[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
Re: [APML] RG200 characteristics. Was: New Images From Last weekend
Hi Bert,
The summed sensitivity gives you density on film - true. But for accurate
OIII response the parts of that sum need to be 50% green and 50% blue
(approximately). Now look again at the curves for RG200...
http://www.kodak.com/global/en/consumer/products/pdf/e42.pdf
... and you can see that that's not the case. Follow the 500 line
vertically and look carefully at the blue / green crossover - you can see
that the green component is much higher than the blue component. The mix
of green and blue is nothing like 50/50 - the film records the 500.7nm
spike almost wholly in the green with hardly any blue. That is the whole
issue here. It doesn't matter what the summed sensitivity is - you can
have the summed sensitivity as high as you like but if the _components_ of
that sum are not 50/50 then it is not accurate OIII response.
A good place to read up on this is Chuck's article...
http://www.aa6g.org/Astronomy/Articles/emission.html
Look at figure 5 for the "ideal" crossover. Some films can come very close
to this ideal - see again the crossover for Fuji NHGII 800 towards the
bottom of this page:
http://www.astrocruise.com/xppfvnhg.htm
Now look back at the RG200 crossover and you will see what I mean by
"wierd". It is nowhere near accurate. They get away with it because these
films are designed to record broadband light, but the blue / green
crossover makes this film inappropriate for any kind of scientific
work. From a scientific perspective this film cannot record the visible
spectrum accurately. The 500.7nm OIII spike merely highlights the problem
but the problem is there anyway. Fortunately for astrophotography,
hypering seems to correct the crossover problem to some degree.
--Philip
At 21:42 18/10/01 -0700, you wrote:
>Hi Philip:
>OK, we're definitely communicating on the H-alpha line: it's falling in the
>middle of the cyan-forming layer where the sensitivity of the other layers
>is zero, so the sensitivity of that layer alone will determine what
>registers on the film. More cyan, I guess, will "print" as more (brighter)
>red.
>
>Where I lose my grip is in your line immediately after the Kodak URL. When I
>look at the curves, I see a certain quantity of sensitivity in the yellow
>(minus-blue) and a certain quantity of sensitivity in the magenta
>(minus-green) layer. What I DON'T see is the weird result you refer to. I
>just see two curves and I would like to *predict* from those curves whether
>I will get good registration of bluish-green emission lines from that film.
>
>Taking a different tack, perhaps the point I'm missing is in your comment
>about the *relative* quantity of blue and green response. Are you saying
>that if the yellow and magenta curves crossed at exactly 500.7 (like NHG800)
>rather than at 485 (or whatever that RG200 cross-over wavelength is), then
>the OIII color would be more accurate? That would be great, because then I
>could simply look to see where the curves for any given film crossed. I
>still wouldn't understand *why* the cross-over point wavelength was more
>important than the summed sensitivity at 500.7, but at least I would have a
>predictive tool I could use.
>
>I hope I'm making my problem clear. Sometimes just formulating the question
>makes the answer obvious, but that hasn't worked for me so far.
>
>Thanks for your thoughtful input....
>Bert
>
>Bert Katzung
>katzung1@home.com
>www.astronomy-images.com
>
> > Hello Bert,
> >
> > I think I understand what you are saying. In the case of 656.3nm it's
>only
> > quantity that matters, because the whole of that spike can only fall in
>the
> > red (cyan forming) layer. It can only evoke a response in the red
> > layer. But in the case of the 500.7nm spike it is more complicated
>because
> > that does not fall right in the middle of one of the colour forming
> > layers. In fact it falls almost exactly mid way between the blue and
>green
> > layer. So if the film is responding correctly, that 500.7nm spike should
> > produce a colour that is mid way between green and blue (sort of 'aqua
> > marine' or whatever you want to call it:-) but if you look at that curve
> > for RG 200 again...
> > http://www.kodak.com/global/en/consumer/products/pdf/e42.pdf
> > ... you can see something wierd - that 500.7nm spike will produce a
> > response that is almost wholly in the green with very little blue. It
>will
> > produce a predominantly green colour, and when green is mixed with red
> > (656.3nm) it produces a colour that veers towards the yellow (minus
>blue) -
> > in other words Ha regions that are orangey-red. That's simply wrong and
>is
> > very typical of Kodak films. If the 500.7nm spike produced the correct
> > blue-green colour then, when mixed with the red, it would de-saturate it,
> > producing a pinkish red. That is correct, and it is the colour that we
>see
> > from tri-colour work, both CCD and film.
> >
> > So it is not just a case of quantity with the 500.7nm spike - you can't
> > simply add up green and blue, and because the total is OK, say that the
> > response is correct. In this case it's also the quality that's critical -
> > the _relative_ quantity of green and blue response. And as mentioned
> > above, they should be about equal.
> >
> > OTOH Fuji films often have a good OIII response and in the case of NHG II
> > 800 it is terrific. See here (scroll down the
> > page): http://www.astrocruise.com/xppfvnhg.htm But just as if to deny
>the
> > astrophotographer his perfect film, these films often have a weak Ha
> > response, as is the case with NHG II 800. The exact opposite of the Kodak
> > problem! If only Kodak and Fuji could get together... there is no doubt
> > whatsoever that technology exists to produce an astro film with perfect
> > characteristics.
> >
> > (Much of the above owes nothing to me and everything to Chuck - I hope I
> > have it reasonably correct)
> >
> > --Philip
> >
>
>
>
>-- APML Archives at <http://astro.umsystem.edu/apml/> ---
> Unsubscribe at <majordomo@seds.org>
Philip Perkins - philip@astrocruise.com
Wiltshire UK & Luberon France
Astrocruise - http://www.astrocruise.com
-- APML Archives at <http://astro.umsystem.edu/apml/> ---
Unsubscribe at <majordomo@seds.org>