A little history and tutorial in Equalization...
I realized the other night that there is a lot more I wanted to share then time would permit on a work night...
...So before I continue with the speaker comparison, I want to share some of the discoveries I have had trying different DSP tools in JRiver. (Do not worry, I am off tonight, so I will get to the comparison). I think that it is important for others who are following this thread, maybe considering building a full-range line array, some of the different methods available to them, as it is a well known fact that a full-range array is going to need some Equalization. Here is my history...
Adventures with Analog:
I have used both a graphic equalizer (Audio Source and SoundCraft) and a two band parametric equalizer (SAE 1800) in the past. In both cases, I found I could flatten the Frequency Response, but at the cost of wrecking the 3D aspect of soundstage and adding either "grain" or "hiss" to the sound. I got a flatter response with the graphic equalizer then I did with the parametric, but the parametric sounded better. (Maybe less is more
) But in all cases the compromise in soundstage was unacceptable. I always believed the above mentioned side effects where due to the age and quality of the components I was using...
A Hope for Digital:
My first digital experiences were using the graphic equalizers built into Foobar and the Realtek Audio Suite. FooBar only went down to 55 Hz (why?) but it was better then the Realtek which added "grain". Foobar was limited to music, where as the Realtek worked on everything, including streaming and games. The soundstage was held in tact a little better over the analog counterparts, but I was still not happy with the perceivable tiny drop in quality. Yet there seemed to be potential here, maybe a better DSP chip, like a MiniSharc, placed before the DAC...
A turning point: JRiver...
Then Wesayso introduced me to JRiver... Now the rubber hits the road...
If You have not experimented much with Digital Equalization, Cross-overs, FIR filters, etc, I would encourage You to try JRiver. They have a 30 day trail and You can learn a lot about how DSP works, even if You plan on going with Behringer, MiniDSP or a Najda pre-amp. There was a lot I did not know. I could have spent a lot of money on expensive hardware to only have my audio be 2 seconds behind the picture on my screen due to something called "latency". (more on this below).
I would even add to the above statement that if You are into designing Active Analog cross-overs, JRiver would be a good investment for You... (more below )
I have used three kinds of Equalization with-in JRiver:
I started out using the Graphic Equalizer, as I already had the curve figured out from the Realtek. Much better. No added grain (only the slight grain inherent to the Realtek DAC). First time I have used an equalizer that kept a soundstage! We are moving in the right direction...
The Parametric EQ. set up:
Now we are really moving forward. Nice thing about JRiver's Digital Rarametric EQ is that it is very customizable. Frequency, Q and large gains and cuts can be programmed in, not to mention a lot of bands can be added. I could also set up cross-overs. Now I could finally set up the 80 Hz high pass to the arrays the way I wanted to. As the analog cross-over build into my plate amps is 12 dB/octave, I wanted to sync up the cross-over point phasing by using a 36 dB/octave on the arrays. Talk about a new found mid-range freedom in the arrays!
There was quite a few nights I sat in my room in awe at the dynamics and width of the soundstage. I could not believe I was still listening to my Cheap and Cheerfuls! To this day, I still use this set-up for Live Streaming, Skype, and Video Games.
The Final Frontier - FIR filters and convolution:
But it does get better, indeed it does! All of JRiver's (and most digital DSP) DSP use IIR filters, which means all the filters still suffer the phase shift that their analog counterparts suffer. But digital is so nice, because You can try different cross-over slopes and frequencies without having to buy new components. The trick is You need to make sure You have enough DAC outputs to support all of this. A 4-way system needs 8 DAC outputs, which most half-way decent motherboards provide with their on-board audio. This is way I recommend JRiver for those wanting to build active analog filters, as You can figure everything out first on your PC! Anyway, I regress...
JRiver has a "convolution" engine built in, which means You can create FIR based digital filters and plug them in. There are a lot of different software packages to use. I like to do things manually, so I am using rePhase. I started by putting in the settings I used in the Parametric EQ, but lo and behold, I got different measurements! As it turns out, the Phasing of filters DOES make a difference! Needless to say, I had to re-figure my EQ settings. At first, I continued using the 12 dB/octave low pass on the sub and the 36 dB/octave high pass on the arrays, except, I eliminated the phase shift. This was tricky with the subs, because I was stuck with the phase shift in the analog cross-over built into the amp. I overcome this by using the "Phase-Linearization" feature in rePhase. I also fed the subs a straight full-range signal, so only the amps cross-over would be in effect.
This set-up took everything up to a whole new level, esp in bass. This does not happen often, but this was one of them times when there was no doubt in my mind that the change I was hearing was definitely more "right" then anything I have heard in the past. You may have heard the expression "bass is supposed to be felt not heard", well that is the best way to describe this. And I go to enough of my daughter's band recitals to know that is the way it is. The other huge plus about this set-up, is that it has proven to me that a full-range array crossed at 80 Hz is indeed a very good target to pursue for this project. Freeing up the array of the sub-sonic stuff really does open up the mid-range and give more dynamic headroom. And with the subs, I can focus more on the felt aspect verses the blending...
Wesayso, I am sure You will be reading this at some point, if You get the chance, borrow a sub or two from a friend and try this, I (as well as many others) would be very interested in what it does to widening your soundstage on your arrays!
But it gets better... I was still getting a little overlap in the 60 to 120 Hz area with the 12 dB/octave on the subs, so I added a 24 dB/octave linear phase to the subs, yielding a 36 dB/octave overall low pass. There is still a little overlap addition going on, but upon listening, I can say that percussion is Magical, esp hand drums! I would go as far to say - REAL-LIKE! Brings me back to my college days when I used to perpetrate in large drumming circles with my jinbei. This is what it is all about.
I would like to try a brick wall filter on both the arrays and subs, to see how far this can go...
There is one compromise with this kind of digital filtering, and it can be a show stopper in some cases: LATENCY!
To run FIR filters, esp those with higher Taps, due to sample rate timings and the such, they "need" an extra second or two to process. As I am crossing-over at bass frequencies, I have to use higher Taps. I have also found that higher taps have a more open sound. Of course more taps equals more latency. I use 65536 taps/channel, which is pretty high, but that has the best sound stage. I could go higher, but I think that would be overkill. Less then that does not sound as good. I heard this could be a rePhase thing, and I have heard others using less Taps in other software while still keeping their sound stage together. I have not explored other software yet...
Latency is not a problem for music or audio only sources. It is not a problem for ripped or blu-ray movies played in JRiver. JRiver will adjust the video to sync with the audio and their are manual settings to fine tune this if necessary. Please note if You are playing back DVD on a Windows based PC, that windows will prevent JRiver from syncing the video. (Why windows built that limitation into the OS is well beyond me, Blu-ray does not have this limitation.) This is overcome by ripping the DVD's into MKV format. MKV format is completely loss less and loading up movies to watch using this format is fast!
Where Latency IS a problem is Streaming Live Video online, Skype, and Video Games, where instant audio feedback is a requirement. I hope one day that the Online Streaming limitation can be overcome, as that is all buffered anyway (this is more of an Adobe flashplayer issue). There is not too big of a loss with Skype and Video Games, as the quality of the sound is sub-par to begin with, yet I still prefer the IIR digital set-up over the analog set-up.
As You can tell, I have learned a lot. And what I have learned here, will help in both this project and my brother's project.
My brother needs "plug and play" so...
Analog and IIR cross-overs work best when the phasing is "naturally" lined up at the cross-over point. I tried running 12 dB/octave on both the subs and the arrays and reversing the phasing on the subs to "in theory" sync of the phasing, but that did not work as well as running 12 dB with 36 dB. Not sure if it is because there was more overlap or not, but there where more frequencies cancelling each over out with the 12 dB/octave filters on both subs and arrays. I can see why 24 dB/octave filters are so popular.
On very quest for the largest, most awesome three-dimensional Soundstage...
Soundstage seems to be a very delicate thing. Phasing, room reflections, out of time inter-driver interference can totally wreck it. Also, a driver needs to run with-in its limits to not "mask" out these delicate nuances. (I will be addressing this more during the comparison).
Time a-ticking away here... Next post: Stereo, Dolby Surround, 5.1/7.1
I realized the other night that there is a lot more I wanted to share then time would permit on a work night...
...So before I continue with the speaker comparison, I want to share some of the discoveries I have had trying different DSP tools in JRiver. (Do not worry, I am off tonight, so I will get to the comparison). I think that it is important for others who are following this thread, maybe considering building a full-range line array, some of the different methods available to them, as it is a well known fact that a full-range array is going to need some Equalization. Here is my history...
Adventures with Analog:
I have used both a graphic equalizer (Audio Source and SoundCraft) and a two band parametric equalizer (SAE 1800) in the past. In both cases, I found I could flatten the Frequency Response, but at the cost of wrecking the 3D aspect of soundstage and adding either "grain" or "hiss" to the sound. I got a flatter response with the graphic equalizer then I did with the parametric, but the parametric sounded better. (Maybe less is more
) But in all cases the compromise in soundstage was unacceptable. I always believed the above mentioned side effects where due to the age and quality of the components I was using... A Hope for Digital:
My first digital experiences were using the graphic equalizers built into Foobar and the Realtek Audio Suite. FooBar only went down to 55 Hz (why?) but it was better then the Realtek which added "grain". Foobar was limited to music, where as the Realtek worked on everything, including streaming and games. The soundstage was held in tact a little better over the analog counterparts, but I was still not happy with the perceivable tiny drop in quality. Yet there seemed to be potential here, maybe a better DSP chip, like a MiniSharc, placed before the DAC...
A turning point: JRiver...
Then Wesayso introduced me to JRiver... Now the rubber hits the road...
If You have not experimented much with Digital Equalization, Cross-overs, FIR filters, etc, I would encourage You to try JRiver. They have a 30 day trail and You can learn a lot about how DSP works, even if You plan on going with Behringer, MiniDSP or a Najda pre-amp. There was a lot I did not know. I could have spent a lot of money on expensive hardware to only have my audio be 2 seconds behind the picture on my screen due to something called "latency". (more on this below).
I would even add to the above statement that if You are into designing Active Analog cross-overs, JRiver would be a good investment for You... (more below
)I have used three kinds of Equalization with-in JRiver:
I started out using the Graphic Equalizer, as I already had the curve figured out from the Realtek. Much better. No added grain (only the slight grain inherent to the Realtek DAC). First time I have used an equalizer that kept a soundstage! We are moving in the right direction...
The Parametric EQ. set up:
Now we are really moving forward. Nice thing about JRiver's Digital Rarametric EQ is that it is very customizable. Frequency, Q and large gains and cuts can be programmed in, not to mention a lot of bands can be added. I could also set up cross-overs. Now I could finally set up the 80 Hz high pass to the arrays the way I wanted to. As the analog cross-over build into my plate amps is 12 dB/octave, I wanted to sync up the cross-over point phasing by using a 36 dB/octave on the arrays. Talk about a new found mid-range freedom in the arrays!
There was quite a few nights I sat in my room in awe at the dynamics and width of the soundstage. I could not believe I was still listening to my Cheap and Cheerfuls! To this day, I still use this set-up for Live Streaming, Skype, and Video Games. The Final Frontier - FIR filters and convolution:
But it does get better, indeed it does! All of JRiver's (and most digital DSP) DSP use IIR filters, which means all the filters still suffer the phase shift that their analog counterparts suffer. But digital is so nice, because You can try different cross-over slopes and frequencies without having to buy new components. The trick is You need to make sure You have enough DAC outputs to support all of this. A 4-way system needs 8 DAC outputs, which most half-way decent motherboards provide with their on-board audio. This is way I recommend JRiver for those wanting to build active analog filters, as You can figure everything out first on your PC! Anyway, I regress...
JRiver has a "convolution" engine built in, which means You can create FIR based digital filters and plug them in. There are a lot of different software packages to use. I like to do things manually, so I am using rePhase. I started by putting in the settings I used in the Parametric EQ, but lo and behold, I got different measurements! As it turns out, the Phasing of filters DOES make a difference! Needless to say, I had to re-figure my EQ settings. At first, I continued using the 12 dB/octave low pass on the sub and the 36 dB/octave high pass on the arrays, except, I eliminated the phase shift. This was tricky with the subs, because I was stuck with the phase shift in the analog cross-over built into the amp. I overcome this by using the "Phase-Linearization" feature in rePhase. I also fed the subs a straight full-range signal, so only the amps cross-over would be in effect.
This set-up took everything up to a whole new level, esp in bass. This does not happen often, but this was one of them times when there was no doubt in my mind that the change I was hearing was definitely more "right" then anything I have heard in the past. You may have heard the expression "bass is supposed to be felt not heard", well that is the best way to describe this. And I go to enough of my daughter's band recitals to know that is the way it is. The other huge plus about this set-up, is that it has proven to me that a full-range array crossed at 80 Hz is indeed a very good target to pursue for this project. Freeing up the array of the sub-sonic stuff really does open up the mid-range and give more dynamic headroom. And with the subs, I can focus more on the felt aspect verses the blending...
Wesayso, I am sure You will be reading this at some point, if You get the chance, borrow a sub or two from a friend and try this, I (as well as many others) would be very interested in what it does to widening your soundstage on your arrays!
But it gets better... I was still getting a little overlap in the 60 to 120 Hz area with the 12 dB/octave on the subs, so I added a 24 dB/octave linear phase to the subs, yielding a 36 dB/octave overall low pass. There is still a little overlap addition going on, but upon listening, I can say that percussion is Magical, esp hand drums! I would go as far to say - REAL-LIKE! Brings me back to my college days when I used to perpetrate in large drumming circles with my jinbei. This is what it is all about.
I would like to try a brick wall filter on both the arrays and subs, to see how far this can go...
There is one compromise with this kind of digital filtering, and it can be a show stopper in some cases: LATENCY!
To run FIR filters, esp those with higher Taps, due to sample rate timings and the such, they "need" an extra second or two to process. As I am crossing-over at bass frequencies, I have to use higher Taps. I have also found that higher taps have a more open sound. Of course more taps equals more latency. I use 65536 taps/channel, which is pretty high, but that has the best sound stage. I could go higher, but I think that would be overkill. Less then that does not sound as good. I heard this could be a rePhase thing, and I have heard others using less Taps in other software while still keeping their sound stage together. I have not explored other software yet...
Latency is not a problem for music or audio only sources. It is not a problem for ripped or blu-ray movies played in JRiver. JRiver will adjust the video to sync with the audio and their are manual settings to fine tune this if necessary. Please note if You are playing back DVD on a Windows based PC, that windows will prevent JRiver from syncing the video. (Why windows built that limitation into the OS is well beyond me, Blu-ray does not have this limitation.) This is overcome by ripping the DVD's into MKV format. MKV format is completely loss less and loading up movies to watch using this format is fast!
Where Latency IS a problem is Streaming Live Video online, Skype, and Video Games, where instant audio feedback is a requirement. I hope one day that the Online Streaming limitation can be overcome, as that is all buffered anyway (this is more of an Adobe flashplayer issue). There is not too big of a loss with Skype and Video Games, as the quality of the sound is sub-par to begin with, yet I still prefer the IIR digital set-up over the analog set-up.
As You can tell, I have learned a lot. And what I have learned here, will help in both this project and my brother's project.
My brother needs "plug and play" so...
Analog and IIR cross-overs work best when the phasing is "naturally" lined up at the cross-over point. I tried running 12 dB/octave on both the subs and the arrays and reversing the phasing on the subs to "in theory" sync of the phasing, but that did not work as well as running 12 dB with 36 dB. Not sure if it is because there was more overlap or not, but there where more frequencies cancelling each over out with the 12 dB/octave filters on both subs and arrays. I can see why 24 dB/octave filters are so popular.
On very quest for the largest, most awesome three-dimensional Soundstage...
Soundstage seems to be a very delicate thing. Phasing, room reflections, out of time inter-driver interference can totally wreck it. Also, a driver needs to run with-in its limits to not "mask" out these delicate nuances. (I will be addressing this more during the comparison).
Time a-ticking away here... Next post: Stereo, Dolby Surround, 5.1/7.1
Last edited:
Wesayso also got my to like JRiver... after some growing pains on my system.
On the other hand, I tried the graphic EQ in JRiver, and didn't like what it was doing. Just keeping it flat and clicking it on and off, it took the life away when that EQ was turned on...
I have had much better luck doing everything with the Parametric EQs. As you've mentioned, more precise and more control.
If you keep the DRC (or RePhase in your context) computations to a minimum, it is possible to use it without so much latency. I had a setting that was very intensive at one point, and yes, it was adding a lot of latency. But, I've been using some more gentle settings, and I can't hear the latency anymore.
Right now, DRC is limited to the Left and Right speakers only. I have yet to implement DRC to all my channels (5.1) but it looks like a daunting task, using a config file.
My DAC is a studio interface that has 8 outputs. USB. Works well enough.
On the other hand, I tried the graphic EQ in JRiver, and didn't like what it was doing. Just keeping it flat and clicking it on and off, it took the life away when that EQ was turned on...
I have had much better luck doing everything with the Parametric EQs. As you've mentioned, more precise and more control.
If you keep the DRC (or RePhase in your context) computations to a minimum, it is possible to use it without so much latency. I had a setting that was very intensive at one point, and yes, it was adding a lot of latency. But, I've been using some more gentle settings, and I can't hear the latency anymore.
Right now, DRC is limited to the Left and Right speakers only. I have yet to implement DRC to all my channels (5.1) but it looks like a daunting task, using a config file.
My DAC is a studio interface that has 8 outputs. USB. Works well enough.
You're welcome.... I can honestly recommend that jewel to anyone interested in audio/video playback.
Any wider soundstage at my end would be insane
. Not saying I won't experiment with adding subs. I'm sure it will happen in time.Did you by any chance record an impulse in REW? would love to see it.
Just tried this on off click on the EQ, no worries here... seems strange to me. Did you test with someone else clicking it on and off while you are listening and unaware of it's setting? The power of the mind is a strong thing. That way you'd know if it really affects things. Be sure to have the first slider on zero too (gain).
Last edited:
Stereo, Dolby Surround, 5.1/7.1...
Since DVD has come out, I have been wondering why most movies on them sound so flat. What I am referring to here is how some movies seem to be more three-dimensional in sound then others. Being I had Laser Disc, I was used to movies surrounding me, even being run in stereo. The effect was not as discreet as it was in Dolby Surround/Pro logic, but it was still there to a certain degree. Most 5.1 Dolby Digital soundtracks sound pretty flat compared to their 2.0 Dolby Surround counterparts. I automatically assumed that the "matrixing" used in Dolby Surround would be "encoded in" during a 5.1 mix down. As I started digging into this, I found out that is not the case at all. The surround channels are basically added to the fronts. I have been looking into finding a plug-in that may include some "matrixing" processing for 2 channel playback, but I have only found stuff for headphones, which I can not see working too good for a speaker layout. From what I have read, this is being addressed with the new Dolby Atmos Systems, where the "object based" audio stream will be mixed down and "matrixed" to the playback system. But it would appear that the best way to playback 5.1 and 7.1 soundtracks, is with at least one pair of back speakers. So needless to say, I may need to consider building back speakers as part of this "All aspiring array project".
I would even consider doing up nice cabinets for the NSB arrays I have now, and using them in the back. The setback with them, is the three damaged drivers. I can fix one of them, but ideally I would prefer to replace them IF I can find replacements. Anyone?
The reason I bring this up is because I have recently acquired a Blu-ray player. What I can say, is it is as great as they boast. The sound is comparable to Laser Disc. In fact, compared to DVD, the best way to describe it, is that there are sounds in-between the sounds (in both tonality and space) when comparing Blu-ray to DVD. (I have some movies in both formats). Blu-ray has a much wider and deeper soundstage in the front and the overall sound is dynamic, smooth and relaxed compared to its DVD counter part. There are more gradations in the tones also. The bummer is, the stereo playback does not have the "sound in the back" like Laser Disc had, due to the mix down issue above. Hence my desire for back arrays! Like I say, the frontal stereo image is amazing wide and deep, and this is a aspect I enjoy very much with DTS Master and Dolby True HD Audio.
The funny thing, if the back channels where Matrixed, I would be in no hurry for back speakers, as the arrays throw an amazing "surround effect" on the older Dolby Stereo Surround soundtracks. Maybe I should combine the surround channels off 5.1/7.1 mixes into one mono channel, duplicate them, then phase one 90 degrees positive and add that to the left front channel; and the other 90 degrees negative and add that to the right front channel. (I could just reverse the phase on one channel also). I think that is how they matrixed Dolby Surround, but I do not know for sure. I am sure they phased degrees between 0 and 90 to "pan" from front to back.
Any thoughts?
Since DVD has come out, I have been wondering why most movies on them sound so flat. What I am referring to here is how some movies seem to be more three-dimensional in sound then others. Being I had Laser Disc, I was used to movies surrounding me, even being run in stereo. The effect was not as discreet as it was in Dolby Surround/Pro logic, but it was still there to a certain degree. Most 5.1 Dolby Digital soundtracks sound pretty flat compared to their 2.0 Dolby Surround counterparts. I automatically assumed that the "matrixing" used in Dolby Surround would be "encoded in" during a 5.1 mix down. As I started digging into this, I found out that is not the case at all. The surround channels are basically added to the fronts. I have been looking into finding a plug-in that may include some "matrixing" processing for 2 channel playback, but I have only found stuff for headphones, which I can not see working too good for a speaker layout. From what I have read, this is being addressed with the new Dolby Atmos Systems, where the "object based" audio stream will be mixed down and "matrixed" to the playback system. But it would appear that the best way to playback 5.1 and 7.1 soundtracks, is with at least one pair of back speakers. So needless to say, I may need to consider building back speakers as part of this "All aspiring array project".
I would even consider doing up nice cabinets for the NSB arrays I have now, and using them in the back. The setback with them, is the three damaged drivers.
I can fix one of them, but ideally I would prefer to replace them IF I can find replacements. Anyone?The reason I bring this up is because I have recently acquired a Blu-ray player. What I can say, is it is as great as they boast. The sound is comparable to Laser Disc. In fact, compared to DVD, the best way to describe it, is that there are sounds in-between the sounds (in both tonality and space) when comparing Blu-ray to DVD. (I have some movies in both formats). Blu-ray has a much wider and deeper soundstage in the front and the overall sound is dynamic, smooth and relaxed compared to its DVD counter part. There are more gradations in the tones also. The bummer is, the stereo playback does not have the "sound in the back" like Laser Disc had, due to the mix down issue above. Hence my desire for back arrays!
Like I say, the frontal stereo image is amazing wide and deep, and this is a aspect I enjoy very much with DTS Master and Dolby True HD Audio.The funny thing, if the back channels where Matrixed, I would be in no hurry for back speakers, as the arrays throw an amazing "surround effect" on the older Dolby Stereo Surround soundtracks. Maybe I should combine the surround channels off 5.1/7.1 mixes into one mono channel, duplicate them, then phase one 90 degrees positive and add that to the left front channel; and the other 90 degrees negative and add that to the right front channel. (I could just reverse the phase on one channel also). I think that is how they matrixed Dolby Surround, but I do not know for sure. I am sure they phased degrees between 0 and 90 to "pan" from front to back.
Any thoughts?
In response... EQ phasing...
I forgot to mention this in the above post with EQ. I liked the graphic EQ in JRiver better then the other graphic EQs I have used, but I did not like it as much as the Parametric EQ. I have a theory on this, and it plays into the phasing relationships I am talking about. With a graphic equalizer, we are bringing are creating our target response by moving each frequency band up or down, each frequency band has a set Q and equal spacing between each band. Each band will also induce a phase ripple. When you move two bands next to each other, up, for instance, the phase ripple between those two bands will interact with each other, possibly causing some cancellation somewhere. This will create either a desirable effect or not. With the Parametric EQ, we can possibly make fewer adjustments by varying the Q of each customized frequency band to hit our targeted curve, therefore having less phasing ripples. Less is more, perhaps. This is just a theory, and I have not done the ABX test You mention above. Something I would like to try, is create a new FIR filter that is just the cross-over, and use it with the IIR Parametric EQ curve I like and compair that to the FIR filter I have now. They say phasing on EQ filters does not matter that much, we will see. As I mentioned in the post above, the IIR EQ and the FIR EQ required different frequencies and settings to get the target curve.
I will need to look into REW. I am really enjoying TrueRTA, as I am finding different software has way of helping me achieve something I may have missed with Holm.
Wesayso, I like how well You document your build with the measurements, and I think it could help here too, even if I come across something the measurements can not show.
Now on to the comparison, (golly, it is almost my bedtime!)
I forgot to mention this in the above post with EQ. I liked the graphic EQ in JRiver better then the other graphic EQs I have used, but I did not like it as much as the Parametric EQ. I have a theory on this, and it plays into the phasing relationships I am talking about. With a graphic equalizer, we are bringing are creating our target response by moving each frequency band up or down, each frequency band has a set Q and equal spacing between each band. Each band will also induce a phase ripple. When you move two bands next to each other, up, for instance, the phase ripple between those two bands will interact with each other, possibly causing some cancellation somewhere. This will create either a desirable effect or not. With the Parametric EQ, we can possibly make fewer adjustments by varying the Q of each customized frequency band to hit our targeted curve, therefore having less phasing ripples. Less is more, perhaps. This is just a theory, and I have not done the ABX test You mention above. Something I would like to try, is create a new FIR filter that is just the cross-over, and use it with the IIR Parametric EQ curve I like and compair that to the FIR filter I have now. They say phasing on EQ filters does not matter that much, we will see. As I mentioned in the post above, the IIR EQ and the FIR EQ required different frequencies and settings to get the target curve.
I will need to look into REW. I am really enjoying TrueRTA, as I am finding different software has way of helping me achieve something I may have missed with Holm.
Wesayso, I like how well You document your build with the measurements, and I think it could help here too, even if I come across something the measurements can not show.
Now on to the comparison, (golly, it is almost my bedtime!)
A quick question for the next round, as today's program seems to be filled already.
When playing Blu-ray trough JRiver to 2 channel stereo with JRSS mixing and "Detect stereo sources in Surround" selected, does that make a difference? Usually there are multiple tracks on the disk but I always use the JRSS mixing from JRiver of the multi-channel tracks, playing from MKV, even for Blu-ray.
When playing Blu-ray trough JRiver to 2 channel stereo with JRSS mixing and "Detect stereo sources in Surround" selected, does that make a difference? Usually there are multiple tracks on the disk but I always use the JRSS mixing from JRiver of the multi-channel tracks, playing from MKV, even for Blu-ray.
I forgot to mention this in the above post with EQ. I liked the graphic EQ in JRiver better then the other graphic EQs I have used, but I did not like it as much as the Parametric EQ. I have a theory on this, and it plays into the phasing relationships I am talking about. With a graphic equalizer, we are bringing are creating our target response by moving each frequency band up or down, each frequency band has a set Q and equal spacing between each band. Each band will also induce a phase ripple. When you move two bands next to each other, up, for instance, the phase ripple between those two bands will interact with each other, possibly causing some cancellation somewhere. This will create either a desirable effect or not. With the Parametric EQ, we can possibly make fewer adjustments by varying the Q of each customized frequency band to hit our targeted curve, therefore having less phasing ripples. Less is more, perhaps. This is just a theory, and I have not done the ABX test You mention above. Something I would like to try, is create a new FIR filter that is just the cross-over, and use it with the IIR Parametric EQ curve I like and compair that to the FIR filter I have now. They say phasing on EQ filters does not matter that much, we will see. As I mentioned in the post above, the IIR EQ and the FIR EQ required different frequencies and settings to get the target curve.
I will need to look into REW. I am really enjoying TrueRTA, as I am finding different software has way of helping me achieve something I may have missed with Holm.
Wesayso, I like how well You document your build with the measurements, and I think it could help here too, even if I come across something the measurements can not show.
Now on to the comparison, (golly, it is almost my bedtime!)
I agree with you on the benefits of PEQ. It's almost all I use beside my DRC-FIR filters. But after setting an EQ curve with the FIR filters I have no problem using Graphic EQ for a small cut here or there. It does indeed influence phase, all EQ apart from linear phase EQ does. I use it to fine tune my "House Curve" for lack of a better word. Usually my next round of measurements would include the cuts (which are between 0 to 2 dB at most) in the next target curve. But simply clicking EQ on or off does nothing negative for/in my experience. That's what I was addressing here.
Sorry to hear you lost a post. Now don't stay up too late, that's what I usually do when I want something off my mind
. Can't sleep anyway if it's still bugging me.
Comparison, second try...
Lost my flow, no fun re-typing stuff, I am a slow writer to begin with...
I will do my best to keep my energies high on this one. I can say this, Logitech done a big mistake with the adding a "back" button to the new model of this mouse. I Love the mouse, hate the new "back" button, too easy to hit doing "normal stuff", time to take it apart and remove the button.
Anyway, maybe this one will be better.
So here we have it, the comparison between the Pioneer Vintage Multi-way with passive cross-overs, the Full-range Avebury and the Array and Subs.
Set Up:
I set up the Pioneers on chairs about 18" off the floor, this put the tweeters at ear height and evened out the bass a lot. I tried both IIR and FIR EQ and in this particular case did not have a preference. I set up the EQ to produce a flat response from 60 Hz to 22 kHz. In retrospect, I wish I would have done a 10 dB treble roll off starting at 1 kHz. That is something I would like to revisit in the future. I did not want to push the low bass on this system, as I done that with these speakers growing up, and it did not go well! You know what teenage hormones can do! Instead, I tried these with the sub woofers, rolling the Pioneers at 36 dB/octave at 80 Hz and the subs coming in at 80 Hz at 12 dB/octave. I found this to be a bit much, too much overlap. (This was before I tried 36 dB/octave on the subs with the arrays) Again, I would like to revisit this. But I can say, for the sake of this comparison, I enjoyed these speakers without the subs. I really wanted to explore the mid-range and sound stage.
Avebury was setup with a 48 dB/octave low cut at 33 Hz and EQed flat from 33 to 800 Hz and a steady 10 dB roll off out to 20 kHz. I also created a second curve that had a 5 dB roll off on the bass starting at 200 Hz on down to 33 Hz. I definitely liked the FIR filter best on this system, esp in the bass. With the IIR, the bass did not have as much "impact" as the FIR. With the FIR, it sounded similar to the "unprocessed" version, which is what I was after. I main goal of the low cut was to get rid of the sub sonic stuff Avebury can not do away.
The Arrays and Subs where setup with a 36 dB/octave high pass, with the subs on a 12 dB/octave low pass, so this is one step down from the best of the best. The EQ curve is flat from 80 Hz to 1000 Hz and a steady 10 dB roll off to about 12 kHz. I will be playing with the 15 thru 20 kHz area more in the future, I can not hear much that high, but I am curious.
Bear in mind, when I say "flat" there are peaks and dips throughout this range. The goal was to get it close with-in reason.
Sound:
When it comes to tone, the Pioneers are a very likable speaker. Maybe it is because I grew up with them, but I found myself getting warm fuzzes inside for awhile. The sound is warm, clear, full and present. There was enough bass that I did not feel sub deprived, even though there was no felt sensation at all. The highs were extended and resolved, which I have not heard for awhile. All the tone was there in a very instant gratification sort of way. The presentation is very direct, up front and center, but there was no soundstage at all! For where ever reason, 3D was not happening here. (I have a theory). Also, I found these did not do loud well. They did not seem to handle higher power too well, oddly enough, my 40 watt chip amps were a good match for these. I also found louder levels to be pretty fatiguing, the horn tweeters got to be a bit much. These are pretty interesting speakers, they welcome You in, then that is it, the tone is there but there is nothing to hold your attention. But when listening to all three systems on the short term, I found myself gravitating to the Pioneer CS-605s.
With Avebury, I can say a new level was reached with these. I know in the past, I have mentioned, esp when watching movies, that the main dialog seems to get lost, well when setting these up, they did have a lot of low end, with peaks of 10 to 15 dB above the mid-range! This is due in part to their placement in my smallish room. I had to apply significant EQ below 200 Hz to get these "flat". Boy did these come alive with the bass tamed a bit. Another huge plus with this, is that now these have a bit more power handling. These were not quite as warm sounding compared to the Pioneers, but they had nice low end extension with a little bit of sub sonic sensation at 40 Hz. The presentation was more laid back and background compared to the Pioneers. The sound stage was deep and strong between the speakers, not as wide as the arrays. I actually preferred this system with the bass rolled off 5 dB. I missed the 40 Hz "kick" but again the "kick" got to be a bit much during longer listening sessions. This system seemed more relaxed with the bass roll-off. It could be that I was liking the felt mid-range of the arrays, that the bass of the Avebury's was just not cutting it anymore! At louder levels, I felt this system still struggled a bit having to reproduce so much sound. I still would like to try this system with a more powerful amp, I really do not think the chip amps come close to bringing out the potential of these speakers. I wonder if I can get a more refined high out of these. I find the highs a little rough and "heavy". It is going to be fun pairing these with the drivers for the All Aspiring Array driver blowouts.
Do I need to say much about the arrays? They are still my favorites, with the biggest sound stage. It is one of the few systems I have worked with that sounds good at all volume levels. There is something to be said of the "relaxed" nature of having 15 drivers do the work instead of 1 or 4. The Pioneer and Avebury really sounded stressed even at levels slightly above moderate. I was able to play all the systems at the same levels, as I had TrueRTA on, showing the peak and average dB levels in the room. (These levels may not be dailed in for true SPL, but relatively speaking I was able to listen to each system at the same dB levels this mic read). I would say the weakness with the array system, especially now with how good the bass is set up, is the highs. In fact, overall, highs are not happening on any of these systems. The Pioneers seem the most resolving, but if they are fatiguing it does no good, but this maybe due to the lack of the 10 dB roll off. Highs have always been a challenge for me. This could also be an amp issue, as it does show up in all three of these different systems.
Bottom line here, the speaker that wins the driver blowouts for this project is more then likely the one that has the most listenable highs to me. Still prefer to go full-range (80 Hz on up), but I have not ruled out going the tweeter route, now knowing the power of FIR filters.
I am sure I have forgotten a bunch of stuff. That will come at a later date.
It is time for me to go to bed!
Good Day,
Allen
Lost my flow, no fun re-typing stuff, I am a slow writer to begin with...
I will do my best to keep my energies high on this one. I can say this, Logitech done a big mistake with the adding a "back" button to the new model of this mouse. I Love the mouse, hate the new "back" button, too easy to hit doing "normal stuff", time to take it apart and remove the button.
Anyway, maybe this one will be better.
So here we have it, the comparison between the Pioneer Vintage Multi-way with passive cross-overs, the Full-range Avebury and the Array and Subs.
Set Up:
I set up the Pioneers on chairs about 18" off the floor, this put the tweeters at ear height and evened out the bass a lot. I tried both IIR and FIR EQ and in this particular case did not have a preference. I set up the EQ to produce a flat response from 60 Hz to 22 kHz. In retrospect, I wish I would have done a 10 dB treble roll off starting at 1 kHz. That is something I would like to revisit in the future. I did not want to push the low bass on this system, as I done that with these speakers growing up, and it did not go well!
You know what teenage hormones can do! Instead, I tried these with the sub woofers, rolling the Pioneers at 36 dB/octave at 80 Hz and the subs coming in at 80 Hz at 12 dB/octave. I found this to be a bit much, too much overlap. (This was before I tried 36 dB/octave on the subs with the arrays) Again, I would like to revisit this. But I can say, for the sake of this comparison, I enjoyed these speakers without the subs. I really wanted to explore the mid-range and sound stage. Avebury was setup with a 48 dB/octave low cut at 33 Hz and EQed flat from 33 to 800 Hz and a steady 10 dB roll off out to 20 kHz. I also created a second curve that had a 5 dB roll off on the bass starting at 200 Hz on down to 33 Hz. I definitely liked the FIR filter best on this system, esp in the bass. With the IIR, the bass did not have as much "impact" as the FIR. With the FIR, it sounded similar to the "unprocessed" version, which is what I was after. I main goal of the low cut was to get rid of the sub sonic stuff Avebury can not do away.
The Arrays and Subs where setup with a 36 dB/octave high pass, with the subs on a 12 dB/octave low pass, so this is one step down from the best of the best. The EQ curve is flat from 80 Hz to 1000 Hz and a steady 10 dB roll off to about 12 kHz. I will be playing with the 15 thru 20 kHz area more in the future, I can not hear much that high, but I am curious.
Bear in mind, when I say "flat" there are peaks and dips throughout this range. The goal was to get it close with-in reason.
Sound:
When it comes to tone, the Pioneers are a very likable speaker. Maybe it is because I grew up with them, but I found myself getting warm fuzzes inside for awhile. The sound is warm, clear, full and present. There was enough bass that I did not feel sub deprived, even though there was no felt sensation at all. The highs were extended and resolved, which I have not heard for awhile. All the tone was there in a very instant gratification sort of way. The presentation is very direct, up front and center, but there was no soundstage at all! For where ever reason, 3D was not happening here. (I have a theory). Also, I found these did not do loud well. They did not seem to handle higher power too well, oddly enough, my 40 watt chip amps were a good match for these. I also found louder levels to be pretty fatiguing, the horn tweeters got to be a bit much. These are pretty interesting speakers, they welcome You in, then that is it, the tone is there but there is nothing to hold your attention. But when listening to all three systems on the short term, I found myself gravitating to the Pioneer CS-605s.
With Avebury, I can say a new level was reached with these. I know in the past, I have mentioned, esp when watching movies, that the main dialog seems to get lost, well when setting these up, they did have a lot of low end, with peaks of 10 to 15 dB above the mid-range! This is due in part to their placement in my smallish room. I had to apply significant EQ below 200 Hz to get these "flat". Boy did these come alive with the bass tamed a bit. Another huge plus with this, is that now these have a bit more power handling. These were not quite as warm sounding compared to the Pioneers, but they had nice low end extension with a little bit of sub sonic sensation at 40 Hz. The presentation was more laid back and background compared to the Pioneers. The sound stage was deep and strong between the speakers, not as wide as the arrays. I actually preferred this system with the bass rolled off 5 dB. I missed the 40 Hz "kick" but again the "kick" got to be a bit much during longer listening sessions. This system seemed more relaxed with the bass roll-off. It could be that I was liking the felt mid-range of the arrays, that the bass of the Avebury's was just not cutting it anymore! At louder levels, I felt this system still struggled a bit having to reproduce so much sound. I still would like to try this system with a more powerful amp, I really do not think the chip amps come close to bringing out the potential of these speakers. I wonder if I can get a more refined high out of these. I find the highs a little rough and "heavy". It is going to be fun pairing these with the drivers for the All Aspiring Array driver blowouts.
Do I need to say much about the arrays? They are still my favorites, with the biggest sound stage. It is one of the few systems I have worked with that sounds good at all volume levels. There is something to be said of the "relaxed" nature of having 15 drivers do the work instead of 1 or 4. The Pioneer and Avebury really sounded stressed even at levels slightly above moderate. I was able to play all the systems at the same levels, as I had TrueRTA on, showing the peak and average dB levels in the room. (These levels may not be dailed in for true SPL, but relatively speaking I was able to listen to each system at the same dB levels this mic read). I would say the weakness with the array system, especially now with how good the bass is set up, is the highs. In fact, overall, highs are not happening on any of these systems. The Pioneers seem the most resolving, but if they are fatiguing it does no good, but this maybe due to the lack of the 10 dB roll off. Highs have always been a challenge for me. This could also be an amp issue, as it does show up in all three of these different systems.
Bottom line here, the speaker that wins the driver blowouts for this project is more then likely the one that has the most listenable highs to me. Still prefer to go full-range (80 Hz on up), but I have not ruled out going the tweeter route, now knowing the power of FIR filters.
I am sure I have forgotten a bunch of stuff. That will come at a later date.
It is time for me to go to bed!
Good Day,
Allen
Last edited:
I will do my best to keep my energies high on this one. I can say this, Logitech done a big mistake with the adding a "back" button to the new model of this mouse. I Love the mouse, hate the new "back" button, too easy to hit doing "normal stuff", time to take it apart and remove the button.
Good Day,
Allen
Cure it in software, with Logitech mouse software you can turn that button off.
Look at the second post in this thread: Unable to install Setpoint 6.32 on Windows 7 x64 - Logitech Forums
Choose the appropriate version, 32 or 64 bit and install Logitech's Setpoint software.
In Setpoint you can change the button behaviour, I have 2 side buttons and chose for them to do nothing. I also change my scroll wheel to middle mouse button instead of it's usual scroll wheel function. The scroll wheel will still work but if you happen to press the middle mouse button you won't get the annoying scroll function on screen.
A long time...
Wow, we left off with mouse issues....
I got much bigger fish to fry at the moment, one is getting used to windows 10. The upgrade really is quite buggy, so will be planning a fresh install when time permits, or when it drives me crazy enough.
Time is still a massive challenge for me. I no longer work mid-night shift (Thank God), and work evenings instead. Due to the pay cut I took, I must work more in my Care Giving business. There is not much for good paying jobs where I live, so the options I am left with now is relocate when I send my daughter off to college in three years, or start a second business. My Care Giving business is fulfilling, yet labor intensive, so it is not feasible to expand that without partnering with someone. Been reflecting as much as time permits to decide which direction to go...
In the audio world, as some of You already know, I got my First One v1.4 assembled and working. It really is all I had hoped it would be. The sound on both the Avebury and Cheap and Cheerful arrays have improved significantly. In terms of texture and details, when either system is EQed to the same response, they both perform equally well. The Cheap and Cheerful arrays have the detail of the Avebury, and the Avebury has the loudness and mid-range of the arrays. The preference for either system is now based more in terms of the speaker system's overall physical design, as in full-range single driver verses full-range line array with subs. I still prefer the massive soundstage of the Arrays, but the Aveburys are no slouch. In fact, the First One Amp is a very good match with that system.
You can read about my First One journey at:
Impressions: http://www.diyaudio.com/forums/vend...-mosfet-amplifier-module-303.html#post4746030
Assembly: http://www.diyaudio.com/forums/vend...-mosfet-amplifier-module-305.html#post4751690
All and all, I am very happy with First One v1.4, in fact, I feel pretty confident that it is the strongest piece in my signal chain. It reminds me of my McIntosh MC-150, but this feels new and improved. The sound stage is huge, and it is so nice to have this kind of power again, You know, the kind of Power where You can play it louder then You need. In fact, I have had to re-learn what loud is. It is no longer perceivable distortion, but instead a relatively black background with space, air, and timbre that is LOUD.
More to come...
Wow, we left off with mouse issues....
I got much bigger fish to fry at the moment, one is getting used to windows 10. The upgrade really is quite buggy, so will be planning a fresh install when time permits, or when it drives me crazy enough.
Time is still a massive challenge for me. I no longer work mid-night shift (Thank God), and work evenings instead. Due to the pay cut I took, I must work more in my Care Giving business. There is not much for good paying jobs where I live, so the options I am left with now is relocate when I send my daughter off to college in three years, or start a second business. My Care Giving business is fulfilling, yet labor intensive, so it is not feasible to expand that without partnering with someone. Been reflecting as much as time permits to decide which direction to go...
In the audio world, as some of You already know, I got my First One v1.4 assembled and working. It really is all I had hoped it would be.
The sound on both the Avebury and Cheap and Cheerful arrays have improved significantly. In terms of texture and details, when either system is EQed to the same response, they both perform equally well. The Cheap and Cheerful arrays have the detail of the Avebury, and the Avebury has the loudness and mid-range of the arrays. The preference for either system is now based more in terms of the speaker system's overall physical design, as in full-range single driver verses full-range line array with subs. I still prefer the massive soundstage of the Arrays, but the Aveburys are no slouch. In fact, the First One Amp is a very good match with that system. You can read about my First One journey at:
Impressions: http://www.diyaudio.com/forums/vend...-mosfet-amplifier-module-303.html#post4746030
Assembly: http://www.diyaudio.com/forums/vend...-mosfet-amplifier-module-305.html#post4751690
All and all, I am very happy with First One v1.4, in fact, I feel pretty confident that it is the strongest piece in my signal chain. It reminds me of my McIntosh MC-150, but this feels new and improved. The sound stage is huge, and it is so nice to have this kind of power again, You know, the kind of Power where You can play it louder then You need.
In fact, I have had to re-learn what loud is. It is no longer perceivable distortion, but instead a relatively black background with space, air, and timbre that is LOUD. More to come...
Attachments
Driver Comparison: part 1
As I mentioned above, I feel that the First One Amp is the strongest link in my audio chain. But what do I focus on next? DAC or the next speaker build.
Throughout my listening sessions with both Avebury and the Cheap and Cheerful, there is still something I have not been able to resolve: The mid-range shrillness. I hear it with my Chip Amps also, so I have ruled out the Amps being the issue. It is more frustrating with the First One, as it has a more definite sound, almost like noise or grain in the background (hence my "almost black background" comment above). But what is causing this? Speaker break-up or DAC?
I can say, with the Chip Amps, my Realtek DAC was almost tolerable, but with First One, it was ruled out right away. The mid-range issue above was indeed much worse with the on board DAC. The M-Audio is significantly better, but the issue is not solved...
It is more noticeable in the Arrays, unless I EQ Avebury to the same response as the Arrays, then it is there also. If the Avebury is run at their natural response, the issue is less apparent (more on that later).
So, I decided to give the drivers I acquired for this project a listen.
You know about the Scan Speak 10f and Fostex FF83wk. Some of You know I acquired a Bunch of Vifa TC9s. I also got a pair of Tectonics. I will upload the pictures of the drivers when time permits. Sorry I could not keep up with the thread. Night shift was starting to cause some health problems, which is why I could not keep up with this and other home repair projects. I have been spending the last seven months catching up and getting lots of sunshine to get my health back on track...
Anyway...
For the first comparison, I basically held each driver in my hand and listened to them. As the problem in my system seems to be the upper mid-range 1 - 3 kHz area, I figured this simple test would make the driver selection easy.
When listening, three drivers stood out. I was immediately drawn in to the texture and detail of the fostex. The Vifa was the most listenable, in that it had less of the issue. And the 10F showed most of the issue. The Tectonic was very unrefined compared to the others, with the issue showing up as intensely as it did with the 10F. At this point in the journey, I was thinking that the Vifa was going to be the winner. It still had the issue, but it was less. But, I really wanted the resolved sound of the Fostex or 10F...
As I mentioned above, I feel that the First One Amp is the strongest link in my audio chain. But what do I focus on next? DAC or the next speaker build.
Throughout my listening sessions with both Avebury and the Cheap and Cheerful, there is still something I have not been able to resolve: The mid-range shrillness. I hear it with my Chip Amps also, so I have ruled out the Amps being the issue. It is more frustrating with the First One, as it has a more definite sound, almost like noise or grain in the background (hence my "almost black background" comment above). But what is causing this? Speaker break-up or DAC?
I can say, with the Chip Amps, my Realtek DAC was almost tolerable, but with First One, it was ruled out right away. The mid-range issue above was indeed much worse with the on board DAC. The M-Audio is significantly better, but the issue is not solved...
It is more noticeable in the Arrays, unless I EQ Avebury to the same response as the Arrays, then it is there also. If the Avebury is run at their natural response, the issue is less apparent (more on that later).
So, I decided to give the drivers I acquired for this project a listen.
You know about the Scan Speak 10f and Fostex FF83wk. Some of You know I acquired a Bunch of Vifa TC9s. I also got a pair of Tectonics. I will upload the pictures of the drivers when time permits.
Sorry I could not keep up with the thread. Night shift was starting to cause some health problems, which is why I could not keep up with this and other home repair projects. I have been spending the last seven months catching up and getting lots of sunshine to get my health back on track... Anyway...
For the first comparison, I basically held each driver in my hand and listened to them. As the problem in my system seems to be the upper mid-range 1 - 3 kHz area, I figured this simple test would make the driver selection easy.
When listening, three drivers stood out. I was immediately drawn in to the texture and detail of the fostex. The Vifa was the most listenable, in that it had less of the issue. And the 10F showed most of the issue. The Tectonic was very unrefined compared to the others, with the issue showing up as intensely as it did with the 10F. At this point in the journey, I was thinking that the Vifa was going to be the winner. It still had the issue, but it was less. But, I really wanted the resolved sound of the Fostex or 10F...
Driver Comparison: part 2
As some of You know, I have really long work days during the week, so listening sessions happen only on the weekends. Part 1 was one weekend, part 2 being another.
The midrange shrillness issue was seeming weird, and I was wondering if not having the drivers mounted was causing some diffractions or something. So I built some simple panels and got the measurement system out. I also took out one of my NFS drivers from the Cheap and Cheerful, so I could listen and measure that mono to mono with the others...
I tried to level match using pink noise from True RTA, but found that the midrange test tone in JRiver was easier. I found out why after taking the measurements. The Fostex, 10F and NFS where pretty close in sensitivity. The Vifa was 4.5 dB down from the 10F Might be needing that 240 watt First One after all! I did not have the time to cut a hole yet for the tectonic, I have a feeling that one is going to need some EQ to even be listenable.
I decided to listen to the drivers first, before taking the frequency response and THD measurements. In the mono comparisons, I was surprised how well the NFS stood against the others. It had a rough high compared to the others, but it was still pretty resolved. The Vifa still sounded mellow, in fact too mellow at times. It still had less of the issue. Once again, the 10F showed the issue right away, as did my NFS.
Weirdly enough, I was wondering if something was going on with the DAC. I looked, and sure enough, the Power Supply was disconnected, so it was running on USB Power.
After plugging the power back in, the issue still remained, but less so. It still bothered me, like a tiny splinter You can not dig out. Yet, I think I am starting to figure out the source of the issue. With a little less of the issue, I could at least listen to the 10F now...
So for awhile I listened and compared the three main drivers in stereo. Bear in mind, I was only hearing content above 300 Hz due to the baffle roll-off, and at this time, it was not possible to hook up the Avebury for the lows.
The Vifa was holding its own, being mellow and resolved in the highs. The Fostex seemed a hare distant. And the 10F was just how I liked it. Fostex and 10F were very close, Fostex sounded more resolved in the highs, but I preferred the 10F.
When I measured the drivers, and looked at the Frequency response, I was quite surprised. It is educational to "see" what I was hearing. 10F was pretty flat, Fostex has a small peak in the highs, Vifa has a wide, swallow dip in the upper mids (similar to Alpair 12), and the NFS was rougher in the upper mids and highs. So are the things we hear, like "sounding distant" or "more resolved in the highs" and "mellow" (lack of mid-range shrillness issue), more related to the frequency response then I actually thought? Would the Fostex sound like the 10F if I EQed that High peak flat? That would cut the cost of this project in half!
Or, what if the Vifa could have the upper mids boosted for a response similar to 10F. Those are bought and paid for! The only drawback to the Vifa is the sensitivity. These are things I will be exploring later, as time permits.
The other cool thing I saw in the measurements are the dips caused by the baffles. This will help greatly, as I design baffles and cabinets for these drivers.
As some of You know, I have really long work days during the week, so listening sessions happen only on the weekends. Part 1 was one weekend, part 2 being another.
The midrange shrillness issue was seeming weird, and I was wondering if not having the drivers mounted was causing some diffractions or something. So I built some simple panels and got the measurement system out. I also took out one of my NFS drivers from the Cheap and Cheerful, so I could listen and measure that mono to mono with the others...
I tried to level match using pink noise from True RTA, but found that the midrange test tone in JRiver was easier. I found out why after taking the measurements. The Fostex, 10F and NFS where pretty close in sensitivity. The Vifa was 4.5 dB down from the 10F
Might be needing that 240 watt First One after all! I did not have the time to cut a hole yet for the tectonic, I have a feeling that one is going to need some EQ to even be listenable. I decided to listen to the drivers first, before taking the frequency response and THD measurements. In the mono comparisons, I was surprised how well the NFS stood against the others. It had a rough high compared to the others, but it was still pretty resolved. The Vifa still sounded mellow, in fact too mellow at times. It still had less of the issue. Once again, the 10F showed the issue right away, as did my NFS.
Weirdly enough, I was wondering if something was going on with the DAC. I looked, and sure enough, the Power Supply was disconnected, so it was running on USB Power.
After plugging the power back in, the issue still remained, but less so. It still bothered me, like a tiny splinter You can not dig out.
Yet, I think I am starting to figure out the source of the issue. With a little less of the issue, I could at least listen to the 10F now...So for awhile I listened and compared the three main drivers in stereo. Bear in mind, I was only hearing content above 300 Hz due to the baffle roll-off, and at this time, it was not possible to hook up the Avebury for the lows.
The Vifa was holding its own, being mellow and resolved in the highs. The Fostex seemed a hare distant. And the 10F was just how I liked it. Fostex and 10F were very close, Fostex sounded more resolved in the highs, but I preferred the 10F.
When I measured the drivers, and looked at the Frequency response, I was quite surprised. It is educational to "see" what I was hearing. 10F was pretty flat, Fostex has a small peak in the highs, Vifa has a wide, swallow dip in the upper mids (similar to Alpair 12), and the NFS was rougher in the upper mids and highs. So are the things we hear, like "sounding distant" or "more resolved in the highs" and "mellow" (lack of mid-range shrillness issue), more related to the frequency response then I actually thought? Would the Fostex sound like the 10F if I EQed that High peak flat? That would cut the cost of this project in half!
Or, what if the Vifa could have the upper mids boosted for a response similar to 10F. Those are bought and paid for! The only drawback to the Vifa is the sensitivity. These are things I will be exploring later, as time permits.The other cool thing I saw in the measurements are the dips caused by the baffles. This will help greatly, as I design baffles and cabinets for these drivers.
Attachments
The baffles...
Here are the baffles and the measurements of the three main drivers together. I would flip the baffles when listening to the lower drivers. You noticed I taped the hole where the NFS was, could not watch a movie without my arrays.
You can see what the baffle does to the lower mid-range.
Here are the baffles and the measurements of the three main drivers together. I would flip the baffles when listening to the lower drivers. You noticed I taped the hole where the NFS was, could not watch a movie without my arrays.
You can see what the baffle does to the lower mid-range.
Attachments
THD measurements
Opps, the photos did not stay rotated! That is windows 10 for You, it does not rename and flip pictures like 7 did, I actually have to press "enter", not just click on the next image. Not a bad thing, just not used to it.
I found the THD measurements quite interesting also. 10F is all they say it is, definitely the leader in this area. The Vifa is surprisingly good, with the Fostex scoring second to last, compared to the NFS having the most distortion. These are by no means professional measurements, in fact, I am amazed these came out as good as they did! The biggest take away I got from these graphs is the mid-range issue I am experiencing is not related to these distortion figures at all. The issue is more apparent with the 10F and NFS, both the best and worse in this measurement. And after the little mishap with the power supply, I am thinking the DAC is causing the issue.
I have been digging a little deeper into this. I do know a lot of First One owners do upgrade their DACs, as the Amp hides nothing in this regards. I have read that a USB interface could be an issue too. This is being addressed by the newer DACs, but my M-audio Fast track Pro is a pretty run of the mill ten year old consumer grade DAC...
So, before I can continue with these driver evaluations, I need to get this issue resolved. I have started assembling my Buffalo DAC. We will see if that solves it or not. There are some people who use the Buffalo kit with the First One Amp, most use tube output stages.
So it seems that is going to be a whole new adventure for me. There is no turning back now. I did not know sound could have some much texture, depth, and dimension as it does thru the First One. Just got to bring the rest of my sytem up to the bar it has set...
You can read about my Buffalo build adventure here:
http://www.diyaudio.com/forums/twis...buffalo-iii-building-journey.html#post4783299
Hopefully, I will get back to this thread sooner then a year!
The Aspiring Array Project is indeed alive and well.
Allen
Opps, the photos did not stay rotated! That is windows 10 for You, it does not rename and flip pictures like 7 did, I actually have to press "enter", not just click on the next image. Not a bad thing, just not used to it.
I found the THD measurements quite interesting also. 10F is all they say it is, definitely the leader in this area. The Vifa is surprisingly good, with the Fostex scoring second to last, compared to the NFS having the most distortion. These are by no means professional measurements, in fact, I am amazed these came out as good as they did! The biggest take away I got from these graphs is the mid-range issue I am experiencing is not related to these distortion figures at all. The issue is more apparent with the 10F and NFS, both the best and worse in this measurement. And after the little mishap with the power supply, I am thinking the DAC is causing the issue.
I have been digging a little deeper into this. I do know a lot of First One owners do upgrade their DACs, as the Amp hides nothing in this regards. I have read that a USB interface could be an issue too. This is being addressed by the newer DACs, but my M-audio Fast track Pro is a pretty run of the mill ten year old consumer grade DAC...
So, before I can continue with these driver evaluations, I need to get this issue resolved. I have started assembling my Buffalo DAC. We will see if that solves it or not. There are some people who use the Buffalo kit with the First One Amp, most use tube output stages.
So it seems that is going to be a whole new adventure for me. There is no turning back now. I did not know sound could have some much texture, depth, and dimension as it does thru the First One. Just got to bring the rest of my sytem up to the bar it has set...You can read about my Buffalo build adventure here:
http://www.diyaudio.com/forums/twis...buffalo-iii-building-journey.html#post4783299
Hopefully, I will get back to this thread sooner then a year!
The Aspiring Array Project is indeed alive and well.
Allen
Attachments
You know what would be a cute test in between right?
FIR correct all your sample drivers and listen again. Use gmad's templates with something like 4 or 5 cycles. Correct them to the same curve and listen again.
Even re-measure them after correction, to see what has changed there. View the uncorrected impulse and the corrected one. Browse the filters in the filtered IR tab. Look at early waterfall plots.
See how you do in a blind test by letting someone else switch drivers. All this can be done with the current DAC to see what you get, do you hear differences in cone material? Can you pick a favorite? Keep the test mono, listen on axis (hmmm with your baffle setup blind testing is going to be a bit harder).
Not only will FIR correction alter driver response, it will also change the DAC output to some degree.
You'll notice the more the FR is the same, the more difficult it becomes to identify your drivers. There will still be differences due to construction and cone material, but they will be closer overall. The differences can be found, if you're willing to look "inside" the measurements.
I'd pick the one needing the least correction, you can view the correction itself within REW.
If you need help, I'm here, I can help you with the processing, I can even do it for you and make suggestions. Then you can decide what route to take. You won't believe the different tone colors and sounds and level of detail I've heard throughout my experiments. FR is number one for what we hear. But of coarse there's more. I've come to rely on early waterfall plots as a good indication of differences. The more clean you get that, the more clear the sound. None of them will be as clear as they can be by themselves.
The bad rep for FIR is unfair, it works way better than IIR when used with care. See IIR as a permanent alteration of the FR. Not changing over time. FIR can alter time and let you determine WHEN the sound needs to be correct. The drawback being delayed playback. Not a problem in music or movies when played trough JRiver.
Let me know if you're up for it.
Keep in mind that in an array, you can't correct a "broken" top end due to the drivers sound arriving one after the other due to the separate distance to the listener of the drivers in the array. A ragged IR can be fixed for one driver with FIR. But not an array of them. That's a good reason to look at the IR and keep that fact in mind.
FIR correct all your sample drivers and listen again. Use gmad's templates with something like 4 or 5 cycles. Correct them to the same curve and listen again.
Even re-measure them after correction, to see what has changed there. View the uncorrected impulse and the corrected one. Browse the filters in the filtered IR tab. Look at early waterfall plots.
See how you do in a blind test by letting someone else switch drivers. All this can be done with the current DAC to see what you get, do you hear differences in cone material? Can you pick a favorite? Keep the test mono, listen on axis (hmmm with your baffle setup blind testing is going to be a bit harder).
Not only will FIR correction alter driver response, it will also change the DAC output to some degree.
You'll notice the more the FR is the same, the more difficult it becomes to identify your drivers. There will still be differences due to construction and cone material, but they will be closer overall. The differences can be found, if you're willing to look "inside" the measurements.
I'd pick the one needing the least correction, you can view the correction itself within REW.
If you need help, I'm here, I can help you with the processing, I can even do it for you and make suggestions. Then you can decide what route to take. You won't believe the different tone colors and sounds and level of detail I've heard throughout my experiments. FR is number one for what we hear. But of coarse there's more. I've come to rely on early waterfall plots as a good indication of differences. The more clean you get that, the more clear the sound. None of them will be as clear as they can be by themselves.
The bad rep for FIR is unfair, it works way better than IIR when used with care. See IIR as a permanent alteration of the FR. Not changing over time. FIR can alter time and let you determine WHEN the sound needs to be correct. The drawback being delayed playback. Not a problem in music or movies when played trough JRiver.
Let me know if you're up for it.
Keep in mind that in an array, you can't correct a "broken" top end due to the drivers sound arriving one after the other due to the separate distance to the listener of the drivers in the array. A ragged IR can be fixed for one driver with FIR. But not an array of them. That's a good reason to look at the IR and keep that fact in mind.
Last edited:
The big question: How different are they really?
I am thinking along those lines, doing it manually, as I have so far with rephase. But I have not dug quite as deep as You have with the timing of things, so a lot to learn there yet.
When doing this comparison, I was surprised how similar all the drivers were. They are definitely more similar then they are different. Even my NFS held its own quite well in the group. I believe the First One Amp is partially responsible, as it really "grabs on" to any driver and takes authority over it. (Avebury is not and easy system to drive with complex material, let alone at louder levels - yet First One does). When I do take this to the next step, and attain similar frequency response on the drivers, I am very curious to know what the differences will be. How much of a difference do materials and physical driver design make? I have learned that some driver designers purposely tailor the frequency response either to create a certain kind of sound for a given market and/or to mask and distract the listener's mind from a nasty break-up node or ringing a given design may have. This is going to be an interesting test, to say the least.
I have the impulse response of the drivers also, just have not got that far yet. I still have a lot to learn on the measurement side of things. As I am starting to actually "see" how they can help, my appetite for learning more increases. As You all know, time is a massive limitation for me...
I am pretty anxious to get the DAC built, it has only been sitting around for two years now! Got the first board done, hoping to get the next one tonight.
I am thinking along those lines, doing it manually, as I have so far with rephase. But I have not dug quite as deep as You have with the timing of things, so a lot to learn there yet.
When doing this comparison, I was surprised how similar all the drivers were. They are definitely more similar then they are different. Even my NFS held its own quite well in the group. I believe the First One Amp is partially responsible, as it really "grabs on" to any driver and takes authority over it. (Avebury is not and easy system to drive with complex material, let alone at louder levels - yet First One does). When I do take this to the next step, and attain similar frequency response on the drivers, I am very curious to know what the differences will be. How much of a difference do materials and physical driver design make? I have learned that some driver designers purposely tailor the frequency response either to create a certain kind of sound for a given market and/or to mask and distract the listener's mind from a nasty break-up node or ringing a given design may have. This is going to be an interesting test, to say the least.
I have the impulse response of the drivers also, just have not got that far yet. I still have a lot to learn on the measurement side of things. As I am starting to actually "see" how they can help, my appetite for learning more increases. As You all know, time is a massive limitation for me...
I am pretty anxious to get the DAC built, it has only been sitting around for two years now!
Got the first board done, hoping to get the next one tonight. - Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.