• Welcome to BirdForum, the internet's largest birding community with thousands of members from all over the world. The forums are dedicated to wild birds, birding, binoculars and equipment and all that goes with it.

    Please register for an account to take part in the discussions in the forum, post your pictures in the gallery and more.
ZEISS DTI thermal imaging cameras. For more discoveries at night, and during the day.

Did Merlin Bird ID really hear these in my garden? (1 Viewer)

My experience with Grasshopper Warbler is that BirdNET treats every cricket like it's one. In this case, Merlin seems to score every time, no false hits (not that I know of).
Can you run this through bN? or both if you like. Former won't get this right for me. I do have the species within a few miles of me.

 
But if I walk through a wood for two hours an AI can nail every audible call even if several are calling over each other, without bias. Up to the limits of its accuracy which are probably as good as anyone posting here. I can't throw away that kind of data.
In my experience this is simply not true, or else the posters you describe have very poor hearing indeed. Take for example this recording. To my relatively novice ear there are 14 species. Merlin only identified 12 (I believe the two it missed were BTNW and OVEN). Add to that the fact that a smartphone mic cannot pick up all the sounds that are audible. Even the faintest species in this recording were easily audible and identifiable in the field. So being generous, let's say Merlin identifies 90% of birds in a recording, and the recording picks up 90% of the audible bird sounds. That means that Merlin only has 80% of the effectiveness of the human ear.

That said, I don't care if people use Merlin as a tool to detect birds. The problem is, far too many beginners use it as their authority for identifying birds, and even report birds they never even heard themselves.
 
I’ve been using the app for about 10 days, in rural Shropshire loaded with the Europe: Britain and Ireland pack. It had been identifying species(albeit some rare) which can be expected in my area, but - yesterday - came up with the Common Mynah (Asian) the Roufus-bellied Thrush (national bird of Brazil) and the Large Hawk-Cuckoo (Asian). There have been rare invaders of the first, but the other two seem impossible.

Comments, anyone?
 
I’ve been using the app for about 10 days, in rural Shropshire loaded with the Europe: Britain and Ireland pack. It had been identifying species(albeit some rare) which can be expected in my area, but - yesterday - came up with the Common Mynah (Asian) the Roufus-bellied Thrush (national bird of Brazil) and the Large Hawk-Cuckoo (Asian). There have been rare invaders of the first, but the other two seem impossible.

Comments, anyone?
Was your location active? I find these kind of wild results I only happen if my GPS is off or misfunctioning.
 
Didn't read all comments, but concerning Yellow-legged Gull, it probably was Lesser Black-backed Gull since they call very similar to YLG while Herring Gulls have a clearly higher pitch which Merlin might recognize much easier.
 
I have never had a foreign species alert whilst my correct location is showing. When it 'loses' it (throws it away) I can get anything even tanagers. It can come and go constantly.
 
Just downloaded it and given.it a quick go. It correctly identified some species but also added Blackcap (which I couldn't hear so doubtful) and Swift which are present but I couldn't see any at that time. Personally I wouldn't want to rely on it without getting some kind of look at the bird.

I am rubbish at bird calls so would find it helpful with skulking birds. I don't mind putting the effort in to see something if the app has given me a heads up. Its great when you have spent an age scanning a bush for something very common to pop out!

How does the app cope with say a Great Tit that has a vast range of calls as a species or Starlings when.yhey are mimicking other birds?
 
How does the app cope with say a Great Tit that has a vast range of calls as a species or Starlings when.yhey are mimicking other birds?
Tit calls are not 100% reliably ID'able on sound, whether by app or by ear (for now, from what I know), due to Great Tit mimicry (with the possible exception of Great Tit's signature alarm call). Other than that, I don't think the app has been particularly bad at trying to guess the most probable tit species, and the songs should be OK. It can identify an otherwise absent mimicked species from the mimic's song, not only in Starlings.
 
My two penn'orth - without a clear visual sighting then surely the first port of call should be the use of probability.

Now whilst we can't rule out rarities (there are plenty of megas that abound) we surely must first pause and reflect.

In saying that - my app has just told me that there's a Capercaillie in my back garden in Liverpool so sorry but I've got to go..........
 
I just gave the app a quick try outside work in Littlehampton. It picked up Blue and Great Tit and Greenfinch fine. It missed Magpie and Whitethroat which were both audible on the recording, and magicked up a Robin which wasn't calling or singing at the time
 
Tit calls are not 100% reliably ID'able on sound, whether by app or by ear (for now, from what I know), due to Great Tit mimicry (with the possible exception of Great Tit's signature alarm call). Other than that, I don't think the app has been particularly bad at trying to guess the most probable tit species, and the songs should be OK. It can identify an otherwise absent mimicked species from the mimic's song, not only in Starlings.
No, but, it's a good tool -- just a tool, though. For me, it's been a big aid in learning.
 

Users who are viewing this thread

Back
Top