For this work, a 16bit 44.1/48kHz sample rate is sufficient. Conversion to 320Kbps mp3 will have no impact.
Smallish and predictable effect is had using PEQ around 8kHz to see what added sizzle of H4 response has.
All else being equal in recording process, dominant factor in overall brightness compared to reference tracks is do to room reflections and directivity of the speakers. X's room is bright. After listening to sections <0.5 seconds in looping mode and going through the different drivers in different orders for several minutes, and then switching to reference track, the reference sounds profoundly dull and dry. The ears become rapidly adjusted to the room sound as a reverberation effect that makes transients all sound louder.
Planet10; I somehow get the impression that you have not spent time listening to the clips.
Smallish and predictable effect is had using PEQ around 8kHz to see what added sizzle of H4 response has.
All else being equal in recording process, dominant factor in overall brightness compared to reference tracks is do to room reflections and directivity of the speakers. X's room is bright. After listening to sections <0.5 seconds in looping mode and going through the different drivers in different orders for several minutes, and then switching to reference track, the reference sounds profoundly dull and dry. The ears become rapidly adjusted to the room sound as a reverberation effect that makes transients all sound louder.
Planet10; I somehow get the impression that you have not spent time listening to the clips.
But at least there is a concistency in the test and the results, here they are closer maybe because as noticed : qualities are near and tastes of the listener stays important !
Of course there is a loss, which is hearable, but the exercice is to choose into those limitations and it seems, again, a certain concistency in the results is present ! When someone asks me if I see the color of a flower without showing me infra red picture : it stays a honnest exercice !
We are not totaly blind, there a lot of many good drivers and the role of the designer is too extract the best of it like when you cook a shrimp
o I'm not sure of the picture here, but french reflex, sorry !)
The good news is a sota designer can make a better speaker with a B choice driver (I mean not the B of this test which is not a B choice in my mind ! More an ecrevisse than a shrimp !)
You must stay in the limit of the exercice ! Because of what you said I choosed 3 tracks with low and high peaks to make the exercice easier to my tastes! I see with no sucess : one view and after the member woke up ! I have bad tastes !)
Allez quoi ! Just do it : http://www.diyaudio.com/forums/full...2in-4in-drivers-round-4-a-22.html#post4448327
And I have found the good cymbals track : Of course : Kind of Blue : first track, first 1m30 of "So What"... and I bet everybody has it in his Library !
Of course there is a loss, which is hearable, but the exercice is to choose into those limitations and it seems, again, a certain concistency in the results is present ! When someone asks me if I see the color of a flower without showing me infra red picture : it stays a honnest exercice !
We are not totaly blind, there a lot of many good drivers and the role of the designer is too extract the best of it like when you cook a shrimp
o I'm not sure of the picture here, but french reflex, sorry !)The good news is a sota designer can make a better speaker with a B choice driver (I mean not the B of this test which is not a B choice in my mind ! More an ecrevisse than a shrimp !)
You must stay in the limit of the exercice ! Because of what you said I choosed 3 tracks with low and high peaks to make the exercice easier to my tastes! I see with no sucess : one view and after the member woke up !
I have bad tastes !)Allez quoi ! Just do it : http://www.diyaudio.com/forums/full...2in-4in-drivers-round-4-a-22.html#post4448327
And I have found the good cymbals track : Of course : Kind of Blue : first track, first 1m30 of "So What"... and I bet everybody has it in his Library !
Last edited:
Ah if you travel, like a lot this as well (too compressed , but the direction is one of the best I heard for Appalachian Spring : https://www.youtube.com/watch?v=KTEEykZZwEc
At least You tube ! mega compressed ! I believe it's hard to have an idea on a standalone driver with symphonic music ! Or at least from any speakers : too hard to reproduce with so much bad compressed reccording and so bad listeners who mix volume and dynamic while thinking to the neighboors !
At least You tube ! mega compressed ! I believe it's hard to have an idea on a standalone driver with symphonic music ! Or at least from any speakers : too hard to reproduce with so much bad compressed reccording and so bad listeners who mix volume and dynamic while thinking to the neighboors !
Last edited:
He has measurements of the Vifas and the ScanSpeak ! But he has himself a mic and and not an annechoic-bidule chamber ! (maybe he measures from the ground semi digged far from any obstacles ?
So bad he stoped to publisch ! Sniped by a speaker brand ?
Then look forward because just after revealing xrk971 for previous rounds release 4mS gated plots so we can see what fooled ourselves, more than that we often get IR HD and waterfall plots too. These are not measured on a IEC baffle as Zaph but at actual setup as seen loking into post 1 where TG9FD 8 ohm version is in focus, think measurements were done same spot in space where recording mic was placed. Below FR measurements from round 1/2/3, go to respective threads to have IR HD and waterfalls.
Round 3:
Round 2:
Round 1:
Last edited:
That is a Zoom H4N, a newer and different model than my Zoom H4 used here. Back in post 137 I showed the response of the H4 vs a calibrated UMM-6 with pink noise as the source. I don't think there is a treble lift at 5kHz like the 4N. It is actually a little depressed above 15kHz but within 2dB around the 5kHz range. It maybe that the drivers have a mild naturally rising response and since Round 2, I stopped applying a -3dB low Q high shelf as baffle step compensation. I match SPL between the woofer and full range near the XO point but if the driver has a self rising response - it is left alone and that may be what you are hearing.
Here is the H4 vs calibrated mic with pink noise.
But this is against the raw UMM-6 microphone correct? That means that the correction file has not been applied to it?
The UMM-6 has a naturally rising treble also as seen from typical calibration files. Like these from here (Serious Audio: Should You Get Your Measurement Microphone Professionally Calibrated?)
Plus we can see from your pink noise plot compared to an ideal pink noise 10db/decade line that there is some signature. But this is not bad.
Maybe the brightness we hear in all the recordings (compared to the references) is just the +4db peak of the zoom H4 at 8khz that we see above when compared to the idea 10db/octave line.
The brightness seems universal in all the recordings so it does feel like it is part of the recording process (mic, preamp, room) so it should not effect the relative impressions of the drivers. Maybe next round you can try some test recordings with the UMM-6 and Zoom H4 and see if the UMM-6 looses some of that brightness compared to the built in mics.
But yeah maybe it is just the loss of the -3db baffle step correction from previous setups. But we should see that in your measured FR plots (once you release them), yes?
Attachments
Last edited:
It wasn't me most recently questioning the response of your mic… what you post wrt that is not all that clear. Just asking for a clearer presentation.
dave
Thanks. The TG9 and 10F look the best in round 2 (assuming the measurement mic isn't fooling us). Almost identical.These are not measured on a IEC baffle as Zaph but at actual setup as seen looking into post 1 where TG9FD 8 ohm version is in focus, think measurements were done same spot in space where recording mic was placed. Below FR measurements from round 1/2/3, go to respective threads to have IR HD and waterfalls.
Round 3:
Round 2:
Round 1: