Date: 4/17/26 12:47 pm From: Van Remsen via groups.io <vnremsen...> Subject: Re: [labird] MERLIN audio -- warning!
One of my best Merlin yard "records":
On Thu, Apr 16, 2026 at 5:29 PM Van Remsen via groups.io <vnremsen=
<gmail.com...> wrote:
> Before I get accused of "hating Merlin", I should follow-up by making it
> clear that I think Merlin is great for learning bird songs and detecting
> species one might not have noticed. It's an amazing tool. What I "hate"
> is when Merlin output is uncritically considered the same as eBird input.
> "Trust but Verify" as the saying goes.
>
> Van Remsen
>
> On Thu, Apr 16, 2026 at 12:14 PM Van Remsen via groups.io <vnremsen=
> <gmail.com...> wrote:
>
> > LABIRD: Most people in this venue have heard multiple times that MERLIN
> > frequently makes mistakes and should never be relied on as the sole
> > evidence for an eBird entry. Merlin's own directions emphasize this. Yet
> > many eBirder seem unaware of this and treat Merlin hits as sufficient
> > evidence for presence.
> >
> > Yesterday provided a poignant example. A single Summer Tanager singing
> > was identified as 3 separate species: Am. Robin, Rose-breasted Grosbeak,
> > and Summer Tanager. Yes, the same individual bird. Yeah, the songs are
> > kind of similar but not THAT similar.
> >
> > As I have mentioned previously, I did 30 trials of 30 minutes at my place
> > in St. Gabriel and NOT ONE was without 1 to 5 obvious or highly
> suspicious
> > false positives. Since then, I've done dozens of unofficial trials with
> > the same outcome. You would think just by chance I'd have a "clean"
> trial
> > every so often. The most frequent local false positives are Red-eyed
> > Vireo, House Sparrow, E. Towhee, Baltimore Oriole, Y-b Chat, Am. Robin,
> and
> > Rock Pigeon. On the other hand, Merlin often misses Fish Crow, Y-b
> Cuckoo,
> > Mourning Dove, Barred Owl, and Screech Owl if they are more than couple
> > hundred yards away, and misses Brown Thrasher "Chack" calls even at close
> > range.
> >
> > Van Remsen
> >
> >
> >
> >
> >
> >
>
>
>
>
>
>
