019 by Ags

[vlaiv]

4 minutes ago, Ags said:

Is 1.2 megapixels too low to be useful for DSOs?

Yes and no

You can certainly capture many of the DSOs with that pixel count, people doing EEVA often use that much or even less pixels to capture, so it's possible, however you might not be pleased with the looks of image on today's display devices as most of them utilize at least 2mp displays, so if you look your image at 1:1 (which you should for best detail) you will end up with image that is smaller than the screen. Stretching it on screen will involve resampling and blurring of the details, so in my view - something to avoid.

That many pixels does not mean you need to take "small" image - you can do mosaics to get larger pixel count (and wider FOV) - but it is laborious.

In principle you could use it, I've done it when I started fiddling around with AP, for example this one:

but you'll probably want one with a bit more mega pixels if you are thinking of doing anything more serious.

[Ags]

I made a test image with the resolution of the ASI 224MC, and it's pretty big - big enough to look good at 1:1 on my 1080p monitor. I am tending towards the 224MC now, as in addition to possibly using this for DSOs and planets it is apparently good for live stacking EEVA which might be a good route for me to take given the bad light pollution I have. Also my partner is quite keen on looking at things on a computer screen, so it helps with the sales pitch! 

So now I am looking at:

ZWO ASI 224MC

Teleskop Service Camera Lens Adapter

7mm and 5.5mm T2 extension tubes

1.25" Ha filter (already have UHC) (I want to get the Baader 35nm one because it is cheap and wide enough to work at F1.8, but I suppose it is a bad idea with my light pollution!)

IR pass filter

AZ-GTi

For DSOs I already have 50/1.8 and 100/2.8 lenses. (Might add the 35/2.0 and 85/1.8 lenses down the line)

For planets I am aiming at getting a C6 by the time of the next Mars opposition.

For EEVA, I have an ST80 and 0.5 reducer.

[vlaiv]

Well, if 224 is an option, then maybe 385 is an option too?

It is going to be better for all the things that you mentioned - it has same pixel size, larger pixel count (and sensor size) and it has even less read noise (important for planetary and EEVA).

[Ags]

Ooh, mission creep! 385 would be very nice, but more than I want to spend. I see I can get the 178 for a little less, but I lose IR sensitivity? The 178 has a bigger sensor than the 385, presumably it's small pixels can be compensated for by binning 2x2 (would still leave a 1.5 megapixel image)?

[vlaiv]

1 hour ago, Ags said:

Ooh, mission creep! 385 would be very nice, but more than I want to spend. I see I can get the 178 for a little less, but I lose IR sensitivity? The 178 has a bigger sensor than the 385, presumably it's small pixels can be compensated for by binning 2x2 (would still leave a 1.5 megapixel image)?

With OSC sensors, binning is not quite as straight forward.

When you use OSC to image and keep "native" resolution, due to bayer matrix on sensor you will be effectively filling in missing values (for example red pixels - they are spaced every other pixel in both height and width, so you need to "make" up missing values - this is done by interpolation). If you then bin resulting image - you won't get anything (but a little blur) since pixels that you are binning are already dependent on existing pixels (not separate genuine capture). Result will be the same as just making every image "dense" by removing gaps without putting anything there (this is actually better way to do it as it does not produce additional pixel blur).

Once you remove those gaps and get R, G and B images (there will actually be 2 G subs - which you can just use as regular subs for stacking - you will end up with x2 total number of G subs), then you can bin further with expected results. However this means that on 3000x2000 you will end up with 750x500 images (in terms of pixel count) if you bin. This also means that final resolution after binning will be x4 of what we would calculate as resolution based on single pixel size (if bayer matrix is for example RG - being top row, this means that you will have for red channel R_R_ in top row, you ignore spaces, so you will end up with RR and then bin that to single R - 4 original pixels get you one resulting pixel per channel).

Just for clarity what I mean by additional blur when you try to bin interpolated images:

examine following R1 _ R2 (just 3 pixels of top row), you fill in missing by linear interpolation, you will end up with R1, (R1+R2)/2, R2, and now you bin (add) first two - (3*R1+R2)/2 - you won't have same SNR increase as regular binning, because you take 3 times one pixel and one time other pixel - this is creating correlation between them that creates blur and result for SNR won't be the same as regular binning (less SNR gain).

In order to Bin OSC sensor with good results you really need to oversample by quite a bit and small pixels help there, but these small sensors leave very low pixel count after binning. If you want best possible resolution out of OSC then you should not interpolate missing values, but rather "condense" each channel.

[Ags]

I'm learning a lot from this thread. From the point of view of DSO imaging and EEVA it looks like a mono sensor would be the best.

[Stub Mandrel]

I've got a Touptek mono which has a 'webscam' sized chip. I keep planning to use it to 'try out' RGB imaging. Paired with my 130P-DS or 150PL it should be great for imaging targets like planetary nebulas or smaller galaxies.

[Ags]

I bought an ASI178MM.

[Gina]

Good camera - I've got one.

[Ags]

@Gina what do you use it for? Do you have the cooled or uncooled version?

[Gina]

I used it in my all sky camera for a while.  It's uncooled.

[Ags]

Just realized I need to buy a UV-IR cut filter... It never ends. I think without the UV-IR cut the ST80 will be a bit blurry!

[vlaiv]

5 minutes ago, Ags said:

Just realized I need to buy a UV-IR cut filter... It never ends. I think without the UV-IR cut the ST80 will be a bit blurry!

Even with UV/IR cut filter ST80 is going to be somewhat blurry

 

[Ags]

I hope it will be fine for EEA. Going to get that working Ok with a little dabble in planetary and lunar with the Skymax 102, then will get the bits for "tiny chip, tiny pixel, tiny lens" plan. I hope to get the North America Nebula done this year.

Edited June 26, 2019 by Ags

[vlaiv]

EEA part is quite "relaxed", or so I found - you have a sense of experimenting with something new and there is no pressure of achievement that is for example present in AP - there is so much work already done in AP and when people examine other people's work they form an opinion on what is good / great and such - bar seems to be set at certain height there.

With EAA there is no such thing - you are thrilled to observe, much like with visual - if you see it, you won't mind if it has a bit of coma in the corners or if image is not perfectly sharp - thrill of seeing it will trump all of that.

ST80 will work perfectly fine for EAA to start with. If you after some time start to find CA blur objectionable (it wont be seen as color due to mono sensor - it will be as regular sort of blur / halo around bluish stars) - there is simple fix - either wratten #8 or longpass 495 filter will work wonders to remove much of such artifacts.

[Ags]

Yeah, just don't want to disappoint my family. I hope they are not expecting Hubble.