Well you got the general idea right 
. With manual EQ, broad strokes, I shape the initial measurement in the EQ window from REW. Just to get it within a window of, say plus or minus 10 dB. The EQ settings are then put into use in JRiver's PEQ section. Next I loop REW's signal trough JRiver to make a new measurement. That one is the base for DRC.
To see what it's doing I make another measurement with REW trough JRiver with the convolved.wav added and the basic EQ still on to see if it does what it's supposed to do.
No complaints here.
. With manual EQ, broad strokes, I shape the initial measurement in the EQ window from REW. Just to get it within a window of, say plus or minus 10 dB. The EQ settings are then put into use in JRiver's PEQ section. Next I loop REW's signal trough JRiver to make a new measurement. That one is the base for DRC.To see what it's doing I make another measurement with REW trough JRiver with the convolved.wav added and the basic EQ still on to see if it does what it's supposed to do.
No complaints here
.
The new pre-amp is in. Did a quick session to try it out. Much cleaner results even with a busy street outside.
This is after EQ+Convolver doing their thing. And a house curve applied.
Still need to play more with that though but no more time this session.
Here's the waterfall:
Not bad for a (mostly untreated) living room I'd say.
Still need to find out what is causing the suck-out at 60 Hz. Probably a room effect as it changes quickly with different windows applied.
This is after EQ+Convolver doing their thing. And a house curve applied.
Still need to play more with that though but no more time this session.
Here's the waterfall:
Not bad for a (mostly untreated) living room I'd say.
Still need to find out what is causing the suck-out at 60 Hz. Probably a room effect as it changes quickly with different windows applied.
Last edited:
The new pre-amp is in. Did a quick session to try it out. Much cleaner results even with a busy street outside.
This is after EQ+Convolver doing their thing. And a house curve applied.
Still need to play more with that though but no more time this session.
Here's the waterfall:
Not bad for a (mostly untreated) living room I'd say.
Still need to find out what is causing the suck-out at 60 Hz. Probably a room effect as it changes quickly with different windows applied.
Amazingly clean response
Your cabinet work pays off here in spades !
I did not bypass the psychoacoustic stage but used flat as a basic curve first. Adjusted from that for preference using a custom curve. It is purely coincidence it resembles the B&K curve. The downward slope is exaggerated as this is a left + right response. It is not as severe with left only of right only curves.
Not quite pleased yet though. Something sounds forced. A bit hollow. But that could very well be some fine tuning of the curves.
Have you experimented with dip limiting? I mean the DLMinGain etc. I was kinda hoping this works somewhat similar to Acourate, preventing the correction to put energy into real dips.
In my experience, the PT stage gives a proper tonal balance as long as the PL stage is properly tweaked, so just make sure this is right before going ahead with the addition of a custom curve. Also, you may want to take a listen without the PT stage to get a feel for what's happening there.
I'm not sure why your L+R response would show more of a tilt than either single measurement....
Dip limiting indeed prevents the filter from fighting a losing battle against deep notches in the response caused by the room. I have never played around with this stage however.
Thanks for checking in. You're right about the tilt, it's the same as the left/right responses by themselves.
I checked out the PT stage settings and it is easily bypassed I see. I listened to the PT stage and it is way to bright. Probably because I have multiple sources playing with some delay.
I need to get a grip on the window sizes needed. It seems that might be the cause of my hollow sound. But I want to see what's happening after changing those. I took window sizes from a thread from Bob Katz on Audiolense. Not totally fair of coarse, I think I could shorten them without loosing the focus. Right now I was using values around the "Normal" template and the "Soft". But I think I might want to use shorter window settings in the mid frequencies.
So much variables and so little time to play with them all. I played with the Target Designer to get less bright highs. But it would be easier to start with a flat response. I also need to look at how to interpret the target. Linear or Logarithmic or one of the Splines. That would require extra control points on the target curve to get it to behave.
I've had great sound and I have had less successful runs. The differences between good and great are often very small. I'm trying to change one factor at a time but it's time consuming.
For ease of use I held on to using mostly DRCDesigner but maybe it's time to break loose and manipulate it a bit further.
I checked out the PT stage settings and it is easily bypassed I see. I listened to the PT stage and it is way to bright. Probably because I have multiple sources playing with some delay.
I need to get a grip on the window sizes needed. It seems that might be the cause of my hollow sound. But I want to see what's happening after changing those. I took window sizes from a thread from Bob Katz on Audiolense. Not totally fair of coarse, I think I could shorten them without loosing the focus. Right now I was using values around the "Normal" template and the "Soft". But I think I might want to use shorter window settings in the mid frequencies.
So much variables and so little time to play with them all. I played with the Target Designer to get less bright highs. But it would be easier to start with a flat response. I also need to look at how to interpret the target. Linear or Logarithmic or one of the Splines. That would require extra control points on the target curve to get it to behave.
I've had great sound and I have had less successful runs. The differences between good and great are often very small. I'm trying to change one factor at a time but it's time consuming.
For ease of use I held on to using mostly DRCDesigner but maybe it's time to break loose and manipulate it a bit further.
Well, there's a lot of factors involved and since you can't really focus on more than one parameter at a time it can seem like a daunting task at first. I think you will want to venture away from DRC Designer and use the scripts when you're ready.
I would suggest starting with the "minimal" preset with PLStartFreq and PLEndFreq set to 64 and 8192 respectively and the PT stage bypassed. The response should look pretty flat (with exception to severe room-induced dips). If this is not the case, there may be a problem with the PLMaxGain value (the default value should be enough for the towers if you are applying eq ahead of the measurement). You may want to pause here for a bit to try out the log weighted linear phase PLType to see the advantage that it offers in the phase response, as well as bypassing the optional RT (ringing truncation) stage to obtain a flatter frequency response if desired. When you're confident with the values of the aforementioned parameters, you might then want to try longer windows ("soft" or "normal" preset) to see what effect that has. However, if the response looks flat enough already, these presets may be excessive (I like "minimal" and "erb" personally). When you're happy with the way everything looks, run the filter with the PT stage engaged and give it a listen. Chances are that some recordings will sound too bright (I think that many recordings were possibly monitored with downward tilted eq targets or maybe even made bright on purpose to offset the effects of cutting to vinyl) and if this bothers you, you may want to create an additional filter to compensate in those cases.
I would suggest starting with the "minimal" preset with PLStartFreq and PLEndFreq set to 64 and 8192 respectively and the PT stage bypassed. The response should look pretty flat (with exception to severe room-induced dips). If this is not the case, there may be a problem with the PLMaxGain value (the default value should be enough for the towers if you are applying eq ahead of the measurement). You may want to pause here for a bit to try out the log weighted linear phase PLType to see the advantage that it offers in the phase response, as well as bypassing the optional RT (ringing truncation) stage to obtain a flatter frequency response if desired. When you're confident with the values of the aforementioned parameters, you might then want to try longer windows ("soft" or "normal" preset) to see what effect that has. However, if the response looks flat enough already, these presets may be excessive (I like "minimal" and "erb" personally). When you're happy with the way everything looks, run the filter with the PT stage engaged and give it a listen. Chances are that some recordings will sound too bright (I think that many recordings were possibly monitored with downward tilted eq targets or maybe even made bright on purpose to offset the effects of cutting to vinyl) and if this bothers you, you may want to create an additional filter to compensate in those cases.
Thanks for the advise! Had another session of measurements today, to verify if the dips were room induced or part of my choices in baffle construction. I can now safely say it is the room. Need to undo what I did and re-measure before moving on.
I removed the little O-rings from the baffle bolts and it seems to introduce more distortion. So I'll have to put them back in and verify again to be sure.
After that I can play with settings again. I did find a way to counter my dip in the low frequencies. If I'm successful with the rings I'll do a before and after plot.
I had my first serious listener over today. Right after the measurements (clean slate after removing the O-rings, had to re-do everything). So I was kind of nervous about it.
My listener is an avid classical music fan and played some orchestral tracks and choirs and even some Opera.
He was moved to tears, stating it was the next best thing he heard beside actually being there. (he actually said it was like being there with eyes closed)
He himself performs in one of those choirs so that to me said something! I'm getting there and am on the right path.
I guess the most difficult person to please will be myself.
To add to the bass solution: I'm guessing a better solution is to add a few subs a la Geddes solution but the trick I used works at the listening area, being my couch. I dialled down the left line in the ~65 Hz area and increased the same amount to the right array. This evens out the lines when working together.
I removed the little O-rings from the baffle bolts and it seems to introduce more distortion. So I'll have to put them back in and verify again to be sure.
After that I can play with settings again. I did find a way to counter my dip in the low frequencies. If I'm successful with the rings I'll do a before and after plot.
I had my first serious listener over today. Right after the measurements (clean slate after removing the O-rings, had to re-do everything). So I was kind of nervous about it.
My listener is an avid classical music fan and played some orchestral tracks and choirs and even some Opera.
He was moved to tears, stating it was the next best thing he heard beside actually being there. (he actually said it was like being there with eyes closed)
He himself performs in one of those choirs so that to me said something! I'm getting there and am on the right path.
I guess the most difficult person to please will be myself
.To add to the bass solution: I'm guessing a better solution is to add a few subs a la Geddes solution but the trick I used works at the listening area, being my couch. I dialled down the left line in the ~65 Hz area and increased the same amount to the right array. This evens out the lines when working together.
Last edited:
Todays measurements went well. I found the source of the distortion I was seeing yesterday. I had a ground loop isolator on the outputs of my Dac.
With Isolator:
Without Isolator:
I had bought that ground loop isolator for the Xonar Essence ST soundcard. The power supply of my PC isn't very clean. At really loud levels I could hear a faint 50 Hz on the Musical Fidelity M1 Dac as well. So I figured to just put it on there. But this shows the problem needs to be solved elsewhere.
Don't look at the different slopes, they are not comparable. Taking out the ground loop isolator did change the high frequencies output somewhat.
So I had to re-measure to get a new base line for the convolve steps.
Impulse of todays measurement:
Turns out my first guess was wrong, the rubber O-rings aren't back in yet. I'll think about it, but it sounds very good the way it is.
With Isolator:
Without Isolator:
I had bought that ground loop isolator for the Xonar Essence ST soundcard. The power supply of my PC isn't very clean. At really loud levels I could hear a faint 50 Hz on the Musical Fidelity M1 Dac as well. So I figured to just put it on there. But this shows the problem needs to be solved elsewhere.
Don't look at the different slopes, they are not comparable
. Taking out the ground loop isolator did change the high frequencies output somewhat. So I had to re-measure to get a new base line for the convolve steps.
Impulse of todays measurement:
Turns out my first guess was wrong, the rubber O-rings aren't back in yet. I'll think about it, but it sounds very good the way it is.
Last edited:
I HAVE to like them, after all this time they are part of me .
I want to limit spilling my own listening impressions. After all, we are all proud of what we build right? It's opinions from others that I value. I'm just responsible to get them to sound their best to the max of my abilities. So far I get the impression I'm doing well.
.I want to limit spilling my own listening impressions. After all, we are all proud of what we build right? It's opinions from others that I value. I'm just responsible to get them to sound their best to the max of my abilities. So far I get the impression I'm doing well
.