isitone, Very nice! did you turn it manually with some jig/template? I mean it is pretty accurate work, the measurements are really good.
May I ask how many hours of work went into it and how much experience is needed to achieve such piece? One of my relatives has suitable lathe but I don't have very high hopes turning one yet since I don't have experience and the owner is not into audio nor precision
It could be nice if you gave a quick summary what it takes to turn such horn at home? Thanks!
May I ask how many hours of work went into it and how much experience is needed to achieve such piece? One of my relatives has suitable lathe but I don't have very high hopes turning one yet since I don't have experience and the owner is not into audio nor precision
It could be nice if you gave a quick summary what it takes to turn such horn at home? Thanks!
Last edited:
Do I do this from the same distance as before, which was about 1 meter from throat.
Yes. But I would typically prefer to see 1.5 m from the mouth, but that may not be possible in your setup.
You've got lots of familiar horns there. I spot (2x?) B&C, RCF, SEOS, E-V / JBL-type diffraction horns (or R-H / Community clones) and the asymmetrical Limmer (021?).
I endorse your findings. While some smooth diffraction horns can sound really good when crossed high (the small RCFs, for example), I prefer a bigger horn without slot for crossover points around or below 1000 Hz.
Last edited:
TNT, Image with thousand words
EQ is not relevant to see the waveguide "quality" behind the graph isitone posted. You can use any EQ and as much as you want with horn like this. Do a FIR filter with the listening axis signal and the response fixed on all axis (because the frequency response is mostly the same on all measured axis) which means there is no acoustical problems with the waveguide = as good as it gets. This particular waveguide however shows some narrowing around 1.5kHz which some of Marcel designs don't have. Still not nearly as much problems as most commercial horns available to purchase. Wanted to post it here since for example turk182 continuously questions "the quality", I thought maybe people don't understand what they are looking at in the graphs?
EQ is not relevant to see the waveguide "quality" behind the graph isitone posted. You can use any EQ and as much as you want with horn like this. Do a FIR filter with the listening axis signal and the response fixed on all axis (because the frequency response is mostly the same on all measured axis) which means there is no acoustical problems with the waveguide = as good as it gets. This particular waveguide however shows some narrowing around 1.5kHz which some of Marcel designs don't have. Still not nearly as much problems as most commercial horns available to purchase. Wanted to post it here since for example turk182 continuously questions "the quality", I thought maybe people don't understand what they are looking at in the graphs?
Attachments
Last edited:
Sure I just wanted to understand if it was applied or not in this particular measurement. My view is still, except FIR et al., that the naturally smoother, the better i.e. I welcome local higher Q dips and peaks to be absent. Broader variations are none malign. I do understand that all angels follow EQ.
//
I just wanted to understand if it was applied or not in this particular measurement. My view is still, except FIR et al., that the naturally smoother, the better i.e. I welcome local higher Q dips and peaks to be absent. Broader variations are none malign. I do understand that all angels follow EQ.//
Sure One more sentence for anybody still wondering what the heck. The dips and peaks that appear in all axis are problems of the driver, not the horn and can be EQ:d out. Use smooth driver and the response is smooth without EQ. If the lines were not tracking each other they would indicate acoustic problem in the horn which can't be fixed with EQ. Cheers
One more sentence for anybody still wondering what the heck. The dips and peaks that appear in all axis are problems of the driver, not the horn and can be EQ:d out. Use smooth driver and the response is smooth without EQ. If the lines were not tracking each other they would indicate acoustic problem in the horn which can't be fixed with EQ. Cheers
Last edited:
I have been very interested in the theory of the Klippel Near Field Scanner ever since I became aware of it some time back.
You may remember we discussed this in the "DIY Klippel" thread.
What improvements will your work offer over the way the NFS does this?
Best wishes
David
Last edited:
Performance wise, I doubt that there will be any improvement - they are basically both the same approach.
My interest is in simplifying the technique to be useful for DIY (who can't afford the 100k$+ that a Klippel costs.)
The dips and peaks that appear in all axis are problems of the driver, not the horn and can be EQ:d out. Use smooth driver and the response is smooth without EQ. If the lines were not tracking each other they would indicate acoustic problem in the horn which can't be fixed with EQ. Cheers
I don't believe that this is correct.
If there is a reflection at the mouth that causes a standing wave inside of the device, then these resonances will be as constant throughout the sound field as a resonance in the driver itself.