[photoptimist]

You are right. It may be bright-mode which I used. 512x in-camera averages usually fail with magenta tint as I demonstrated.

That's not that surprising if the color was changing during the sequence. The "average" of a shot with blue sky and one with red-lit clouds would be magenta.

[MossyRocks]

I actually like the one where it failed as it produced a rather weird surrealness. I almost want to do it myself because I can, now to find a scene where that could work.

[TedH42]

Interesting discussion - bookmarking this thread.

[Aaron28]

in intervals for star trails for sure....don't want my sensor burning for an hour without rest....
use hdr or interval shooting on some water features to smooth things out as I do not have a ND filter

[edri]

I once did this on a hike. I forgot the ND filters at home. It was the only time I used this option.

[biz-engineer]

That's not that surprising if the color was changing during the sequence. The "average" of a shot with blue sky and one with red-lit clouds would be magenta.

At least the change of light is something to consider around sunset times.... especially why exposure bracketing doesn't work that well for long exposures (e.g 5 minutes), and why ND grad filters are still usefull.

[MJKoski]

It is not because change of light. That magenta pops up with enough stacking indoors when nothing changes. It is limitation of camera hardware and/or software when doing the calculations. Something gives in there. I also have some material of grey dull rainy evenings with heavy magenta cast as the outcome.

[Madaboutpix]

To me, I must admit, the decision which of the two approaches to use is less influenced by noise level geekiness (nothing wrong with that, BTW) than by the aesthetic of the subject motion blur that is created. I just happen to find the ND filter look more natural and pleasing, at least in the comparisons I have seen.

Here is an older thread featuring a brief comparison of that kind: ND filter or Stacked interval Composite - PentaxForums.com

[MJKoski]

I too find some composites weird looking. Especially if the single frames of which the output is composed of are too short exposures. Either few long ones or huge number of short ones (or, huge number of long ones ). As suggested, using light ND filter of only 1-3 stops and combining averaging is sometimes a good approach.

[BrianR]

You are right. It may be bright-mode which I used. 512x in-camera averages usually fail with magenta tint as I demonstrated.

Bright mode would be the only way to get those crisp snapshots of the ice blobs as they swirled around. Bright should also be easier on the hardware than Average from a computational standpoint, and since you've shown it can handle 512 exposures I'd bet it has no problems with the max number of frames and probably even beyond that (barring the sensor overheating or some other mechanical failure).

[MJKoski]

I have one 1024x composite from the same spot! It looks a total mess Too much stuff going on there.

[BrianR]

I have one 1024x composite from the same spot! It looks a total mess Too much stuff going on there.

Mess how? I wouldn't expect any computation errors from a Bright composite, but that wouldn't prevent it from looking like a visually busy disaster.

[biz-engineer]

It is limitation of camera hardware and/or software when doing the calculations

Interesting. I ran on test comparison between composite average at base iso underexposed, composite average at high iso to compensate for underexposure, composite adding underexposed, composite adding at high iso. And I think I (kind of) understood what the camera does. When averaging, the camera does a bit wise average before storing the result in DNG without rescaling to 14bits, this leads to banding because the small gradations of tones are discarded by the average. While for the adding mode, the camera performs sequential 14bits wide word additions, then store the results in DNG. In the case of average, shadows that code barely over 5bits remain 5bits deep, and as the averaging increases the value can even become corrupted if one bit on a word flips from 0 to 1 but least significant bits don't change, so this is the banding I observed. Now, the adding mode is much more interesting, as it is effectively pushing up least significant bits to higher positions within 14bits words as the value of the stack increases, there is no loss of information, on the contrary as the stack grows the values get more refined until reaching the 14bits range, as which point the image quality is at its max (better than ISO100 quality since the DR is 14bits and the read noise has been cancelled out along the way). So... the best is to lower to ISO100 and count the number of stops of underexposure, then select the number of frames to be added based as 2^(# of stops of underexposure), and this maxes out image quality without any banding or artifacts.

[MossyRocks]

Interesting. I ran on test comparison between composite average at base iso underexposed, composite average at high iso to compensate for underexposure, composite adding underexposed, composite adding at high iso. And I think I (kind of) understood what the camera does. When averaging, the camera does a bit wise average before storing the result in DNG without rescaling to 14bits, this leads to banding because the small gradations of tones are discarded by the average. While for the adding mode, the camera performs sequential 14bits wide word additions, then store the results in DNG. In the case of average, shadows that code barely over 5bits remain 5bits deep, and as the averaging increases the value can even become corrupted if one bit on a word flips from 0 to 1 but least significant bits don't change, so this is the banding I observed. Now, the adding mode is much more interesting, as it is effectively pushing up least significant bits to higher positions within 14bits words as the value of the stack increases, there is no loss of information, on the contrary as the stack grows the values get more refined until reaching the 14bits range, as which point the image quality is at its max (better than ISO100 quality since the DR is 14bits and the read noise has been cancelled out along the way). So... the best is to lower to ISO100 and count the number of stops of underexposure, then select the number of frames to be added based as 2^(# of stops of underexposure), and this maxes out image quality without any banding or artifacts.

Interesting. I congratulate you on tracking this down and digging into it. I'm probably going to use that last bit the next time I try shooting some wide field astro shots as the ultrawide lens I have is really slow.

[jbinpg]

Thanks for this, biz. Will need to try it out. I think the new APS-C sensor is 16-bit which will be a different beast so we will need to experiment all over again.