[Tweeters] Magic man fowls it up again

Steve Hampton via Tweeters tweeters at u.washington.edu
Wed Apr 3 20:49:42 PDT 2024


Some other common Merlin sound mistakes around here to be alert for:

1) junco vs Chipping Sparrow; sometimes Merlin will call out both from one
song!
2) Townsend's vs Hermit vs Black-thr Gray Warblers; it often calls these
Hermit, though Hermit x Townsend's hybrid is far more common.
3) Red Crossbill vs House Finch; there's one call note that House Finches
do that sounds like a crossbill, and I've seen Merlin get fooled as well as
me.

If you're back East, Merlin often mistakes Red-eyed Vireo for the much less
common Philadelphia Vireo.

That said, I'm overjoyed that my son is now learning bird songs via
Merlin! It's a game changer.

good birding,



On Wed, Apr 3, 2024 at 7:20 PM Nagi Aboulenein via Tweeters <
tweeters at u.washington.edu> wrote:


> The totally bonkers out-of-range recording misidentifications can

> sometimes occur when you have multiple bird packs downloaded and all are

> enabled (enabling specific regional bird packs is done via the “Explore”

> page, and carries over to the Sound ID). For example if you happen to have

> the Palearctic bird pack downloaded and “all” bird packs enabled, Merlin

> Sound ID can get confused and do crazy things like identifying a

> Yellowhammer in New Mexico. To eliminate these kinds of

> mis-identifications, it helps to narrow down the enabled bird packs to the

> one specific to the actual region you’re in. Then at least out-of-range

> birds won’t be considered by the Sound ID AI in trying to find a match.

>

> Having said that, Merlin Bird ID is still easily confused - we had Nathan

> Pieplow give a keynote talk at one of the Oregon Birding Association

> membership meetings a few years back, and he explained back then that for

> every 10,000 photos that Merlin’s Photo AI gets trained on, there is a

> single sound recording available for training the Sound ID AI. As more and

> more correctly identified sound recordings get uploaded and verified by

> local experts the AI will improve. But I don’t think it will ever be

> fool-proof against excellent mimics such as the Starling and the

> Mockingbird. Cornell Labs has at various times hired folks in Costa Rica,

> Panama and Colombia to both do recordings, as well as to go over other

> folks’ recording to verify correctness.

>

> All best,

>

> Nagi

> On Apr 3, 2024 at 15:19 -0700, PENNY & DAVID KOYAMA via Tweeters <

> tweeters at u.washington.edu>, wrote:

>

> Ha, ha. I still laugh about the time Merlin "recorded" a Yellowhammer

> (Eurpean bird) when birding in New Mexico, and another time last spring

> when Barb Webster and I were at Stillwater near Carnation, WA and Merlin

> claimed American Robin rather than the Bullock's Oriole we watched calling.

>

> I do admire your persistence as a Merlin detective!

> Penny Koyama, Bothell

>

> On 04/03/2024 10:12 AM PDT Kevin Lucas via Tweeters <

> tweeters at u.washington.edu> wrote:

>

>

> Over the winter Merlin has repeatedly mistaken vocalizations by starlings

> as being from Western Wood-Pewees. There are several places where a

> starling does a GREAT imitation, e.g. Eschbach Nature Area, the Eagle Path

> at the southern end of the Yakima Greenway Path in Union Gap, and at two

> places where we lived. At the Eagle Path and Eschbach, when I've read

> out-of-season Western Wood-Pewee heard-only reports, I've thought it likely

> to be one of the talented Starlings I'd heard there.

>

> In our back yard last week Merlin was fooled that a starling was a robin.

> I was watching and listening to the starling atop the utility pole calling

> like a robin. It sounded good enough I decided to give Merlin a go. Merlin

> failed, said American Robin. Most times with Starlings Merlin correctly

> indicates Starling.

>

> During a winter raptor survey this year in the White Swan area I heard an

> unusual vocalization. I was next to running water, so I could not hear all

> of the quality of the song. I tried Merlin. It said Wood Thrush. I didn't

> know what Wood Thrushes sounded like. Once home, I listened to Wood Thrush

> vocalizations. It had not been a Wood Thrush. I worked on the Merlin

> recorded audio using Audacity -- noise reduction, frequency notch filters,

> and amplification. I then played back the cleaned-up audio on my computer,

> and used Merlin on my phone to try again. This time Merlin nailed it --

> Song Sparrow.

>

> Yesterday I went to one of my favorite spots to look for Greater

> White-fronted Geese. About 10 minutes after I arrived I heard a flock of

> them heading my way. As they got closer I took some photos. Some landed

> with the Cackling and Canada Geese, others flew over. I wondered whether

> Merlin was good at distinguishing between Cackling and Canada Geese, but

> wouldn't be able to test since they were mixed together there. When I heard

> some more Greater White-fronted Geese calling as they flew toward me, I

> pulled out my phone and gave the magic man a go. They flew low, calling,

> right over me. Merlin fowled it up, identifying them as Cackling Geese.

>

> Perhaps Merlin can distinguish between Cackling and Canada Geese, but I'd

> trust him even less after his latest foul up with fowl up.

>

> I'm looking forward to hearing my first Western Wood-Pewee of the year

> while I'm watching it. I'm not annoyed by Merlin. I'm amazed, and sometimes

> I find him incredible.

>

> Good Birding,

> https://www.aba.org/aba-code-of-birding-ethics/

> Kevin Lucas

> Yakima County, WA

>

> *Qui tacet consentire videtur*

> _______________________________________________

> Tweeters mailing list

> Tweeters at u.washington.edu

> http://mailman11.u.washington.edu/mailman/listinfo/tweeters

>

> _______________________________________________

> Tweeters mailing list

> Tweeters at u.washington.edu

> http://mailman11.u.washington.edu/mailman/listinfo/tweeters

>

> _______________________________________________

> Tweeters mailing list

> Tweeters at u.washington.edu

> http://mailman11.u.washington.edu/mailman/listinfo/tweeters

>



--
​Steve Hampton​
Port Townsend, WA (qatáy)
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman11.u.washington.edu/pipermail/tweeters/attachments/20240403/bdd592e2/attachment.html>


More information about the Tweeters mailing list