Jump to content


- - - - -

ISTAR 2-inch Dielectric Diagonal

Discuss this article in our forums

ISTAR 2-inch Dielectric Diagonal

By Douglas Bullis

This review is of an ISTAR 2-inch BK-7 dielectric diagonal, purchased out of curiosity from a Cloudy Nights ad in August 2009 for a decidedly modest $89.00. ISTAR stocks only 2-inch dielectrics; they span the price/precision range of float glass, BK-7, and quartz.

ISTAR Optical is located in Page, Arizona. The company’s owner, Mr. Ales Krivanek, has announced an imposing line of newly designed achros, apos, Petzvals, and full OTA builds, and accessories. This takes fiscal courage in today’s climate of rising expectations and sinking budgets. Mr. Krivanek’s emails and forum posts promise much: a number of achro & apo lens cells from 127mm to 310mm clear aperture; plus 160mm F/5 and 160mm F/8 APO Petzval systems using Bk-7, FPL-51, and FPL-52 glass using a design which apparently is scalable from 140mm to 200mm. Mention also has been made of a 131mm F/5.5 two-element achro and a 152mm F/4.9 apo flat-field astrograph. The spot diagrams for some of these lenses have been posted on the CN Vendor’s Forum. If the final products match the posted spot diagrams and the prices remain approximately those promised on the forums, high-quality—though not necessarily top-end—optics will be achievable by many more enthusiasts.

Many of the designs fall into the category of exactly that: designs. Mr Krivanek’s business plan seems to be keeping a few builds in stock at all times, and fulfilling orders for specific designs as they come in. The results so far are promising. Full-production models of a 210mm f/9 and 152mm f/8 and f/12 started reaching consumers in January this year. Despite a couple of glitches in early models, the present reports coming in to the forums describe high-quality optics at mid-level prices. Everyone seems to feel ISTAR’s after-purchase service to be exceptional. The issue of where exactly these optics are produced, which concerned many at first, has largely become a non-issue: if the optics are good, they are good. Besides, buyers will essentially determine the product catalog, as some designs will simply not catch on while others appeal to many. This is especially true of ISTAR’s promised apos, which are priced roughly half of similar apos from the well-known leaders in this line.


I have no financial or professional interest in ISTAR. I bought and used the diagonal before deciding to review it. I have not personally met Mr. Krivanek, though we have emailed one another re. ISTAR’s optical designs and fabrication methods.

The Product

This review is discursive because ISTAR is a newcomer whose stated aim is to present high quality at low cost. Trying to merge those two ambitions has ended in more corporate epitaphs than probably any other business strategy. Will ISTAR be one of them? Here we examine the product-quality side of the issue.

Five days after the order, the diagonal arrived double-boxed with bubble-wrap between a sturdy cardboard outer box and a second box which itself was fitted on the inside with die-cut foam cut to the exact shape of the diagonal. For an additional $15.00 I bought an ISTAR valise-style case filled with closed-cell foam cut to the shape of the diagonal. According to Mr. Krivanek, the closed-cell construction, “doesn't absorb moisture from the atmosphere like open-cell foam does, and it is more rigid, which helps protect heavier items.” Sturdy it is, and very professional in appearance.

Larger image at http://tinyurl.com/ISTAR1

Visually and mechanically, the ISTAR dielectric differs from many diagonals in the $90 to $150 bracket. The mirror housing is a CNC-machined anodized aluminum block whose mating surface with the mirror is an ellipse with a 44.45mm minor and 69.25mm major axis. Thin felt triangles cushion the mirror from the housing. The nosepiece is chrome-plated hardened steel with smooth sides; it is threaded for 48mm filters and itself threads into the mirror housing. The gap between mirror and housing is the width of a 80-lb sheet of paper on all sides. Then comes a 0.5mm thermal gasket sized slightly larger than the mirror, which keeps dust and humidity from affecting its bottom surface. Beneath this comes a gel-opal finish powder-coated 1mm aluminum mating plate, color-coded red for float glass, blue for BK-7, and gold for quartz. This whole assembly is secured in the housing by a 3mm powder-coated aluminum base plate laser-cut with the ISTAR logo in the center. The entire unit with reducer weighs 658 gm, 1.45 lb.

There is a 44.45mm i.d. internal baffle in the nosepiece—the same diameter as the minor axis—which traps incoming stray light beginning 8mm anterior to the nearest mirror surface. No other diagonal used in the test had an interior baffle—effectively the last baffle in a telescope tube. The ISTAR readily moved back and forth between the three focusers used during tests without hanging up.

The ISTAR comes with brass compression rings on the 2” main body and 1.25“ reducer. The eyepiece end of the reducer featured a sort of “cinder-cone” shoulder—not quite a volcano and not quite flat—which added 2mm to an eyepiece’s height above the flange. The total optical path from the mirror center to the mating surface of a 2-inch eyepiece is 67mm; add 4mm to the 1.25 inch eyepiece base. On eyepieces such as the Series I Speers-Waler brand, which need a lot of infocus, 2mm can make a difference. Some users may need a low-shoulder adapter.

The mirror is 12mm thick and is larger in length and breadth than the other diagonal mirrors tested. It is coated with what Mr. Krivanek specifies as “40 layers of dielectric coating and 99% reflectivity across the full visual spectrum.” Surface accuracy is “1/10 wave peak-to-valley.” 

Comparison Diagonals

The ISTAR (left center) was compared with a 2-inch GSO dielectric (far left); a 2-inch unlabeled Japanese dielectric that is sold under a number of brand-names (right center); and a 1.25-inch Lumicon dielectric (lower far right). The unbranded diagonal at the upper far right was so poorly built that its mirror rattled when shaken. It was not tested. Larger image at http://tinyurl.com/yzgof66

Upon disassembly, the ISTAR’s mirror proved to be the largest and thickest of all. The ISTAR also was the only one to be adequately blackened—a very dense matte that was the most scatter-absorbent of all optics in the test except for the telescope tubes themselves. The GSO housing’s black anodized interior was incongruously shiny—to the point of being a built-in scatter source; its brushed metal exterior was actually darker. Only the Lumicon was as precisely fitted together as the ISTAR.

During reassembly, all the focuser tubes were lined with Protostar flocking. The GSO and Japanese generic housing interiors were painted with Liquitex™ Mars Black, a water-soluble acrylic that can be bought at an art-supply shop. It can be thickened by evaporation or thinned with water, and leaves a dead flat, not matte, finish. All the focuser tube threads also received a daub of loktite silicon gel, which minimized the possibility of a loose tube with a heavy eyepiece or bino on the business end from swiveling upside-down and possibly dumping the load.

Test Equipment

The test scopes were a 127mm f/6.3 Antares, Burgess 127 f/8, and an Anton Jaegers 150 mm f/5. A 127mm lens has a throughput of 307x a naked eye exit pupil of 7 mm. The Jaegers’ actual aperture was reduced to 148mm when the cell surfaces were lined immediately fore and aft of the lenses with ProtoStar flocking to minimize scatter from the unblackened lens edges; the resulting throughput is estimated at 365x.

These scopes are kept scrupulously clean, with practically no dust on the optics. The Antares and Jaegers dewcaps and interiors also were lined with Protostar flocking, including the baffle edges; the Burgess was satisfactorily blackened as is.

Objects viewed with the refractors were cross-checked against my own Intes MN-61 Mak-Newt and a 210 Tak Mewlon operated by fellow enthusiast Tony Bonanno. Throughout the entire test period seeing conditions favored mid-aperture scopes; the same objects viewed with a 12” Lightbridge did not resolve planetary detail with enough distinction to report in this test. The rather tremulous atmospherics necessitated very long observing sessions per object—twenty to forty minutes patiently awaiting every last glimmer of detail at powers ranging from mid-50s to low 200s.

Eyepieces, like the scopes themselves were middle-budget: Astro-Tech ED2 Paradigms, Spears-Waler, Burgess/TMB, and Baader Hyperion.

Seeing Conditions

The tests took place from August through mid-December 2009 at a 6,730 ft. elevation dark site in New Mexico intruded upon by only one 3rd magnitude porch light on one dwelling two miles away. Transparency averaged 3 to 5 (on a scale of 5). Seeing was 2 to 4, though many times a 4 night improved to 5 starting around three hours before dawn. NELM averages mag 6.5 at this site in level 4 transparency conditions. The light dome of Santa Fe rose about 20 degrees high in the NW, and Albuquerque’s reached 15 degrees high in the SW. No observations below 50º elevation were considered adequate for this test series. Non-visual entertainment was provided by an astonishing number of coyotes checking in with each other from dusk till dawn. This setting probably qualifies as astronomical bliss to those accustomed to urban gazing. Five or six hours of the ever-present wind chill and the glee turns to grit.


The greater reflectivity of a dielectric—99% -vs- 90 to 97%—enhances contrast more than magnitude. Hence the test objects were selected to gauge the richest detail in extended objects in addition to the faintest attainable magnitude in stars. The test series comprised multiple observations of the same set of objects on many nights under varying seeing conditions. The diagonals were repeatedly switched back and forth to cross-check every detail worth documenting. Each night’s tally comprised a star-by-star tick chart of the type one sees on prison walls in cartoons, so that eyeball-to-eyepiece concentration was interrupted as little as possible. A mark went down each time a faint detail was glimpsed and the sumtotals tallied later. (It was originally thought to use a pocket voice recorder, but transcribing a dozen or so two-second utterances spread across an hour or two of nothing is tantamount to a property tax on patience.) The apparent angular size of an extended object was estimated by the angle it subtended in a triangulation using one star nearby and two more distant ones whose separation was known.

The shopping list comprised NGC 6822 (Barnard’s Nebula), NGC 6960 and 6992 (Veil & Filamentary) in Cyg, NGC 6572 (Emerald Nebula) in Oph, the NGC 7006 galactic halo globular in Delphinus, NGC 7042&43 (mag 13.5 SGs) in Peg, NG 7760–7771 (both mag 12.2 SGs separated by 6’ of arc) also in Peg, Stephan’s Quintet, NGC 185 in Cass, M35’s companion NGC 2158 Gem, M38’s companion NGC 1907 Auriga, and the lovely juxtaposition of face-on spiral NGC 6946 and nearby OC NGC 6939 Ceph.

The total data set more resembled a monograph than a review, so this report was pared down to two objects: the Deerlick Group in Pegasus and NGC 2419 Lynx. These fulfilled the criteria and are popular visits for most everyone.

Deerlick Group

The grid-like linearity of galaxies NGC 7331, 7335, 7337, and 7340 plus two pair of stars on either side of 7331 make the group a good test of contrast and seeing limits. At mag 10.3, NGC 7331 is one of the more attractive near-edge-on spirals in the sky. About 4’ to the west are two stars of mag 10.3 and 11.2. A third co-linear “star” to the south turns out to be elliptical galaxy NGC 7340, a much more demanding object at visual mag 13.9. In between are NGC 7335 at 13.6 and 7337 at 14.6. Two reference stars lie on the opposite side of 7331, USNOA2 1200-19344308 at 13.5 and 19344923 at mag 14.

NGC 7331 Deerlick Group. All objects with listed magnitudes (decimal points removed) above 14.2; were observed during the course of these tests; “vm” = visual magnitude. For larger image see http://tinyurl.com/ye9yp7t. For detailed further information on the Deerlick objects, see the NED database http://tinyurl.com/ygdrrq8; SIMBAD http://tinyurl.com/yemgr24; and HyperLEDA http://tinyurl.com/7331d. Paul Alsing provides a very useful chart on his Pegasus database http://www.pnalsing.com/pegasus.

Altogether, these keep 5 and 6-inch scopes on their toes. Using the ISTAR, the 127mm Antares could pick up NGC 7335 using averted vision. In the 127mm Burgess the stars were slightly sharper on a darker field, and 7335 was slightly fuzzier. The 148mm Jaegers brightened the fuzz in 7335; 7340 was clearly diffuse and not a star. NGC 7337 presented as a nonstellar wisp above a mag 11 foreground star, looking rather like a very faint NGC 7008.

These fields looked pretty much the same using the unbranded Japanese dielectric, but 7337 and 7340 lost about 20 arcsecs of halo and reverted to star-like points. Using the 1.25” Lumicon, NGC 7335 was very faint but still unmistakable—the same visual impression one gets with a faint, tight planetary. There is a good stellar-limit test near 7331 in the form of a ternary asterism of mag 13.4 and >14 stars on the other side of 7331. The Jaegers and Intes could see the two brightest, even though they are supposed to be beyond a 150mm aperture’s limit. More surprising is that the mag 13.5 star was frequently glimpsed in the Antares and Burgess. [Compare these results with Ron Bee’s TV-102 f/8.6 apo in the IAAC report http://tinyurl.com/yc5xsua.] Perhaps the fact that this observing site’s skies are clear enough that M33 is frequently visible with averted vision and occasionally with direct can explain why such faint objects are attainable. At various times, other naked-eye transparency tests were M52, M35, M46-47, the Rosette, and the breathtaking vm5.8 open cluster NGC 2477 in Puppis.

NGC 2419 (Lynx)

This globular has a certain exotic appeal due to its status as an intergalactic wanderer—NGC 2419 is the 5th most distant globular cluster and one of the 6 extreme halo globular clusters. At magnitude 10.8 it is not hard to spot—it shows faintly in my 80mm Orion ST finder. The globulars NGC 5053 in Coma and 5466 in Bšotes are more challenging, but were on the wrong side of the sky.  Its interest as test object is the pentagon of mag 13.0 to 15.2 field stars surrounding it, whose positions are certain with respect to the GC. Patience rewards, but ever so slowly: observing times for this object averaged 25 minutes each in 11 sessions in November and December 2009.

NGC 2419, vm10.3. All stars above with numerical magnitudes (decimals removed) were multiply observed during the test. For a more detailed image see http://tinyurl.com/yc5xsua. For details of 2419’s surface photometry, see the 22 research reports listed on http://tinyurl.com/29elr9u , especially Natali, G.; Pedichini, F.; Righini, M., “BVI CCD surface photometry of the globular cluster NGC 2419” in Astronomy and Astrophysics, vol. 248, no. 2, Aug. 1991, p. 426-429. For more on 2419’s interesting place in the Milky Way’s morphlogy, see Harris, William E., et. al, “NGC 2419, M92, and the age gradient in the galactic halo”, Astronomical Journal v.114, p. 1030-1042 (1997). For readers keen to keep up with 2419’s latest doings, see http://tinyurl.com/ya8mbs6.

NGC 2419’s trapezium of mag 13 to 15.2 stars resembles the outline of Auriga. The ISTAR-equipped Antares and Burgess easily brought in USNOA2 1275-07006422 (13.9) with averted vision, and mag 14.3 USNOA2 1275-07004126 with a challenging-but-gratifying mix of rarity, ephemera, and certainly. In this cluster, patience is less a virtue and more a duty. The Jaegers picked up the entire pentagon, whose faintest direct-vision member, USNOA2 1275-07006422, was 13.9. The 14.8 field star USNOA2 1275-07005609 on the edge of 2419’s halo was seen quite frequently—and as often with direct vision as averted—despite it’s being 0.9 magnitude fainter than the 13.9 limit calculated by http://www.cruxis.com. (Parameters: clean optics, 1.2mm eye pupil, 3.5mm eyepiece @ 234x, lim. mag 6.5, 70º elevation, atmos. extinction estimated at 0.3 mag.) The most difficult objects were the 14.2 & 15.5 pair on the right side of the image. This pair was ephemeral at all times, glimpsed fleetingly in averted vision as a vague patch, but four times as a clear pair of distinct stellar points during the cumulated 6.6 hours of observing. All the above sightings were confirmed multiple times using the Intes MN-61 Mak-Newt.

Differences between the diagonals had their effects on the globular disk itself. The 14.8 halo-edge field star provided a boundary object to gauge against. A readily available triangulation tool can be mentally drawn by extending straight lines from the 8.1 and 10.7 stars at the upper left of the picture (2.1 arcmins separation) across the 7.2 mag HD 60771 lying 4 arcmins from 2419’s core. The narrow isosceles thus formed can be used to estimate the diameter of the faint halo as seen through the various telescopes. The halo’s core would brighten and its disk fatten from approx 1.5 to 2 arcmins apparent diameter as the eyepieces and scopes were changed. (NGC 2419’s astrometric diameter is listed as 4.1 arcmins). Subjectively, in moments of good seeing the core simply appeared more “there.”

The unbranded dielectric didn’t perform quite as well as the ISTAR—the mag 14.8 field star was seen only infrequently using averted vision. The 14.2/15.5 pair never appeared at all. The Lumicon’s 1.25 tube limited its apparent field, but its superb optics matched the ISTAR in all measures except limiting magnitude. The 14.2/15.5 pair never showed up, but the GC’s halo was more crisply boundaried than in any other diagonal.

Detecting the differences between 99% dielectrics -vs- 90% to 97% nondielectrics is an exercise in ephemera akin to taking the temperature of a raindrop. Half the time the difference was how many times an extended object at the very limit of visibility popped into & out of view, using first one diagonal and then the other. Then switch and repeat, then do it again and again across many observing sessions. The other half of the time, certainty comprised which stars of known magnitude could NOT be seen compared with nearby stars known to be 0.1 to 0.2 magnitudes brighter which were glimpsed consistently, even if fleetingly.

Some of the above are fainter than the calculated seeing limits for the apertures of the scopes. This was not out of the ordinary: in earlier observation notes of, for example, NGC 7006 in Delphinus, the 152mm f6.3 Intes could pick up the pair USNOA2 1050-19061377 and -190961997 at mag 14.15 and 14.2. Another example: the two 127mm refractors could detect USNOA2 1275-07004126 (13.5) inside the planetary disc of NGC 2348 adjacent to M46.

The phenomenon of seeing well beyond theoretical limiting magnitudes occurred too often on too many objects on too many nights to be chalked up to the Starcebo Effect—”I see it because I know it is there.” (Also known as “delusions of glimmer.”) Perhaps air as turbulent as that above the dark site, where seeing was usually 3/5 and transparency 4/5, a transient lensing effect can occur which concentrates a particular stellar wavefront across the full air column of a scope’s aperture into an energy density high enough to end up at the eyepiece seeming brighter than the object actually is. Throughout the entire observing period, field stars shimmered through the eyepiece in the same tremulous manner as starlight reflecting off a zephyred pond.


Jupiter was the only planet in the sky high enough to test for contrast and detail. Here the tests were a draw—the three dielectrics throughput images that were indistinguishable from one another. Transiting moons looked like dewdrops on a beige baseball, though their shadows were as tack-black on the cloud surface as nearby field stars were tack-bright. Seeing conditions were never really good enough for rigorous planetary tests, e.g., poor belt definition and a lukewarm Red Spot. Live downwind from Albuquerque and see what you get.


Given the modest size and price point of the scopes used in this test, it is not surprising that none of the diagonals stood head and shoulders above the others in the way a TV Maxbrite would. Using the admittedly casual eyeball & tick-chart method of data recording, the dielectrics sampled here detected stars approx 0.1 to 0.2 magnitude fainter than the non-dielectric in the 127 and 150mm scopes used. The ISTAR performed slightly better at the limits of seeing than the other dielectrics, and significantly better than the 90% reflectivity Al-SiO2 diagonal.

On price point, the ISTAR delivered the best performance by far. The ISTAR slightly bettered the others on extremely faint objects—but really, how many times do we chase after the faintest dot in the field? The ISTAR’s outstanding quality was its construction and price/performance ratio. Its thicker mirror, drawtube baffle, low-scatter interior, and quality of metal finishing testifies that it was custom-designed and custom-produced. How ISTAR does this for $89.00 is beyond me.

The Future

Does the one lone ISTAR product assayed here say anything about the company’s grander visions down the road? One can’t generalize from a low-end diagonal to an OTA assembly five magnitudes more demanding. If ISTAR remains as price/quality sensitive as the company is now, good tidings are on the way. As of this writing, at least twelve 152mm f/10 achros in cell and one 210mm f/9 full-production OTA have gone to customers for testing. The demo builds and manufacturing details ISTAR has released to forum readers show some innovative thinking, e.g., helicoid lens-cell venting. The most reassuring thing about ISTAR’s lenses is that they will be made one by one and not ground out by the thousands. Mr. Krivanek states that all ISTAR lenses will be hand-figured; this includes modest-aperture achros. The CN and AM forums are worth bookmarking as ISTAR products reach experienced users.

Mr. Krivanek has forwarded details of ISTAR’s planned OTAs and optical designs, and who has done those designs. It is an impressive lineup that contains a few surprises. For example, the f/6.2 paired doublet Petzval design is scalable across the aperture range of 127mm to 20mm; its multi-analysis plots some interesting wavefront curves for lateral color and chromatic focal shift. ISTAR’s 150mm and 160mm apo design specs hit all the high Cs on the opera score written by librettists Seidel/, RMS, & Strehl, but it is best to see products in glass instead of commenting about products on paper. In my view, Strehl is just a number till it comes out of an eyepiece.

Beyond ISTAR, everyone watching the variously-branded 2XXmm refractors now being announced is keen to see how ISTAR stacks up in the larger-aperture market of two-element glass bearing four-digit price tags. We’ve all seen instances of a business devising wondrous things for which there proves insufficient buyer base. And no matter what the prices, markets take more time to develop than hopes take to dash.

The Joy of Seeing

On a personal observer’s note, there were rewards to this exercise beyond knowing which diagonal has the best price/performance value. One reward was comprehending just how difficult it is for scheduling directors of major observatories and the HST to allot the most efficient amount of time needed to attain a specific limiting magnitude or spectral line. (And pity for the poor Ph.D. candidate whose long-awaited observing slot has just been swallowed by a New Mexican snow squall.)

This test occasioned many waits of five-plus minutes to confirm the sighting of a single star at or below the aperture limiting magnitude. This led to the second reward: getting familiar with exceedingly small patches of sky. One can grow rather affectionate toward a circle in the sky 15 or so arcminutes in diameter when it becomes as familiar as one’s children. The third reward is the joy of subtlety: a detail glimpsed for a few fleeting instants across the span of half an hour can acquire the tang of suspense and reward one has while watching the last at-bat in a no-hitter or a 60-yard pass caught with only one cleat left on the grass inside the end zone. Or to put it in loftier esthetic terms, visual astronomy is looking past the beauteous face of the Mona Lisa to notice the really fascinating weather that was going on that day.

Add coyotes, and it’s the Howlalluja Chorus out there.


Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics