View Full Version : Petzval nodal point ???

29-Apr-2010, 23:34
So I have this lens and would like to know 3 things:

1) Where the nodal (iris) point is so I can correctly measure the focal length and figure the speed, from all the drawings on the net it looks like in the center is where they put it.

2) Is there any reason to cut a waterhouse stop slot in a projection lens, or am I just making problems for myself & the lens??

3) How do you check to see if the elements are in the right configuration ? What happens if you rearrange the glass?


Emmanuel BIGLER
30-Apr-2010, 03:26
Hello from France !

Interesting task to make reverse-engineering of a lens designed in the XIX-st century !

To determine the focal length, I can suggest three different methods, without guarantee for precision, but with guarantee that you'll not have to care for any prior knowledge about the nodal points at all !!

The first method is based on the following optics law, which is common to all lenses of any design, even the most exotic and assymmetric ones.
- A - finding the focal length by the 2f-2f method
- first find where the focal point F' is located. This is eaay, mount the lens on your camera and focus on a distant object. Measure the distance between the ground glass and a reference plane on the lens, for example the lensboard. Note that this distance is _not_ the focal length. But the setting of the camera is now the infinity-focus position
- second, prepare a grid on paper as a test target, and try to focus so that the image has exactly le same dimension as the object. You can measure the image of the grid on the ground glass with a ruler. In order to reach this position you have to increase the bellows draw by _exactly_ one focal length with respect to the infinity-focus position. This rule is valid for any lens design.

- B - finding the focal lenth by the "x = f. tan(Theta)" method
A second method, which may sound really unusual, will give you an estimate of the focal length without any prior knowledge about the cardinal points : foci F, F' and principal (or nodal) points H=N and H'=N'.
Mount the camera on a good tripod head with a panoramic movement. Most tripod heads have a panoramic movement so you do no need anything special to to this measurement. No "nodal slide" required here !
Now, aim at a distant object like a spire or any element in the landscape and measure how much you have to rotate the whole camera + ground glass in order to get a given displacement of the image on the ground glass, for example measure the rotation angle that will move the image by plus or minus 2" (plus or minus 5 cm) on your ground glass. The actual image size on a 4x5" film holder is something like 94x120 mm so it is easy to reach plus or minus 5 cm horizontally.
In order to get a more precise measurement of the angle, you can attach a wooden bar to the camera and measure how much the end of the bar moves with respect to a piece of paper glued on a table when tou rotate the camera.
Now you'll estimate the focal length as follows.
Il the bar is 1m long, and if the end of the bar moves by, say 10cm when the image moves by 10 cm on the ground glass, then the focal length of the lens is equal to the length of the bar ! Obviously the end of the bar will not move by 10 cm, since you do not have a focal length of 1 m (40 inches) ; may be it will move by 2 cm, in this condition the focal length is 1/5-th of the bar length, i.e. about 20 cm.

You can compare both methods and have a reasonable idea of the focal length ; not very precise but sufficient to determine the f-number.

A third method, more precise requires a device named "nodal slide"
- C - by the nodal slide method

The method may seem strange as well. You should mount the lens on something that can rotate like a tripod head, but you should be able to slide the lens by an amount, of, say twice as long as the lens length itself. A piece of paper, or your ground glass should be mounted on a separate support. Separate from the lens, in the operation, the distant object is fixed, the ground glass stays fixed after focusing, only the lens rotates.
You cannot use the bellows, hence you'll get a lot of parasitic light in the image but this is not very important.
Make the image of a distant object, and try to find, by sliding the lens back and forth on the nodal slide (the front standard of a monorail camera plus rail is perfect), a position where the image on the fixed ground glass no longer moves when you rotate the lens around a vertical axis.
In this position, where the image is stationary on a fixed ground glass, with respect to rotating the lens, you have in fact placed the rear nodal point N' just above the rotation axis of the tripod head.
To do this you need a monorail camera or a nodal slide device.
Measure the distance between the rotation axis and the sharp image of a distant object on the ground glass, this is your focal length, f = N'F'.


In order to get the f-number, you need to know the diameter of the entrance pupil, which is the diameter of image of the iris as seen from the front of the lens. The f-number = focal length divided by diameter of the entrance pupil.
To do this, place a small bulb, a L.E.D, or a small flashlight in the focal plane, at the centre of the field ; simply remove the ground glass after proper focusing at infiniy and illuminate the lens from the back from there ; place a piece of translucent paper in front of the lens, you'll see the projection of the iris, measure the diameter of this illuminated circle, this is the diameter of the entrance pupil.

If you find : 2 cm for this diameter, and if the focal length is 20 cm, your f-number is f/10.
If I remember well, Voigtländer's Petzval opened at something like f/3.3 which was a record a the time ; so if your lens has a focal length of 20 cm - 8 inches, the diameter of the entrance pupil for f/3.3 should be 20/3.3 = approx 6 cm as measured on the translucent paper.

Good luck !

30-Apr-2010, 03:30
1. From: The First Principles of Photography... By Clement J. Leaper. 1892 "....of the Petzval type, I find that the nodes are between the lenses, but are this time crossed, i.e., the node of emission is about 1/3rd inch in front of the stop, and that of admission 1/4th inch behind it.

In a portrait lens of the Petzval type, the node of admission is behind the stop, that of emission in front of it."

2. A stop will increase DOF, improve image quality at the edges, slow down the lens, remove swirl....you should decide if those are things you want. I wouldnt hack a lens ( because of lack of talent )...oh, and you can improvise by using cardboard/handmade stops inserted inside the lens or in front......Or, just look for a relatively cheap Petzval on ebay already with stops.

3. Most Petzval lenses are either based on the original Petzval design, or the modified Dallmeyer (reversed rear group) type. On late Voigtlander Petzvals (post 1878) the rear group is cemented like the front and has no air space in the rear group like the original and Dallmeyer designs... So, first check your lenses existing configuration against these designs in the image attached. The front group is typically cemented ( except for some later Petzvals made by Morrison ), so the front group is usually not the issue, unless screwed in backward. Lastly, just shoot and test. The right configuration will produce an image with good central sharpness and a progressively less sharp image in the field...usually can see curvature to field as well.... if you are getting no sharp focus in the center, check the rear group's order and orientation...

You might even find that the lens is mounted correctly and in right the sequence, but that a spacer in the rear group is missing... this can also effect image quality... in most cases you can tell just by gentle shaking of the rear group and feeling the elements move a bit.

You'll get more good answers from others.


30-Apr-2010, 03:31
most petzval water house stops are not in the middle of the lens. they are slightly rearward.

an easier thing to do would be to make a washers stop and add it to the front or rear of the lens.....like a cap with a hole in in. doing this will give you variuos afect but it is easy to do and reversible.


Paul Fitzgerald
30-Apr-2010, 06:41
"most petzval water house stops are not in the middle of the lens. they are slightly rearward."

you can make a washer stop from card stock to the barrel ID and leave 4 little tabs to bend over, then slide it thru the barrel to the center and keep moving it backward until you get the best image. Try it keep it square in the barrel, make the center hole about 1/2 glass diameter.

for FL just compare to a known lens or lenses

for speed, lay a transparent ruler as close to the front glass as practical and view from the rear with a petzval type lens.

Have fun with it

Steven Tribe
30-Apr-2010, 11:19
"for FL just compare to a known lens or lenses" - that is, by measuring something of a constant length (building width, window frame) by using different known lenses without changing the distance of the camera to the objective.

Replies are super good. A picture of the lens would help practical advice giving. Cheap ND filters are another way to control exposure if you are after maximum "petzval" image breakdown.

30-Apr-2010, 14:18
I just have to say that you are a very smart group of people. I think you explained that very very well.
thanks from me too.
Robert N.

30-Apr-2010, 22:47
Geez ... I Super agree .... What a wealth of info.

Thanks to everyone!!


Mark Sawyer
1-May-2010, 10:08
Keep in mind that the aperture doesn't necessarily go where either of the two nadal points are. In a Petzval, its placement is used to correct barrel/pincushion distortion and influence field curvature. Because of changes in element strength, placement, glass, etc, the aperture may have its "optimum" position in somewhat different places in different Petzvals, at the designer's determination.