Two CPLs = Zoomable ND?


Status
Not open for further replies.
V

vince123123

Guest
#1
An inspiration just hit me today.

In the past, I used to play with linear polarizers - when you hold them at 90 degrees to each other, the light gets totally blocked (ie totally opaque). By adjusting them at varying angles to each other, you can get adjust to get a range of light transmissions.

Would the same work for circular polarizers? I was thinking if that is possible, we could actually get a zoomable ND, from ND2 to ND8 to ND100 with no problems.
 

Artosoft

Senior Member
Aug 31, 2005
3,710
0
0
Tanjong Katong
#2
There is a product called variable ND filter, which is use two CPL (or 1 CPL and 1 LPL IIRC). You can try using 2 CPL (or 1 CPL and 1 LPL IIRC), but your result is not as good as (consider bad) the variable ND filter.

IIRC, the price of 1 variable ND filter is more expensive than few (at least 3 or 4) different ND filters.

Regards,
Arto.
 

azul123

Senior Member
Dec 4, 2004
2,776
0
0
Eastern Bloc
#3
When I did a search thru the web sometime back, some forumers in FredM or some other site, mentioned that the Variable ND can be achieved by using a LP and CPL stacked.

../azul123
 

V

vince123123

Guest
#4
The problem with using a Linear Polarizer as one of the components is that Autofocus cannot work with Linear Polarizers.

Artosoft: how come you say that using two CPLs is worse than using the variable ND - I thought you said the variable ND is indeed using two CPLS?
 

azul123

Senior Member
Dec 4, 2004
2,776
0
0
Eastern Bloc
#6
Yes, that's the one to get and it's pricy...

../azul123
 

azul123

Senior Member
Dec 4, 2004
2,776
0
0
Eastern Bloc
#7
Alternative... Hoya has the NDx400 or maybe cheaper option is to stack several ND8s together?

But not turn variable.

../azul123
 

Artosoft

Senior Member
Aug 31, 2005
3,710
0
0
Tanjong Katong
#8
The problem with using a Linear Polarizer as one of the components is that Autofocus cannot work with Linear Polarizers.

Artosoft: how come you say that using two CPLs is worse than using the variable ND - I thought you said the variable ND is indeed using two CPLS?
It is looks like variable ND use unusual Polarizer filter (whether one is Linear, one is Circular, or both is Circular polarizer filter). That's explain why it is so expensive. If you use 2 CPL (found in market), most probably you will get non linear ND filter (some part of ND filter is black than other part), or some changes in color ND filter. There is a lot of discussion I read before when I come to this idea too.

It is more economical and better to use several ND filter than variable ND filter, IMHO.

Regards,
Arto.
 

roygoh

Senior Member
Jan 18, 2002
5,011
0
0
Northwest
Visit site
#12
For variable ND using polarizers, the first one (closest to subject) should be linear. The second one (closest to lens) can be linear or circular, depending on the camera. Most modern cameras would require the second polarizer to be circular in order to let the metering work correctly.

If the first one is circular then the second one cannot achieve the variable ND effect, because the first CPL would produce "depolarized" light that will pass through the second one with the same transmission regardless of alignment.
 

azul123

Senior Member
Dec 4, 2004
2,776
0
0
Eastern Bloc
#13
roygoh said:
For variable ND using polarizers, the first one (closest to subject) should be linear. The second one (closest to lens) can be linear or circular, depending on the camera. Most modern cameras would require the second polarizer to be circular in order to let the metering work correctly.

If the first one is circular then the second one cannot achieve the variable ND effect, because the first CPL would produce "depolarized" light that will pass through the second one with the same transmission regardless of alignment.
Thanks for your input Roy, saw Sammy888's and your postings on CPL/LP filters and they are indeed very technical and detailed.

../azul123
 

lsisaxon

Senior Member
Nov 29, 2004
11,941
0
0
#14
An inspiration just hit me today.

In the past, I used to play with linear polarizers - when you hold them at 90 degrees to each other, the light gets totally blocked (ie totally opaque). By adjusting them at varying angles to each other, you can get adjust to get a range of light transmissions.

Would the same work for circular polarizers? I was thinking if that is possible, we could actually get a zoomable ND, from ND2 to ND8 to ND100 with no problems.
Unfortunately, it doesn't work with circular polarizers. Variable optical attenuators/ND actually use 2 linear polarizers and ideally, the cosine of the difference of the polarization angles gives the transmission coefficient.
 

lsisaxon

Senior Member
Nov 29, 2004
11,941
0
0
#15
When I did a search thru the web sometime back, some forumers in FredM or some other site, mentioned that the Variable ND can be achieved by using a LP and CPL stacked.

../azul123
Lin pol and Cir pol stacked do not give as good an attenuation as 2 lin pol stacked. To some extent it can work but not very well. :)
 

V

vince123123

Guest
#16
I was wondering, the CPL converts the LP's polarised light into a type of light that can work with AF? Cos I do know that LP alone won't work with AF.

For variable ND using polarizers, the first one (closest to subject) should be linear. The second one (closest to lens) can be linear or circular, depending on the camera. Most modern cameras would require the second polarizer to be circular in order to let the metering work correctly.

If the first one is circular then the second one cannot achieve the variable ND effect, because the first CPL would produce "depolarized" light that will pass through the second one with the same transmission regardless of alignment.
Azul - could you drop a link to Roy and Sammy888's threads on CP/LP?

Thanks for your input Roy, saw Sammy888's and your postings on CPL/LP filters and they are indeed very technical and detailed.

../azul123
 

azul123

Senior Member
Dec 4, 2004
2,776
0
0
Eastern Bloc
#17
Status
Not open for further replies.
Top Bottom