That's very cool and I think we all have moments where we listen to our creations and get chills or memories of something much more than the music. I remember hearing my synergy reference system and being stunned is probably the word. Did not want to stop listening but of course life gets in the way and we have to go to work etc. 
What was the setting you changed that made it all good again?
What was the setting you changed that made it all good again?
Well, that's a simple question with possibly a longer answer than expected
.To answer it I have to tell a bit about the history of the things I tried.
As I mentioned earlier I started my quest to find good sound by mimicking the Curve I found from Bob Katz. Basically flat from 20 Hz to 1 KHz and a dropping HF response to -6 dB at 20 KHz. I actually had a bit more drop in the HF curve, probably room and reflection related.
After that curve I altered the low response and added a couple of dB's to the low frequencies. A bit more impact there felt better. Though some things sounded quite good, other material did sound coloured. I changed the target a lot, going from pure flat to something like the B&K curves, the Bob Katz did sound more open than the B&K curves, the B&K sounded more veiled to me. Yet it was more pleasant on some recordings.
(don't forget, I also changed the room by adding damping panels)
The next step in learning was to adjust the graphical EQ while playing a lot of different songs. Preferably with a lot of (different) voices. While doing that just listening to tonal balance, nothing else mattered at that point. Get voices to sound as real and natural as possible was the goal.
I ended up with a few cuts and boosts and measured the end result with REW to see the curve I ended up with.
The resemblance from that REW measurement to a downward straight line was striking, one not unlike Sean Olive from JBL proposes. So that was my new target in measurements from then on. A downwards slope from 20 HZ to 20 KHz. This gave me a very good tonal balance. It also made more sense than the knee at 1 KHz from Bob Katz if you try to figure out why this would work compared to flat response. I never could figure out why one would deviate at a certain point only.
Not saying it wouldn't work (for some room speaker combo's) but I bet the room response has as much to do with it as the actual curve. Yet the open sound of the Bob Katz curve never left my thoughts.
In one of my tries/measurement sessions I had a bit of a drop around 900 - 1 KHz. I decided to use EQ to counter that but didn't measure the actual response. That's when I noticed the open quality again, excellent depth of the stage and no loss in the width of the stage. I thought I had a flat downward line from ~ 40 Hz to 20 KHz but didn't know of coarse. I didn't measure it in REW to know for sure. At this time my bass response was off due to the room induced dip I experience in my room between 60 and 70 Hz. At the same point I switched computers so that's part of the reason I didn't get that measurement.
I decided to start over clean. Measure with REW to get as even response between left and right channel as possible, straight line (though dropping towards HF) from ~40 HZ down to 20 KHz as the target. Figuring out the filling of the dip between 60-70 Hz and confirming with REW to fix the Bass. That's when I had the most straight line ever with left and right response as mirror images to maximise imaging and stage (or so I thought).
The result was balanced sound. Very nice tonal balance, but a bit "flat" if you will. It sounded excellent, placing of instruments and voices was good but in a 2D manner. I tried all kinds of sliding windows in DRC to see if that could give me back the depth. Short version: it didn't. It did influence focus and "texture" of voices (highly subjective term I know). But it didn't give me back the depth of stage I perceived in my previous session.
That's when I stumbled over this review, mentioned in a thread I followed:
SoundStage! Ultra | SoundStageUltra.com (UltraAudio.com) | Unconventional Success: Kaiser Kawero! Classic Loudspeakers
The following passage jumped out to me:
Soundstage! Ultra said:
I remember thinking to myself, could it be that easy? The depth I had with the Bob Katz curve immediately came to mind. As did the openness of the sound at that time, it just mist a bit of tonal balance.
So here comes the easy answer: I adjusted the curve to have a knee at 1.5 KHz. It's only 2dB higher at 1.5 KHz Still a downward line overall but with that slight knee at 1.5 KHz. Tonal balance did not really change. But the soundstage deepened, width remains the same as before, but more presence to voices.
So, a small change, big effect. For now implemented with PEQ; Frequency is 1500 Hz, Gain 2 dB and Q of 0,2 on left and right channel.
Now I get it why so many targets have a knee at 1 to 2 KHz. It also confirms the open sound of the Bob Katz curve, but for me this one works without the drawbacks. Tonal balance still feels right but the imaging and stage comes alive. One could call this similar to the Bob Katz curve, only slightly tilted down.
Probably the difference between a "live room" vs "Studio".
(sorry for the long winded response
)
Last edited:
Woah, that's a very interesting find. Thanks for sharing! Did you experiment much with the parameters?
I'm thinking I'm going to have to start design work on a peak filter soon.
I plan to do a lot more research, but first I'm going to enjoy this sound for a while. For me it's easy to play with in JRiver. The key is to not make a small hump at 1.5 KHz but have the knee there as a gradual change.
Of coarse you can play with varying points between 1 to 2 KHz. A reasonably flat response to begin with helps.
Usually a graph shows more than words can express:
Showing the Left channel response as measured in REW with a flattering 1/3 octave smoothing and the applied PEQ filter as predicted.
The target is a Sean Olive "like" downwards curve with in my case a 1.2 dB/octave downwards slope.
Scary smooth response when viewed in 1/3 octave
Maybe I should use that as a standard! (lol)
Of coarse you can play with varying points between 1 to 2 KHz. A reasonably flat response to begin with helps
.Usually a graph shows more than words can express:
Showing the Left channel response as measured in REW with a flattering 1/3 octave smoothing and the applied PEQ filter as predicted.
The target is a Sean Olive "like" downwards curve with in my case a 1.2 dB/octave downwards slope.
Scary smooth response when viewed in 1/3 octave
Maybe I should use that as a standard! (lol)
Last edited:
I plan to do a lot more research, but first I'm going to enjoy this sound for a while. For me it's easy to play with in JRiver. The key is to not make a small hump at 1.5 KHz but have the knee there as a gradual change.
Of coarse you can play with varying points between 1 to 2 KHz. A reasonably flat response to begin with helps
Well, I'm looking forward to your results. I hope you'll include measurements.
No, that's not intentional and has more to do with the chosen smoothing.
I usually use REW's variable smoothing but for the purpose of following the target I showed my response with the 1/3 oct so you can follow it better.
Actual response is flat to about 15 KHz and response drops a bit from there and falls off above 17 KHz. Still not bad for a line of full range drivers measured off axis... I should try pointing the mic at the speakers one of these days.
My usual smoothing as seen in other graphs confirms that.
You have to admit this looks way different:
Just a change in smoothing to 1/48, now show me a manufacturer that would like to show that.
I have no anechoic room, the room you see is what is used.
My speakers are toed in 15 degree and measurement microphone is pointed straight forward, the measurement is (about 15 degree) off axis. See:
More toe-in (to the point of on axis) had a huge detrimental effect on my perceived stage width. I like this better.
I usually use REW's variable smoothing but for the purpose of following the target I showed my response with the 1/3 oct so you can follow it better.
Actual response is flat to about 15 KHz and response drops a bit from there and falls off above 17 KHz. Still not bad for a line of full range drivers measured off axis... I should try pointing the mic at the speakers one of these days.
My usual smoothing as seen in other graphs confirms that.
You have to admit this looks way different:
Just a change in smoothing to 1/48, now show me a manufacturer that would like to show that
.I have no anechoic room, the room you see is what is used.
My speakers are toed in 15 degree and measurement microphone is pointed straight forward, the measurement is (about 15 degree) off axis. See:
More toe-in (to the point of on axis) had a huge detrimental effect on my perceived stage width. I like this better.
Last edited:
I was under the impression that your system had smooth frequency response upto 20KHz.
Funny how chosen smoothing can apparently change the frequency response!
Which make me want to ask you this: Do you miss some air?
Given, the high horizontal directivity of this speaker, your curves would get quickly messed up if you are not sitting at the spot where the microphone is placed like in a group listening session. So that would mean you are adjusting your system for single user critical listening, right?
Funny how chosen smoothing can apparently change the frequency response!
Which make me want to ask you this: Do you miss some air?
Given, the high horizontal directivity of this speaker, your curves would get quickly messed up if you are not sitting at the spot where the microphone is placed like in a group listening session. So that would mean you are adjusting your system for single user critical listening, right?
I'm glad I never made that claim of response out to 20 KHz, nor can I hear that high myself. I do not miss air, on the contrary, very few systems I've listened to had the same convincing "you are there" quality with eyes closed. With full out orchestral music it has that floating presence of the scene. I wouldn't dare to get higher response out of the Vifa TC9.
A couch measurement from a while ago shows it actually behaves quite well across a bigger area:
Bob Katz like curve at that time with the bass bump, prior to adding extra damping panels in the room
This was a measurement with both speakers playing in stereo. You can see it's quite stable output up to that ~17 KHz point even with a stereo measurement.
I measure in one (ideal listening) spot. The whole couch area benefits as the multi point measurement above shows.
Now guess who's sitting in that one special spot (lol). The speakers have quite an even response horizontal, It might have helped that I put the front baffle on there with it's rounded holes as a short waveguide. I have mentioned looking in to that at some point during the Design and there used to be a thread on a German forum showing the difference in off axis behaviour between a flush mounted Vifa TC9 and one with a similar "wave guide" as I have. The one with the "wave guide" had better off axis response and more even response over the 0 to 30 degree area. Sadly I didn't save that picture and it is gone. I can't easily redo those kind of measurements with these huge Towers. But from what I have seen it is more even than I had hoped for. Regardless of the fact the baffle helped with that or not.
I'll admit that I have thought about adding an ambience tweeter, much like one discussed by Lynn Olsen in his huge "Beyond the Ariel" thread. I've posted about that idea here: http://www.diyaudio.com/forums/full-range/203356-cloning-ids-25s-25.html#post4113889
After hearing the arrays (at that point they weren't up and running) I've played with that thought but current funds don't allow me to pursue. I haven't missed it but who knows, it might work very well...
A couch measurement from a while ago shows it actually behaves quite well across a bigger area:
Bob Katz like curve at that time with the bass bump, prior to adding extra damping panels in the room
This was a measurement with both speakers playing in stereo. You can see it's quite stable output up to that ~17 KHz point even with a stereo measurement.
I measure in one (ideal listening) spot. The whole couch area benefits as the multi point measurement above shows.
Now guess who's sitting in that one special spot (lol). The speakers have quite an even response horizontal, It might have helped that I put the front baffle on there with it's rounded holes as a short waveguide. I have mentioned looking in to that at some point during the Design and there used to be a thread on a German forum showing the difference in off axis behaviour between a flush mounted Vifa TC9 and one with a similar "wave guide" as I have. The one with the "wave guide" had better off axis response and more even response over the 0 to 30 degree area. Sadly I didn't save that picture and it is gone. I can't easily redo those kind of measurements with these huge Towers. But from what I have seen it is more even than I had hoped for. Regardless of the fact the baffle helped with that or not
.I'll admit that I have thought about adding an ambience tweeter, much like one discussed by Lynn Olsen in his huge "Beyond the Ariel" thread. I've posted about that idea here: http://www.diyaudio.com/forums/full-range/203356-cloning-ids-25s-25.html#post4113889
After hearing the arrays (at that point they weren't up and running) I've played with that thought but current funds don't allow me to pursue. I haven't missed it but who knows, it might work very well...
Last edited:
If it were me, I wouldn't bother with anything above 15KHz.... my ears betray me past that... too many nights with loud music in my youth.
I am curious... you pointed out to an Effects channel in JRiver, but I don't remember if you posted before about which one you are using. Care to share?
I am curious... you pointed out to an Effects channel in JRiver, but I don't remember if you posted before about which one you are using. Care to share?
Uploaded for a previous post:
Do NOT use the room effect. Only the Surround Field.
I have that on Medium Enhancement at the moment. Seems to function a bit like cross talk cancelation with arrays. Based on Blumlein Shuffle principle.
Get it to the top right below Adaptive volume because it splits in Mid/Side signal processing, you don't want to disrupt that with EQ etc.
Do NOT use the room effect. Only the Surround Field.
I have that on Medium Enhancement at the moment. Seems to function a bit like cross talk cancelation with arrays. Based on Blumlein Shuffle principle.
Get it to the top right below Adaptive volume because it splits in Mid/Side signal processing, you don't want to disrupt that with EQ etc.
Last edited:
Wow, what did You change?
Opps, found it! it is way past my bedtime. 14 hour day working graveyard...
Thanks for sharing. I have tried that "surround" effect also. It is by far one of the most effective effects I have used to exact sound field ambient information without wrecking the clarity or precise 3D placement of the recorded music. Well done JRiver. I am going to get some sleep, and add some of my experiences later.
Allen
Last edited:
Give it a shot. I suspect that the "brightness" that people generally associate with a "flat" in-room response may be partially due to the use of an off-axis mic for the measurement.
That had crossed my mind gmad, you might be right there.
But while you're here, care to try that small knee at ~ 1.5 KHz with your setup?
Just to see (or actually hear) if "Depth" perception changes. I'd advice to not have the speakers on axis for this test. Aim them in a way that they cross behind you. That should maximise the perception I.M.H.O.
But while you're here, care to try that small knee at ~ 1.5 KHz with your setup?
Just to see (or actually hear) if "Depth" perception changes. I'd advice to not have the speakers on axis for this test. Aim them in a way that they cross behind you. That should maximise the perception I.M.H.O.
Hi there, have been lurking around all the time from day one, but finally feel it's worth sharing my experience regarding measurement microphones.
Have built my own set of line sources 2 years ago with 25 Visaton FRS8M each, digitally eq them, and finally sold my long time reference Magnepan T4A's ... says enough about the DIY effort, doesn't it?
If using small capsule 1/4" or below types, they usually are of the cardioid variety.
With my calibrated mic from the DAAS32+ measurement suite, there's no appreciable change when pointing it straight on or upwards.
It's manual even recommends the latter orientation in order to catch all horizontal sound reflections from 360° when using longish sample sequences to measure broad band room responses.
When pointing at the sound source, you'd skew the room response by not recording all from the side- and even less from the rearward response.
If you take a gated first arrival measurement, pointing straight at the source is the ticket.
Just my 2 cts.
Best of wrestling the last bit of perfection from your beauties, Wesayso!
Have built my own set of line sources 2 years ago with 25 Visaton FRS8M each, digitally eq them, and finally sold my long time reference Magnepan T4A's ... says enough about the DIY effort, doesn't it?
If using small capsule 1/4" or below types, they usually are of the cardioid variety.
With my calibrated mic from the DAAS32+ measurement suite, there's no appreciable change when pointing it straight on or upwards.
It's manual even recommends the latter orientation in order to catch all horizontal sound reflections from 360° when using longish sample sequences to measure broad band room responses.
When pointing at the sound source, you'd skew the room response by not recording all from the side- and even less from the rearward response.
If you take a gated first arrival measurement, pointing straight at the source is the ticket.
Just my 2 cts.
Best of wrestling the last bit of perfection from your beauties, Wesayso!
Thank you for dropping in, marin. I'd say that truly does say something about the DIY effort. What were the listening impressions if you compare those two?
Do you use (P)EQ only? Have you tried FIR based processing?
I'll go experiment with mic positions one of the next times. Although I must say my microphone did pick up on the reflections coming from behind it. I usually point it a bit upwards and have it at the (referring to me) listeners ear height.
Do you use (P)EQ only? Have you tried FIR based processing?
I'll go experiment with mic positions one of the next times. Although I must say my microphone did pick up on the reflections coming from behind it. I usually point it a bit upwards and have it at the (referring to me
) listeners ear height.
I donn´t think this is (as much anyway) the case when it is a line array we are talking about.
The dispersion patterne is quite uniform through the freq. spectrum in line sources , whereas pointsource speakers tend to be more and more directive as the frequency rises.
Koldby
I kind of agree with you on that Koldby, but there's no harm in trying it to see if there are (huge) differences...
I also think that the perceived brightness has a lot to do with the room itself. No house curve could make everyone happy. But general trends do apply. It's up to us in our own room to find the right balance.
I also think that the perceived brightness has a lot to do with the room itself. No house curve could make everyone happy. But general trends do apply. It's up to us in our own room to find the right balance.
Absolutely no harm in trying.
By the way you said pointing the mic , not toe in the speakers..!
So it would be the mic. pattern that would change the recieved result, not the dispersion of the speakers..
And that does make it even more unlikely to make any difference as the measuring mic tend to have a nice omni directivity.
Koldby