The Making of a 255mm f/6 AstroReflect Mirror
One of the most common questions we receive is:
“What actually happens between a polished mirror blank and a finished premium optical surface?”
To answer that question, we decided to document the complete fabrication process of one of three custom 255mm f/6 mirrors manufactured for a customer in the United States.
Rather than presenting only the final test report, this article follows the entire evolution of the optical surface—from the first interferometric measurements after polishing, through astigmatism correction, and finally to the completed paraboloid.
All images and measurement results shown below are from the actual production process of this mirror.
Fine Grinding and Polishing
The mirror blank is 38 mm thick and was generated to the required f/6 curvature. After fine grinding to 5 microns, machine rotary polishing was performed using a 75% tool with optical resin and cerium oxide.
During polishing, the surface was periodically inspected under a microscope at 150× magnification in order to monitor the progress of the polish. Polishing normally continues until all microscopic pits left by the final 5-micron grinding stage disappear completely, leaving a smooth and continuous optical surface.
After approximately 8–10 hours of polishing, we begin interferometric testing.
At this stage, the purpose of the measurements is twofold.
First, interferometry allows us to determine with very high precision how closely the surface approaches an ideal sphere. While Foucault and Ronchi testing can provide valuable information about the overall figure, interferometry offers a quantitative assessment with significantly greater sensitivity.
The second objective is even more important: accurate measurement of astigmatism.
Our experience shows that nearly every mirror blank—particularly thinner blanks—contains some degree of residual astigmatism. In many cases the amount is small enough to have little impact during visual observations. However, when the goal is to produce a truly exceptional mirror, even very small amounts of astigmatism deserve attention.
To accurately measure the mirror’s intrinsic astigmatism, it is necessary to eliminate deformation caused by the support system during testing. Every mirror bends slightly under its own weight when mounted vertically, even relatively thick mirrors.
To separate support-induced deformation from the mirror’s actual figure, the mirror is rotated through several angular positions and measured repeatedly. Typically, measurements are taken in four or eight orientations. The software then rotates all measurements back to a common reference frame and removes the effects introduced by the support stand.
This process reveals both the true magnitude and orientation of the mirror’s astigmatism.
Why We Rely on Interferometry
Many manufacturers provide FigureXP reports derived from Foucault testing with a Couder mask. Such reports often indicate excellent spherical correction and impressive PV and Strehl values.
However, these methods evaluate only a limited number of zones along a single profile — or at best a few averaged profiles across the mirror surface.
As a result, they primarily evaluate spherical aberration, or in other words, how closely the surface matches the ideal paraboloid.
They cannot reliably detect or quantify:
- Astigmatism
- Trefoil
- Tetrafoil
- Surface roughness
- Localized asymmetrical errors
- Higher-order aberrations
Most importantly, they cannot accurately quantify astigmatism.
Over the years we have examined mirrors produced by respected manufacturers whose owners were dissatisfied with their real-world performance despite impressive test reports. In several cases, interferometric testing revealed significant astigmatism that made precise focusing and high-contrast observing impossible.
While a Foucault test samples a limited number of points along selected zones, an interferometer evaluates hundreds of thousands of points across the entire optical surface, providing a far more complete picture of the mirror’s true optical quality.
Many optical workshops would consider the level of astigmatism typically encountered at this stage acceptable for practical visual use. In fact, in many telescopes the observer would never notice it directly.
Our philosophy is different.
When the objective is to produce an outstanding mirror, every measurable error matters.
Astigmatism is particularly undesirable because it affects the ability to achieve precise focus and reduces image contrast. Even when the wavefront error appears small, residual astigmatism can prevent a mirror from delivering the sharp, clean stellar images expected from a premium optic.
For that reason, we treat astigmatism correction as a dedicated stage of the manufacturing process rather than merely an incidental part of figuring.
The procedure is deliberately gradual. A small amount of astigmatism is removed, followed by additional polishing intended to smooth the asymmetrically worked surface. The mirror is then measured again and the process repeated as necessary.
This cycle continues until the astigmatism is reduced to negligible levels and the surface exhibits excellent rotational symmetry.
Only then do we proceed toward final figuring.
Preparing for Parabolization
Once polishing quality, surface smoothness, and astigmatism are fully under control, parabolization can begin.
Parabolization is performed using a unique method developed by us, through years of experimentation and refinement.
The process combines interferometric feedback, CNC-controlled motion, specially designed pitch tools, and extremely fine Opaline cerium oxide.
Our approach differs significantly from industrial small-tool CNC polishing methods and fast speed, which can leave residual ripple, tool marks, and localized surface structure.
Instead, we use:
- Large-diameter pitch tools
- Slow polishing speeds
- Extremely fine polishing compounds
- Broad tool coverage across the entire mirror surface
The CNC machine does not replace craftsmanship. Its purpose is to provide exceptional control over where and how much correction is applied.
The combination of high-resolution interferometric measurements and precise machine motion allows us to modify specific regions of the surface by extremely small amounts while preserving overall smoothness.
The result is a continuously refined optical surface that remains smooth, uniform, and free of zonal structure throughout the figuring process.
The following measurements illustrate the actual progression of this mirror from polished sphere to finished paraboloid.
Session 0
After approximately eight hours of polishing, the mirror exhibited the surface figure shown above.
The overall correction measured approximately 13%, where 0% corresponds to an ideal sphere.
Because this is a relatively long-focus f/6 mirror, the deviation from a perfect sphere is not particularly significant at this stage. For mirrors of this focal ratio, substantial departures from the final figure can still produce acceptable optical performance after parabolization phase.
The more important issue revealed by the measurement is the presence of residual astigmatism.
Although the surface already appears promising, the astigmatism prevents the mirror from achieving the level of symmetry and optical quality required for a premium instrument.
For this reason, the next step is not further correction toward a paraboloid, but rather targeted work aimed at reducing the astigmatism.
Asymmetric corrective figuring was applied to the high regions responsible for the measured astigmatism. Following this intervention, the mirror underwent an additional thirty minutes of smoothing rotary polish before a new interferometric measurement was performed.
Session 1
The results show that the astigmatism has been reduced to a very low level and no further dedicated correction appears necessary at this stage.
In this particular case, the initial astigmatism was not severe. Many manufacturers would likely consider it entirely acceptable and proceed directly with figuring.
However, our goal is not just to produce a very good mirror, our goal is to produce an excellent mirror with the highest level of optical performance.
It is worth noting that astigmatism removal is not always accomplished in a single cycle. In some cases, several hours of corrective work and subsequent smoothing may be required before the desired result is achieved.
With the astigmatism now under control, polishing continues for approximately two additional hours in order to complete the polishing process and further refine the smoothness of the surface.
Throughout this stage, microscopic inspection and interferometric measurements are performed periodically until both polishing quality and surface uniformity meet our requirements.
Only then is the mirror ready for the next phase: precision parabolization.
Parabolization Begins
At this stage the mirror is fully polished, astigmatism has been reduced to negligible levels, and the surface is ready for precision figuring.
From this point onward, our focus shifts from polishing quality to wavefront quality.
To see the remaining errors on the surface more clearly, spherical aberration is mathematically excluded from the interferometric analysis. In practice, this means that the software displays the residual deviations as if the mirror had already reached perfect paraboloidal correction. This allows subtle hills and valleys to become clearly visible and makes it easier to determine the next corrective action.
Since astigmatism has now been eliminated, subsequent measurements are performed in one direction with 16 interferograms. This approach introduces a certain amount of environmental noise caused by air currents, vibration, and thermal fluctuations. However, the noise level is low enough that the overall error distribution and surface topology remain clearly visible and suitable for decision-making.
The objective now becomes straightforward in theory, but considerably more challenging in practice:
To increase the overall correction and at the same time improve and maintain the smoothness of the surface.
Session 5
At this stage the residual surface structure becomes clearly visible.
Two elevated regions can be seen, one near the 50% zone and another close to the 95% zone. These areas stand above the surrounding surface and must eventually be blended smoothly into the overall figure.
The task sounds deceptively simple. In reality, modifying one region of a mirror inevitably influences neighboring areas. Every correction carries the risk of introducing new errors elsewhere on the surface.
A CNC figuring stroke was selected specifically to reduce these elevated regions while preserving the excellent overall smoothness already achieved.
After completion of the figuring run, a new interferometric measurement was performed.
Session 6
The surface responded exceptionally well to the applied correction.
The elevated zones identified during Session 5 have been reduced significantly, and the overall figure appears noticeably smoother.
A slight residual elevation remains near the 50% zone. In addition, shallow depressions are now visible near the center and around the 85% zone.
Such behavior is entirely normal during precision figuring. Every correction alters the balance of the surface, and the next polishing action must take into account the newly emerging topology.
A new CNC stroke was applied to affect precisely these residual irregularities, while at the same time maintaining the trend towards increasing the overall correction. The mirror was then measured again.
Session 7
The surface quality improved further.
At this point, if the mirror had already reached full 100% parabolic correction, most observers and opticians would describe the mirror as excellent.
However, the overall correction remains only 32%, meaning that a substantial amount of figuring work still lies ahead.
This stage represents one of the most delicate phases of mirror making.
Creating a smooth surface is challenging. Maintaining that smoothness while progressively increasing correction is considerably more difficult.
The objective now is to preserve the quality already achieved while steadily advancing toward the final paraboloidal figure.
A new CNC figuring stroke was selected to improve the remaining surface structure while simultaneously increasing the overall correction.
Following the figuring run, another interferometric measurement was performed.
Session 8
The surface remained remarkably stable.
Only minor changes are visible in the surface structure, and the overall smoothness remains essentially unchanged from the previous session.
More importantly, the global correction increased from 32% to approximately 42%.
This represents another significant step toward the final figure.
One of the most satisfying outcomes during precision figuring is seeing correction increase while surface quality remains preserved. Every successful session reduces the distance to the final paraboloid without sacrificing smoothness.
Another carefully chosen stroke was applied.
The mirror was then measured again.
Session 9
The result is particularly satisfying.
The surface remains virtually unchanged in terms of smoothness, indicating that the previous corrective actions have not introduced new zonal errors or localized defects.
Meanwhile, the overall correction has increased to approximately 53%.
This is precisely the kind of progression every optician hopes to see during figuring: steady movement toward the target figure while preserving a highly refined surface.
At this point the process becomes increasingly psychological as well as technical.
The temptation is always present to make larger corrections in order to reach the target more quickly. Experience teaches the opposite lesson.
Patience is often the fastest path to success.
Another carefully calculated CNC stroke was therefore applied, prioritizing stability and control over speed.
The mirror was measured once again.
Session 10
The surface quality improved yet again.
Not that there was much room for improvement.
The figure now exhibits an exceptionally smooth character with no obvious zonal structure.
At the same time, the overall correction has increased to approximately 62%.
The goal is to increase correction without disturbing the remarkably smooth surface that has been preserved through multiple figuring cycles.
A stroke was therefore selected that would distribute its effect broadly across the surface, minimizing the risk of introducing localized errors.
At this point, the optical surface has progressed from a fully polished sphere with measurable astigmatism to a highly refined figure with approximately 62% correction and outstanding smoothness.
The remaining work will require even greater restraint, as the mirror enters the final stages of its transformation into a precision paraboloid.
The challenge now is preserving what has already been achieved.
We apply a suitable stroke that would give an even change across the entire surface. We don’t want to ruin it with inappropriate manipulation, but we can’t stop because we need 100% correction.
Session 11
The overall surface structure remained essentially unchanged, indicating that the selected figuring stroke had acted uniformly across the mirror without introducing any measurable zonal artifacts.
Meanwhile, the overall correction increased from approximately 62% to 72%.
Nevertheless, experience teaches caution.
It’s very easy for an excellent surface to be damaged during the final stages if you try to progress too quickly.
The next figuring stroke was therefore kept conservative and carefully controlled.
A new interferometric measurement followed.
Session 12
Once again, the surface remained remarkably stable.
The smoothness achieved during the previous sessions was preserved, while the overall correction increased to approximately 82%.
The mirror is steadily approaching the desired paraboloid, yet the surface continues to exhibit the same smooth and well-behaved character seen several sessions earlier.
Now, small changes that would have been insignificant earlier in the process can now determine the final quality of the optic.
The next corrective stroke was selected with particular care.
Following completion of the figuring stroke, the mirror was measured again.
Session 13
The trend continued.
The surface remained essentially unchanged and the total correction increased to approximately 90%.
However, the last ten percent often required more attention than the previous ninety. The closer the surface approached its target value, the more risky any small error became.
Exceeding the desired correction could eventually compromise the surface quality achieved so far. For this reason, the duration of the figuring stroke was reduced and the corrections became increasingly gentle.
The next parabolizing stroke was intentionally shortened.
At this point, we usually measure the mirror using the full multi-orientation procedure to make sure that nothing has changed.
We measure the focal length precisely again, because it shortens slightly during the parabolization process.
Session 14
The mirror was measured in eight different rotational orientations to eliminate the deformation caused by the support and to ensure the most accurate representation of the true optical figure.
It may seem to us that the surface has deteriorated, but a visual comparison with previous measurements in one direction can be misleading, since the visualization scales are different.
In fact, the wavefront values show that the surface quality has improved.
The overall correction had already reached approximately 94%.
But more importantly, let’s see what the results say about astigmatism.
The result shows a slight increase in astigmatism. We could leave it as it is and it will be fine, but since we are looking for perfect results, we decide to influence it very slightly.
We will do this in the next session.
A carefully selected figuring stroke was applied and the mirror was measured again.
Session 15
The overall surface smoothness remained unchanged and the correction was increased to approximately 97%.
At this point, the mirror had already reached a level of performance that would satisfy even the most demanding observer.
If the process had stopped there, the mirror would undoubtedly have qualified as a first-class optic.
But we will continue a little further, to the limit of the impossible.
Only a small correction is necessary. The final figuring was therefore carried out with extreme care. No load was applied to the instrument. The duration was intentionally short. The astigmatism correction was accordingly quite light. The figuring was closer to a gentle touch than to a conventional session. The mirror was then allowed to reach full thermal equilibrium before the final measurements were made.
For the final evaluation, measurements were made in eight orientations, with sixteen interferograms recorded and averaged for each orientation.
This provides the highest level of credibility for the finished optics.
From the DFTFringe processing software, we include the display of all optical aberrations.
Session 16 — Final Result
The final measurements showed the result of the patient and careful efforts.
The surface smoothness reached:
RMS = 0.009 waves
The calculated optical performance reached:
Strehl Ratio = 0.997
The Peak-to-Valley wavefront error measured:
PV ≈ 1/30 wave
The final correction was:
100.5%
The other two mirrors in the set have the following characteristics:
• Strehl Ratio: 0.998
• RMS Wavefront Error: 0.008 waves
• Peak-to-Valley Error: approximately 1/30 wave
• Final Correction: 99.9%
• Strehl Ratio: 0.992
• RMS Wavefront Error: 0.014 waves
• Peak-to-Valley Error: approximately 1/20 wave
• Final Correction: 100.4%
Most importantly, these values were obtained from a full aperture interferometric analysis of hundreds of thousands of points, encompassing the entire optical surface. These are not values derived from a small number of zonal measurements along a single profile like FigureXP will give you. The results include the influence of:
- Spherical aberration
- Astigmatism
- Trefoil
- Tetrafoil
- Higher-order aberrations
- Surface irregularities across the full aperture
They represent the true measured behavior of the entire optical system.
Here is some more information from the final results
Verification with Classical Optical Tests
Although interferometry is our primary quantitative tool, we always compare the final results with direct visual examination using Foucault and Ronchi testing.
In the Foucault test, the finished mirror showed an exceptionally smooth and uniform surface.
The surface had a smooth, continuous, and velvety appearance. There were no zones, tool marks, or the characteristic artifacts that often accompany aggressive machine shaping.
The mirror edge was well-shaped, balanced, and visibly luminous on both sides of the mirror.
Final Thoughts
A 255mm f/6 focal length mirror is generally considered relatively easier to manufacture.
Producing a mirror of this focal ratio does not require the extreme tolerances associated with today’s fastest Newtonians, and many skilled amateur telescope makers are capable of producing excellent results at f/6.
The real challenge begins below approximately f/4.0.
As focal ratios become shorter, tolerances shrink dramatically. The acceptable margin for error decreases, figuring sessions multiply, and the risk of generating zones, ripple, or surface roughness increases substantially.
Progress becomes slower, corrections grow in small increments, and maintaining a perfectly smooth surface becomes increasingly difficult.
Yet the fundamental principles remain unchanged.
Success comes from accurate measurements, careful analysis, patience, persistence, and accumulated experience.
This mirror represents only one example of that process.
What ultimately matters is not a Strehl number, an RMS value, or a test report.
What matters is that under the night sky the mirror disappears, leaving only the pure image.
And that is the goal of every optical surface we produce.
Yordan Stoykov
Owner&CEO
AstroReflect Ltd