R5 - Long Exposure Noise Reduction

[Quarkcharmed]

I made this 'discovery' after >1 year of the camera use, so just sharing it here in case it helps someone else.

Apparently the long exposure noise reduction on the R5 affects RAW. It works pretty well and leaves only few hot pixels, unlike long exposures without the noise reduction. Pretty impressive.

I recommend (and the manual does) to set it to Auto if you use it for landscape photography, so it's only applied for exposures longer than 1 sec.

The caveat: the camera apparently takes a dark frame after the actual exposure, and it lasts about the same time, so the effective exposure time becomes two times longer.

 

[Franz Kerschbaum]

Its exactly the same as in all digital eos cameras the last two decades...

 

[Quarkcharmed]

Its exactly the same as in all digital eos cameras the last two decades...

I had a pretty strong impression it only worked on jpegs.

In fact the long exposure noise wasn't hugely bad on my previous 5DIV so I never needed it. With the R5, the long exposure noise is quite heavy in exposures longer than 30s.

 

[Franz Kerschbaum]

Its the same approach since my first 20D in 2005. The larger number of hot pixels in the R5 is an interesting observation. I hope the number is not increasing over time. Maybe you should have them remapped by cps. My current R5 shows about the same as my Ra and much less than my 7d2. Unfortunately the sensorcleaning remapping seems not to work with the r5.

 

[Quarkcharmed]

Its the same approach since my first 20D in 2005. The larger number of hot pixels in the R5 is an interesting observation. I hope the number is not increasing over time. Maybe you should have them remapped by cps. My current R5 shows about the same as my Ra and much less than my 7d2. Unfortunately the sensorcleaning remapping seems not to work with the r5.

Up to 30s the R5 is perfectly bearable without LENR. Now if you go for a few minutes, it becomes a real problem. With the LENR it works fine. I even tried it at ISO 800 and 4-minute exposure at night, looked usable.

Before recently, I was really convinced it only worked on jpegs, perhaps I'd read something misleading at some point, so those who already know it and use this mode can disregard this topic.

 

[Tristimulus]

Its the same approach since my first 20D in 2005. The larger number of hot pixels in the R5 is an interesting observation. I hope the number is not increasing over time. Maybe you should have them remapped by cps. My current R5 shows about the same as my Ra and much less than my 7d2. Unfortunately the sensorcleaning remapping seems not to work with the r5.

Up to 30s the R5 is perfectly bearable without LENR.

Not in a hot climate.

During the winter LENR is not needed at my location but warm summer days are a very different story. Thermal noise increase with increasing temperature.

Now if you go for a few minutes, it becomes a real problem. With the LENR it works fine. I even tried it at ISO 800 and 4-minute exposure at night, looked usable.

Before recently, I was really convinced it only worked on jpegs, perhaps I'd read something misleading at some point, so those who already know it and use this mode can disregard this topic.

 

[ProDude]

I've been using LENR on virtually every camera I've owned over the last 20 years. It's ALWAYS been a wonderful feature and very effective on ALL cameras. Even a 1" sensored RX10MkIV I had. So indeed I highly recommend it for longer exposures. I've never seen a hot pixel resulting from it either. Love the utter lack of noise it renders which simply saves any issues needed for post processing.

 

[elroyie]

LENR is affecting the RAW in all Canon cameras, it is not like High ISO speed noise reduction, that affects the JPEG. LENR is highly recommended if you are doing long or extremely long exposures to get rid of the 'burned' pixels and the phenomena of the fixed pattern noise, the last one is very hard to correct in post processing.

 

[CameraCarl]

I have seen some recommendations to turn it off for astrophotography because of the long second exposure. What does everyone here do?

 

[Quarkcharmed]

I have seen some recommendations to turn it off for astrophotography because of the long second exposure. What does everyone here do?

LENR is for only specific cases I think, and not a very good fit for the night sky. However as above I used it for long exposures for the night foreground and it worked well. With astroscapes, you normally compose the foreground and the sky anyway.

For the sky, I've recently started using stacking of multiple 10-15s exposures and DeepSkyStacker. You may shoot a separate dark frame for it so no in-camera LENR required in this case.

 

[G Dickson]

I have seen some recommendations to turn it off for astrophotography because of the long second exposure. What does everyone here do?

I don't use it. With astro the total integration time is hugely important. I would far rather use that time gathering data than doubling my exposure time. I will then either shoot separate dark frames or often just not bother if I am using my R6 or 6Dii.

 

[Franz Kerschbaum]

If temperature is constant or only changing slowly astrophotographers typically reuse darkframes to correct their many, many light exposures. Typical example You expose 5 dark frames of eg 60s and average (or median combine) them. With this well "exposed" master dark D60 you correct ALL your maybe hundreds of 60s light exposure (I) at the same ISO setting and approximate temperature. You can even use it for exposures of other lenghts (t) if you also use "0s" Bias Frames (B) and calculate your own dark frames for any given exposure time following this easy equations: d1=(D60-B)/60 and then Icorr=I-B-d1*t. This all saves precious observing time.

 

[ChrisBeere]

My R6 suffers from thermal noise on long exposures that rears its ugly head when the scene is underexposed and needs shadow recovery, as is sometimes the case in nightscape scenes.

If the scene is well exposed and has strong signal to noise ratio then you can shoot 8 minute exposures with no problem as the thermal noise simply gets swamped by the signal.

Here is my R6 thermal noise footprint


R6 2 minutes ISO 1600

This a 2 minute capped shot - you can see the thermal noise at the edges of the frame.



Here is the same exposure with LENR applied.


R6 2 minutes ISO 1600

Again, this a 2 minute capped shot - you can see the thermal noise has been subtracted out by the LENR dark frame.

 

[John Sheehy]

Its exactly the same as in all digital eos cameras the last two decades...

I had a pretty strong impression it only worked on jpegs.

In fact the long exposure noise wasn't hugely bad on my previous 5DIV so I never needed it. With the R5, the long exposure noise is quite heavy in exposures longer than 30s.

The R5 runs a lot hotter than the 5D4 does, especially when the 5D4 is using OVF mode.