Ah, i was triggered by the 13cm. Floor reflections is for sure present in that measurement. Hmmm, why not so present in the felt version, interesting. Will try to set up the vituixcad setup like you did.
Tmuikku - your post was very interesting. It caused me to go back and looked at some of my past measurements to check for the same phenomenon.
I have said it so many times I am probably becoming annoying, but I will say it again; One of the most technically challenging aspect of DIY speaker design and construction is making accurate repeatable measurements.
I have said it so many times I am probably becoming annoying, but I will say it again; One of the most technically challenging aspect of DIY speaker design and construction is making accurate repeatable measurements.
Indeed 
its kind of flying blind until consistency is achieved, sometimes they are accurate enough for good results and sometimes they have error that can lead to wrong conclusions. Although wrong conclusions can be done with good data as well. Increasing understanding whats on the graphs is very important I think.
its kind of flying blind until consistency is achieved, sometimes they are accurate enough for good results and sometimes they have error that can lead to wrong conclusions. Although wrong conclusions can be done with good data as well. Increasing understanding whats on the graphs is very important I think.It is still there in the felt one just reduced and more visible further off axis. With such a pronounced peak dip combination it is easy to see as diffraction/reflection entering. Go back to look at the no felt IR, you should be able to modify the gate to get rid of it which would also help to isolate the time it is occurring.
Follow-up: I did a fresh start and used the same ref-time for both measurements. Set up in VituixCAD also listening distance to 1000mm (measure-dist), and highlighted the 40' off-axis response (nice feature of VituixCad, i change the selected measurments of driver, and the highlighted response remains).
FELT:
NOFELT:
Ideal-Mid-105mm-90Difhor-nornd:
Ideal-Mid-105mm-90Difhor-40mmrnd:
The directivity increases with felt, so also the amount of floor-reflection reduces a bit
Specifically the response in the first 40 degrees hor is much better, reduced diffraction effect i assume for the moment.
At higher of-axis angles of course the floor reflection is relatively stronger present in measurements.
The dip at around 1700 Hz most likely a interference from the 2cm felt creating a "open ended cylinder" and resonance calc for such cylinder get a ref-freq of ~1700 Hz. The peak at 2500 Hz is presumably a negative effect of the protective grille.
So all in all the difference with Felt (absorption sheet) on the baffle is not trivial, but more testing needed to get the most out of it.
I will stop hijacking Tmuikku's thread, and start my own.
Must thank the input of Tmuikku, Hifijim and Fluid for pointing me in more analysis use of VituixCad for role of reflections, the need to be meticulous with IR2FR, and to get consistent results. Fist now to get rid of the DC disturbance in the dual-channel set up with Arta, as it appears also to disturb the correct timing,
On to get consistent and correct measurements!
FELT:
NOFELT:
Ideal-Mid-105mm-90Difhor-nornd:
Ideal-Mid-105mm-90Difhor-40mmrnd:
The directivity increases with felt, so also the amount of floor-reflection reduces a bit
Specifically the response in the first 40 degrees hor is much better, reduced diffraction effect i assume for the moment.
At higher of-axis angles of course the floor reflection is relatively stronger present in measurements.
The dip at around 1700 Hz most likely a interference from the 2cm felt creating a "open ended cylinder" and resonance calc for such cylinder get a ref-freq of ~1700 Hz. The peak at 2500 Hz is presumably a negative effect of the protective grille.
So all in all the difference with Felt (absorption sheet) on the baffle is not trivial, but more testing needed to get the most out of it.
I will stop hijacking Tmuikku's thread, and start my own.
Must thank the input of Tmuikku, Hifijim and Fluid for pointing me in more analysis use of VituixCad for role of reflections, the need to be meticulous with IR2FR, and to get consistent results. Fist now to get rid of the DC disturbance in the dual-channel set up with Arta, as it appears also to disturb the correct timing,
On to get consistent and correct measurements!
The numbers might add up but to me it just looks like the basic baffle diffraction response due the placement of the driver on the baffle. A similar thing is seen in the idealized response just not as deep. The felt modifies it but does not cause it. What the felt does do is clean up the on / off axis differences around the baffle diffraction signature, something that creating a smoother baffle transition also does.
Could well be, need to test. Btw if i change the cylinder length to 0.9-1cm (suspension + grille) it coincides with the dip of the no-felt. Or the felt effect is that of a longer distance to edge, together with some radius, thus lowering the frequency of the dip. Test goals defined
Last edited: