Puzzling Colour

[Aster]

This is my third test with the 400D and with my inexperience I am a bit puzzled with the difference in background colour in the converted TIF frames.

All images are 5x10 minutes shot with a unmodded 400D, shooting Raw+Jpg large, seeing was lousy.

RAW converted to TIF with Canon Digital Photo Pro.

All images stacked in DSS with darks, No flats or bias.

Could someone please explain in simple terms why the difference in background colour, also, is it better to use the Original RAW frames or to convert the RAW into TIF ?

PS: First image is a untouched JPG stack

[RB]

Stack the RAW files directly in DSS.
Then save the stacked image as a tiff and open it in photoshop or something similar.
From here you can start to stretch the data to bring out the detail in your image (work in 16 bit mode).

Adjust your levels, curves and colour balance till you have a nice 'bell' shaped histogram without any clipping at either end.

I had a quick go at one of your frames as an example to bring out some detail.

[jase]

RB's on the money.

As you're converting data between file formats, you must consider the source and destination bit space and the impact of your actions. Clearly changing data from say 12-bit down to 8-bit results in a series of calculations to determine how to scale the data with minimal data loss. Some calculations you have control over, while others are automated. What you need to determine is how the data is scaled between the bit spaces. Going from 12-bit to 16-bit will not result in any data scaling, but the image will probably look rather dark until you stretch it. Going the other direction 16-bit to 12-bit or 8-bit, the data needs to be scaled to continue showing the full dynamic range.

[Aster]

Should have been clearer with my question.

All 4 images are as originally stacked, no levels or curves or anything.

The main thing which had me puzzled was why the change in the background hue/colour in the converted RAW to TIF Stack ?

Meaning, I converted the Original RAW frames into TIF before stacking in DSS.

Jase, you made me look at the different bits, have to think a while about that :(

Andrew, you used the converted RAW frames Stack, the one with the different background, to fiddle with. I found that harder to get the right balance in CS2 then the original RAW stack.

So, it is best to stack the original RAW frames, not convert them to TIF and than stack.?

Thanks to both of you

[RB]

Yep we realise that and that's why the two raw stacks are darker, it's what Jase was talking about with the bit depth stretch etc.
"Going from 12-bit to 16-bit will not result in any data scaling, but the image will probably look rather dark until you stretch it."


When you convert the RAW's to tiff and then stack the tiff's you end up stacking the stretched converted frames.
Not what you want to happen.


Yes I used this one because it had more range, don't forget these images have been down-converted for the web so I chose the jpg which had more room to play with before clipping.
Yes it's best to stack the original RAW frames and not convert them to tiff first then stack.
The final RAW stacked file will be darker but this will give you a bigger dynamic range to manipulate.

:)

[Aster]

Thanks Andrew, restacked the 6x10min RAW frames in DSS and did some beginners work in CS2.

Another question if I may. Is it better to use no interpolation whilst stacking RAW frames in DSS, OR does that depend on the DeepSpaceObject ?


Thanks Again

[RB]

Good work Aster ! :thumbsup:

As for DSS, sorry but I seldom use it so not sure about that.
Hopefully someone else can answer that.

:)