Colour camera versus mono – price of comfort

Pretty many of astroimaging amateurs at some point need to choose between colour (OSC – One Shot Colour) or monochrome camera. This choice is not easy and many aspects needs to be considered. Colour cameras do not need additional investments (like filters and filter wheel for example), they are quite easy to use and also requires less work during processing. Monochrome cameras on the other hand are more flexible, provide bit better resolution and are more sensitive. Basing on specifications of QHY183 mono and colour version I checked how this sensitivity difference looks like.

QHY183 cameras are modern CMOS based devices – very low noise and also very sensitive. Quantum efficiency peak for mono version is well above 80% and for colour model it is still respectable more than 70%. QE distribution for both cameras can be analysed at the plot below:

QHY183 mono and colour cameras sensitivity plot

RGB sensitivity distribution is pretty standard for colour camera. Mono version of QHY183 is also sensitive in near UV and near IR range as well. But for this example we consider only visible range between 400 and 700nm. In this range average QE of QHY183M camera is 69%. It is extraordinary value. Average QE of QHY183M is better than peak QE for many other dedicated monochrome cameras.

QHY183M average QE in visual range is 69%

When you take a look on filter transmission for colour camera you may think, that these colour plots are all together not much different than black one for mono version. And if you sum all colours up to one area, its average sensitivity will be quite high – it converts to 51%. But you cannot do it like this for OSC camera!

You cannot sum three colours curves to calculate OSC camera sensitivity!

Pixels in OSC camera are not all the same. Quarter of them are with red filter, quarter with blue filter, and half of pixels are with green filter. Once you average these values you get mean QE for QHY163 colour version as 23.6%.

 

All right – it does not look well. But in this case we have instant colour image, and for monochrome camera we need to use filters! So when we add filters to mono camera it will definitely look different. And it does look different. For mono camera when we add standard Baader CCD RGB filters we will get values like this:

These data averages to mean sensitivity of 23.7%. Ha! So there is no benefit here, and colour cameras wins!

Not really. Considering sensitivity we assumed that we use mono camera with RGB filters only, and this is very rare case. Most monochrome camera users expose luminance with L filter (that has 69% average sensitivity), then colours with RGB filters, and then merge it all together. It is LRGB technique. 

In this technique, when we expose half of the time with L filter, and another half with RGB filters, then average sensitivity is 49%.

When we gather more L signal, so the ratio will be 2/3 to 1/3, then average sensitivity raises to 53%.

You can approximately say, that using LRGB method you are able to gather twice as much photons, comparing to OSC camera or RGB technique. This is the case for imaging object that radiates more less the same in the whole spectrum. For edge cases (like emission nebulae) camera efficiency may be analysed at the plot below:

QHY183 cameras QE plot with annotated H alpha and Oiii lines

For the mono camera version the situation is pretty simple: for H alpha band we have about 50% and for Oiii over 80%. For OSC version of camera it is not do obvious anymore. Red pixels catch about 45% of H alpha signal, but only 1/4 of pixels are red, so mean QE of QHY183C camera in H alpha band is about 11%. For Oiii we have more however. Half of pixels are green and gather Oiii photons with 70% efficiency, and additionally 1/4 of pixels are blue and gather about 40% of Oiii signal. It gives average QE for Oiii at level of 45%.

In real life there is quite lot of another criteria we need to consider before we choose camera. Resolution with monochrome cameras is a bit better – depends on sampling. For undersampled images difference in resolution can be clearly visible. Also colour distribution differs between RGB filters in OSC cameras and dedicated CCD filters for mono cameras. But the most important aspects to consider will be: camera application, sky quality and sensitivity.

And decision definitely will not be easy 🙂

Below few images for comparison. These images were made with exactly the same setup under similar conditions with cameras QHY163C and QHY163M. These are 180s frames made with 130mm refractor. 

M106 area image made with color (left) and mono cameras.
Zoom in of the frame above
Star intensity and SNR comparison. At image from monochrome camera (left) star intensity is 133,000 ADU, and SNR is 170. At image from colour camera the same star intensity is 59,000 ADU and SNR is 101.

And this is the price to pay. And comfort? With OSC camera you don’t need to invest in filters. You do not need to plan session, or split it over more nights, when you was not able to complete all RGB filter frames. Just attach camera to the telescope and shot as many frames as you want 🙂  

Clear skies! 

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.