Re: Optical filters

I'm trying to make compatible the filter quality with the souche wave: 1.
Filter for souce (depends on exitation-Ex) 2. Filter for Emission (depend on
the emission of the partikle (Em)
Usually the different are about 15-5- nm.  The problem of point 1 is that
the source is low bell shape with broad spectrum like mercury lamp we need
as monochromatic as possible.  The problem with point 2 is we need a
compatible filter depends on from the source dand the extrinsic dye we use
for our particle/ organelle.  Another problem are nowadays we are looking
forward on using intrinsic dye for living cells like chlorophyle in
plankton.  The source could be NIR, FIR,  UVA, B, C and the filter for the
source & Emision should be compatible eq the Ex Em NADH using mercury lamp.
My opinion, you should scan all the Invitrogen Ex/ Em and one cell system to
advance your filter custome.  Also the one signal syst to catch the signal
from your Em filter (due to Image processing).	If you think of chromo dan
omega, why not also in the mw & hp spectrum .  We buy filters from china, it
is cheaper.
All the best,
peni

----- Original Message -----
From: "Howard Shapiro" <hms@shapirolab.com>
To: cyto-inbox
Sent: Tuesday, November 07, 2006 11:59 PM
Subject: Re: Optical filters


> James Marvin wrote:
>
> > Has anybody looked very hard into quality of optical filters from
> > either Chroma or Omega?  Or I guess some other company you buy your
> > filters from.
>
> Problems with filters from the major manufacturers are more likely to
> be due to improper specification of the filter than to "quality"
> issues. The specification should include angle of incidence and
> passband; note that even short and long pass filters effectively have
> a passband in addition to a cut-on or cut-off wavelength. This was
> straightforward for older instruments; the angle of incidence for a
> filter was typically at or near zero degrees, and that for a dichroic
> at or near 45 degrees. For newer instruments, particularly those from
> BD that use octagons or trigons, the angles of incidence are
> different, and one cannot simply swap filters from an older
> instrument into the optical system of a newer one.
>
> Typical passband specifications only go so far. For example, suppose
> you have a filter with a 525 nm center wavelength and a 30 nm width
> at half maximum. That tells you how the filter responds in the
> spectral region in which you want to use it, but says nothing about
> the range in which you *don't* want light to get through. You need to
> specify the "blocking" on the out-of-passband regions to be sure that
> interfering light at wavelengths you don't want will not get through
> the filter along with the light in the passband region that you do want.
>
> Manufacturers generally supply transmission curves with their
> filters, with transmission on a linear scale. This is relatively
> useless; you could get pretty much all the information on such a
> curve from the maximum percentage transmission value. You would
> ideally like your filters to be blocked to O.D. (optical density) 6
> or so outside the passband. O.D. is a logarithmic quantity; a filter
> with O.D. 1.0 transmits 10% of the incident light, one with O.D. 2.0
> transmits 1%, and so on. So, what you want from the manufacturer is
> an O.D. curve. If you can see any little warts along the baseline of
> a transmission curve, they will almost certainly correspond to
> spectral regions in which the O.D. of the filter is considerably
> lower than 6.
>
> Although you can do a transmission measurement on a filter in a
> spectrophotometer if you happen to have one handy, you really can't
> measure an O.D. curve unless you have a much more expensive
> instrument, with a much higher dynamic range; that's what the filter
> manufacturers use. In my experience, all of the manufacturers will be
> happy to show you both transmission and O.D. curves, and some even
> have the tabulated values on line in Excel or an equivalent format.
> They'll work with you, but you have to tell them what you want.
>
> I've been planning to write up a technical note on this topic, but
> this should do for now.
>
> -Howard
>