37 replies to this topic

[EdwardMH]

I am a beginner amateur astrophotographer but a thought hit me this afternoon. I know mono cameras collect more details than color due to the color ccd Bayer-Filter array. My thought came from TV’s rgb emitters getting smaller and smaller from 720, 1080, 4K, 8K and even 16K... Couldn’t that shrinking work with color ccd chips and make them 2-4x smaller than the Mono equivalent thereby making them more sensitive to smaller details? Or would the smaller sizes absorb less light? Would it work if they found a naturally more light sensitive ccd material?

 

And also what is the detail difference between a color CCD and a mono used with rgb filter wheels? Does the pure mono no filters still gather more details than a mono with filter wheel?

 

Sorry if this was a dumb thought.

[kathyastro]

The sensor chips are the same in mono and OSC (colour) cameras.  Smaller sensors gather less light.

 

An OSC camera has a filter over each pixel sensor.  In each group of four pixels, one has a red filter, two have green filters and one has a blue filter.  Each pixel only records the colour of its sensor.  The other two colours for that pixel are calculated by interpolating from neighbouring pixels.  The interpolation is pretty good, so it's not like you are getting a full 67% reduction in resolution, but there is a loss of actual data.  The mono camera using the same sensor uses every pixel for every colour.  There is no interpolation: all the colour data is actual sensed data.

 

A mono camera with a filter gathers the same number of pixels as the camera without the filter.  The filter just restricts which wavelengths the sensor will record.  However, RGB imaging requires exposures with three filters, so you end up recording three times as much data.

[Madratter]

It isn't a dumb thought at all. And in fact, decreasing the size of the pixels does decrease the detail advantage of mono somewhat since at some point, additional resolution doesn't actually help to any practical degree.

 

As for mono itself, that is a rather complicated question. Part of the reason is that the detail achievable isn't strictly a matter of resolution only. It also is a matter of how contrast. This is most obviously seen when comparing something like an Ha image of a emission nebula with a luminosity image of the same nebula, or even a red image of that nebula.

 

That in turn is also true of the mono vs RGB camera and detail debate. And again, you can recover some of that detail (coming from contrast) by using various enhance filters on an RGB camera.

 

BTW there is a cost of smaller pixels (besides being harder to manufacture). They tend to suffer from more noise/unit area.

 

In other words, this who topic is somewhat complicated and parties on one side or the other tend to over simplify when discussing (arguing) about it.

Edited by Madratter, 28 November 2020 - 01:48 PM.

[idclimber]

Making pixels smaller does not always help especially with larger focal lengths. It is about seeing conditions and sampling data beyond that limit is called over sampling. 

[Madratter]

Making pixels smaller does not always help especially with larger focal lengths. It is about seeing conditions and sampling data beyond that limit is called over sampling. 

In some sense, that is his point. If you make the pixels small enough, the resolution advantage of the mono over the color is false gain. They are both limited by seeing, tracking accuracy, etc.

Edited by Madratter, 28 November 2020 - 01:51 PM.

[idclimber]

Current sensors are reaching incredible quantum efficiencies of around 90%. That is 9/10 electrons are detected. You simply can't detect much more than that as 10/10 is really only marginal improvement. What will continue to happen is decreases in noise and increases in pixel depth.. 

Edited by idclimber, 28 November 2020 - 02:08 PM.

[EdwardMH]

I kinda expected these answers as I was typing the question but it was an intriguing thought to me so I posted it. So it looks like no color ccd will ever reach the details a mono can assuming "perfect" (yes I know there are not perfect conditions only ideal ones) imaging conditions.

[sn2006gy]

I kinda expected these answers as I was typing the question but it was an intriguing thought to me so I posted it. So it looks like no color ccd will ever reach the details a mono can assuming "perfect" (yes I know there are not perfect conditions only ideal ones) imaging conditions.

Depends...

 

In "color" they should be pretty on par if you process accordingly and have matched filters that aren't notched.

 

In NB, the Mono could collect image data slightly faster, but in perfect conditions its more about how you process what you get out of it.  

 

To me, it comes down to less about the perfect camera - something better will always come out - but do you get what you want out of it?  Do you absolutely love Hubble Palette or variation there of - go mono. No questions asked.  Do you enjoy color and broadband and chasing dark nebula and star clusters and stuff like that? I can't say one is any better than the other as the real output is how more correlated to the effort put into it.

 

Get dark skies over anything to have the most increase of data.  If you can't get to dark skies then image with as narrow of a narrowband as you can get to bypass light pollution.

[RedLionNJ]

I kinda expected these answers as I was typing the question but it was an intriguing thought to me so I posted it. So it looks like no color ccd will ever reach the details a mono can assuming "perfect" (yes I know there are not perfect conditions only ideal ones) imaging conditions.

I wouldn't put it quite like that.

 

To a large extent, this depends on the stacking method used (and number of frames stacked).

 

If you use something like AutoStakkert for stacking your frames (quite popular for hi-res DSO imaging), then there is no resolution difference between mono and OSC.  Other free apps (e.g. DSS) do use the interpolation referred to above and "lose" resolution in the context of OSC.

[EdwardMH]

If mono and filter wheel are 3-4 (with luminance) times the data collection would it be possible to get a mono and one shot color of the same camera model and combine them to replace lost details from the color ccd?

[EdwardMH]

I use both autoStakkert, and Registax.

[Madratter]

If mono and filter wheel are 3-4 (with luminance) times the data collection would it be possible to get a mono and one shot color of the same camera model and combine them to replace lost details from the color ccd?

First off, that 3-4 times more data is rather misleading. That is NOT 3-4x as much time. It isn't usually even 3-4x as many exposures.

 

Second, yes, what you are contemplating is very possible. It doesn't even need to be the same camera model.

 

Third, as I explained above (and which you seem to have taken the wrong way), it is not necessarily the case you lose any detail at all in the OSC camera.

 

Fourth, even if you don't, there may be efficiencies to be gained in using a mono camera.

 

Fifth, rather than trusting so call experts around here (including myself), I recommend you get a book by someone who at least was able to convince a publisher they were an expert. There are good ones by Bracken and Woodhouse:

 

https://www.amazon.c...06607734&sr=8-1

https://www.amazon.c...06607776&sr=8-1

 

Other authors depending on your interests:

 

Michael Covington

Jerry Lodriguss

 

Both of them occasionally are to be found in this forum.

 

Oh and sixth, sometimes the best way to really learn about something is to just do the experiment.

Edited by Madratter, 29 November 2020 - 02:32 AM.

[decep]

If mono and filter wheel are 3-4 (with luminance) times the data collection would it be possible to get a mono and one shot color of the same camera model and combine them to replace lost details from the color ccd?

This is a relatively common strategy.  Use the OSC for the color data and use the mono for luminance.  Most stacking software will also correct for images with different scale, so the same model is not necessary.

[freestar8n]

It takes time for the CN community to adapt to new ways of processing and how they impact performance with different devices - but when it comes to OSC imaging, if you use Bayer drizzle instead of deBayering, there is no reason to expect any loss of detail compared to mono.

 

It used to be that when aligning and stacking osc images, you would first debayer each exposure into a single color frame - and that process of debayering involves some kind of interpolation over the nearby R, G, B pixels.  And there is likely to be some bloating in that process.  You would then align and stack those color frames.

 

But if you use Bayer drizzle instead of debayering, you just align each raw OSC exposure and map the R, G, B pixels directly into separate R, G, B frames.  This ends up being no different from mono alignment and stacking except each frame contains a mixture of R, G, B pixels.  But the way they get mapped into the destination frame is no different from mono and there is no blurring interpolation going on.

 

The only real differences are that there is an extra G pixel for every R, G, B triad - and the passbands or the OSC pixels are broader than mono.

 

Frank

[EdwardMH]

Thank you all again for your information.

[bobzeq25]

I kinda expected these answers as I was typing the question but it was an intriguing thought to me so I posted it. So it looks like no color ccd will ever reach the details a mono can assuming "perfect" (yes I know there are not perfect conditions only ideal ones) imaging conditions.

The Bayer matrix filter, designed for terrestrial photography, pretty much ensures that.  But modern OSC cameras do an excellent job.

 

Just not as good as mono, particularly in light polluted skies.  The drawback to mono is mostly the expense.

 

One can argue the relevant merits (particularly "how much difference is there?") endlessly, but there is no magic technology on the horizon.  <smile>

Edited by bobzeq25, 29 November 2020 - 03:59 PM.

[sn2006gy]

 

One can argue the relevant merits (particularly "how much difference is there?") endlessly, but there is no magic technology on the horizon.  <smile>

And there will always be something newer/better/bigger/faster the second you do choose something.

[TelescopeGreg]

The Bayer matrix filter, designed for terrestrial photography, pretty much ensures that.  But modern OSC cameras do an excellent job.

 

Just not as good as mono, particularly in light polluted skies.  The drawback to mono is mostly the expense.

 

One can argue the relevant merits (particularly "how much difference is there?") endlessly, but there is no magic technology on the horizon.  <smile>

Eyes open.  Besides the cost, there is also a base assumption (or, need) in the level of capability and automation in the rest of the equipment.  Mono becomes much more cumbersome if one doesn't have, for example, an electronic focuser.  The filter wheel takes up back-focus, so cannot be used with all scopes.  And walking back even further, a complete compute infrastructure with an automatic sequencer (SGP, NINA, Voyager, et al) is assumed, along with a setup that one can let run on its own for hours on end.  Overall, it fundamentally has a lot more moving parts, the sum of which makes it well more than 4x (L, R, G, B) more tedious.  Only through automation is the process one can execute on a regular basis, and that's just the acquisition.  Processing assumes higher-end (higher cost, higher complexity, higher learning curve) software.  Only then can the advantages be realized.  If one already has that infrastructure in place, then the rest is simply "a matter of money".  Otherwise, remember to look at the whole picture.

 

I decided that it wasn't worth it for me, especially when my imaging sessions are only a couple of hours long (including setup and tear down).  I stayed with OSC when upgrading from my DSLR, and went with the ASI2600MC-Pro.  Very happy.
 

[WadeH237]

Mono becomes much more cumbersome if one doesn't have, for example, an electronic focuser.

It depends on how you think about this.

 

With a one shot color camera, it is impossible to change focus between colors.  In that way, a mono without an electronic focuser will show exactly the same chromatic aberration as a one shot color.  The difference is that a mono camera with electronic focuser can refocus on filter changes to mitigate this.

[WadeH237]

And there will always be something newer/better/bigger/faster the second you do choose something.

So what?

 

It's not as if the release of a new technology somehow breaks your existing gear.  The only reason for this to be an issue is if you feel that you must always have the "latest and greatest".  If you understand your requirements before you buy, and choose accordingly so that you are meeting them, then you don't really care about what happens after your purchase.

[WadeH237]

One can argue the relevant merits (particularly "how much difference is there?") endlessly, but there is no magic technology on the horizon.  <smile>

I would agree with this 100% for optics and mounts.

 

For cameras and sensors, I would argue that we are in the middle of seeing huge leaps in disruptive technology.  Just look at how fast CMOS is advancing.  Five years ago, CMOS cameras were the exception, and most of the threads were discussing their shortcomings (real or imagined) vs CCD.  Today, we're talking about the "death of CCD".

 

Beyond that, the "price vs performance and size" ratio of the latest crop of CMOS cameras is totally turning the astrophotography imaging world on its head.  I suspect that you were talking about the underlying physics and technology behind how one shot color sensors work, but if you look at the value available today in astro cameras, it seems almost like "magic technology" to me.

 

There is some irony in this for me, I just bought a new camera a little over a year ago.  I ended up going with CCD for two reasons:  First, I wanted mono, and there are still somewhat limited CMOS offerings here.  And second, the mono CMOS cameras that do exist have sensors that are much larger than either my optics and filters will support.  If I went with a current mono CMOS camera, I would have needed to upgrade my scope and filters - and that would have cost much more than the premium I paid for CCD.  So essentially, CMOS has leap-frogged other parts of my system to the point that I can't make effective use of the latest and greatest CMOS cameras.

[Madratter]

So what?

 

It's not as if the release of a new technology somehow breaks your existing gear.  The only reason for this to be an issue is if you feel that you must always have the "latest and greatest".  If you understand your requirements before you buy, and choose accordingly so that you are meeting them, then you don't really care about what happens after your purchase.

Definitely can be true if, as you say, you know your requirements when you buy, and are careful about that purchase. I have a SBIG STF-8300m CCD camera that I purchased back in early 2013 that I still use and which still serves my purposes very well.

 

As I am fond of saying, there are exactly zero CMOS cameras that I would want to trade in my STF-8300m in order to get.

 

Let me give an analogy. Certainly much has changed on the music scene since Jimmy Hendrix used a Fender Strat to blow everyone's mind. You can and people still do make incredible music with those guitars. I have an American Strat I purchased quite a while ago. I have more expensive guitars. But that is the guitar I gravitate towards and the one I am mostly likely to use when playing in front of an audience.

[TelescopeGreg]

It depends on how you think about this.

 

With a one shot color camera, it is impossible to change focus between colors.  In that way, a mono without an electronic focuser will show exactly the same chromatic aberration as a one shot color.  The difference is that a mono camera with electronic focuser can refocus on filter changes to mitigate this.

Fair point.  I am blessed with a high-end refractor, where the CA is not a problem.  But for someone without an APO that can be an issue.  I'm not sure it would push me to go Mono, given that a lower-end OTA is probably not decked out with all the rest of the automation.  It might be something to consider if one were to want to experiment with IR imaging, where an APO isn't so well corrected.

[bobzeq25]

I used mono plus filters long before I got auto focus.  When the filter changed I checked FWHM.  If it went South, I refocused.  It usually didn't, my Astronomik filters were quite "parfocal".

 

Focus is at least as likely to change with temperature or other reasons.  This is not a reason to not get mono.

 

Again, nothing wrong with OSC.  But OSC guys often tie themselves into knots trying to claim that mono isn't better.  It is. 

 

The dyed glass Bayer matrix filter was engineered to make terrestrial cameras with pleasing color at minimal cost.  It maybe costs the camera maker $10.  That it competes so well with expensive interference filters specifically designed for astro, that send data to all the pixels, is what's surprising.  Part of it is sophisticated Debayering (appropriate phrase <smile> ) techniques, that interpolate data to recover a lot of resolution.  Interpolation is not necessary with mono plus RGB filters.

 

How much better mono is can be debated.  Whether it's "worth it" is just a personal decision, not cosmic right and wrong.

Edited by bobzeq25, 29 November 2020 - 10:56 PM.

[TelescopeGreg]

Both "better" and "worth it" are judgement calls based on both technical and non-technical factors.  One cannot claim "better" without specifying conditions, just as one cannot evaluate "worth it" without understanding one's values and circumstances.

 

All this said, I just noticed that Edward said he uses Autostakkert and Registax (post #11).  If he's doing planetary imaging, I believe most of what has been discussed in this thread doesn't apply.  One does not practically use mono imaging for many of the planetary targets.  Jupiter, especially, rotates fast enough that an entire imaging session takes maybe two minutes.  The other planets and the Moon could use it, but going back to the first post, the focus of the thread should be on getting more detail, and getting more detail in planetary imaging is a matter of telescope aperture, magnification, and atmospheric effects (seeing, transparency).  The camera technology - OSC vs Mono (when one can use it) - is not the primary driver of better results.

 

We seem to love to debate the OSC vs Mono topic so much that I suspect we missed asking what sorts of targets he's actually trying to image...  Edward?

Edited by TelescopeGreg, 30 November 2020 - 01:10 AM.