Please do no emergency fixes, this is caused ( simulation or real measurement) by the lowpass filter in the Nycuist theoreum of the sample rate.
Nyquist–Shannon sampling theorem - Wikipedia
Nyquist–Shannon sampling theorem - Wikipedia
Hi everybody,
I have been tinkering with my pair of F503´s triamping my passive spakers ( kef ref 205) using REW and U-MIK 1 ( no crossovering duties for the time being). I was not happy with the sound in general, sounded kind of dead..I had made some measurements and there was the expected downtilt towards high frequencies to wich I didn´t really pay that much attention, as I was focusing in correcting stuff under 450Hz and wasn´t applying any filters on higher frequencies, so they would remain unchanged ( I knew that the output filters might roll the very high treble just a tiny bit, but I can´t hear past 14,5k). HOWEVER I decided to do a comparison, and hook my Musical fidelity KW 250s again to the speakers to drive it and then compare the same speaker with the mic in the same position with the Fusion driving it with no filters on, my non processed, flat preset, and was shocked to discover that from 2Khz on up to 20000KHz, levels go progressively down in a straight line up to -4,5 db at 20K, no filters applied post 1Kz as I said ( 5,0 Q, -3 db peak filter at 1080 Hz, ). So in short the Fusion, in totally flat mode comes up 4,5 db lower at 20k, starting that descent at 2k.
The setup was : REW measuring running, laptop headphone 3,5 mm out - 3,5mm jack into rca cable - MF KW 250S aux in - KW250S volume control - loudspeakers-UMIK into USB on the same laptop, versus REW measuring running, laptop headphone 3,5 mm out - 3,5mm jack into rca cable - RCA of FA 503 - FA driving loudspeakers ( HF channel on treble, and each module on mid and bass, as they are triwirable) - loudspeakers-UMIK into USB on the same laptop. Laptop output levels were constant You will notive that the levels are very different...Why that is so I don´t know, I tried to check levels but they were always displayed as 20-30> dbs different, althoug they sounded pretty much the same after using each preamp ( MF and Hypex) function and checking the measurement volume by ear ( MF always plotted much lower, and Fusion too high, actual leves were at normal 80-85 db listening levels..). I had conducted measurements before, and encountered no problems at all, checking levels and getting curves plotted beteween 75 and 95 db with right away, with either solution, don´t know what is happening here..
Anyway, apart from that, the aditional droop post 2Khz is very visible on the Hypex, some 4,5 db) on top of the 4 db droop seen with the MK (wich is 2 db shy of Toole target but I dont mind a little brightness) What is the reason for this, I don´t think this should be happening, measurements of the Hypex barely showed any frequency load dependent variaton in the extreme treble, so why am I getting this deviation ( and the definetely sound like that too) ?? Do I have a defective output choke or something? Only tested one channel so far. Any suggestions are welcomed as to waht is causing this are welcomed, thanks in advance.
I have been tinkering with my pair of F503´s triamping my passive spakers ( kef ref 205) using REW and U-MIK 1 ( no crossovering duties for the time being). I was not happy with the sound in general, sounded kind of dead..I had made some measurements and there was the expected downtilt towards high frequencies to wich I didn´t really pay that much attention, as I was focusing in correcting stuff under 450Hz and wasn´t applying any filters on higher frequencies, so they would remain unchanged ( I knew that the output filters might roll the very high treble just a tiny bit, but I can´t hear past 14,5k). HOWEVER I decided to do a comparison, and hook my Musical fidelity KW 250s again to the speakers to drive it and then compare the same speaker with the mic in the same position with the Fusion driving it with no filters on, my non processed, flat preset, and was shocked to discover that from 2Khz on up to 20000KHz, levels go progressively down in a straight line up to -4,5 db at 20K, no filters applied post 1Kz as I said ( 5,0 Q, -3 db peak filter at 1080 Hz, ). So in short the Fusion, in totally flat mode comes up 4,5 db lower at 20k, starting that descent at 2k.
The setup was : REW measuring running, laptop headphone 3,5 mm out - 3,5mm jack into rca cable - MF KW 250S aux in - KW250S volume control - loudspeakers-UMIK into USB on the same laptop, versus REW measuring running, laptop headphone 3,5 mm out - 3,5mm jack into rca cable - RCA of FA 503 - FA driving loudspeakers ( HF channel on treble, and each module on mid and bass, as they are triwirable) - loudspeakers-UMIK into USB on the same laptop. Laptop output levels were constant You will notive that the levels are very different...Why that is so I don´t know, I tried to check levels but they were always displayed as 20-30> dbs different, althoug they sounded pretty much the same after using each preamp ( MF and Hypex) function and checking the measurement volume by ear ( MF always plotted much lower, and Fusion too high, actual leves were at normal 80-85 db listening levels..). I had conducted measurements before, and encountered no problems at all, checking levels and getting curves plotted beteween 75 and 95 db with right away, with either solution, don´t know what is happening here..
Anyway, apart from that, the aditional droop post 2Khz is very visible on the Hypex, some 4,5 db) on top of the 4 db droop seen with the MK (wich is 2 db shy of Toole target but I dont mind a little brightness) What is the reason for this, I don´t think this should be happening, measurements of the Hypex barely showed any frequency load dependent variaton in the extreme treble, so why am I getting this deviation ( and the definetely sound like that too) ?? Do I have a defective output choke or something? Only tested one channel so far. Any suggestions are welcomed as to waht is causing this are welcomed, thanks in advance.
Thanks for the quick suggestions will try all of the above. The preset I was using was my number 3, wich was actually the factory no pre equing or nothing, The hi freq channel is driven direct with no bass/mid filtering. I did not want to put any filtering there so as to preserve the original fase of my loudspeaker not having to resort to Rephase until I master REW and the HFd ( difficult software very incomplete manual, haven´t used it for measurement as REW is so much easier). Because of this I tested using my laptop headphone jack out, and the analog RCA in ogf the Hypex, don´t no if that is in play in here too ( ADC problems inside the Hypex?) Have software limiters with 300/150/100 watts enabled in each channel in case something goes wild, but at the levels I was using they could not be kicking in, no way
Hi,
Already did some of the proposed. Fired REW with the same setup as before, and measured my right module on the right speaker and got the exact same curve as the left (just some very minor variation below 500hz, as this speaker doesn´t have a nearby boundary like the left one): a steady droop in level 2khz onwards, totalling 9,5-10 db at 20Khz mirroring exactly what I got on the other channel, that is, some 4,5-5 db down at 20Khz compared to my MF output. Same problems with REW too, getting solid 88-91db in reality ( checked with spl meter app, reads 90 db minimum) but Rew complaining levels were too low ( UMIK issue?). Anyway, even if I have some issue with the mic or REW are acting on me, its the same mic and settings in every measurement, both the modules and MF.
Checked filters and pre EQ in my HFD 4.95, firmware 1.41, everything is flat on the preset I am measuring. Have recentely redownloaded the firmware to them too, becaus e of remote control issues, as advised by Hypex support ( although they talked about vs 1.42, only vs 1.41 was available in their download page and that is what is running). So as both modules display the same behaviour, I then, decided to try and load a totally new project and see what happens: on "file" chose "new FA503", a blank, flat factory preset, I thoughy.. With that, when you look at magnitude window, no filters selected, no measurements or any change performed yet, I get the same displayed curve in every channel, right from the bat: a blue curve with a +2 db peak centered at 30hz spanning 15-60hz, and another one, +3,5 to 4 db slightly asymetric starting at 2KHz, and peaking at 10Khz, almost returning to base line at 20Khz. There is also a very slight peak, like 1 db at 80 something Hz till 110 hz. Biquads available: 15 in every channel, all channels displaying magnitude curves exactly the same without touching anything, no filters appear on yhe display, no pre EQ selected, just opening a new project... What the heck is happening here..?
Already did some of the proposed. Fired REW with the same setup as before, and measured my right module on the right speaker and got the exact same curve as the left (just some very minor variation below 500hz, as this speaker doesn´t have a nearby boundary like the left one): a steady droop in level 2khz onwards, totalling 9,5-10 db at 20Khz mirroring exactly what I got on the other channel, that is, some 4,5-5 db down at 20Khz compared to my MF output. Same problems with REW too, getting solid 88-91db in reality ( checked with spl meter app, reads 90 db minimum) but Rew complaining levels were too low ( UMIK issue?). Anyway, even if I have some issue with the mic or REW are acting on me, its the same mic and settings in every measurement, both the modules and MF.
Checked filters and pre EQ in my HFD 4.95, firmware 1.41, everything is flat on the preset I am measuring. Have recentely redownloaded the firmware to them too, becaus e of remote control issues, as advised by Hypex support ( although they talked about vs 1.42, only vs 1.41 was available in their download page and that is what is running). So as both modules display the same behaviour, I then, decided to try and load a totally new project and see what happens: on "file" chose "new FA503", a blank, flat factory preset, I thoughy.. With that, when you look at magnitude window, no filters selected, no measurements or any change performed yet, I get the same displayed curve in every channel, right from the bat: a blue curve with a +2 db peak centered at 30hz spanning 15-60hz, and another one, +3,5 to 4 db slightly asymetric starting at 2KHz, and peaking at 10Khz, almost returning to base line at 20Khz. There is also a very slight peak, like 1 db at 80 something Hz till 110 hz. Biquads available: 15 in every channel, all channels displaying magnitude curves exactly the same without touching anything, no filters appear on yhe display, no pre EQ selected, just opening a new project... What the heck is happening here..?
Yes, good suggestion ...but my Toshiba laptop doesn´t have a spdif out, so I have no easy way to output a digital sign from my laptop into the Hypexs, ( don´t own a USB converter either..) nor do I really know how to output a digital sweep into the hypex spdif and capture it on REw or HFD for now..
I did eliminate one of the problems, and sorry for spreading some confusion here, noob mistake
: the blue superimposed curve I was alluding before, in every channel after starting the project from new is actually the superimposed UMIK correction curve that I had loaded into HFD, and was undully selected in tools to be overlayed, sorry about that ( never had seen it before)..
But the other issues remain... Could the preamp ( its a valve hybrid thing, using two ECC 88 I think, Telefunkens in my case) in my MF somehow be rising the response from 2KHk on up that much? Or my laptop headphone out measure so different attached to diferent devices? My speakers were never ever been described as dull, and 10 db down at 18Khz/20Khz on the analog input of the Hypexs is clearly wrong, and mirrors my listening experience ( I though my earing had taken another shot..) whereas 4 db down at 20Khz on the MF sounds about right, a slightly bright speaker, and what you would expect given the specs.
I think I am gonna erase the whole thing in both units, reset, ,reinstall REW, HFD and firmware again and see if that changes something along the way..
I did eliminate one of the problems, and sorry for spreading some confusion here, noob mistake
: the blue superimposed curve I was alluding before, in every channel after starting the project from new is actually the superimposed UMIK correction curve that I had loaded into HFD, and was undully selected in tools to be overlayed, sorry about that ( never had seen it before)..But the other issues remain... Could the preamp ( its a valve hybrid thing, using two ECC 88 I think, Telefunkens in my case) in my MF somehow be rising the response from 2KHk on up that much? Or my laptop headphone out measure so different attached to diferent devices? My speakers were never ever been described as dull, and 10 db down at 18Khz/20Khz on the analog input of the Hypexs is clearly wrong, and mirrors my listening experience ( I though my earing had taken another shot..) whereas 4 db down at 20Khz on the MF sounds about right, a slightly bright speaker, and what you would expect given the specs.
I think I am gonna erase the whole thing in both units, reset, ,reinstall REW, HFD and firmware again and see if that changes something along the way..
