The Forecast Said High. The Map Logged Five Reports.

The Independent’s sound forecast gave the September 25 concert at the Ford Amphitheater its highest rating, a band in which past nights have drawn a median of twenty-four reports to the resident-run noise map. The map logged five, three of them the following day. The forecast has now misjudged three shows in a row, twice low and once high.

A little over three hours before the Commodores and the Spinners played the Ford Amphitheater on Friday, September 25, the Independent’s sound forecast told readers to expect a busy night. It rated the show High, the top of its five bands. The five comparable nights in its record had drawn a median of twenty-four reports to the resident-run noise map, and none had drawn fewer than six.

Five reports arrived, two that night and three the next day. The forecast was wrong for two reasons, and both were its own. The breeze it expected to carry sound toward the neighborhoods north of the venue died before the show began, and this newspaper had filed a soul and R&B double bill in its top band for bass. Each error raised the rating by one band. This is the third consecutive show the tool has misjudged. It rated the previous two low–moderate, and they drew twenty-four and thirty-three reports. A forecast that misses twice in one direction and once in the other is not biased. It is imprecise, which is harder to fix.

What the page told readers

The forecast recalculates twice a day and keeps every run. The last one before the show was computed at 4:11 p.m. It expected a 7 mph wind carrying sound toward the neighborhoods at 4.3 mph through the evening, and it classed the bill as high in low-frequency content. Together those score four out of four.

The page had been saying much the same for two weeks. Of the thirty-two runs made in the sixteen days before the concert, twenty-two said High, five said Elevated and five said Moderate. None said anything lower. It did carry a warning that the forecast had reversed direction twice in its last eight runs. Both runs on Thursday had the wind blowing away from the neighborhoods, and both on Friday had it blowing toward them. What arrived was neither. There was almost no wind at all.

The wind left early

The southerly breeze the forecast was counting on was there at six o’clock. By seven it had dropped below 5 mph and swung to the northeast, and it stayed light for the rest of the evening. Over the five hours the page scores, the forecast put the wind toward the neighborhoods at +4.3 mph. The weather archive puts it at +1.2 mph. The Air Force Academy airfield, whose instruments the forecast does not use, puts it at −0.5.

At those speeds direction carries little information. The archive has the wind arriving from the east-northeast, the south-southwest, the east-northeast again and the southeast in successive hours. This newspaper reached the same diagnosis about the Brantley Gilbert show on September 5, another evening of light and wandering wind. That night drew twenty-four reports. This one drew five. A calm evening, on this evidence, can go either way.

A soul bill in the top bass band

The forecast’s second input is the act. The Independent assigns each performer to one of five genre categories, four of them taken from published acoustic research with a measured bass differential for each, and groups them into three bands. It filed the Commodores and the Spinners under the category the page labels “Urban / hip-hop,” which sits in the top band alongside dance music. The Gazette’s preview of the show described the evening as R&B and soul.

That assignment was this newspaper’s own judgment, and it was wrong. The bill has been refiled as pop, the middle band.

Scored against the weather that occurred, the show rates Elevated as it was filed and Moderate as it is filed now. The Moderate band’s ten nights have a median of four reports, so the night enters the record looking like a success. It was not one. A reader who opened the page on Friday afternoon was told High, and a classification corrected after the fact is hindsight.

Five reports, and a phone reading

Two reports reached the map on the night. One came from 6.3 miles east of the venue and was rated mild: “We could hear the rumbling of the concert while we were outside.” The other came from 2.6 miles north and gave the evening the map’s maximum score outdoors. Three more were filed on Saturday. Under the Independent’s standing rule a report filed the day after a show is credited to that show when the venue has nothing on that following night, which makes five. None carried a decibel figure.

The map records who chose to report, not how loud a night was, and there is some indication that the concert could be heard. Murray Relf, a retired radiologist active in residents’ advocacy over the venue’s noise, wrote to City of Colorado Springs officials on September 27 and copied this newspaper. Relf lived about a mile from the amphitheater until he moved to Monument last year, and he has supplied documents to the Independent before. His email attached screenshots from a phone sound-meter app, made during the concert on Narrowleaf Drive in the Northgate Highlands neighborhood, by his account. They show ten-minute A-weighted averages of 54.2 decibels beginning at 9:32 p.m. and 53.2 beginning at 10:12 p.m., against the 44 he gives as typical for the area without a concert. The app’s calibration field reads 0.0 dB.

Relf has since told the Independent that the readings were made with Sound Level Analyzer PRO, an iPhone app, and has sent video of a handheld sound meter which he says was recorded in the same neighborhood. Data embedded in the files date two clips to 9:32 and 9:35 p.m. on September 25, inside the app’s first ten-minute trace, and put the camera about nine-tenths of a mile north-northeast of the venue. In them the meter, set to A-weighting, reads mostly between 52 and 60 decibels, with brief peaks in the 60s. A third clip is dated 10:30 p.m. on September 30, when the venue had no show, and in it the same meter reads between 41 and 45.

Relf points to a 2016 study in the journal Applied Acoustics, which tested 100 smartphones against a calibrated meter and found one app, SLA Lite, within one decibel of the true level in each of four test conditions. The developer of Sound Level Analyzer says SLA Lite is its app. The tests used steady white noise in a laboratory, and the study found that a phone’s age affected its accuracy. Relf’s readings are of amplified music outdoors.

The Independent has verified the calibration of neither the phone nor the handheld meter, and it does not treat their readings as measurements. They are reported here because they are the only readings of any kind this newspaper holds for the night, and because they agree with the report from 2.6 miles north: there was sound, and few people said so. It was a cool evening, in the 50s during the show, though temperature has not tracked complaint volume in this record.

The map heard Bruno Mars too

Saturday’s three reports come with a complication. Bruno Mars played Falcon Stadium at the Air Force Academy on Saturday and Sunday, September 26 and 27, KOAA reported. The stadium is about 1.9 miles southwest of the amphitheater. One of Saturday’s three reports names him. The other two name no performer, and none of the three mentions Friday’s concert. The standing rule credits them to Friday’s show, and that is what is done here. With them or without them, the night falls below anything the High band had recorded.

Three further reports arrived on Sunday, which attach to no Ford show. Two of them name Bruno Mars. “We could hear Bruno Mars loud and clear in the house until 10:04pm,” one resident wrote from a home about 2.3 miles from the stadium.

These are the first reports the Independent has found on the map that name another event. The map is named for one venue, and some of the people who use it can evidently hear two.

One night in four

The Independent published the forecast on August 18. Four shows have been played since.

The page describes itself as a ranking that “is wrong a meaningful share of the time.” On the four nights it has been tested in public, the share is three in four.

There is a reason the public record is worse than the one the page advertises. Its stated accuracy comes from past nights scored with the weather already known. A forecast has to guess the weather as well, and on September 25 that guess cost it a band. The other two misses owe nothing to the weather forecast: scored against what actually happened, September 5 and September 19 still come out low–moderate.

Four nights prove little on their own. They are, however, the only four on which the tool has had to commit in advance. The forecast page now carries a note recording all three misses and the refiling.

The Independent is still working on the tool. One change has been made already. The second of its two daily readings, which was taken at 4:10 p.m., is now taken one hour before each show, so that the last forecast before a concert rests on later weather data. Whether that would have caught the wind dying on September 25 is not known. The record behind the forecast is fifty-six nights long and each show adds one. With more of them, the paper hopes the forecast will become more accurate.

Two shows left

The season’s last two shows are for KING + COUNTRY on Saturday, October 3, and Dwight Yoakam on Friday, October 9. As of Thursday morning the forecast rates the first Low–moderate, with the wind carrying sound away from the neighborhoods at 6.9 mph, and the second Elevated, with the wind toward them at 6.0 mph. Both carry the page’s “unsettled” flag, and the Yoakam forecast is eight days out, where wind direction is least reliable. Both acts have played the venue before: for KING + COUNTRY drew nine reports in September 2024 and Dwight Yoakam nineteen in May 2025.

VENU Holding Corp., the company behind the venue, has not replied to questions the Independent sent on September 6 about the Brantley Gilbert show. It has not been asked about the two shows since.

On its public record so far, the forecast has ranked one night in four correctly. Saturday is the fifth.

Sources & further reading

The Independent’s own coverage: The Wind Blew the Other Way, and the Complaints Came Anyway — the September 19 show and the second miss · Sound Reached Six Miles on a Nearly Windless Night — the September 5 show and the first · At the Ford Amphitheater, the Complaints Track the Wind — the three-season analysis the forecast came out of · The Independent Builds a Sound Forecast for the Ford Amphitheater — how the tool was built · Ford Amphitheater sound forecast — the live page, updated twice daily.

Complaint data: Ford Amphitheater noise map, all 836 records pulled October 1, 2026.

Weather: Open-Meteo forecast, as stored by the forecast page at each run, and its historical forecast archive, hourly, at the venue coordinate · Iowa Environmental Mesonet ASOS archive, observations from the U.S. Air Force Academy airfield.

Venues and events: Ford Amphitheater events calendar · KOAA, “Bruno Mars will perform at Falcon Stadium this weekend,” September 23, 2026 · The Gazette, “Q&A with Jessie Peck: The Spinners and The Commodores head to Ford Amphitheater,” September 24, 2026 · VENU Holding Corp.

Residents: Murray Relf, email of September 27, 2026, to City of Colorado Springs officials, copied to the Independent, with three screenshots, and two emails to the Independent of October 1, 2026, with an annotated chart and three video clips · Enda Murphy and Eoin A. King, “Testing the accuracy of smartphones and sound level meter applications for measuring environmental noise,” Applied Acoustics, volume 106, 2016, abstract · Quiet Communities, “People over profit: Residents challenge outdoor concert noise,” June 2026.

Methodology

Complaint counts come from the resident-run noise map, a self-selected, anonymous, advocacy-aligned record of complaints rather than a measurement of sound. It cannot be read as decibels. Reports are credited to the show on their own date where one exists, otherwise to the previous evening’s. The count of five for September 25 is two reports dated that day and three dated September 26, when the venue’s calendar lists no show; one of those three names a different concert, and the rule is applied without exception. The three reports dated September 27 follow a night with no Ford show and are credited to none. The September 19 count of thirty-three is twenty-one reports dated that day and twelve dated the next. The statement that no earlier report names another event rests on a keyword search of the 729 records that carry a comment and on the Independent’s weekly reading of new records, not on a fresh reading of every comment.

Forecast figures are the forecast page’s own stored runs, which record the hourly weather each run was given. The run quoted for September 25 is the last before the show, made at 4:11 p.m. Mountain. Its window covers the five hourly readings from 6 p.m. to 10 p.m. for a show advertised at 7:30. Band medians and ranges quoted for forecasts are those of the fifty-four performance dates the live page was using when they were made. The record as regenerated on October 1 holds fifty-six, and the Moderate band’s figures are from that record.

Actual wind comes from two records. The Open-Meteo historical forecast archive was read at the venue coordinate on October 1, 2026. The Air Force Academy airfield reports a few minutes before each hour; each reading is assigned to the hour that follows and converted from knots. The component toward the neighborhoods is the wind projected onto the bearing from the venue to the vector-mean position of all cleaned complaint locations, 18.4 degrees on the page as it stood on September 25. Over the four hours from 7 p.m. alone the archive gives −0.2 mph and the station −1.8.

The Commodores and the Spinners were refiled from the forecast’s urban category to pop on October 1, after the show. The change lowers the rating that night would receive and is disclosed for that reason. It leaves the tool’s published correlations almost unchanged.

Distances to reports are measured from a venue coordinate the Independent derived by trilateration from residents’ own stated distances, accurate to about 0.09 miles. The distance between the amphitheater and Falcon Stadium is measured between the coordinates OpenStreetMap gives for each.

The phone readings are reproduced from three screenshots attached to Murray Relf’s email of September 27 and from the annotated chart he sent on October 1, which is shown as he supplied it. The screenshots show an A-weighted, slow-response setting and a ten-minute measurement period. After publication Relf named the app and the street and supplied three video clips of a handheld meter, a TopTes TS-501B set to A-weighting and slow response. The Independent read the meter’s display at one-second intervals through each clip and disregarded frames caught while the display was changing. The dates, times and approximate position of the clips come from data embedded in the files and were not otherwise verified. That data was removed from the two clips published here, which were also reduced in size; they are otherwise as supplied. A wind gauge beside the meter reads zero through the concert clip shown and rises above zero for part of the clip without a concert. The Independent has not verified the calibration of either device. The 44-decibel ambient figure on the chart is Relf’s, attributed by him to Hankard Environmental, the firm the City of Colorado Springs engaged to measure the venue’s sound; the Independent has not confirmed it from a published report. Of the study Relf cites, the Independent read the abstract, not the full paper, and the statement that SLA Lite is the same developer’s app rests on the developer’s own web page. The Independent’s standing practice is not to publish decibel trends from uncalibrated devices, and none is drawn here. Biographical details for Relf come from the moderator’s introduction to a Quiet Communities webinar published on YouTube in June 2026, read from YouTube’s automatic captions. Those captions are generated by automatic speech recognition rather than written by a person and misrender names; Relf’s name is taken from his own email.

Resident comments are reproduced exactly as submitted, including spelling and punctuation.