Yes, chromatic aberration is a frequently encountered problem with the Optosplit, but actually the cause is in the microscope, specifically the objective! 

It’s all to do with “longitudinal magnification,” which is a potential problem with any high-magnification system such as a microscope.

Longitudinal magnification is the SQUARE of the lateral magnification, and this has potentially nasty consequences!!!

Consider, for example, a x40 objective looking at a one-micron cube. 

Clearly its lateral size is 40 times bigger, i.e., 40 by 40 microns.

But to the extent to which the depth of field allows the image to be in focus more or less through its entire depth, then the size of the image in z is not 40 but 1600 microns!

In 3D we thus have a very elongated block instead of a cube.

Well, the chromatic aberration means that the focal length changes a bit across the wavelength range.

The focal length of the objective is given by the focal length of the tube lens divided by the magnification of the objective (of course!), and a change in focal length is basically equivalent to moving the object by the same amount.

But because of the enormous longitudinal magnification, the effect on the image is correspondingly greater. In other words, any chromatic aberration in the objective is ENORMOUSLY magnified.

For a 1:1 relay like the Optosplit, this effect just doesn’t happen, and for the lenses we use the chromatic aberration should be down at the diffraction limit anyway.

However, once you split an image between two wavelength ranges, the objective’s chromatic aberration hits you right in the face.

Especially true as you go out towards the red, where cameras see images far better than we do, and the objective’s chromatic performance tends to be getting worse in any case!

So, our corrector lenses act like “spectacle lenses,” to compensate for a problem elsewhere. They therefore need to be selected by straightforward trial and error!

And as for flatness of field, EXACTLY the same argument applies. A tiny error in the objective across the field is enormously magnified, and in practice will swamp any such errors in the relay.

So finally, this isn’t a specific problem with anyone's objectives! Everyone’s do it to some extent.

Customers who do accurate spectral scans have to use something like a PIFOC focuser to move the objective during the scan to compensate for all this!

There are several ways to achieve bypass with the OptoSplit II:

  1. Remove the dichroic cube, center the image with the split control, open the aperture, and adjust the trim control on the bottom to remove any vignetting. 

  2. Leave the dichroic cube in place, block the long-wavelength path with the shutter plate provided, center the image with the split control, open the aperture, and adjust the trim control on the bottom to remove any vignetting. 

  3. Leave the dichroic cube in place, block the short-wavelength path with the shutter plate provided, center the image with the split control, open the aperture, and adjust the trim control on the bottom to remove any vignetting.


This allows non-split mode using either filter in the Optosplit cube, or with neither in place. With a little practice, this bypass will only take a few seconds.

For smaller sensors, it may not be necessary to adjust the Trim control routinely, but we would suggest observing the effects of this control as it allows the beam separation to be optimized. 

The mirror coatings in the OptoSplit II are optimized to transmit visible light and in ‘bypass’ mode (following removal of the dichroic cube), transmission efficiency is an average 87% (across blue, green, and red emission).

If imaging low-light samples ‘full field,’ we would recommend removal of the Optosplit from the imaging path to maximize throughput. 

For the very best results, we would recommend using a 1X microscope C Mount (with no optics) and introducing the magnification in the splitter itself.

We do, however, have many customers who get excellent results from magnifying and demagnifying C Mounts and also using standard C Mount camera lenses.

If the two channels of your OptoSplit are not parallel, then it may be the case that you are adjusting the two channels using the wrong controls.

When the 2 channels are superimposed, you should only need to make adjustments using the split control and aperture controls.

If you are using the V1 and V2 controls to split the image along the vertical axis, it can result in the channels becoming misaligned.

If this occurs, you should refer to the manual to realign the OptoSplit. If you need a manual for an older version of the OptoSplit, then don’t hesitate to contact us. 

This is a routine application for the Optosplit, as the product has had provision for corrector lenses in one or other pathway since its inception.

This facility was originally provided for the correction of any chromatic aberration in the preceding optics but rapidly found a further application for deliberately defocusing one or other imaging pathway in order to allow different depths to be in focus at the same time!

Obviously, for z-plane splitting, a beam splitter is used rather than a dichroic. You may also be interested in our Optosplit III, which allows three different depths!

Cairn filter cubes accept standard 25mm diameter emission filters.

We strongly recommend the Chroma ET range to maximize levels of light transmitted.

For dichroic beamsplitters, we recommend 26mm x 38mm x 2mm (for other sizes, please get in touch with us, and we will be happy to advise).

After extensive testing, we only recommend Chroma 2mm thick ‘Ultra-Flat’ dichroic mirrors for minimal image distortion across all of our image splitter range. For more technical detail, see our Chroma Filters and Beamsplitters page.

For any filter or mirror advice (not limited to image splitters), please get in touch as we are UK stockists for Chroma – we are happy to help!

Both models are designed for two-channel image splitting on a single camera, but the OptoSplit II Bypass model has a simple lever to switch between split mode and single-channel operation.

By introducing a bypass mirror into the pathway, none of the splitter controls are moved, so there is no need to realign when you return to split mode. In addition, we use longer focal length lenses for improved image quality and use magnetically held cubes for improved registration.

Optosplit II Bypass Misalignment

A lesser degree of misalignment can occur if you turn the split control clockwise when splitting the image along the horizontal axis.

The OptoSplit II is designed to work optimally when the split control is turned anticlockwise. 

OptoSplit II Controls

Here is a diagram showing the controls for the most recent OptoSplit. Older versions may have different control names, but all the controls are located in the same place.

The rough edges in the image are a classic case of dirt on the aperture blades and can be remedied by cleaning with a lint-free lens tissue.

Learn how we helped 100 top brands gain success