Are chip specifications fairly universal, or do they vary greatly via the hardware they get paired with?

[Trippelforge]

I picked up a D5300 and always wondered what the typical related AP centric specifications were for it. Luckily the chip it uses has been featured in a few dedicated AP cameras. As the Nikon D5300 uses the Sony IMX193, which was featured in an older AP model under QHY247C. With that in mind you can actually find the specifications on it, things like well capacity, read out noise etc. However I was curious if those specifications transferred over to other devices that use the same chip.

For instance the chip is listed at having a 36ke full well capacity on the dedicated camera's specs. It also lists the peak QE at 74. Does that mean that the Nikon does as well? I know things like readout noise are greatly effected via cooling, so obviously the DSLR isn't going to retain the values. But I am still curious if cooling would be the only reason?

Thanks for filling me in!

 

 

Edited January 6, 2023 by Trippelforge

[Richard_]

When you look at the read noise, full well and dynamic range charts you'll notice that the x-axis shows gain. With astro cameras, you can usually set the gain in the image acquisition software to whatever you like which will affect read noise etc.

With DSLR's you cannot control gain but instead can set the ISO value which has a similar effect. Personally, I don't know how to convert an ISO value to gain, so unless someone else has created a readout noise/full well/dynamic range chart against ISO, you'll struggle to use the charts expressed in gain on the x-axis. 

Finally, the QE chart should be somewhat similar between DSLR and dedicated astro camera but there will be a difference in the transmission values (ie the light which reaches the sensor). Assuming the DSLR hasn't been astro-modified, it will likely have an IR filter placed in front of the sensor which blocks certain wavelengths from hitting the sensor. I've provided an example image below showing the transmission of light at certain wavelengths for a canon camera (red) against a baader acf filter (blue). When imaging deep space objects, one of the most abundant wavelengths emitted is those from Hydrogen-alpha at 656.3nm. This is shown as a vertical green line in the picture below. This is in the red wavelength region and is what contributes to the beautiful reds you see in nebulae images. As shown by the graph, the transmission value at H-alpha wavelength is about 25% so the IR-block has essentially blocked 75% of that signal from hitting the sensor! The advantage of dedicated astro cameras is that this filter is not placed on the sensor and the transmission line is more like the blue one. You can see the transmission at H-alpha is just over 95% so hardly any of this signal is blocked. 

Source: https://www.dslrastromod.co.uk/technical.html

 

Here's a real world example. I captured ~4hrs of the rosette nebula using my stock Canon 600D a few years ago. I then bought an ASI533MC and re-imaged the rosette nebula for ~2.5hrs using all the same equipment aside from the camera used. Ignoring my mediocre processing skills at the time, you can see a huge difference in the amount of red H-alpha captured between the cameras despite having less imaging time with the ASI533MC. This is due to the IR block present on the Canon per the above comment which limited the amount of H-alpha captured. 

[The Lazy Astronomer]

Dark current is the only parameter affected by cooling. QE is a constant regardless of camera settings, and read noise, full well, etc. will remain constant for a specific gain/ISO setting. As mentioned above though, you don't know how your ISO setting relates to the gain setting on the graph. The scales even very between astro cam manufactures - examples below of the read noise graphs for zwo2600mc and qhy268c. Both use the imx571 sensor, but different gain scales. E.g. zwo quote ~1.5e- read noise at gain 100, but for qhy, it's at gain 60 (in mode 1).