Reference sonar systems

Confidence

• CMRE MUSCLE: high. Frequency, bandwidth, element count and array length come from the designers' journal paper and are repeated by three independent sources; only the range figure varies by source and definition.

How each value was reached

• Channels: Gerg 2018 says thirty-two channels; Bellettini 2009 (the design paper) and Hansen 2011 say 36.
• Channel spacing: 1.2 m / 36 elements = 33.3 mm [Bellettini 2009]; cross-check: Fig. 10 gives the pitch as 6.66 wavelengths (the lambda glyph is lost in the text as '6:66') and lambda = 5 mm at 300 kHz [Bellettini 2009], so 6.66 x 5 mm = 33.3 mm
• Along-track resolution: 2.5 cm is the nominal figure [Reed 2019]; the 2006 trial images measured 5 cm cross-range [Bellettini 2009]. 2.5 cm is not d/2 of the 33.3 mm pitch (16.7 mm); it matches Bellettini 2009's design example of 24 elements over 1.2 m (50 mm), the 36 elements being 3/2 spatial oversampling [Bellettini 2009].
• Across-track resolution: 1.25 cm = c/(2B) = 1500/(2 x 60 kHz), stated by Bellettini 2009 as the maximum range resolution; 1.6 cm measured in the 2006 trial [Bellettini 2009]
• Maximum range: 170 m: design far range for a 1.2 m array at SAS oversampling 4/3 and 2 m/s [Bellettini 2009], also the demonstrated range [Bellettini 2009]. 140 m: nominal imaging range of the fielded system [Gerg 2018]. Others: about 150 m maximum effective range for 6 deg grazing at 20 m depth [Bellettini 2009]; 150 m [Reed 2019]; 200 m at 2 m/s [Hansen 2011].

Where the sources disagree

• Channel count: Bellettini 2009 and Hansen 2011 give 36 elements over 1.2 m; Gerg 2018 calls MUSCLE 'a thirty-two channel side looking SAS'. The design paper is the primary source, so 36 stands; whether the fielded array differs is not settled by any source read.
• Maximum range has five values by source and definition: 170 m design far range at 2 m/s and 170 m demonstrated (Bellettini 2009), about 150 m maximum effective range from the 6 deg grazing criterion at 20 m water depth (Bellettini 2009), 140 m nominal imaging range (Gerg 2018), 150 m (Reed 2019), and 200 m at 2 m/s (Hansen 2011). Hansen 2011's 200 m matches neither c L/(4v) = 1500 x 1.2/8 = 225 m nor that value over the 4/3 oversampling ratio (169 m).
• Resolution: nominal 2.5 cm along-track x 1.25 cm across-track (Reed 2019; Bellettini 2009 gives 1.25 cm as the maximum range resolution) versus 1.6 cm range x 5 cm cross-range measured in the June 2006 trial (Bellettini 2009).
• The 2.5 cm along-track figure is not the d/2 limit of the 33.3 mm element pitch (16.7 mm); Bellettini 2009's design example reaches 2.5 cm with 24 elements over 1.2 m (50 mm) and describes 36 elements as 3/2 spatial oversampling.

Where the simulator's file differs from the sources

• along track resolution cites only Reed 2019 for 2.5 cm; the value is a nominal figure, and Bellettini 2009 measured 5 cm cross-range in 2006 (already noted in the source string).

Sources and quotations

Simulator file: sascore/systems/cmre_muscle.json

Confidence

• Kongsberg HISAS 1030: high. Every cell is stated by an FFI paper (Synnes 2010, Hansen 2011) and matches the Kongsberg sheets (Kongsberg 2006, Kongsberg 2010); only the element pitch is computed, and it agrees with the published element length.

How each value was reached

• Center frequency: default 100 kHz is the centre of the typical 85-115 kHz configuration [Hagen 2010], inside the 50-120 kHz configurable band [Hagen 2010; Hansen 2011]
• Bandwidth: default is the typical 30 kHz (115 - 85 kHz [Hagen 2010]); the system bandwidth is up to 50 kHz [Kongsberg 2006; Kongsberg 2010]
• Channel spacing: 1.2 m receiver array / 32 elements = 37.5 mm [Hansen 2011; Hagen 2010]; equals the published element length 3.75 cm of a dense (contiguous) array [Synnes 2010]
• Along-track resolution: check: theoretical d/2 = 37.5 mm / 2 = 1.9 cm [Kongsberg 2006; Hansen 2011], consistent with the published 2 cm
• Across-track resolution: check: c/2B = 1500/(2 x 30 kHz) = 2.5 cm at the typical 30 kHz; 1.5 cm at the 50 kHz maximum
• Maximum range: 2 m/s = 3.9 kn; Hagen 2010 gives 175-200 m at 4 knots and 230-260 m at 3 knots, consistent with 200 m at 2 m/s

Where the sources disagree

• Frequency band: Hagen 2010 (2010) says frequency and bandwidth are configurable within 50-120 kHz with 85-115 kHz as the typical configuration, which reconciles Hansen 2011's 50-120 kHz total range with the 100 kHz / 30 kHz typical values (Synnes 2010, Hansen 2011). The Kongsberg sheets give 60-120 kHz (Kongsberg 2006, Kongsberg 2010), the 2012 slides 60-110 kHz (Hancock 2012 slide 32) and 70-100 kHz (Hancock 2012 slide 6). The same Holmengraa/U-735 image is captioned 80-120 kHz in Hagen 2010 slide 6 and 85-115 kHz in Lleches 2016.
• Receive array length: 1.27 m in the Kongsberg sheets (Kongsberg 2006, Kongsberg 2010) versus 1.2 m with 32 elements in Hansen 2011 and Hagen 2010; Synnes 2010 gives 32 elements of 3.75 cm (1.20 m). Three FFI-side sources agree on 1.2 m of active elements; the 1.27 m figure is most likely the array including its end margins, but no source says so.
• Range at 1.5 m/s: 275 m in Kongsberg 2006 (2006 sheet) versus 260 m in Kongsberg 2010 (2010 brochure). The 200 m @ 2 m/s point agrees in Kongsberg 2006, Hansen 2011, Kongsberg 2010 and Hagen n.d. (2x200 m swath), and Hancock 2012 gives ~200 m. Kongsberg 2015 (2015) gives 300 m either side with no speed, and Lleches 2016 shows HISAS 1030 imagery out to 325 m at 2.3 knots.
• Transmitter length: 0.35 m in Kongsberg 2006 versus 0.32 m in Kongsberg 2010.
• Maximum instantaneous area coverage rate: 750 m2/s in Kongsberg 2006 versus 730 m2/s in Kongsberg 2010; typical 2 km2/h in Hansen 2011 and more than 2 km2/h in Hagen n.d..
• Range versus speed: Hagen 2010 gives 175-200 m @ 4 knots and 230-260 m @ 3 knots (or 10 x altitude); Kongsberg 2006 and Kongsberg 2010 give 200 m @ 2 m/s (3.9 kn) and 275 m (Kongsberg 2006) or 260 m (Kongsberg 2010) @ 1.5 m/s (2.9 kn). These agree within the Hagen 2010 spans. Hagen 2010 slide 16 shows a pipeline survey with sonar range approx. 180 m at 4 knots.
• Area coverage rate: typical 2 km2/h in Hansen 2011, more than 2 km2/h in Hagen n.d.; Hagen 2010 slide 16 reports 2 nmi x 500 m covered in 1 hour with heavy overlap.

Where the simulator's file differs from the sources

• channel spacing = 0.040 m (derived from '1.27 m housing length [Kongsberg 2006] / 32'): Kongsberg 2006 calls 1.27 m the 'Primary receiver array' length, not a housing, and Synnes 2010 states 'Each element is 3.75 cm long' (Hansen 2011: 1.2 m / 32 = 37.5 mm). The sources support 0.0375 m, the value the catalog only records as 'proposed'. tx_element_azimuth and rx_element_azimuth inherit the same 0.040 m.
• maximum range is tagged not_public with value null: Kongsberg 2006 lists 'Maximum range' 200 m @ 2 m/s, Hansen 2011 Table 2 lists 'Maximum range @ 2 m/s 200 m' and Kongsberg 2010 lists 'Maximum range (each side of vehicle): 200 m @ 2 m/s'. It should be 200 m, published.
• sources.H10 ('Synthetic aperture sonar processing for the HUGIN AUV' and FFI summaries) is the wrong source for '175-200 m @ 4 knots, 230-260 m @ 3 knots': Hansen et al., Oceans Europe 2005, read in full, covers the SENSOTEK prototype and the EdgeTech 4400-SAS and gives no HISAS 1030 range. The pairs are verbatim in Hagen, UI 2010 (Hagen 2010, slide 8). Re-cite image far range to Hagen 2010 and drop H10.
• area coverage gives '750 m2/s maximum instantaneous' from Kongsberg 2006 only; the later Kongsberg brochure (Kongsberg 2010) gives 730 m2/s.
• review.notes band entry ('60-120 kHz [Kongsberg 2006] versus 50-120 kHz [Hansen 2011]') is largely settled by Hagen 2010: 50-120 kHz is the configurable range and 85-115 kHz the typical configuration.
• review.notes says FFI gives 'the acoustic array as 1.2 m [Hansen 2011]' while the catalog uses 1.27 m: Hagen 2010 also gives '1.2 m, 32 Rx elements', so the 1.2 m / 37.5 mm reading has a second independent source.
• transmit element length source gives the transmitter as 0.35 m [Kongsberg 2006]; Kongsberg 2010 gives 0.32 m.

Sources and quotations

Simulator file: sascore/systems/kongsberg_hisas_1030.json

Confidence

• Kongsberg HISAS 1032: medium. Centre frequency, resolution and range are now published by Kongsberg (Kongsberg n.d., Lleches 2016, Kongsberg 2015); bandwidth and array geometry are still assumed from the HISAS 1030, though the published ranges fit a 1030-size array.

How each value was reached

• Center frequency: Kongsberg 2016 product table lists HiSAS 1032 at 100 kHz; no band or centre is given elsewhere
• Bandwidth: no 1032 bandwidth is published; the HISAS 1030 typical 30 kHz [Hansen 2011] is used because the 1032 shares the 100 kHz centre [Lleches 2016] and the family designation
• Channels: no 1032 array data is published; HISAS 1030 value [Hansen 2011; Hagen 2010]. The 1032 ranges in Kongsberg n.d. (240 m @ 3 kn, 185 m @ 4 kn) fall inside the HISAS 1030 spans in Hagen 2010 (230-260 m @ 3 knots, 175-200 m @ 4 knots), and the published ranges imply a receive array of about 1.0 m minimum (see max_range_m), consistent with a 1030-size 1.2 m array
• Channel spacing: HISAS 1030 element length 3.75 cm [Synnes 2010]; 1032 element size is not published
• Maximum range: consistency check with Rmax = cL/(4v) [Hansen 2011]: L >= 4 x 1.543 m/s x 240 m / 1500 m/s = 0.99 m; at 4 kn, 4 x 2.058 x 185 / 1500 = 1.02 m; the HISAS 1030 rating gives 4 x 2 x 200 / 1500 = 1.07 m against its 1.2 m array

Where the sources disagree

• The HISAS 1032 ranges (240 m @ 3 knots, 185 m @ 4 knots; Kongsberg n.d.) fall within the HISAS 1030 spans given in Hagen 2010 (230-260 m @ 3 knots, 175-200 m @ 4 knots). This supports, but does not prove, the assumption that the 1032 uses a 1030-size array.
• Range: 240 m @ 3 kn and 185 m @ 4 kn per side (Kongsberg n.d., HUGIN 3000 data sheet) versus 'Swath up to 300 m either side' (Kongsberg 2015, shared 1030/1032 line) and '300m' in the 'Depth' column of the 2016 product table (Lleches 2016). Naval Technology 2019 gives a '1,000m swath at 2.5kt' for the HUGIN Superior; if that is 500 m per side, Rmax = cL/(4v) needs L >= 4 x 1.286 x 500 / 1500 = 1.71 m, far more than the array implied by Kongsberg n.d. (about 1.0 m), so either the Superior's array is longer or the figure is not a full-resolution SAS swath.
• Resolution: 5x5 cm in Kongsberg n.d. and approximately 5 x 5 cm in Naval Technology 2019, versus 4x4 cm in practice and 2 x 2 cm theoretical in Kongsberg 2015.
• Area coverage rate: 2-2.5 km2/hour (Kongsberg n.d.) versus typically 2 km2/hr (Lleches 2016).
• The 200, 300 and 400 kHz figures that appear next to HISAS on Kongsberg payload slides (Kongsberg 2015 p. 4) and in the HUGIN Superior article (Naval Technology 2019) belong to the EM 2040 multibeam, not to the HISAS 1032.

Where the simulator's file differs from the sources

• center frequency is tagged assumed ('as the HISAS 1030'); Lleches 2016 (Kongsberg, 2016) lists 'HiSAS 1032 100 kHz', so it can be tagged published [Lleches 2016].
• sources.Kongsberg n.d..read says 'unreachable (HTTP 500); figures from a search-engine excerpt': the data sheet was read in full from an Internet Archive copy, and it confirms 240 m @ 3 knots and 185 m @ 4 knots. The document number printed on that copy is 110-0047668/A, not -b as in the citation.
• review.notes says a Kongsberg payload slide lists 200, 300 and 400 kHz 'for the 1030/1032': on the slide that was found (Kongsberg 2015 p. 4) those frequencies belong to the EM2040 multibeam entry, so there is no contradiction with the 1030 data sheet.
• summary says 'No acoustic specification is public': Kongsberg n.d., Lleches 2016 and Kongsberg 2015 publish the centre frequency, resolution, range and coverage rate.

Not yet obtained

• A Kongsberg HISAS 1032 product data sheet (none found). Would settle: bandwidth, element count and element length, and whether the HUGIN Superior 1000 m swath is a SAS figure

Sources and quotations

Simulator file: sascore/systems/kongsberg_hisas_1032.json

Confidence

• Kongsberg HISAS 2040: medium. Band, range and resolution are published (Hydroid 2018, Kongsberg 2015); centre, bandwidth and pitch are computed from them, and only the channel count is assumed.

How each value was reached

• Center frequency: centre of 220-280 kHz [Hydroid 2018]: (220 + 280)/2 = 250 kHz
• Bandwidth: full band 280 - 220 = 60 kHz [Hydroid 2018], assuming all of it is used; check: the 1.2 cm theoretical across-track resolution [Kongsberg 2015] needs c/(2 x 0.012 m) = 62.5 kHz
• Channels: not published; the HISAS 1030 count [Hansen 2011]. Lower bound from Rmax = cL/(4v) [Hansen 2011]: L >= 4 x 2 m/s x 130 m / 1500 m/s = 0.69 m [Hydroid 2018], so N >= 0.69 m / 26 mm = 27
• Channel spacing: element d = 2 x delta_x [Hansen 2011] with delta_x = 1.3 cm along-track theoretical [Kongsberg 2015]; reading '1.2 x 1.3 cm' as across x along follows Kongsberg's HISAS 1030 convention [Kongsberg 2006], Kongsberg 2015 does not state the order

Where the sources disagree

• Practical resolution: 5cm x 5cm (Hydroid 2018, 2018) versus 4 x 4 cm (Kongsberg 2015, 2015, a line that may cover both HISAS models) versus 'up to 2cm by 2cm resolution across a 300-meter swath' (Hydroid 2019, 2019).
• Range: 130 m each side at 2 m/s (Hydroid 2018) versus 150 m either side with no speed (Kongsberg 2015) and a 300-meter swath (Hydroid 2019, i.e. 150 m per side if symmetric).
• The 1.2 x 1.3 cm theoretical resolution and 4 x 4 cm practical figure come from the Kongsberg AUT 2015 Perth presentation (Kongsberg 2015), not from the MACHC13 deck (H9). H9 is a 2012 deck that mentions only the HiSAS 1030.

Where the simulator's file differs from the sources

• sources.H9 is cited as 'Kongsberg presentation to IHO MACHC13 (2015)' and as the source of the 1.2 x 1.3 cm theoretical and 4 x 4 cm practical resolution: the MACHC13 deck (La Antigua, 19-22 November 2012, read in full) does not mention the HISAS 2040. Those figures appear in the Kongsberg 'HUGIN AUV Sensor Suite' presentation, AUT 2015 Perth (Kongsberg 2015). Re-cite channel spacing and resolution to Kongsberg 2015, and correct H9's year to 2012.
• bandwidth source note cites the 1.2 cm range resolution to [H9]; it is in Kongsberg 2015.
• review.notes says the element length 'rests on a 2015 Kongsberg slide (search excerpt only)': the slide (Kongsberg 2015) has now been read in full. It does not say which number is across-track and which is along-track.

Not yet obtained

• A Kongsberg HISAS 2040 product data sheet (none found). Would settle: element count and length, bandwidth actually used, and the speed for the 150 m range

Sources and quotations

Simulator file: sascore/systems/kongsberg_hisas_2040.json

Confidence

• Kongsberg HISAS2020 ACR 80: low. Only the band, the resolution bounds and the module length are published; bandwidth, pitch and channel count are computed from bounds or assumed, and no range is given.

How each value was reached

• Center frequency: centre of 120-210 kHz [Kongsberg 2026a; Kongsberg 2026b]: (120 + 210)/2 = 165 kHz
• Bandwidth: B = c/(2 delta_r) = 1500 / (2 x 0.010 m) = 75 kHz from the <10 mm premium resolution [Kongsberg 2026a]; a lower bound, since the stated resolution is an upper bound. The full band is 210 - 120 = 90 kHz [Kongsberg 2026a]
• Channels: not published. 796 mm TRX module length [Kongsberg 2026a] / 20 mm pitch = 39.8, if the module is one contiguous receive array and the pitch equals 2 x the premium resolution; both are assumptions
• Channel spacing: d = 2 x delta_x [Hansen 2011] with delta_x = 10 mm premium [Kongsberg 2026a]; an upper bound on the element size, since <10 mm is a bound. The 13 mm trial estimate [Kongsberg 2026c] would allow up to 26 mm
• Maximum range: no range is published. If the 1.2 km2/h rate [Kongsberg 2026a] covers both sides, R = ACR/(2v) = 333 m2/s / (2v): 162 m at 2 kn, 81 m at 4 kn; the speed it refers to is not stated

Where the sources disagree

• Resolution: the data sheet gives '<10x10mm' for SAS premium (Kongsberg 2026a); Kongsberg's sea-trial note reports 'approximately 13 mm × 13 mm' (Kongsberg 2026c).
• Power: '<90W incl processing' in the specification table versus 'a total power draw of less than 100W' in the same data sheet's text (Kongsberg 2026a).
• Area coverage: 1.2 km2/h (Kongsberg 2026a) is about 56 % of the SAS limit cL/2 = 1500 x 0.796 / 2 = 597 m2/s = 2.15 km2/h for two 796 mm arrays [Hansen 2011], so either the acoustic aperture is shorter than the module or the rated mode uses large ping overlap.

Where the simulator's file differs from the sources

• summary: 'HISAS2020 is a model number, not a year [Kongsberg 2026b]': Kongsberg 2026b does not say this. It says the launch represents 'the fourth generation' of HISAS, and gives no account of the name.
• channel count source and review.notes treat an 80 cm array as an unconfirmed reading of 'ACR 80': Kongsberg 2026a publishes 'Transducer setup 2×800mm TRX modules' (drawn as 796 mm long). The module length is published, though no source says the whole module is receive aperture.
• power w '< 90 W including processing' is from the table; the same data sheet's text says less than 100 W.

Not yet obtained

• A HISAS2020 user or integration manual, or an FFI/Kongsberg paper on the HISAS2020 array. Would settle: element count and length, the bandwidth used in premium mode, the meaning of 'ACR 80', and range versus speed

Sources and quotations

Simulator file: sascore/systems/kongsberg_hisas2020.json

Confidence

• Kraken MINSAS 60: medium. Frequency, bandwidth, resolution and range are published, but the two Kraken sheets disagree on range (100 m vs 118 m at 3 kn) and the channel count and pitch are derived.
• Kraken MINSAS 120: high. All values except the channel pitch are published in two Kraken sheets (Kraken 2025, Kraken 2024a); the channel count needs only the module-length match, and the pitch is an upper bound.
• Kraken MINSAS 180 (KATFISH): high. The KATFISH 180 sheet (Kraken 2024b) publishes frequency, bandwidth, resolution and range-versus-speed; the channel count and pitch are derived from its 192-channel total and Steele 2024's 1.6 m rows.
• Kraken MINSAS 240: low. The only source is a 2019 press release that gives resolution, ACR (5.3 km2/h) and tow speed (up to 20 kn) but no frequency, array length, channel count or range.

How each value was reached

• Channels, Kraken MINSAS 60: One 16-channel receiver module per row: Dillon 2020 gives 16 channels per module (channels 0-15 aft, 16-31 forward); Kraken 2024a gives the MINSAS 60 receiver as 53.0 cm against 109.0 cm for the MINSAS 120 (about two modules), so the 60 is one module.
• Channel spacing, Kraken MINSAS 60: d ≤ receiver housing length / channels = 530 mm [Kraken 2024a] / 16 [Dillon 2020, derived] = 33.1 mm; an upper bound, since the elements sit inside the housing.
• Channels, Kraken MINSAS 120: Dillon 2020 publishes 32 channels (two 16-channel modules, aft and forward) along one side of a MINSAS array but does not name the array length; Kraken 2024a gives the MINSAS 120 receiver as 109.0 cm, about two 53.0 cm MINSAS 60 modules, so the two-module array of Dillon 2020 is the 120.
• Channel spacing, Kraken MINSAS 120: d ≤ receiver housing length / channels = 1090 mm [Kraken 2024a] / 32 [Dillon 2020] = 34.06 mm. If the 120 uses the same 16-element module as the 60, d ≤ 530/16 = 33.1 mm, because 16 x 34.06 = 545 mm does not fit in the 530 mm MINSAS 60 housing. (Dillon 2019 images on a 1.67 cm along-track pixel, which would match d/2 for d = 33.4 mm; not used.)
• Channels, Kraken MINSAS 180 (KATFISH): 192 channels total, dual sided [Kraken 2024b] / (2 sides x 2 interferometric rows [Steele 2024]) = 48 per row; consistent with three 16-channel modules [Dillon 2020].
• Channel spacing, Kraken MINSAS 180 (KATFISH): row length approximately 1.6 m [Steele 2024] / 48 channels (derived above) = 33.3 mm; approximate because Steele 2024 gives the length only as approximately 1.6 m (Kraken 2024b lists the array as 180 cm x 7 cm, a housing dimension).
• Center frequency, Kraken MINSAS 240: Not published for the 240 [Kraken 2019]; assumed equal to the rest of the MINSAS family (Kraken 2025, Kraken 2024a, Kraken 2024b all give 337 kHz).
• Bandwidth, Kraken MINSAS 240: Not published for the 240 [Kraken 2019]; assumed equal to the MINSAS family's 40 kHz (Kraken 2024a, Kraken 2024b).
• Channels, Kraken MINSAS 240: Not published. The array names step by one 60 cm, 16-channel module (60 = 1 module, 120 = 2, 180 = 3: Kraken 2024a, Dillon 2020, Kraken 2024b), so the 240 is assumed to be four modules, 64 channels per row.
• Channel spacing, Kraken MINSAS 240: Not published [Kraken 2019]; assumed to be the MINSAS module pitch, taken at its upper bound 530 mm / 16 = 33.1 mm from the MINSAS 60 housing [Kraken 2024a] (Steele 2024 puts it near 33.3 mm, 1.6 m / 48).

Where the sources disagree

• MINSAS 60 range: Kraken 2025 (2025) gives "100 m per side" for the SAS-60 at 2 to 3.5 kn, while Kraken 2024a (2024) gives 118 m per side at 3.00 kn; both give 87 m at 4 kn.
• Pulse length: Kraken 2025 says "1 - 16 ms configurable"; Kraken 2024a and Kraken 2024b say "configurable 1 ms -> 10 ms".
• Operating speed: Kraken 2025 gives 2-4 kn (SAS-60) and 2-9 kn (SAS-120); Kraken 2024a gives 2-5 kn (MINSAS 60) and 2-6 kn (MINSAS 120).
• Kraken 2025's area-coverage table prints both "2 kn (1.54 m/s)" and "3 kn (1.54 m/s)"; 2 kn is 1.03 m/s, so the first row is a typo.
• Across-track resolution: Kraken 2025, Kraken 2024a and Kraken 2024b give 3.0 cm real time; Dillon 2019 measures 2.6 cm (complex image) on an AquaPix it calls "300 kHz"; Steele 2023 states "1.5 cm across track", which matches its 16-sample, 24 cm averaging window and is probably the pixel spacing.
• KATFISH 180 resolution: Kraken 2024b (2024) gives 3.3 cm along x 3.0 cm across; Kraken 2026 (2026) rounds this to "3 cm x 3 cm" real time.
• KATFISH altitude: Kraken 2026 says 5-25 m on one page and 5-30 m on another; Kraken 2024b says min 5 m, max 30 m.
• KATFISH swath: Kraken 2024b gives ">400 m at 4 kn" dual sided; the catalog's "440m at 4 kts" came from a search excerpt of an unidentified listing and was not found in any document.
• Module pitch: Kraken 2024a's receiver lengths (53.0 cm for 16 channels, 109.0 cm for 32) imply different upper bounds (33.1 mm vs 34.1 mm); 16 x 34.1 mm = 545 mm does not fit in the 530 mm MINSAS 60 housing, so the 120's L/N figure overstates the pitch if the modules are identical.
• Dillon 2019 describes AquaPix as "a wideband 300 kHz interferometric SAS"; Kraken 2025, Steele 2024, Kraken 2024a and Kraken 2024b all give a 337 kHz centre frequency (300 kHz appears to be a nominal class label).
• Steele 2024 gives the MINSAS 180 rows as "approximately 1.6 m in length"; Kraken 2024b lists the KATFISH array as "180 cm x 7 cm" (housing).

Where the simulator's file differs from the sources

• modes[120 cm].platform speed: still null (proposed 1.80); Kraken 2024a publishes 200 m per side up to 3.50 kn (1.80 m/s), so the 200 m range is not tied to a speed in the simulator file.
• modes[120 cm].sample rate: keeps 50 kHz (1.25 x bandwidth) with an unadopted proposal of 60 kHz; fs is a simulator convention, not a source value.
• pulse length: Kraken 2024a and Kraken 2024b give only a configurable 1 to 10 ms span (Kraken 2025: 1 to 16 ms). The 60 and 180 cm modes default to an assumed 10 ms (the most common pulse in Steele 2024 Table 1); the 120 cm mode stays null with 10 ms proposed, because its preset field set is pinned by tests/test_system_catalog.py; the 240 cm mode has no published span.

Not yet obtained

• The KATFISH listing quoted as "Dual sided max swath 200m at 8 kts (440m at 4 kts)" (seen only as a search-engine excerpt; not found on krakenrobotics.com or in any archived Kraken PDF). Would settle: Nothing the table needs now (Kraken 2024b gives the KATFISH 180 range-speed table); it would only explain the older 440 m figure.
• A Kraken MINSAS-240 datasheet (none found; only the 2019 press release Kraken 2019 is public). Would settle: MINSAS 240 centre frequency, bandwidth, receiver length, channel count and maximum range per side.
• A Kraken statement of the MINSAS receiver element pitch (or active row length) for any array. Would settle: Channel spacing for all four rows; the sources give only housing lengths and an approximate 1.6 m row.

Sources and quotations

Simulator file: sascore/systems/kraken_minsas.json

Confidence

• Thales SAMDIS: low. Only the resolution and the three-view design are published in the sources obtained; frequency, channel count and range are not public and the two technical papers that might give them (Chabah 2016, Burlet 2023) could not be retrieved.

How each value was reached

• Center frequency: Not published in any source obtained. 300 kHz is borrowed from the Thales-built MUSCLE SAS, which has the same 2.5 cm x 1.25 cm resolution pair [Reed 2019]; plausible range 200 to 450 kHz.
• Bandwidth: B = c / (2 dr) = 1500 / (2 x 0.0125 m) = 60 kHz, with dr = 1.25 cm across-track [Le Gall 2023] and c = 1500 m/s assumed; a lower bound if the 1.25 cm includes window broadening
• Channels: Not published; Defence Procurement International 2017 says only 'a combination of several array lengths' []. 32 is a placeholder within the 16 to 48 channels typical of AUV SAS.
• Channel spacing: element length d = 2 x dx = 2 x 25 mm = 50 mm from the 2.5 cm along-track resolution [Le Gall 2023], assuming dx = d/2 and contiguous elements. This is an upper bound on the pitch: the Thales-built MUSCLE reaches 2.5 cm with a 33.3 mm pitch.
• Along-track resolution: Le Gall 2023: 'up to 2.5cm' in real time; Defence Procurement International 2017: 1 inch
• Across-track resolution: Defence Procurement International 2017 gives 1 inch (2.54 cm) for both axes
• Maximum range: Defence Procurement International 2017 says only 'long-range'; DGA 2023 gives operating depth 7 to 300 m, not range

Where the sources disagree

• Across-track resolution: Le Gall 2023 (Thales authors, 2023) gives 1.25 cm; Defence Procurement International 2017 (trade press, 2017) gives 1 inch x 1 inch, i.e. 2.54 cm on both axes. Le Gall 2023 is the more specific and more recent source.
• Squint angles: no source obtained states the +30, 0, -30 degree beams. Defence Procurement International 2017 says three views 'under different angles', Le Gall 2023 says performance equivalent to three surveys with different track orientations, and DGA 2023 labels the three images Forward, Broadside and Backwards View.

Where the simulator's file differs from the sources

• platform lists 'REMUS 620 (SAMDIS 600)'; no retrieved source mentions REMUS 620 or SAMDIS 600 (DGA 2023 mentions only a 'Smaller SWaP SAMDIS payload for the next generation of platform').

Not yet obtained

• M. Chabah, N. Burlet, J.-P. Malkasse, G. Le Bihan and B. Quellec, "SAMDIS: A New SAS Imaging System for AUV," Springer 2016, pp. 107-118, doi:10.1007/978-3-319-32107-3_10 (paywalled; no public copy found). Would settle: centre frequency, bandwidth, array length and channel count, maximum range
• N. Burlet et al., "Intensive Resolution Measurement with the SAMDIS Multi-Aspect SAS," Proc. IOA 45(1), 2023 (catalogue URL in the catalog returns 404; no public copy found). Would settle: measured along- and across-track resolution, and likely frequency, bandwidth and squint angles
• S. Delayes, N. Burlet, Y. Le Gall and J.-P. Malkasse, "SAMDIS a multiview SAS," UDT 2022, Rotterdam, 9 June 2022 (cited by Le Gall 2023; not searched for a copy beyond the Chabah 2016/Burlet 2023 searches). Would settle: squint angles of the three views, array configurations

Sources and quotations

Simulator file: sascore/systems/thales_samdis.json

Confidence

• Northrop Grumman μSAS-SV24, iSAS / SAS (430 kHz): low. Frequency, range and the 0.25 x 1 in resolution/pixel entry are published, but bandwidth, channel count and pitch are not; the pitch rests on an effective aperture inferred from the RAS table.
• Northrop Grumman μSAS-SV24, LRRAS (600 kHz): low. Frequency, range and resolution are published; the datasheet does not say whether this mode uses the same receive array, and bandwidth and channels are unknown.
• Northrop Grumman μSAS-SV24, HRRAS (900 kHz): low. Frequency, range and resolution are published; array, bandwidth and channel count are not.

How each value was reached

• Bandwidth, Northrop Grumman μSAS-SV24, iSAS / SAS (430 kHz): Not public. If the 0.25 in cross-track entry is a -3 dB resolution, c / (2 δy) = 1500 / (2 x 0.00635) = 118 kHz before shading; if it is only the pixel size, 60 kHz (1.25 cm resolution, two pixels) fits. Plausible 60 to 120 kHz.
• Channels, Northrop Grumman μSAS-SV24, iSAS / SAS (430 kHz): Not public. 16 is a placeholder for a 0.66 m receive aperture at a few wavelengths' pitch; plausible 12 to 24.
• Channel spacing, Northrop Grumman μSAS-SV24, iSAS / SAS (430 kHz): Effective receive length from the RAS along-track resolution δx ≈ R λ / L [Northrop Grumman 2023a]: at 600 kHz (λ = 1500/600e3 = 2.5 mm) 1.5 in = 38.1 mm at 10 m gives L = 10 x 0.0025 / 0.0381 = 0.656 m; at 900 kHz (λ = 1.667 mm) 1.0 in at 10 m gives 0.656 m; the 30 m and 60 m entries give the same. c = 1500 m/s and the δx = R λ / L beamwidth convention are assumptions; the drawing labels the receiver "30 in." (0.762 m), probably the housing. Spacing = 0.656 m / 16 (assumed channels) = 41 mm.
• Bandwidth, Northrop Grumman μSAS-SV24, LRRAS (600 kHz): Not public; assumed as the 430 kHz mode (the SAS entry is the same 0.25 x 1 in).
• Channels, Northrop Grumman μSAS-SV24, LRRAS (600 kHz): Not public. 16 is a placeholder for a 0.66 m receive aperture at a few wavelengths' pitch; plausible 12 to 24.
• Channel spacing, Northrop Grumman μSAS-SV24, LRRAS (600 kHz): Effective receive length from the RAS along-track resolution δx ≈ R λ / L [Northrop Grumman 2023a]: at 600 kHz (λ = 1500/600e3 = 2.5 mm) 1.5 in = 38.1 mm at 10 m gives L = 10 x 0.0025 / 0.0381 = 0.656 m; at 900 kHz (λ = 1.667 mm) 1.0 in at 10 m gives 0.656 m; the 30 m and 60 m entries give the same. c = 1500 m/s and the δx = R λ / L beamwidth convention are assumptions; the drawing labels the receiver "30 in." (0.762 m), probably the housing. Spacing = 0.656 m / 16 (assumed channels) = 41 mm.
• Bandwidth, Northrop Grumman μSAS-SV24, HRRAS (900 kHz): Not public; assumed as the 430 kHz mode (the SAS entry is the same 0.25 x 1 in).
• Channels, Northrop Grumman μSAS-SV24, HRRAS (900 kHz): Not public. 16 is a placeholder for a 0.66 m receive aperture at a few wavelengths' pitch; plausible 12 to 24.
• Channel spacing, Northrop Grumman μSAS-SV24, HRRAS (900 kHz): Effective receive length from the RAS along-track resolution δx ≈ R λ / L [Northrop Grumman 2023a]: at 600 kHz (λ = 1500/600e3 = 2.5 mm) 1.5 in = 38.1 mm at 10 m gives L = 10 x 0.0025 / 0.0381 = 0.656 m; at 900 kHz (λ = 1.667 mm) 1.0 in at 10 m gives 0.656 m; the 30 m and 60 m entries give the same. c = 1500 m/s and the δx = R λ / L beamwidth convention are assumptions; the drawing labels the receiver "30 in." (0.762 m), probably the housing. Spacing = 0.656 m / 16 (assumed channels) = 41 mm.

Where the sources disagree

• Northrop Grumman 2023a's 430 kHz rows pair "3.0 [kts]" with "1.4 [m/s]"; 3.0 kn is 1.54 m/s (1.4 m/s is 2.7 kn). The other rows convert correctly (4.0 kn = 2.06, printed 2.0; 4.7 kn = 2.42, printed 2.4).
• Northrop Grumman 2023a rounds feet to metres loosely: 400 ft is 121.9 m (printed 120 m), 300 ft is 91.4 m (printed 90 m), 100 ft is 30.5 m (printed 30 m).
• The RAS along-track entries imply an effective receive length of about 0.656 m (λ R / δx at 10, 30 and 60 m), but the drawing labels the receiver "30 in." (0.762 m), and the 600 kHz entry at 90 m (15.0 in) does not follow the same line (13.5 in expected).
• The 0.25 x 1 in figure is called "2D Pixel Size" in the specification list and "Resolution & Pixel Size (in)" in the mode table header, so it is not clear whether it is a resolution or a pixel spacing; the current ROWS labels it a pixel only.
• Northrop Grumman 2023a lists "Operating Frequency 175-950 kHz" for the system, wider than the three default modes (430, 600, 900 kHz); no 175 kHz mode is described.

Where the simulator's file differs from the sources

• modes[430 kHz].platform speed: 1.4 m/s is the datasheet's printed value, but the same Northrop Grumman 2023a row says 3.0 kn (1.54 m/s).
• modes[600 kHz] and modes[900 kHz]: Northrop Grumman 2023a calls these "Real Aperture SONAR or SAS" modes; the simulator models them as SAS on the 430 kHz array, which Northrop Grumman 2023a does not confirm.

Not yet obtained

• Any Northrop Grumman publication giving the μSAS-SV24 bandwidth, channel count or receive-array active length (none found). Would settle: Bandwidth, channels and element spacing for all three modes (all currently assumed).

Sources and quotations

Simulator file: sascore/systems/northrop_grumman_usas_sv24.json

Confidence

• Exail A18-M (UMISAS 120), HF row: high. Every value is stated in Exail's own paper and repeated in Exail's slides; the long range is explicitly the HF row's (100 m to 220 m at 22 m altitude).
• Exail A18-M (UMISAS 120), VHF row: medium. Band, array and resolution are published, but no maximum range is given for the VHF row; Pinto 2023a reports high VHF coherence only from about 40 m to more than 120 m (the 220/180 m figures belong to the HF row).
• Exail T18-M (UMISAS 240), HF row: medium. Range and resolution are published for the T18-M, but nothing public gives the UMISAS 240's bands, element count or pitch; those are carried over from the UMISAS 120.
• Exail T18-M (UMISAS 240), VHF row: low. Only the system-level T18-M figures are public; the VHF row's band, array and range are all carried over or unknown (Exail 2021's > 180 m at 8 kn is a system figure, and on the A18-M the VHF row is coherent only to about 120 m).

How each value was reached

• Center frequency, Exail T18-M (UMISAS 240), HF row: Not stated for the UMISAS 240. Pinto 2023a describes the UMISAS dual-row HF 250 / VHF 330 kHz design for a sonar "designed for the A18-M ... and T18-M"; the T18-M datasheet gives only "ISAS frequency: 300kHz" [Exail 2021], read here as a nominal label.
• Bandwidth, Exail T18-M (UMISAS 240), HF row: Not stated for the UMISAS 240; assumed equal to the UMISAS 120's 60 kHz per row [Pinto 2023a; Pinto 2023b].
• Channels, Exail T18-M (UMISAS 240), HF row: Not published. Assumed twice the UMISAS 120's 32 [Pinto 2023a] from the 120/240 naming [Pinto 2023a; Pinto 2023b]. Consistency check: the sampling limit V T ≤ L/2 with R = c T / 2 gives L ≥ 4 V R / c = 4 x 4.12 m/s x 180 m / 1500 m/s = 1.98 m for "> 180 m @ 8 kts" [Exail 2021], i.e. at least 58 elements at 34.3 mm; 64 x 34.3 mm = 2.20 m.
• Channel spacing, Exail T18-M (UMISAS 240), HF row: Not published for the UMISAS 240; assumed equal to the UMISAS 120 element pitch d = 34.3 mm [Pinto 2023a; Pinto 2023b].
• Center frequency, Exail T18-M (UMISAS 240), VHF row: Not stated for the UMISAS 240; assumed from the UMISAS dual-row design [Pinto 2023a; Pinto 2023b]. Exail 2021 gives only a nominal 300 kHz.
• Bandwidth, Exail T18-M (UMISAS 240), VHF row: Not stated for the UMISAS 240; assumed equal to the UMISAS 120's 60 kHz per row [Pinto 2023a; Pinto 2023b].
• Channels, Exail T18-M (UMISAS 240), VHF row: Not published. Assumed twice the UMISAS 120's 32 [Pinto 2023a] from the 120/240 naming [Pinto 2023a; Pinto 2023b]. Consistency check: the sampling limit V T ≤ L/2 with R = c T / 2 gives L ≥ 4 V R / c = 4 x 4.12 m/s x 180 m / 1500 m/s = 1.98 m for "> 180 m @ 8 kts" [Exail 2021], i.e. at least 58 elements at 34.3 mm; 64 x 34.3 mm = 2.20 m.
• Channel spacing, Exail T18-M (UMISAS 240), VHF row: Not published for the UMISAS 240; assumed equal to the UMISAS 120 element pitch d = 34.3 mm [Pinto 2023a; Pinto 2023b].

Where the sources disagree

• Source level: Pinto 2023a gives "219dB at HF"; Exail's CNE 2023 slides (Pinto 2023b) give "220dB HF source level".
• Centre frequency: Exail 2021 (T18-M datasheet) gives a single "ISAS frequency: 300kHz"; Pinto 2023a and Pinto 2023b give dual rows at 250 kHz (HF) and 330 kHz (VHF).
• Array length: Pinto 2023a describes the UMISAS 120 assembly as "approximately 8” tall and 42” long" (1.07 m), while 32 x 34.3 mm = 1097.6 mm (43.2 in), as Pinto 2023b states ("L = 32*d = 1097.6 mm").
• Across-track resolution: Pinto 2023a and Pinto 2023b give 1.5 cm (range) for the UMISAS 120; Exail 2021 and Exail n.d. give 3 cm x 3 cm for the T18-M (UMISAS 240) with the same 60 kHz bands, so the T18-M figure is probably a rounded or conservative specification.
• Max range per row: Pinto 2023a Table 1 gives 220 m at 3 kn and 180 m at 4 kn as UMISAS figures, but Sec. III-A attributes 100 to 220 m to the HF row and about 40 to 120 m (high coherence) to the VHF row; the current ROWS entry gives the same range to both rows.
• Pinto 2023a was read as the authors' conference draft ("This is a DRAFT. As such it may not be cited in other works."); the citable version is on IEEE Xplore (paywalled) and was not checked against the draft.

Where the simulator's file differs from the sources

• modes[T18-M UMISAS 240, VHF 330 kHz].image far range: keeps Exail 2021's 180 m system figure as the scene default while max_range_m is null; probably optimistic for the VHF row (Pinto 2023a).
• modes[A18-M UMISAS 120, VHF 330 kHz].maximum range: 120 m is Pinto 2023a's VHF coherent range, a lower bound ("more than 120m"), not a stated maximum range.
• modes[A18-M UMISAS 120, HF 250 kHz].source level db: 219 dB from Pinto 2023a; Pinto 2023b says 220 dB.

Not yet obtained

• An Exail specification of the UMISAS 240 array (element count, length, bands), e.g. a UMISAS or T18-M technical datasheet newer than V0002. Would settle: T18-M channels, spacing, centre frequencies and bandwidth (all currently assumed from the UMISAS 120), and per-row maximum range.
• The IEEE Xplore version of Pinto et al., OCEANS 2023 (paywalled). Would settle: Confirms that the draft's numbers (Table 1, 219/211 dB, d = 34.3 mm) are unchanged in the citable version.

Sources and quotations

Simulator file: sascore/systems/exail_umisas.json

Confidence

• Sonardyne / Wavefront Solstice S3000L (-300 / -600): medium. Beam, channel count, array length and range are published by the maker; centre frequency and bandwidth come from the band edges and the beam's one-/two-way convention is unstated.
• Sonardyne / Wavefront Solstice S4000-300: medium. Only the brochure's model table covers the S4000; frequency and bandwidth are band-edge derivations as for the S3000L.

How each value was reached

• Center frequency, Sonardyne / Wavefront Solstice S3000L (-300 / -600): (725 + 775) / 2 = 750 kHz, centre of the published 725-775 kHz band [Wavefront 2021]
• Bandwidth, Sonardyne / Wavefront Solstice S3000L (-300 / -600): 775 - 725 = 50 kHz, width of the published band [Wavefront 2021]; assumes the pulse fills the band (pulse bandwidth itself is not published)
• Channels, Sonardyne / Wavefront Solstice S3000L (-300 / -600): per side; Wavefront 2021 does not label the +4, its bathymetry comes from 'a vertical hydrophone array on each flank' []
• Channel spacing, Sonardyne / Wavefront Solstice S3000L (-300 / -600): 682 mm hydrophone array / 32 channels = 21.3 mm; assumes a uniform pitch
• Along-track resolution, Sonardyne / Wavefront Solstice S3000L (-300 / -600): 100 m x tan(0.15°) = 0.262 m. One-way vs two-way not stated; 0.886 x 2.0 mm / 0.682 m = 0.149° matches a one-way beam of the full receive array
• Across-track resolution, Sonardyne / Wavefront Solstice S3000L (-300 / -600): c / 2B = 1500 / (2 x 50 kHz) = 1.5 cm with c = 1500 m/s; Wavefront n.d. gives a 3 cm x 3 cm pixel [], which is a grid size, not a resolution
• Maximum range, Sonardyne / Wavefront Solstice S3000L (-300 / -600): 100 m per side at up to 6 kn [Wavefront n.d.]; the simulator mode flies at 3.09 m/s = 6 kn
• Center frequency, Sonardyne / Wavefront Solstice S4000-300: (725 + 775) / 2 = 750 kHz [Wavefront n.d.]
• Bandwidth, Sonardyne / Wavefront Solstice S4000-300: 775 - 725 = 50 kHz [Wavefront n.d.]; assumes the pulse fills the band
• Channels, Sonardyne / Wavefront Solstice S4000-300: per side
• Channel spacing, Sonardyne / Wavefront Solstice S4000-300: 1344 mm hydrophone array / 64 channels = 21.0 mm; assumes a uniform pitch
• Along-track resolution, Sonardyne / Wavefront Solstice S4000-300: 100 m x tan(0.075°) = 0.131 m. One-way vs two-way not stated; 0.886 x 2.0 mm / 1.344 m = 0.076° matches a one-way full-array beam
• Across-track resolution, Sonardyne / Wavefront Solstice S4000-300: c / 2B = 1500 / (2 x 50 kHz) = 1.5 cm
• Maximum range, Sonardyne / Wavefront Solstice S4000-300: 100 m per side at up to 6 kn [Wavefront n.d.]; the simulator mode flies at 3.09 m/s = 6 kn

Where the sources disagree

• Wavefront 2021 calls 0.15° both the "along-track resolution" and the "Azimuth Beam Width"; Wavefront n.d. gives 0.075° for the S4000. No source says one-way or two-way; 0.886 lambda / L for the 682 mm and 1344 mm arrays at 750 kHz gives 0.149° and 0.076°, so the figures match one-way full-array beams.
• Power differs by source: Wavefront 2021 18 W, Tena 2017 21 W including data acquisition, Wavefront n.d. table 26 W with on-board real-time processing, and Wavefront n.d.'s customer quotation about 30 W. Not a table column.
• Wavefront n.d. says Solstice focuses on a 3 cm x 3 cm pixel; the 1.5 cm across-track value is c / 2B from the 50 kHz band. The pulse bandwidth is not published, so 1.5 cm is a best case.
• Wavefront 2021 and Wavefront n.d. give the receiver channels as 2 x (32 +4) without saying what the +4 are; the current ROWS label "32 (+4 bathymetry)" infers it from Wavefront 2021's vertical bathymetry array on each flank.
• The current ROWS have no S4000 row, though Wavefront n.d. publishes it (64 +4 channels, 1344 mm array, 0.075° beam) and the catalog has it as a mode.
• Wavefront n.d. lists S3000L-300 and S3000L-600 with identical acoustics (only depth rating differs); the ROWS label "S3000L" covers both.
• Wavefront n.d. describes Multi-Ping Multi-Look processing combining several pings; no source says whether this gives coherent along-track gain beyond the real-aperture beam.

Where the simulator's file differs from the sources

• modes[S3000L].channel count source says "(4 for bathymetry)": Wavefront 2021 does not label the +4 channels.
• modes[*].bandwidth tagged derived as the width of the 725-775 kHz band: the sources give the frequency band, not the transmitted pulse bandwidth, so 50 kHz is an upper bound.
• modes[S4000] has no azimuth_beam_deg fact although Wavefront n.d. publishes 0.075°.
• sources.Tena 2017 citation names no author; the page text is signed by Ioseba Tena (metadata says Alice Lee).

Not yet obtained

• Any Wavefront/Sonardyne technical paper or manual for Solstice giving the transmit pulse bandwidth and the beamwidth convention. Would settle: b_khz, across_res, whether the 0.15°/0.075° beams are one-way, and whether MPML adds along-track gain

Sources and quotations

Simulator file: sascore/systems/sonardyne_solstice.json

Confidence

• SAS12 (12.75 in AUV): high. Every value is stated in the developers' own OCEANS 2006 paper (public DTIC copy), and the range, advance per ping and speed are mutually consistent.

How each value was reached

• Bandwidth: Matthews 2006 gives the transducer bandwidth as 'a minimum of 30 kHz'; c/(2 x 2.5 cm) = 30 kHz matches the stated 2.5 cm range resolution []
• Channels: 16-channel receiver on each side
• Channel spacing: d = 2 x dx = 2 x 25 mm = 50 mm, since 'the individual element aperture determines the ultimate resolution limit' and that limit is 25 mm []; cross-check: the nominal 35.5 cm advance per ping [] with 2 overlapping phase centres [] gives a phase-centre spacing of 35.5 / (16 - 2) = 2.54 cm, i.e. a 50.7 mm element pitch
• Maximum range: consistency: 2 x 150 m / 1500 m/s = 0.2 s ping interval; 35.5 cm per ping / 0.2 s = 1.78 m/s = 3.45 kn, inside the stated 3 to 4 kn

Where the sources disagree

• Bandwidth: Matthews 2006 states the transducers are designed for 'a minimum of 30 kHz of bandwidth' and that waveforms vary (CW, LFM, arbitrary); it does not state a single transmitted bandwidth. 30 kHz is consistent with the stated 2.5 cm range resolution.
• Speed: the 150 m range is stated for 3 to 4 knots (1.5 to 2.1 m/s), while the Bluefin12 vehicle cruises at 1.5 m/s (about 2.9 kn); Matthews 2006 says range falls as speed rises (constant area coverage rate).

Where the simulator's file differs from the sources

• maker 'Bluefin Robotics / Penn State ARL (AOFNC program)' omits NSWC Panama City, which developed the synthetic aperture image processing []; platform is named Bluefin12 / 'AOFNC 12 Vehicle' in the paper.

Sources and quotations

Simulator file: sascore/systems/arl_psu_sas12.json

Confidence

• MASTODON sample SAS: high. The signal and array values are read directly from the MASTODON system file; only the maximum range is an estimate.

How each value was reached

• Channel spacing: the file sets no elem_spacing_hor, and MASTODON then uses the element width as the spacing [NSWC PCD n.d.b]; receiver elem_width = 2.1 cm [NSWC PCD n.d.a]
• Along-track resolution: dx = d/2 = 2.1 cm / 2 = 1.05 cm (receiver element); the 2 cm projector element gives 1.0 cm
• Across-track resolution: dr = c / (2B) = 1500 / (2 x 50 kHz) = 1.5 cm; c = 1500 m/s assumed (the system file sets no sound speed)
• Maximum range: The MASTODON file gives no range, ping interval or speed, so no maximum range is published and the simulator holds none. The scene's far range, 60 m, is 10 x the 6 m altitude []: the project's 10 percent altitude rule, a convention.

Where the sources disagree

• The projector element is 2 cm wide and the receiver elements 2.1 cm [NSWC PCD n.d.a]; the catalog sets both to 2.1 cm (a documented deliberate departure).
• MASTODON's sas_overlap = 0 moves half the array length each ping [NSWC PCD n.d.c], i.e. no shared phase centres; the catalog replaces it with 2 shared phase centres (documented departure).

Where the simulator's file differs from the sources

• sources.NSWC PCD n.d.a url points to github.com/Sonar-Sim/MASTODON; the release's CITATION.cff gives repository-code https://bitbucket.di2e.net/projects/MASTODON. The GitHub URL was not verified here.

Sources and quotations

Simulator file: sascore/systems/mastodon_sample_sas.json

Confidence

• Klein System 5900 (600 kHz): medium. Frequency, beam, resolutions and range are in the maker's Rev 09/23 datasheet; the channel count and pitch the simulator needs are estimates.

How each value was reached

• Bandwidth: c / (2 x 3.75 cm) = 1500 / 0.075 = 20 kHz with c = 1500 m/s, from the published 3.75 cm across-track resolution
• Channels: Not published. Klein 2023 gives up to 20 beams per side from a 1.8 m array and says only that the 5900 has more than twice its predecessors' channels. A focused array of N elements forms at most about N independent beams within one element's beam, so 20 is the smallest channel count that supports 20 beams; the simulator uses 20.
• Channel spacing: Not published. 1.8 m array / 20 assumed channels = 90.0 mm. Check: each 90 mm element has a 0.886 x 2.5 mm / 0.09 m = 1.41° beam, which holds 20 of the published 0.07° array beams.
• Along-track resolution: 100 m x tan(0.07°) = 0.122 m. Klein 2023 publishes 6.2 cm at the 50 m range setting (constant), 9.3 cm at 75 m, 15.5 cm at 125 m [0.07° beamwidth]; one-way vs two-way not stated, and 0.886 x 2.5 mm / 1.8 m = 0.070° matches a one-way full-array beam

Where the sources disagree

• Current ROWS give 28 channels and 65 mm spacing; Klein 2023 states neither, only "more than twice the number of acoustic channels available on its predecessors" and up to 20 beams per side.
• Current ROWS give bandwidth 40 kHz (assumed); the published 3.75 cm across-track resolution implies 20 kHz (c / 2 delta_r).
• Current ROWS print along-track, across-track and maximum range as "not in catalog"; Klein 2023 publishes 0.07° (6.2 cm at 50 m), 3.75 cm and 150 m per side.
• Klein 2023 gives the array as "182cm" in the text and "1.8 m" in the specification table.
• Klein 2023 calls the 5900 a multi-beam side scan with dynamic focus and says Klein avoids "a full SAS implementation"; the catalog and ROWS model it as a SAS.

Where the simulator's file differs from the sources

• channel count = 28 tagged published: not in Klein 2023.
• channel spacing = 0.065 m derived from 28 channels: the channel count is not public, so the pitch is unsupported.
• modes[600 kHz].bandwidth = 40 kHz (assumed): Klein 2023's 3.75 cm across-track resolution implies 20 kHz.
• type = "sas" and summary "modelled as a 28-channel array per side": Klein 2023 describes a dynamically focused multi-beam side scan, not a full SAS, and gives no 28.
• sources.Klein 2023 has no URL and says "not re-read"; the Rev 09/23 datasheet is now read (URL above).
• operation (max_range_m 150 m) and along/across-track facts are missing although Klein 2023 publishes them; modes[600 kHz].pulse length is "not_public", which is right, but the pulse type "CW, FM-Chirp" is published.

Not yet obtained

• Klein System 5900 user or technical manual (or a Klein paper on the 5900 array). Would settle: channels per side and element pitch (needed to model it as an array), pulse bandwidth

Sources and quotations

Simulator file: sascore/systems/klein_5900.json

Confidence

• EdgeTech 4125i, 400 kHz (400/900 kHz towfish): medium. Beamwidth, footprint and across-track resolution are published, but EdgeTech's own brochure (200 m) and manual (150 m) disagree on range and no source says whether the beamwidth is one-way or two-way.
• EdgeTech 4125i, 900 kHz (400/900 kHz towfish): high. Every printed value except the bandwidth is stated identically in both EdgeTech documents; only the bandwidth is derived.
• EdgeTech 4125i, 600 kHz (600/1600 kHz towfish): high. Beamwidth, footprint, across-track resolution and range are published and agree across both EdgeTech documents; only the bandwidth is derived.
• EdgeTech 4125i, 1600 kHz (600/1600 kHz towfish): high. Every printed value except the bandwidth is stated identically in both EdgeTech documents; only the bandwidth is derived.

How each value was reached

• Bandwidth, EdgeTech 4125i, 400 kHz (400/900 kHz towfish): B = c / (2 dr) with c = 1500 m/s (assumed) and dr = 2.3 cm published across-track resolution; assumes the resolution is the unweighted c/2B (no window factor). No chirp bandwidth is published.
• Channels, EdgeTech 4125i, 400 kHz (400/900 kHz towfish): Neither EdgeTech document gives an element or channel count; EdgeTech 2025b says only that the towfish mounts port and starboard transducers. One receive channel per side per frequency is the conventional real-aperture reading.
• Channel spacing, EdgeTech 4125i, 400 kHz (400/900 kHz towfish): Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / fc and the published horizontal beamwidth read as one-way (the sources do not say).
• Along-track resolution, EdgeTech 4125i, 400 kHz (400/900 kHz towfish): Beamwidth published (one-way or two-way not stated); EdgeTech 2025b publishes the footprint 80 cm @ 100 m, equal to 100 m x 0.46° in radians (0.803 m).
• Bandwidth, EdgeTech 4125i, 900 kHz (400/900 kHz towfish): B = c / (2 dr) with c = 1500 m/s (assumed) and dr = 1.0 cm published across-track resolution; assumes the resolution is the unweighted c/2B (no window factor). No chirp bandwidth is published.
• Channels, EdgeTech 4125i, 900 kHz (400/900 kHz towfish): Neither EdgeTech document gives an element or channel count; EdgeTech 2025b says only that the towfish mounts port and starboard transducers. One receive channel per side per frequency is the conventional real-aperture reading.
• Channel spacing, EdgeTech 4125i, 900 kHz (400/900 kHz towfish): Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / fc and the published horizontal beamwidth read as one-way (the sources do not say).
• Along-track resolution, EdgeTech 4125i, 900 kHz (400/900 kHz towfish): 100 m x 0.28° (0.004887 rad) = 0.489 m; EdgeTech 2025b publishes 24 cm @ 50 m for 900 kHz, which scales to the same 0.48 to 0.49 m.
• Bandwidth, EdgeTech 4125i, 600 kHz (600/1600 kHz towfish): B = c / (2 dr) with c = 1500 m/s (assumed) and dr = 1.5 cm published across-track resolution; assumes the resolution is the unweighted c/2B (no window factor). No chirp bandwidth is published.
• Channels, EdgeTech 4125i, 600 kHz (600/1600 kHz towfish): Neither EdgeTech document gives an element or channel count; EdgeTech 2025b says only that the towfish mounts port and starboard transducers. One receive channel per side per frequency is the conventional real-aperture reading.
• Channel spacing, EdgeTech 4125i, 600 kHz (600/1600 kHz towfish): Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / fc and the published horizontal beamwidth read as one-way (the sources do not say).
• Along-track resolution, EdgeTech 4125i, 600 kHz (600/1600 kHz towfish): Beamwidth published; EdgeTech 2025b publishes the footprint 58 cm @ 100 m, equal to 100 m x 0.33° in radians (0.576 m).
• Bandwidth, EdgeTech 4125i, 1600 kHz (600/1600 kHz towfish): B = c / (2 dr) with c = 1500 m/s (assumed) and dr = 0.6 cm published across-track resolution; assumes the resolution is the unweighted c/2B (no window factor). No chirp bandwidth is published.
• Channels, EdgeTech 4125i, 1600 kHz (600/1600 kHz towfish): Neither EdgeTech document gives an element or channel count; EdgeTech 2025b says only that the towfish mounts port and starboard transducers. One receive channel per side per frequency is the conventional real-aperture reading.
• Channel spacing, EdgeTech 4125i, 1600 kHz (600/1600 kHz towfish): Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / fc and the published horizontal beamwidth read as one-way (the sources do not say).
• Along-track resolution, EdgeTech 4125i, 1600 kHz (600/1600 kHz towfish): 100 m x 0.20° (0.003491 rad) = 0.349 m; EdgeTech 2025b publishes 9 cm @ 25 m for 1600 kHz, which scales to 0.36 m.

Where the sources disagree

• Maximum range at 400 kHz: the brochure [EdgeTech 2025a, 10-22-25] says 200 m ("200m @ 400 kHz"), while EdgeTech's user hardware manual [EdgeTech 2025b, 0004823 Rev O, 15 Aug 2025] says 150 m ("150 m (400 kHz)"). The other three ranges (75, 120, 35 m) agree.
• One-way versus two-way beamwidth: neither 4125i document labels its horizontal beamwidths [EdgeTech 2025a; EdgeTech 2025b]. EdgeTech's 6205s2 manual labels its along-track beamwidths "(2-way)" [EdgeTech 2024a], and a 2205 paper calls its 0.26° two-way [Sethuraman 2024]. The 4125i manual's along-track resolutions equal range x full beamwidth at every listed point (80 cm and 58 cm @ 100 m, 24 cm @ 50 m, 9 cm @ 25 m) [EdgeTech 2025b], so the table's footprint is unaffected; only the catalog's derived element lengths depend on the reading (two-way gives about 0.72 x).
• EdgeTech 2025b captions its 4125i towfish table "Table 2-2: 4205i Towfish Specifications", a typo for 4125i; the section heading is "2.2 4125i Towfish Specifications".
• EdgeTech 2025b gives towfish length 114.8 cm and weight 20.4 kg; EdgeTech 2025a gives 112 cm and 20 kg (not used in the table).

Where the simulator's file differs from the sources

• sources.EdgeTech 2025a: citation "EdgeTech 4125i brochure" with empty url. The file is EdgeTech's "4125i-Brochure-10-22-25.pdf" at https://www.edgetech.com/wp-content/uploads/2023/04/4125i-Brochure-10-22-25.pdf (byte-identical); dated 10-22-25 by file name and PDF creation date, with no part number or revision printed.
• channel count: source "real-aperture side-scan: one element per side [EdgeTech 2025a]". EdgeTech 2025a does not state an element or channel count; the value is an assumption.
• modes[400 kHz].maximum range = 200 m: supported by EdgeTech 2025a but contradicted by EdgeTech's manual EdgeTech 2025b (150 m). The catalog does not list EdgeTech 2025b.
• review.notes: the open one-way/two-way question can now cite EdgeTech 2025b, which publishes along-track resolutions equal to range x beamwidth for all four frequencies (it still does not label the beamwidth one-way or two-way).
• review.notes: "the EdgeTech 2300 sheet's along-track resolutions equal range x beamwidth [EdgeTech 2023]" holds at 120, 410 and 540 kHz but not at 230 kHz (1.4 m @ 150 m vs 1.31 m) or 850 kHz (0.15 m @ 50 m vs 0.175 m).
• modes[*].bandwidth (tag derived): the derivation silently assumes c = 1500 m/s and dr = c/2B with no window factor; the source note should say so.

Not yet obtained

• An EdgeTech statement (engineering note, test report, or a later 4125i manual revision) of whether the 4125i horizontal beamwidths are one-way or two-way -3 dB widths. Would settle: the catalog's derived element lengths (tx/rx_element_azimuth) for all four modes; not needed for the printed table cells
• EdgeTech, Discover 4125i Software Manual (named in EdgeTech 2025b Section 1.2.2), or any EdgeTech pulse table for the 4125i. Would settle: chirp bandwidth and pulse length per frequency (b_khz, currently derived from across-track resolution; pulse_length not public)

Sources and quotations

Simulator file: sascore/systems/edgetech_4125i.json

Confidence

• Klein D3500TF 100 / 400 kHz: high. All values are in the maker's current datasheet; only the bandwidth and the 100 m footprint are derived.

How each value was reached

• Bandwidth: c / (2 delta_r): 1500 / (2 x 9.6 cm) = 7.81 kHz, 1500 / (2 x 2.4 cm) = 31.25 kHz, c = 1500 m/s
• Channels: "Single Beam" per side
• Channel spacing: Single beam per side, so the simulator's spacing is the element length L = 0.886 lambda / theta with lambda = 1500 / fc: 0.886 x 15 mm / 0.7° = 1087.8 mm; 0.886 x 3.75 mm / 0.3° = 634.6 mm. Klein 2024 does not say one-way or two-way; read as one-way.
• Along-track resolution: 100 m x tan(0.7°) = 1.22 m; 100 m x tan(0.3°) = 0.52 m. Klein 2024 does not say one-way or two-way

Where the sources disagree

• Klein 2024 does not say whether the 0.7° / 0.3° horizontal beamwidths are one-way or two-way; the catalog reads them as one-way.

Sources and quotations

Simulator file: sascore/systems/klein_d3500tf.json

Confidence

• Klein UUV-3500 455 / 900 kHz: medium. The 2014 datasheet is a third-party copy (matched by NERC's 2024 entry) and gives one beamwidth and one resolution for both frequencies.

How each value was reached

• Bandwidth: c / (2 x 2.4 cm) = 31.25 kHz, c = 1500 m/s; L-3 Klein 2014 gives one 2.4 cm figure for both frequencies
• Channels: L-3 Klein 2014 lists one port and one starboard transducer but never states the number of receive channels per side
• Channel spacing: One element per side (assumed), so the simulator's spacing is the element length L = 0.886 lambda / 0.34°: 0.886 x 3.30 mm / 0.34° = 492.2 mm at 455 kHz; 0.886 x 1.67 mm / 0.34° = 248.8 mm at 900 kHz, which assumes the single 0.34° figure applies at 900 kHz. One-way vs two-way not stated; read as one-way.
• Along-track resolution: 100 m x tan(0.34°) = 0.59 m. L-3 Klein 2014 gives one horizontal beamwidth with no frequency and no one-/two-way convention; it also lists "Single-frequency operation to maximize along track resolution" [u6]
• Maximum range: "Range (typical)"

Where the sources disagree

• L-3 Klein 2014 and NERC 2024 give a single 0.34° horizontal beamwidth and a single 2.4 cm across-track resolution for both 455 and 900 kHz; with one transducer the 900 kHz beam would be about half as wide, and L-3 Klein 2014's "Single-frequency operation to maximize along track resolution" suggests the along-track figure depends on the operating mode. Neither source says one-way or two-way.
• The catalog derives a 0.49 m element from 0.34° at 455 kHz; L-3 Klein 2014 gives the transducer housing as 22 in (0.559 m) long, so the active length is not published.

Where the simulator's file differs from the sources

• channel count = 1 tagged published ("single-beam side-scan [L-3 Klein 2014]"): L-3 Klein 2014 never says single beam; it lists one port and one starboard transducer.
• modes[900 kHz].transmit element length = 0.2488 m derived by applying 0.34° at 900 kHz: L-3 Klein 2014 does not assign 0.34° to 900 kHz.

Not yet obtained

• Current Klein Marine Systems UUV-3500 datasheet (klein.com) or user manual. Would settle: per-frequency horizontal beamwidth and its one-/two-way convention, per-frequency across-track resolution, and confirmation of the third-party L-3 Klein 2014 copy

Sources and quotations

Simulator file: sascore/systems/klein_uuv3500.json

Confidence

• EdgeTech 2205, 75 kHz: low. Range is published (1000 m datasheet, 800 m manual estimate); beamwidth is extrapolated from another product and bandwidth is unpublished.
• EdgeTech 2205, 120 kHz (125 kHz ping): medium. Range and bandwidth are published for a 2205; the beamwidth comes from the 2300 sheet, which EdgeTech 2000 shows uses the same arrays, but its one-way or two-way basis is unstated.
• EdgeTech 2205, 230 kHz: medium. Bandwidth and range are published for a 2205 and the 2300-sheet resolution matches the widest pulse; the beamwidth is borrowed from the 2300 sheet.
• EdgeTech 2205, 410 kHz: medium. Bandwidth and range are published for a 2205; the beamwidth is borrowed from the 2300 sheet.
• EdgeTech 2205, 540 kHz: medium. Bandwidth, range and a two-way beamwidth are published for 2205 systems, but the 6205s2 manual lists a wider beam for its 2205-based array and the centre frequency is given as 520 or 540 kHz.
• EdgeTech 2205, 850 kHz: medium. Bandwidth and range are published for a 2205; the beamwidth still depends on whether the 2300 or the 6205s2 array is the better match.
• EdgeTech 2205, 1600 kHz (1610 kHz): medium. Range is published for the 2205 and three sister products agree on 0.20° and 6 mm, but no 2205 pulse table covers 1600 kHz.

How each value was reached

• Bandwidth, EdgeTech 2205, 75 kHz: Not published: EdgeTech 2000 has no 75 kHz pulse table. Estimate: the 120 kHz max-range ping's fractional bandwidth (8 kHz at 125 kHz, ping 2300_DDPA_SS125Q_8BW_16MS) scaled to 75 kHz, 8 x 75 / 125 = 4.8 kHz.
• Channels, EdgeTech 2205, 75 kHz: EdgeTech 2000 documents one transmit and receive section per frequency per side for its 120/410/850 and 230/540/850 kHz arrays; this frequency is not among them, so the same layout is assumed.
• Channel spacing, EdgeTech 2205, 75 kHz: Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / 75 kHz; the beamwidth 1.088° is read as one-way (the source does not say).
• Along-track resolution, EdgeTech 2205, 75 kHz: Nothing published at 75 kHz. The catalog assumes the 2300's 120 kHz aperture [EdgeTech 2023, 0.68°] at 75 kHz: 0.68° x 120/75 = 1.088°; 100 m x 1.088° in radians = 1.90 m.
• Across-track resolution, EdgeTech 2205, 75 kHz: No source gives a 75 kHz bandwidth or across-track resolution.
• Maximum range, EdgeTech 2205, 75 kHz: Datasheets EdgeTech 2021 and EdgeTech 2024b give the ranges; EdgeTech 2000 Table 9-1 gives an estimate for the longest pulse with altitude at least 10% of range. EdgeTech 2000 uses a 50.0 ms pulse for this range.
• Center frequency, EdgeTech 2205, 120 kHz (125 kHz ping): Sold as 120 kHz [EdgeTech 2021; EdgeTech 2024b]; every 120 kHz (SSL) pulse in EdgeTech 2000 Table 9-2 has a 125 kHz ping frequency.
• Bandwidth, EdgeTech 2205, 120 kHz (125 kHz ping): Published span 8 to 14 kHz over the pings EdgeTech 2000 lists for this channel; the simulator's default is the ping EdgeTech 2000 Table 9-1 pairs with the maximum range, so bandwidth, pulse length and range come from one EdgeTech pairing. Shorter pings give the wider bandwidths. The 16 ms ping is named 2300_DDPA_SS125Q_8BW_16MS but its BW column prints 16. The name is used: every other row's name matches its BW column, the printed 16 equals that row's pulse width, and 8 continues the table's fall from 14 to 10 kHz as pulses lengthen. No other source confirms it; if the column is right the span is 10 to 16 kHz.
• Channels, EdgeTech 2205, 120 kHz (125 kHz ping): EdgeTech 2000 (a 2205 fitted to a 2300 vehicle) states one transmit and receive section per frequency on each port and starboard array, each sensor operating independently. The AUV datasheets also advertise dynamically focused and dynamic-aperture arrays [EdgeTech 2021], which would be segmented; EdgeTech 2000 lists only non-DF pulses for 850 kHz.
• Channel spacing, EdgeTech 2205, 120 kHz (125 kHz ping): Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / 125 kHz; the beamwidth 0.68° is read as one-way (the source does not say).
• Along-track resolution, EdgeTech 2205, 120 kHz (125 kHz ping): 2300 sheet beamwidth [EdgeTech 2023]; EdgeTech 2000 shows a 2205 driving EdgeTech 2300 side-scan arrays (drawing 0018633) in a 120/410/850 or 230/540/850 kHz set, so the 2300 figure applies to that configuration. No source states one-way or two-way. 100 m x 0.68° in radians = 1.19 m.
• Across-track resolution, EdgeTech 2205, 120 kHz (125 kHz ping): dr = c / (2B), c = 1500 m/s assumed: 8 kHz (simulated ping 2300_DDPA_SS125Q_8BW_16MS) gives 9.38 cm; the widest published ping, 14 kHz, gives 5.4 cm.
• Maximum range, EdgeTech 2205, 120 kHz (125 kHz ping): Datasheets EdgeTech 2021 and EdgeTech 2024b give the ranges; EdgeTech 2000 Table 9-1 gives an estimate for the longest pulse with altitude at least 10% of range. EdgeTech 2000 uses a 16.0 ms pulse.
• Bandwidth, EdgeTech 2205, 230 kHz: Published span 20 to 25 kHz over the pings EdgeTech 2000 lists for this channel; the simulator's default is the ping EdgeTech 2000 Table 9-1 pairs with the maximum range, so bandwidth, pulse length and range come from one EdgeTech pairing. Shorter pings give the wider bandwidths.
• Channels, EdgeTech 2205, 230 kHz: EdgeTech 2000 (a 2205 fitted to a 2300 vehicle) states one transmit and receive section per frequency on each port and starboard array, each sensor operating independently. The AUV datasheets also advertise dynamically focused and dynamic-aperture arrays [EdgeTech 2021], which would be segmented; EdgeTech 2000 lists only non-DF pulses for 850 kHz.
• Channel spacing, EdgeTech 2205, 230 kHz: Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / 230 kHz; the beamwidth 0.5° is read as one-way (the source does not say).
• Along-track resolution, EdgeTech 2205, 230 kHz: 2300 sheet beamwidth [EdgeTech 2023], applicable to a 2205 on 2300 arrays [EdgeTech 2000]. 100 m x 0.5° in radians = 0.873 m; the 2300's own 1.4 m @ 150 m implies 0.53°.
• Across-track resolution, EdgeTech 2205, 230 kHz: dr = c / (2B), c = 1500 m/s assumed: 20 kHz (simulated ping 2300DDPA_SS230Q_20BW_8_0MS) gives 3.75 cm; the widest published ping, 25 kHz, gives 3.0 cm.
• Maximum range, EdgeTech 2205, 230 kHz: Datasheets EdgeTech 2021 and EdgeTech 2024b give the ranges; EdgeTech 2000 Table 9-1 gives an estimate for the longest pulse with altitude at least 10% of range. EdgeTech 2000 uses an 8.0 ms pulse.
• Bandwidth, EdgeTech 2205, 410 kHz: Published span 25 to 60 kHz over the pings EdgeTech 2000 lists for this channel; the simulator's default is the ping EdgeTech 2000 Table 9-1 pairs with the maximum range, so bandwidth, pulse length and range come from one EdgeTech pairing. Shorter pings give the wider bandwidths.
• Channels, EdgeTech 2205, 410 kHz: EdgeTech 2000 (a 2205 fitted to a 2300 vehicle) states one transmit and receive section per frequency on each port and starboard array, each sensor operating independently. The AUV datasheets also advertise dynamically focused and dynamic-aperture arrays [EdgeTech 2021], which would be segmented; EdgeTech 2000 lists only non-DF pulses for 850 kHz.
• Channel spacing, EdgeTech 2205, 410 kHz: Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / 410 kHz; the beamwidth 0.3° is read as one-way (the source does not say).
• Along-track resolution, EdgeTech 2205, 410 kHz: 2300 sheet beamwidth [EdgeTech 2023], applicable to a 2205 on 2300 arrays [EdgeTech 2000]. 100 m x 0.3° in radians = 0.524 m, matching the 2300's 0.5 m @ 100 m. The array drawing calls this channel 400 kHz.
• Across-track resolution, EdgeTech 2205, 410 kHz: dr = c / (2B), c = 1500 m/s assumed: 25 kHz (simulated ping 2300_SS410Q_25BW_4_90MS) gives 3 cm; the widest published ping, 60 kHz, gives 1.25 cm.
• Maximum range, EdgeTech 2205, 410 kHz: Datasheets EdgeTech 2021 and EdgeTech 2024b give the ranges; EdgeTech 2000 Table 9-1 gives an estimate for the longest pulse with altitude at least 10% of range. EdgeTech 2000 uses a 4.9 ms pulse.
• Center frequency, EdgeTech 2205, 540 kHz: Sold as 540 kHz [EdgeTech 2021; EdgeTech 2024b]; EdgeTech 2000's 540 kHz pulses ping at 540 kHz (one at 530 kHz). Sethuraman 2024 lists a 520 kHz centre for its 2205 and EdgeTech 2024a calls the 2205-based channel 520 kHz.
• Bandwidth, EdgeTech 2205, 540 kHz: Published span 30 to 60 kHz over the pings EdgeTech 2000 lists for this channel; the simulator's default is the ping EdgeTech 2000 Table 9-1 pairs with the maximum range, so bandwidth, pulse length and range come from one EdgeTech pairing. Shorter pings give the wider bandwidths.
• Channels, EdgeTech 2205, 540 kHz: EdgeTech 2000 (a 2205 fitted to a 2300 vehicle) states one transmit and receive section per frequency on each port and starboard array, each sensor operating independently. The AUV datasheets also advertise dynamically focused and dynamic-aperture arrays [EdgeTech 2021], which would be segmented; EdgeTech 2000 lists only non-DF pulses for 850 kHz.
• Channel spacing, EdgeTech 2205, 540 kHz: Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / 540 kHz; the published 0.26° is two-way (Sethuraman 2024), converted to one-way 0.26 x sqrt(2) = 0.368°.
• Along-track resolution, EdgeTech 2205, 540 kHz: Sethuraman 2024 lists 0.26° (two-way) for its 2205; the 2300 sheet gives 0.26° and 0.45 m @ 100 m [EdgeTech 2023] for the arrays EdgeTech 2000 shows a 2205 using. 100 m x 0.26° in radians = 0.454 m. The 6205s2 (2205 electronics and modular arrays) lists 0.36° (2-way) at 520 kHz [EdgeTech 2024a], giving 0.63 m. The simulator converts the two-way 0.26° to one-way (x sqrt(2) = 0.368°) for its element length.
• Across-track resolution, EdgeTech 2205, 540 kHz: dr = c / (2B), c = 1500 m/s assumed: 30 kHz (simulated ping 2300_SS540Q_30BW_3_7MS) gives 2.5 cm; the widest published ping, 60 kHz, gives 1.25 cm.
• Maximum range, EdgeTech 2205, 540 kHz: Datasheets EdgeTech 2021 and EdgeTech 2024b give the ranges; EdgeTech 2000 Table 9-1 gives an estimate for the longest pulse with altitude at least 10% of range. EdgeTech 2000 uses a 3.7 ms pulse.
• Bandwidth, EdgeTech 2205, 850 kHz: Published span 55 to 75 kHz over the pings EdgeTech 2000 lists for this channel; the simulator's default is the ping EdgeTech 2000 Table 9-1 pairs with the maximum range, so bandwidth, pulse length and range come from one EdgeTech pairing. Shorter pings give the wider bandwidths.
• Channels, EdgeTech 2205, 850 kHz: EdgeTech 2000 (a 2205 fitted to a 2300 vehicle) states one transmit and receive section per frequency on each port and starboard array, each sensor operating independently. The AUV datasheets also advertise dynamically focused and dynamic-aperture arrays [EdgeTech 2021], which would be segmented; EdgeTech 2000 lists only non-DF pulses for 850 kHz.
• Channel spacing, EdgeTech 2205, 850 kHz: Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / 850 kHz; the beamwidth 0.2° is read as one-way (the source does not say).
• Along-track resolution, EdgeTech 2205, 850 kHz: 2300 sheet 0.2° [EdgeTech 2023] for the 2300 arrays EdgeTech 2000 pairs with a 2205: 100 m x 0.00349 rad = 0.35 m (the 2300's own 0.15 m @ 50 m implies 0.17°). 6205s2 on 2205 modular arrays: 0.29° (2-way) [EdgeTech 2024a], giving 0.51 m.
• Across-track resolution, EdgeTech 2205, 850 kHz: dr = c / (2B), c = 1500 m/s assumed: 55 kHz (simulated ping 2300_SS850_55BW_2_8MS) gives 1.36 cm; the widest published ping, 75 kHz, gives 1.0 cm.
• Maximum range, EdgeTech 2205, 850 kHz: Datasheets EdgeTech 2021 and EdgeTech 2024b give the ranges; EdgeTech 2000 Table 9-1 gives an estimate for the longest pulse with altitude at least 10% of range. EdgeTech 2000 uses a 2.8 ms pulse.
• Center frequency, EdgeTech 2205, 1600 kHz (1610 kHz): Datasheets say 1600 kHz [EdgeTech 2021; EdgeTech 2024b]; EdgeTech 2000 Table 9-1 and Sethuraman 2024 say 1610 kHz.
• Bandwidth, EdgeTech 2205, 1600 kHz (1610 kHz): EdgeTech 2000 delivered no 1600 kHz pulses. B = 1500 / (2 x 0.006 m) = 125 kHz from 6 mm / 0.6 cm (6205s2 [EdgeTech 2024a], 4125i [EdgeTech 2025a], TRACE [EdgeTech 2026]); c = 1500 m/s assumed.
• Channels, EdgeTech 2205, 1600 kHz (1610 kHz): EdgeTech 2000 documents one transmit and receive section per frequency per side for its 120/410/850 and 230/540/850 kHz arrays; this frequency is not among them, so the same layout is assumed.
• Channel spacing, EdgeTech 2205, 1600 kHz (1610 kHz): Not published. The simulator sets channel spacing equal to its single element's length, 0.886 lambda / theta with lambda = 1500 m/s / 1600 kHz; the beamwidth 0.2° is read as one-way (the source does not say).
• Along-track resolution, EdgeTech 2205, 1600 kHz (1610 kHz): All three sister products list 0.20°; EdgeTech 2024a labels it two-way. 100 m x 0.20° in radians = 0.349 m.
• Maximum range, EdgeTech 2205, 1600 kHz (1610 kHz): Datasheets EdgeTech 2021 and EdgeTech 2024b give the ranges; EdgeTech 2000 Table 9-1 gives an estimate for the longest pulse with altitude at least 10% of range. EdgeTech 2000 lists this channel as 1610 kHz with a 1.12 ms pulse.

Where the sources disagree

• Maximum range: the datasheets [EdgeTech 2021; EdgeTech 2024b] give 1000 / 500 / 300 / 200 / 150 / 75 / 35 m for 75 / 120 / 230 / 410 / 540 / 850 / 1600 kHz; the 2205 manual's range estimate [EdgeTech 2000] gives 800 / 500 / 250 / 180 / 150 / 75 / 40 m (the last at 1610 kHz), for the longest pulse with altitude at least 10% of range. The manual also lists a 105 kHz channel (600 m) that no datasheet offers.
• Bandwidth: EdgeTech 2000's pulse tables give 10 to 14 kHz (120), 20 to 25 (230), 25 to 60 (410), 30 to 60 (540) and 55 to 75 kHz (850). The 2300 sheet's across-track resolutions [EdgeTech 2023] match the widest pulse at 230 and 850 kHz (3 cm, 1 cm) but not at 410 kHz (1.8 cm vs 1.25 cm), 540 kHz (1.5 cm vs 1.25 cm) or 120 kHz (6.5 cm vs 5.4 cm), and the long-range pulses are 2 to 2.4 times narrower than the widest.
• EdgeTech 2000 Table 9-2: the 16 ms 120 kHz ping is named "8BW" but its BW column prints 16. The page and simulator use 8 kHz (the name, the falling trend, and the printed 16 matching the pulse width); unconfirmed by any other source.
• 120 kHz channel: sold as 120 kHz, but every EdgeTech 2000 120 kHz pulse pings at 125 kHz; the 540 kHz table has one pulse at 530 kHz.
• 540 kHz centre frequency: EdgeTech 2000's pulses ping at 540 kHz, while Sethuraman 2024 lists a 520 kHz centre for its 2205 and EdgeTech 2024a calls the 2205-based channel 520 kHz in Table 2-1 (540 kHz in its spares kits).
• 1600 kHz: datasheets say 1600 kHz; EdgeTech 2000 Table 9-1 and Sethuraman 2024 say 1610 kHz.
• 410 kHz: EdgeTech 2000's text and pulses say 410 kHz; its attached array drawing 0018633 lists the standard combination as "120KHZ / 400KHZ / 850KHZ".
• 540 kHz beamwidth: Sethuraman 2024 lists 0.26° two-way (equal to the 2300's 0.26° [EdgeTech 2023]); the 6205s2 manual, built on 2205 electronics and modular arrays, lists 0.36° (2-way) at 520 kHz [EdgeTech 2024a].
• 850 kHz beamwidth: 0.2° on the 2300 sheet [EdgeTech 2023], whose arrays EdgeTech 2000 pairs with a 2205; 0.29° (2-way) in the 6205s2 manual [EdgeTech 2024a].
• One-way versus two-way: EdgeTech 2000 gives no beamwidths. Its array drawing 0018633 carries a 38.00 in [965.20 mm] dimension along the array, apparently the housing length (the drawing's dimension labels are not attributable from the text layer). A one-way reading of the 2300's 0.68° at 120 kHz needs a 0.93 m element, close to that full length; the two-way reading needs 0.67 m. This points toward two-way but does not settle it.
• Sethuraman 2024 is internally inconsistent: Eq. (1) is Rx = R x sin(θh) with θh = 0.26°, but its Table 3 values (13.61 cm at 60 m) equal R x sin(0.13°).
• Sethuraman 2024 Table 2 gives sonar ranges 30-150 m, but Table 3 lists survey ranges up to 320 m at 540 kHz, beyond the published 150 m per side; Table 3 may give the two-sided swath (not stated).
• 2300 sheet internal inconsistencies [EdgeTech 2023]: 230 kHz "1.4 m @ 150 m" implies 0.53°, not 0.5°; 850 kHz "0.15 m @ 50 m" implies 0.17°, not 0.2°.
• EdgeTech 2017 prints its side-scan range table only as an image (120 to 1600 kHz, same ranges as EdgeTech 2021/EdgeTech 2024b, no 75 kHz row), although its option list still offers 75/120 and 75/410 kHz.
• Look-down angle: EdgeTech 2000 installs the 2300-type arrays at a nominal 24-27°; the 2300 sheet says "Tilted down 26°" [EdgeTech 2023]; the 6205s2 gives 35° [EdgeTech 2024a] (not a table column).

Where the simulator's file differs from the sources

• channel count: source "real-aperture side-scan, one element per side [EdgeTech 2021]". EdgeTech 2021 does not say this; EdgeTech 2000 does ("one transmit and receive section, respective of frequency, on each port and starboard array"), for a 2205 on 2300 arrays. Cite EdgeTech 2000.
• modes[*].bandwidth (120, 230, 410, 850 kHz), tagged assumed from the 2300 sheet's resolutions: EdgeTech 2000 publishes 2205 pulse bandwidths of 10 to 14, 20 to 25, 25 to 60 and 55 to 75 kHz. The catalog's 11.5 kHz sits inside the 120 kHz range; 25 kHz equals the widest 230 kHz pulse; 41.7 kHz is a 410 kHz mid-value; 75 kHz equals the widest 850 kHz pulse. They should be tagged published and cite EdgeTech 2000.
• modes[540 kHz].bandwidth = 63 kHz and fc = 520 kHz (from Sethuraman 2024): EdgeTech 2000's 540 kHz pulses ping at 540 kHz with 30 to 60 kHz bandwidth.
• modes[*].pulse length = null (not_public) for 75 to 1600 kHz: EdgeTech 2000 publishes per-frequency pulse lengths (Table 9-1 long-range pulse: 50.0 / 12.0 / 16.0 / 8.0 / 4.9 / 3.7 / 2.8 / 1.12 ms; Tables 9-2 to 9-6 give 0.4 to 16 ms options with matching bandwidths).
• modes[*].maximum range / image_range_far_m cite EdgeTech 2021 only; EdgeTech 2000 Table 9-1 gives lower estimates at 75 kHz (800 m), 230 kHz (250 m) and 410 kHz (180 m), and 40 m at 1610 kHz.
• modes[120 kHz].center frequency = 120 kHz: EdgeTech 2000's 120 kHz pulses ping at 125 kHz.
• modes[1600 kHz].center frequency = 1600 kHz: EdgeTech 2000 and Sethuraman 2024 call it 1610 kHz.
• modes[120/230/410/850 kHz].transmit element length tagged assumed "sister product" from the 2300 [EdgeTech 2023]: EdgeTech 2000 shows a 2205 delivered with EdgeTech 2300 side-scan arrays (drawing 0018633), which strengthens but does not replace the borrowing; the sources should add EdgeTech 2000.
• review.notes "is the 2205 transducer set the same as the 2300's?": EdgeTech 2000 answers yes for one delivered configuration (2205 electronics, 2300 arrays, 120/410/850 or 230/540/850 kHz); AUV 2205 arrays may still differ.
• review.notes "Would resolve: a 2205 hardware manual." EdgeTech 2000 is that manual; it settles bandwidth, pulse length, channel layout and ranges, but gives no beamwidth.
• sources.EdgeTech 2024b: url points to the Macartney copy; EdgeTech hosts the byte-identical file at https://www.edgetech.com/wp-content/uploads/2023/04/2205-1-AUV-based-sonar-5-7-24.pdf.
• sources.EdgeTech 2021: "PN 0022368, 9/30/21" appears only in the distributor file name; the datasheet text prints neither.
• sources.EdgeTech 2017: dated 2023, but the PDF was created 8 Feb 2017; 2023-01-25 is the distributor's upload stamp; "FS ROV" is from the file name.
• sources.EdgeTech 2023: "PN 052421, 12/23/23" comes from the file name; the PDF was created 7 May 2024.
• sources.EdgeTech 2025a: url empty; EdgeTech's URL is https://www.edgetech.com/wp-content/uploads/2023/04/4125i-Brochure-10-22-25.pdf (date 10-22-25).
• modes[540 kHz].transmit element length.source: "measured on a 2205". Sethuraman 2024 lists 0.26° as a sensor parameter, not a measurement; likewise review.notes "the measured 520 kHz centre".
• modes[*].transmit element height / rx_element_elevation: tagged derived, but the 50° input is borrowed from the 2300 [EdgeTech 2023]; no 2205 source (EdgeTech 2000 included) gives a vertical beamwidth, so the tag should be assumed.
• modes[850 kHz] and modes[540 kHz] beamwidths ignore EdgeTech 2024a (0.29° and 0.36°, 2-way); modes[1600 kHz] cites EdgeTech 2025a and EdgeTech 2026 but not EdgeTech 2024a (0.20°, 2-way).
• modes[75 kHz].bandwidth = 7.5 kHz and its 10 cm resolution have no source (EdgeTech 2000 has no 75 kHz pulse table); tx_element_azimuth = 1.088° is an extrapolation, not a borrowed published value.

Not yet obtained

• An EdgeTech datasheet, ICD or test report for the 2205 / 2300 side-scan arrays giving horizontal beamwidth (one-way or two-way), element length and vertical beamwidth. EdgeTech 2000's array drawing 0018633 gives dimensions only, and EdgeTech's 2205 product page links no array sheet. Would settle: along_res for all seven channels, the one-way vs two-way basis, the 540 kHz 0.26° vs 0.36° and 850 kHz 0.2° vs 0.29° conflicts, element lengths, vertical beamwidth
• A 2205 pulse table (delivery-specific technical guide like EdgeTech 2000) for a system with 75 kHz and 1600 kHz channels, e.g. a 75/410 kHz AUV set or a 540/1600 kHz set. Would settle: bandwidth, pulse length and across-track resolution at 75 kHz and 1600 kHz, and any beam data for those arrays

Sources and quotations

Simulator file: sascore/systems/edgetech_2205.json

Confidence

• Marine Sonic Sea Scan HDS 600 / 1200 kHz: medium. One third-party listing gives frequency, resolutions and range; its along-track figures match the definition on the maker's ARC Scout MK-II sheet exactly, but no HDS beamwidth or pulse is published.
• Marine Sonic Sea Scan HDS 300 / 900 / 1800 kHz: low. Only the existence of these frequency options is published (Offshore Technology n.d.); every other value is assumed.

How each value was reached

• Bandwidth, Marine Sonic Sea Scan HDS 600 / 1200 kHz: c / (2 delta_r): 1500 / (2 x 1.5 cm) = 50 kHz, 1500 / (2 x 0.75 cm) = 100 kHz, c = 1500 m/s; the HDS pulse type and bandwidth are not published
• Channels, Marine Sonic Sea Scan HDS 600 / 1200 kHz: no source states the HDS receive channel count; a conventional single-element side-scan is assumed
• Channel spacing, Marine Sonic Sea Scan HDS 600 / 1200 kHz: Not published. One element per side (assumed), so the simulator's spacing is the element length L = 0.886 lambda / theta with theta = 0.4° one-way, taken from the ARC Scout MK-II [Marine Sonic n.d.c], whose along-track figures equal the HDS listing's: 317.3 mm at 600 kHz, 158.6 mm at 1200 kHz. Cross-check: Marine Sonic n.d.c's near-field table implies L = 2 x along-track resolution = 304.8 / 152.4 mm, 4 percent shorter.
• Along-track resolution, Marine Sonic Sea Scan HDS 600 / 1200 kHz: The values are T&T Marine Salvage n.d.'s "Resolution Along Track". T&T Marine Salvage n.d. does not define them; Marine Sonic's ARC Scout MK-II sheet [Marine Sonic n.d.c] lists the same 15.24 / 7.62 cm at 600 / 1200 kHz as "Along Track Resolution @ End of Near Field (two way)", with the near field ending at 9.3 / 4.6 m. Applying that definition to the HDS is an inference from the identical values. Marine Sonic n.d.c's table satisfies delta = sqrt(lambda x R_nf) at c = 1500 m/s (600 kHz: sqrt(2.5 mm x 9.3 m) = 15.2 cm). Beyond the near field the footprint grows with range; the HDS beamwidth is not public, so no 100 m footprint is printed
• Maximum range, Marine Sonic Sea Scan HDS 600 / 1200 kHz: "Maximum Freq. Range"; per side is not stated but implied
• Bandwidth, Marine Sonic Sea Scan HDS 300 / 900 / 1800 kHz: nothing published for these channels; scaled with frequency from the 600 / 1200 kHz pair (c / 2 delta_r with delta_r 3 / 1 / 0.5 cm)
• Channels, Marine Sonic Sea Scan HDS 300 / 900 / 1800 kHz: not published
• Channel spacing, Marine Sonic Sea Scan HDS 300 / 900 / 1800 kHz: Not published. Element length L = 0.886 lambda / 0.4° (one-way beam assumed from the ARC Scout MK-II [Marine Sonic n.d.c]) at 300, 900 and 1800 kHz.
• Maximum range, Marine Sonic Sea Scan HDS 300 / 900 / 1800 kHz: nothing published for these channels; the simulator leaves max_range_m empty (null)

Where the sources disagree

• T&T Marine Salvage n.d.'s along-track figures (15.24 / 7.62 cm at 600 / 1200 kHz) equal Marine Sonic n.d.c's "Along Track Resolution @ End of Near Field (two way)" exactly; Marine Sonic n.d.a rounds them to 15.2 / 7.6 cm and Marine Sonic n.d.b gives 15.2 / 7.5 cm, neither with a definition. The current ROWS text "definition unclear" can be replaced by Marine Sonic n.d.c's definition, with the caveat that Marine Sonic n.d.c is for a different product.
• Marine Sonic's horizontal beam figures disagree between sheets: Marine Sonic n.d.a and Marine Sonic n.d.b give 0.34° two-way at 300-1800 kHz, Marine Sonic n.d.c gives "0.4 (one-way), < 0.3 (two-way)" (no unit printed; degrees implied). The catalog reads 0.34° as one-way, which none of the sheets supports. None of these is published for the HDS.
• Marine Sonic n.d.c's near-field table is internally consistent with a transducer whose length scales as 1/f: R_nf = L^2 / (4 lambda) with L = 2 x delta gives 18.6, 9.3, 6.2, 4.6 and 3.1 m at 300-1800 kHz, matching the published values.
• HDS maximum ranges (75 / 25 m) [T&T Marine Salvage n.d.] are about half the ARC Hunter's at the same frequencies (140 / 45 m) [Marine Sonic n.d.a], so ARC figures should not stand in for the HDS.
• The derived 50 / 100 kHz bandwidths fall inside the ARC Hunter's 35-150 kHz pulse bandwidth [Marine Sonic n.d.a]; the HDS pulse type is not published.
• The current ROWS confidence "low-medium" is not one of high / medium / low.

Where the simulator's file differs from the sources

• channel count = 1 tagged published [T&T Marine Salvage n.d.]: T&T Marine Salvage n.d. does not state a channel count or a single element.
• modes[*].transmit element height and rx_element_elevation tagged derived: they rest on the ARC Hunter's 24° vertical beam, not on an HDS figure, so they are assumed.
• modes[300/900/1800 kHz].center frequency tagged published with no citation: the source is Offshore Technology n.d..
• summary quotes [Offshore Technology n.d.] as "Our fully digital entry-level towed system": Offshore Technology n.d. reads "is our fully digital entry-level towed system" (lower-case "our").
• sources.T&T Marine Salvage n.d. citation dates the listing 2015: the document shows no date; 2015 comes from its file name.
• modes[*].transmit element length / rx_element_azimuth built from 0.34° read as one-way: Marine Sonic n.d.a and Marine Sonic n.d.b state 0.34° two-way, and Marine Sonic n.d.c gives 0.4 one-way, < 0.3 two-way; none of it is for the HDS.
• review.notes says the 15.24 / 7.62 cm along-track figures have unclear meaning: Marine Sonic n.d.c defines identical figures as along-track resolution at the end of the near field (two-way).

Not yet obtained

• Marine Sonic Technology, "Sea Scan HDS Spec Sheet 2014" (mail.marinesonic.com does not resolve; Internet Archive was offline/rate-limited when tried). Would settle: HDS horizontal and vertical beamwidths and their convention, pulse type and bandwidth, ranges and resolutions for the 300 / 900 / 1800 kHz channels, and confirmation that the HDS along-track figures use Marine Sonic n.d.c's end-of-near-field (two-way) definition

Sources and quotations

Simulator file: sascore/systems/marine_sonic_sea_scan_hds.json