Corrections

Errors we found, and what we changed

Every data error found and fixed since the first commit, newest first. Most were ours. Some were in a manufacturer’s own material and would have become ours if we had copied them. Each record gives what was published or assumed, what we found, what changed and the source that settled it. The date is the day the fix went in.

30 corrections since 16 August 2026: 16 were on the site, 11 were caught before publication, 3 were errors in a maker’s material

Index

Every correction, newest first
Id	Date	Correction	Status
C-30	12 Oct	The method page's exposure example kept the old read noise	Was on the site
C-29	12 Oct	The Quattro 150P's corrector is in the box	Was on the site
C-28	12 Oct	A reducer named for a factor its maker never printed	Was on the site
C-27	12 Oct	Reducers did not change the focal length	Was on the site
C-26	12 Oct	Cameras with a guide sensor were sold a guide camera	Was on the site
C-25	12 Oct	Ten mono cameras were priced as color	Was on the site
C-24	12 Oct	A tube holder matched on a retailer's title	Caught before publication
C-23	12 Oct	Two Celestron reducers were given the wrong output thread	Was on the site
C-22	10 Oct	The Bortle 2 preset sat in class 1	Was on the site
C-21	10 Oct	The home page promised a resale estimate	Was on the site
C-20	6 Oct	The observatory guide described the design as the build	Was on the site
C-19	28 Sep	An SSD priced at an out-of-stock sale price	Caught before publication
C-18	28 Sep	Figures in the prototype that had no source	Caught before publication
C-17	27 Sep	A false “don't buy” at $3,000	Was on the site
C-16	27 Sep	Mounts priced without the tripod they ship with	Was on the site
C-15	27 Sep	Sky-Watcher USA's EDX specs repeat the discontinued ED line	Error in a maker's material
C-14	27 Sep	A pound-to-kilogram conversion on the HEQ5-R Pro page	Error in a maker's material
C-13	27 Sep	Three EdgeHD reducers in one row, and a solar filter's part number	Was on the site
C-12	27 Sep	The natural sky was a textbook minimum	Was on the site
C-11	27 Sep	Worked examples that did not reproduce	Caught before publication
C-10	26 Sep	The observatory draft disagreed with its own spec	Caught before publication
C-09	26 Sep	Satellite data stored at ten times its scale	Caught before publication
C-08	26 Sep	A spacer in front of the off-axis guider	Was on the site
C-07	26 Sep	A focus range stored as a target	Was on the site
C-06	25 Sep	ZWO's 55 mm charts do not add up	Error in a maker's material
C-05	24 Sep	The solver chose your thread adapter	Caught before publication
C-04	24 Sep	Spacer kits were lists of lengths	Caught before publication
C-03	26 Aug	ZWO's 17.5 mm means two different things	Caught before publication
C-02	16 Aug	Two forms of the Dawes limit presented as one	Caught before publication
C-01	16 Aug	The critical focus zone shorthand was for another wavelength	Caught before publication

Records

C-3012 October 2026Was on the site

The method page's exposure example kept the old read noise

Published or assumed: The worked example on the method page ran on 1.0 e⁻ of read noise, described as the physics reference's round figure.
What we found: On 27 September the reference moved that example to the 1.5 e⁻ ZWO's chart gives for the ASI533 at gain 100 (C-11). The page kept 1.0, so its table matched neither the reference nor what the Exposure Planner computes for the same rig.
What changed: The table now runs at 1.5 e⁻: the shortest sub without a filter is 8.1 s, where it showed 3.6 s, and 147.6 s with the dual-band filter, where it showed 65.6 s. The section carries a dated note.
Settled by: ZWO's read-noise chart for the ASI533 Pro
Page: firstrig.com/method#exposure

C-2912 October 2026Was on the site

The Quattro 150P's corrector is in the box

Published or assumed: The Sky-Watcher Quattro 150P linked no corrector, so its builds in the Rig Reality Check carried an unpriced corrector line.
What we found: Sky-Watcher ships it with “an included coma corrector/reducer” and gives the imaging focal length with it fitted as 518 mm.
What changed: Its builds list the corrector as included, at no extra cost, and are computed at 518 mm. The corrector still counts toward spacing and complexity.
Settled by: Sky-Watcher USA, Quattro 150P
Page: firstrig.com/tools/rig-reality-check

C-2812 October 2026Was on the site

A reducer named for a factor its maker never printed

Published or assumed: The Sky-Watcher reducer for the Evostar 72EDX was listed as a 0.85x reducer.
What we found: Sky-Watcher prints no factor for it: the 72EDX manual gives two back-focus figures and an image height. The 0.85x belongs to the reducer for the discontinued 72ED.
What changed: The row is named “factor not published”, and no reduction is applied to builds that use it.
Settled by: Sky-Watcher Evostar 72EDX manual
Page: firstrig.com/tools/backfocus

C-2712 October 2026Was on the site

Reducers did not change the focal length

Published or assumed: A build with a reducing corrector counted the corrector's price, but was shown and computed at the telescope's native focal length.
What we found: A 0.8x reducer such as William Optics' FLAT 6AIII changes the image scale, the field and the framing. The native figure overstated the focal length by 25%.
What changed: 25 corrector rows carry the reduction factor their maker states. Builds are computed at the reduced focal length, and the card shows both figures.
Settled by: William Optics, FLAT 6AIII: “The reducing factor is 0.8X.”
Page: firstrig.com/tools/rig-reality-check

C-2612 October 2026Was on the site

Cameras with a guide sensor were sold a guide camera

Published or assumed: Builds with ZWO's ASI2600 DUO cameras were given a separate guide camera and guide scope, and builds with the Air cameras a control computer as well.
What we found: The DUO has a guide sensor in the camera body; ZWO: “You don't need a separate OAG and guide camera.” The Air cameras carry a guide sensor and an ASIAIR.
What changed: Those builds list built-in guiding, and for the Air cameras a built-in ASIAIR, as included lines with no guiding mass. ZWO lists four USB ports on the Air cameras for the focuser, filter wheel and mount, so no hub is added either.
Settled by: ZWO, ASI2600MC Duo
Page: firstrig.com/tools/rig-reality-check

C-2512 October 2026Was on the site

Ten mono cameras were priced as color

Published or assumed: The Rig Reality Check read mono or color only from the Exposure Planner's sensor profiles, and treated a camera without a profile as color.
What we found: Ten mono cameras fell through and were priced with a dual-band filter instead of a filter wheel and a narrowband set: the ASI2600MM DUO, the QHY268M, both QHY600M rows, ToupTek's ATR2600M, ATR533M and ATR585M, and the SkyEye26AM plus, 62AM and 62AM plus.
What changed: Each camera row carries its maker's own statement of mono or color, with the quote and the page. A camera with neither a statement nor a profile is shown as “sensor type not catalogued”, and no filter option is priced for it.
Settled by: Each maker's page or manual, quoted in the row; catalogue sourcing notes
Page: firstrig.com/tools/rig-reality-check

C-2412 October 2026Caught before publication

A tube holder matched on a retailer's title

Published or assumed: The first pass at ring sets matched the Takahashi FC-76DCU to a holder sold as TKA20420, on the strength of a retailer's product title.
What we found: Takahashi Europe lists TKA20420 as the 80S, for the FS-60CB only. The FC-76DC takes the 80QS, TKA20420Q.
What changed: The ring row is the 80QS, priced at $167, with Takahashi Europe's page as its stated fit.
Settled by: Takahashi Europe, tube holder for the FC-76DC and FS-60Q
Page: firstrig.com/tools/rig-reality-check

C-2312 October 2026Was on the site

Two Celestron reducers were given the wrong output thread

Published or assumed: Celestron's f/6.3 reducer (#94175) and its EdgeHD 8 reducer were catalogued with an M42 output thread, assumed from the usual T-adapter.
What we found: Celestron's knowledge base gives the SCT rear-cell accessory thread as 2″ × 24, and both reducers' sheets thread the visual back or T-adapter straight onto the reducer's rear, so they exit on that thread.
What changed: Both rows exit on the SCT 2″ × 24 thread. Celestron publishes no optical length for its T-adapters, so the Backfocus Solver reports trains through these reducers as blocked until a length is published.
Settled by: Celestron knowledge base, SCT visual back thread
Page: firstrig.com/tools/backfocus

C-2210 October 2026Was on the site

The Bortle 2 preset sat in class 1

Published or assumed: Choosing Bortle 2 in the Exposure Planner or the Rig Reality Check ran the tools on a sky of 21.9 mag/arcsec².
What we found: 21.9 is the lower bound of class 1 in the same table, so a Bortle 2 choice was computed as a class 1 sky.
What changed: Class 2 uses 21.75, the middle of its range. The bounds did not change, and a test now checks that every class's value falls inside its own class.
Settled by: The Bortle table on the sky brightness method page
Page: firstrig.com/tools/exposure

C-2110 October 2026Was on the site

The home page promised a resale estimate

Published or assumed: The second of the three commitments on the home page said every build carries a resale estimate.
What we found: No build carries one. There is no data source for resale prices yet, and every build card says “not yet estimated”.
What changed: The commitment now says every build states what it cannot do and says when a figure is missing. Resale estimates will be added when there is a source for them.
Settled by: The Rig Reality Check's own build cards
Page: firstrig.com

C-206 October 2026Was on the site

The observatory guide described the design as the build

Published or assumed: The build guide, published on 26 September, described the plan: a 2×10 skirt with a fascia, gray electrical conduit for both cable runs, deck blocks at 2.16 lb, and a photo of the finished deck in the slot for the decking going down.
What we found: As built, the rim is 2×8 with the front face wrapped in a deck board and no fascia. The runs are white plumbing PVC carrying only Ethernet and 12 V DC. The blocks are 2.2 lb by the box, and the pier top is 28⅞″ above the deck, ¾″ under the design.
What changed: Each chapter keeps the design and adds what was built. The photo slot shows the first boards going down, and the conduit section says when listed electrical conduit is required.
Settled by: The builder's measurements of the finished observatory, 6 October 2026
Page: firstrig.com/observatory

C-1928 September 2026Caught before publication

An SSD priced at an out-of-stock sale price

Published or assumed: The first pass at essential prices put the storage line at Samsung's own $479.99 for the T7 2TB.
What we found: That was a sale price on an item Samsung showed as out of stock, on both 27 and 28 September.
What changed: The row is priced at Newegg's own in-stock offer, $399.99, sold and shipped by Newegg. It remains the least stable price in the set and is among the first rows due a check by hand.
Settled by: Newegg, Samsung T7 2TB
Page: firstrig.com/tools/rig-reality-check

C-1828 September 2026Caught before publication

Figures in the prototype that had no source

Published or assumed: The prototype the Rig Reality Check was built from showed “$1,670 you haven't thought of yet” for accessories, guided RMS figures, and clear-night presets labeled by region (“40, typical temperate”).
What we found: The accessory prices and the RMS figures were invented, and the regional labels had no source. No mount maker publishes a guided RMS.
What changed: Accessories are priced from 30 sourced rows. Clear nights are your own number. The tool states the guided RMS each build needs, and scores guiding only when you enter your own figure.
Settled by: Dated retailer and maker listings for every essential; catalogue sourcing notes
Page: firstrig.com/tools/rig-reality-check

C-1727 September 2026Was on the site

A false “don't buy” at $3,000

Published or assumed: The unlisted draft of the Rig Reality Check told a $3,000 buyer that nothing modular fits.
What we found: No camera or corrector in the catalogue carried a price yet, so no build could be fully priced. A missing figure was being presented as advice not to buy.
What changed: 28 cameras and 17 correctors were priced. When nothing ranks only because prices are missing, the tool says the catalogue is incomplete, and offers neither a smart telescope nor “you need more budget”.
Settled by: The tool's rule that a missing figure is counted and named, never estimated
Page: firstrig.com/tools/rig-reality-check

C-1627 September 2026Was on the site

Mounts priced without the tripod they ship with

Published or assumed: The mount catalogue's first pass put head-only prices on the rows for packages that include a tripod: the CEM40, CEM26, GEM28, Star Adventurer GTi and CQ350.
What we found: The engine adds a tripod line only when a mount ships without one, so the unlisted draft of the Rig Reality Check priced those builds with no tripod at all.
What changed: One row is one package: its tripod, its counterweight and every price on it describe the same thing, and a test fails when a price's contents contradict the row.
Settled by: Each maker's own store listing; catalogue sourcing notes
Page: firstrig.com/tools/rig-reality-check

C-1527 September 2026Error in a maker's material

Sky-Watcher USA's EDX specs repeat the discontinued ED line

Published or assumed: Sky-Watcher USA's spec tables for the Evostar and Esprit EDX refractors.
What we found: Their weights (7.3 lb, 8.4 lb) and flattener figures (63/40 mm, 66/33 mm) match the old ED manuals exactly. The Evostar 100EDX is 700 mm f/7, where the 100ED was 900 mm f/9, and the current Esprits ship without a flattener.
What changed: The EDX rows come from Sky-Watcher's global pages and the EDX manuals, and no EDX row links a corrector made for the old ED scopes.
Settled by: Sky-Watcher Esprit 100EDX manual
Page: firstrig.com/tools/rig-reality-check

C-1427 September 2026Error in a maker's material

A pound-to-kilogram conversion on the HEQ5-R Pro page

Published or assumed: Sky-Watcher USA's HEQ5-R Pro spec page gives the payload as “33 pounds (9.97kg)”, and its prose says 30 pounds.
What we found: 33 lb is 14.97 kg, and the page's two US figures disagree with each other. Sky-Watcher's global page and the manual both say 15 kg.
What changed: The row is keyed at 15 kg from the manual. Where Sky-Watcher's US spec data disagrees with a manual or the global page, the manual wins; the same data prints an EQ8-R head weight with an inch mark.
Settled by: Sky-Watcher HEQ5-R Pro manual
Page: firstrig.com/tools/rig-reality-check

C-1327 September 2026Was on the site

Three EdgeHD reducers in one row, and a solar filter's part number

Published or assumed: One Backfocus Solver row covered Celestron's 0.7x reducers for the EdgeHD 9.25, 11 and 14, listed as #94243, #94241 and #94240.
What we found: They are three products. The 9.25's is #94245, a 4-element design; the 11 and 14 take 5-element reducers. #94243 is Celestron's EclipSmart solar filter for 6″ SCTs.
What changed: Three rows, each from its own instruction sheet, each at 146 mm, with its own price.
Settled by: Celestron, 0.7x reducer for the EdgeHD 9.25Celestron, EclipSmart solar filter for 6″ SCTs
Page: firstrig.com/tools/backfocus

C-1227 September 2026Was on the site

The natural sky was a textbook minimum

Published or assumed: The sky-brightness model, published the day before, took the natural zenith sky as 22.0 mag/arcsec².
What we found: 22.0 is a minimum: solar minimum, away from the Milky Way and the zodiacal light. The National Park Service's measured median over 295 nights, with the artificial light removed, is 21.83. Against the NPS total-sky measurements the model read 0.15 mag too dark on average.
What changed: The natural sky is 21.83, the kernel was re-fitted for it, and the bias fell to +0.02. The model can no longer return Bortle 1, and the method page says to read its darkest answer as Bortle 2 or better.
Settled by: NPS Night Skies Program measurements
Page: firstrig.com/method/sky-brightness

C-1127 September 2026Caught before publication

Worked examples that did not reproduce

Published or assumed: The physics reference's exposure examples gave read noise as 1.0 e⁻ but quoted a 148 s shortest sub at 7 nm, a nebula rate of 0.258 e⁻/px/s and 18.4 hours to SNR 40.
What we found: Only 1.5 e⁻ gives 148 s, and ZWO's chart puts the ASI533 at about 1.5 e⁻ at gain 100 with HCG on. The reference's own formula gives the nebula rate as 0.270, and the closed form gives about 18.5 hours.
What changed: The examples use 1.5 e⁻ throughout: 8.1 s without a filter and 147 s at 7 nm. Three pending tests built on them now run.
Settled by: ZWO's read-noise chart for the ASI533 Pro
Page: firstrig.com/method#exposure

C-1026 September 2026Caught before publication

The observatory draft disagreed with its own spec

Published or assumed: The observatory guide's draft gave a 10″ pier in two chapters and about 100 lb for the dome, and the spec put the deck surface 12¼″ above grade.
What we found: The spec has a 12″ pier and a 250 lb dome, and its own joist height plus 1″ decking puts the deck surface at 11¾″.
What changed: The prose and the deck height were brought into line with the spec before the guide was published. The spec's derived numbers are recomputed from its inputs in the tests.
Settled by: The build spec the guide's tables render from
Page: firstrig.com/observatory

C-0926 September 2026Caught before publication

Satellite data stored at ten times its scale

Published or assumed: Every year of the mirrored VIIRS night-light files was read as radiance in nW/cm²/sr.
What we found: The mirror stores 2012 to 2021 at ten times that scale and 2022 to 2024 at one. Read as nW, a 2016 file overstates radiance tenfold; it surfaced as a calibration constant that jumped by a factor of 8 between years. The mirror's 2000 to 2011 files are extrapolated.
What changed: The pipeline declares the scale for each year and checks it against the file every time one is opened. The extrapolated years are never used.
Settled by: The files themselves: every lit value from 2012 to 2021 is a multiple of 10OpenLandMap VIIRS mirror, Zenodo record 17294744
Page: firstrig.com/method/sky-brightness

C-0826 September 2026Was on the site

A spacer in front of the off-axis guider

Published or assumed: When a spacer could sit on either side of an off-axis guider with the same back focus, the solver put it on the corrector side.
What we found: That placed ZWO's 21 mm spacer in front of the OAG, the opposite of ZWO's own charts. The guide camera focuses at a fixed distance from the prism, so the length between the OAG and the sensor is set by the OAG's design.
What changed: Among equally good stacks the solver keeps spacing behind the guider, and reports the guider-to-sensor distance.
Settled by: ZWO OAG manualDiffraction Limited, OAG-8300 manual
Page: firstrig.com/tools/backfocus

C-0726 September 2026Was on the site

A focus range stored as a target

Published or assumed: Correctors that focus over a range, such as Askar's FRA400 and FRA500, stored the longest train they take as a target, with the default ±0.5 mm tolerance.
What we found: A maximum cannot be exceeded at all. The solver accepted a train 0.5 mm past a length the scope cannot focus, and asked for spacers the scope did not need.
What changed: A published range is stored as a range, and no train past the maximum is accepted. Where the maker gives only a maximum, the result says how far the focuser racks out, for you to check against your own focuser.
Settled by: ZWO FF65 APO manualAskar FRA400
Page: firstrig.com/tools/backfocus

C-0625 September 2026Error in a maker's material

ZWO's 55 mm charts do not add up

Published or assumed: ZWO's “ASI Camera 55mm Back Focus Solution” charts, the obvious place to copy a total from.
What we found: Page 4 sums to 55 under a “56mm” caption, page 7 draws the camera at 16.5 mm instead of 17.5, and page 17 sums to 56 under a “55mm” caption. The component figures are consistent across the charts. The captions are wrong.
What changed: The catalogue takes each part's own length and sums the parts. No ZWO total is copied.
Settled by: ZWO, ASI Camera 55mm Back Focus Solution (2025)
Page: firstrig.com/tools/backfocus

C-0524 September 2026Caught before publication

The solver chose your thread adapter

Published or assumed: When a corrector's thread and a camera's differed, the solver picked the shortest adapter chain from the whole catalogue.
What we found: The M48 to T2 conversion alone is sold at 0, 1, 3.5, 5.3 and 6 mm of optical length. Picking one assumed you owned it, and a train built on the 6 mm ring would come out 6 mm long.
What changed: The solver never picks an adapter. When your kit cannot join the threads, it asks which ring you have and offers only rings that complete the path.
Settled by: Manufacturer lengths in the adapter catalogue
Page: firstrig.com/tools/backfocus

C-0424 September 2026Caught before publication

Spacer kits were lists of lengths

Published or assumed: Spacer kits were modeled as lengths, solved with unlimited repeats, and thread adapters as a separate step.
What we found: ZWO's own diagrams show the ASI2600 Pro kit as a 0 mm M42 to M48 ring, a 16.5 mm M48 to M42 extender and a 21 mm M42 spacer, so an M48 corrector screws straight onto the extender. The old model sent that owner to buy an adapter they did not need, then could not reach 55 mm. Repeats let the Baader RCC I solve with two of each extender from a box that holds one.
What changed: Every spacer, extender and adapter is one threaded part with an optical length. A kit holds one of each piece, and no ring is used more often than you own it.
Settled by: ZWO ASI2600 manual, connection diagrams
Page: firstrig.com/tools/backfocus

C-0326 August 2026Caught before publication

ZWO's 17.5 mm means two different things

Published or assumed: The catalogue's first design held one back-focus figure per camera.
What we found: ZWO quotes 17.5 mm across its cooled cameras. On the ASI533, ASI585, ASI294, ASI183 and ASI1600 it is a 6.5 mm body plus the 11 mm ring in the box; on the ASI2600, ASI6200 and ASI071 it is the bare body. ZWO's manuals headline the with-ring figure for some models and the bare one for others on identical arrangements. Guessing wrong is an 11 mm error.
What changed: Every camera row states which configuration its figure describes. Where a camera ships in more than one, the solver asks which you have before it answers.
Settled by: ZWO ASI2600 manual: 17.5 + 21 + 16.5 = 55 with the two extenders in the box
Page: firstrig.com/tools/backfocus

C-0216 August 2026Caught before publication

Two forms of the Dawes limit presented as one

Published or assumed: The physics reference gave 116 / D in millimeters and the widely quoted 4.56 / D in inches as the same formula.
What we found: 116 / 25.4 is 4.567, so the two forms differ by 0.15%, about 0.0009″ at 200 mm. Found the same way, by unit-testing against the reference.
What changed: The physics package implements the millimeter form only. Any inch form shown on the site is derived from it.
Settled by: Arithmetic: 116 / 25.4 = 4.567
Page: firstrig.com/method

C-0116 August 2026Caught before publication

The critical focus zone shorthand was for another wavelength

Published or assumed: The physics reference gave the critical focus zone as about 2.2 × f² micrometers at 550 nm.
What we found: 4.88 × 0.550 is 2.684, so the shorthand at 550 nm is 2.68 × f². The 2.2 coefficient is the 450 nm case and is often quoted without its wavelength. At f/5 that is 67 µm against 55 µm. Found by unit-testing the shipped function against the reference.
What changed: The formula itself was always right. The reference now gives the shorthand as 2.68 × f², and a unit test checks the function against it.
Settled by: Arithmetic: CFZ = 4.88 × λ × f², with λ = 0.550 µm
Page: firstrig.com/method

Reporting an error

A public error tracker is coming.

Last updated 2026-10-12

01Index