or Connect
AVS › AVS Forum › Audio › Receivers, Amps, and Processors › The Audyssey Pro Installer Kit Thread (FAQ in post #1)
New Posts  All Forums:Forum Nav:

The Audyssey Pro Installer Kit Thread (FAQ in post #1) - Page 113

post #3361 of 5281

Here's another question for the Curve Editor gurus. The graph below (1/12 smoothing instead of my normal 1/24th smoothing) shows a dip of up to 5dB between about 300Hz and 1kHz. I can correct this to some extent (+3dB) by using the Curve Editor. Question is, do we think it will be worthwhile and is that the best way to deal with it?  Seems to me that 300-1000Hz is a significant portion of the response.

 

At the same time, I am wondering about bringing down that little peak at 55Hz. If I did all of the above I would be pretty flat from 20 all the way through to 15kHz (ignoring the mid-range comp dip which is, of course, deliberate).

 

What do you guys think?

 

 

1000

 

 

Separate issue:

 

The graph below shows the effect of running Audyssey and then engaging Pgm2 on the Submersives - a 3-4dB boost starting from about 50Hz down and really making its impact felt from 20Hz down. The red line shows Pgm1 and the blue line shows Pgm2 of course. As you can seem by using Pgm2, the response is only 8dB down at 7Hz, compared with the response at about 20Hz, and less than 3.5db down at 9Hz.

 

Question is this: Craig (John) advised me to turn off DEQ if using Pgm2 for fear of overblowing the bass. At the time this seemed to make sense, but since I have been wondering why it is necessary to disengage DEQ.  If we accept the curve below (with DEQ off) as 'flat' for the purposes of this question, then all that DEQ does is restore the perception of flatness when I am listening at below Reference level. IOW, engaging DEQ gives me the perception of the curve below. So why would it be best to disengage DEQ, even when running Pgm2? I usually listen at around -7.5dB on the MV. Thanks for any thoughts offered.

 

 

1000

post #3362 of 5281
Quote:
Originally Posted by kbarnes701 View Post

Here's another question for the Curve Editor gurus. The graph below (1/12 smoothing instead of my normal 1/24th smoothing) shows a dip of up to 5dB between about 300Hz and 1kHz. I can correct this to some extent (+3dB) by using the Curve Editor. Question is, do we think it will be worthwhile and is that the best way to deal with it?  Seems to me that 300-1000Hz is a significant portion of the response.

At the same time, I am wondering about bringing down that little peak at 55Hz. If I did all of the above I would be pretty flat from 20 all the way through to 15kHz (ignoring the mid-range comp dip which is, of course, deliberate).

What do you guys think?

I have an almost identical dip. I have a target curve that adds 3 db back in covering that range and it certainly appears much flatter after running it. There is no easy way to do an A/B but I'm not sure I could hear the difference. But I'm OCD so that's OK.

Quote:
Separate issue:

The graph below shows the effect of running Audyssey and then engaging Pgm2 on the Submersives - a 3-4dB boost starting from about 50Hz down and really making its impact felt from 20Hz down. The red line shows Pgm1 and the blue line shows Pgm2 of course. As you can seem by using Pgm2, the response is only 8dB down at 7Hz, compared with the response at about 20Hz, and less than 3.5db down at 9Hz.

Question is this: Craig (John) advised me to turn off DEQ if using Pgm2 for fear of overblowing the bass. At the time this seemed to make sense, but since I have been wondering why it is necessary to disengage DEQ.  If we accept the curve below (with DEQ off) as 'flat' for the purposes of this question, then all that DEQ does is restore the perception of flatness when I am listening at below Reference level. IOW, engaging DEQ gives me the perception of the curve below. So why would it be best to disengage DEQ, even when running Pgm2? I usually listen at around -7.5dB on the MV. Thanks for any thoughts offered.



I have used Pgm2 and DEQ and sometimes it works and sometimes not so now I don't use it.
post #3363 of 5281
Quote:
Originally Posted by kbarnes701 View Post

Here's another question for the Curve Editor gurus. The graph below (1/12 smoothing instead of my normal 1/24th smoothing) shows a dip of up to 5dB between about 300Hz and 1kHz. I can correct this to some extent (+3dB) by using the Curve Editor. Question is, do we think it will be worthwhile and is that the best way to deal with it?  Seems to me that 300-1000Hz is a significant portion of the response.

At the same time, I am wondering about bringing down that little peak at 55Hz. If I did all of the above I would be pretty flat from 20 all the way through to 15kHz (ignoring the mid-range comp dip which is, of course, deliberate).

What do you guys think?


[/URL

Try it!

Other than with the Curve Editor, I don't know how else to deal with it. And try tweaking the 55Hz peak, too.

Jeff
post #3364 of 5281
Quote:
Originally Posted by audioguy View Post

I have an almost identical dip. I have a target curve that adds 3 db back in covering that range and it certainly appears much flatter after running it. There is no easy way to do an A/B but I'm not sure I could hear the difference. But I'm OCD so that's OK.

 

 

Thanks. I'll have a go then - I'm not sure I'll hear the difference either, but since when has that ever stopped us? :)

 

Quote:
I have used Pgm2 and DEQ and sometimes it works and sometimes not so now I don't use it.

 

I've been auditioning, and I think it may sound just a little bloated with DEQ on, so I think I will try it for a while and see how I get on. I was very happy with Pgm1 + DEQ TBH anyway.

post #3365 of 5281
Quote:
Originally Posted by pepar View Post

Try it!

Other than with the Curve Editor, I don't know how else to deal with it. And try tweaking the 55Hz peak, too.

Jeff

 

OK Jeff - thanks. I will go and have a go right now. Will report back.

post #3366 of 5281
Quote:
Originally Posted by pepar View Post

Try it!
Other than with the Curve Editor, I don't know how else to deal with it. And try tweaking the 55Hz peak, too.
Jeff

FWIW, I once had a -10 db sharp dip at about 600 Hz with 1/24th smoothing, and I ultimately found that getting rid of a USB extension cable and moving the laptop closer helped because it straightened the USB cable directly connected to the mic, and the position of the mic itself, enough to remove the distortion. Or at least I think it did:).

More seriously, I'd try something in the Curve Editor like (approximately):
400 Hz 0
600 Hz +3
800 Hz 0

At least that's what my eyeball glance at the curve would make me think...I'm at work and don't have Pro handy.

More a question for Pepar/Jeff: do you eyeball differences, or apply an octave distance rule (that is, a calculated Hz distance from an origin, using either the raw Hz or transforming them to logs) when you edit the .CSV file to make tweaks?
Edited by sdrucker - 11/26/12 at 10:31am
post #3367 of 5281
Quote:
Originally Posted by kbarnes701 View Post

Yes I did. It didn't matter to me as I have a record anyway and I change the sub distances by using my OmniMic to perform the sub distance tweak. From the earlier discussions, I was expecting it to happen, and it did.

Sorry to be persistent, but did you check the trims as well?  For example, my sub1 and sub2 trims are -4.5 and -2.0.  after re-loading the measurements, both trims get set to -2.0 and require manual intervention.

 

Regarding the EQ2 vs. DEQ question, it's an interesting one.  DEQ does more than just compensate for the bass.  If you turn off DEQ in favor of the EQ2 setting, don't you run the risk of losing other things that DEQ is doing?  If I were running a calibration with EQ1 engaged, I would be more inclined to leave EQ1 engaged, since I have learned to "trust Audyssey" (except for the distance tweak, of course...).  But then I don't have Submersives, which just leaves me green with envy.

post #3368 of 5281
Quote:
Originally Posted by AustinJerry View Post

Sorry to be persistent, but did you check the trims as well? 

It doesn't hurt to remind everyone that it is both levels and distances that need to be checked and (maybe) re-entered.

Jeff
post #3369 of 5281
Quote:
Originally Posted by sdrucker View Post

Quote:
Originally Posted by pepar View Post

Try it!
Other than with the Curve Editor, I don't know how else to deal with it. And try tweaking the 55Hz peak, too.
Jeff

FWIW, I once had a -10 db sharp dip at about 600 Hz with 1/24th smoothing, and I ultimately found that getting rid of a USB extension cable and moving the laptop closer helped because it straightened the USB cable directly connected to the mic, and the position of the mic itself, enough to remove the distortion. Or at least I think it did:).

More seriously, I'd try something in the Curve Editor like (approximately):
400 Hz 0
600 Hz +3
800 Hz 0

At least that's what my eyeball glance at the curve would make me think...I'm at work and don't have Pro handy.
 

 

 

Thanks Stuart. I have just been making some final measurements with OmniMic (centre channel alone, left channel alone, right channel alone, centre+subs) to see what is going on in more detail. It seems if I apply your suggestion above to the L, C and R channels, it will be as good as I can get (IOW no need to create separate filters for those 3 channels).  I will have a go and report back.

 

Quote:
More a question for Pepar/Jeff: do you eyeball differences, or apply an octave distance rule (that is, a calculated Hz distance from an origin, using either the raw Hz or transforming to logs) when you edit the .CSV file to make tweaks?

 

If I could even understand that I'd be a happy guy ;)

post #3370 of 5281
Quote:
Originally Posted by kbarnes701 View Post


Thanks Stuart. I have just been making some final measurements with OmniMic (centre channel alone, left channel alone, right channel alone, centre+subs) to see what is going on in more detail. It seems if I apply your suggestion above to the L, C and R channels, it will be as good as I can get (IOW no need to create separate filters for those 3 channels).  I will have a go and report back.


If I could even understand that I'd be a happy guy wink.gif

Simple version: I was thinking that you if look at where the 'dip' is, you'd apply the adjustment at the midpoint between the two anchors, or as close as you can get given the limits of the Curve Editor. In essence you're assuming a kind of symmetry (even width) to where the adjustment should be, and linear interpolation. If not, ideally I'm thinking that if there's some formula we could use, we could plug in values in Excel, and all you'd need to do is edit the Curve Editor's underlying .CSV files to specify where the anchors and tweaks actually should be. Not as simple as dragging and dropping anchors, but probably more accurate - and likely within the limits of what many of us are comfortable doing in Office.

More gobbly-gook version:
OTOH, depending on how the algorithm actually works, linear interpolation might be an artificial assumption, if the underlying statistical model (some sort of piecework regression or polynomial regression model) isn't using the straight Hz in its calculations. In that case, the question is where can we do the adjustment where it would be 'optimal' for the channels involved. Put into English, if the range you're interested in is 400 to 800 hz (in my example), what's the optimal point of adjustment to get the impact you want (the 3 db boost in the range of your dip)? And what should the proper anchors be to maximize the tweak?

There was some discussion either here or the main Audyssey thread about how octaves are calculated, but I never quite understood if that can be hooked into the Curve Editor per se.

I'm no Feri, but one of these days I have to read the Audyssey technical paper. It can't be too much worse than some of the Journal of Marketing Research stuff I've read for professional education:eek:
post #3371 of 5281
Quote:
Originally Posted by AustinJerry View Post

Quote:
Originally Posted by kbarnes701 View Post

Yes I did. It didn't matter to me as I have a record anyway and I change the sub distances by using my OmniMic to perform the sub distance tweak. From the earlier discussions, I was expecting it to happen, and it did.

Sorry to be persistent, but did you check the trims as well?  For example, my sub1 and sub2 trims are -4.5 and -2.0.  after re-loading the measurements, both trims get set to -2.0 and require manual intervention.

 

No problemo mi amigo. Better to be persistent than negligent :)  Yes, I checked and corrected both distance and trim for each sub.

 

 

 

Quote:
Regarding the EQ2 vs. DEQ question, it's an interesting one.  DEQ does more than just compensate for the bass.  If you turn off DEQ in favor of the EQ2 setting, don't you run the risk of losing other things that DEQ is doing?  If I were running a calibration with EQ1 engaged, I would be more inclined to leave EQ1 engaged, since I have learned to "trust Audyssey" (except for the distance tweak, of course...).  But then I don't have Submersives, which just leaves me green with envy.

 

:)  You make a good point about losing the other beneficial effects of DEQ (eg on the surround levels). I could apply a RLO of 5dB I guess. This would mean at my typical listening level of -7.5 DEQ would be just 'on tickover'. I will try it with DEQ on and off - it's just that I have a lot of faith in guys like Craig and Audioguy and they both say to disengage DEQ if you do the PGM2 thing.

 

There'd be no point in engaging Pgm2 and then running Audyssey - Audyssey would just attempt to correct the boost applied by Pgm2 and force it into the target curve. The idea is to EQ with Audyssey with Pgm1 and get the Audyssey curve and then engage Pgm2 to get the benefit of the Submersive low end boost. You can see the effect on the graph in my earlier post - a 3dB or so boost where it really matters - below 20Hz. It is very easy to go back to Pgm1 - it's just a button press on each sub (after powering off first - the DSP is only engaged on power up).

 

Don't be 400 Jerry ....

 

Be 400

 

Get the Hsus in the classifieds and bust out the pocketbook.... you know it makes sense :)

post #3372 of 5281
Quote:
Originally Posted by sdrucker View Post

Quote:
Originally Posted by kbarnes701 View Post


Thanks Stuart. I have just been making some final measurements with OmniMic (centre channel alone, left channel alone, right channel alone, centre+subs) to see what is going on in more detail. It seems if I apply your suggestion above to the L, C and R channels, it will be as good as I can get (IOW no need to create separate filters for those 3 channels).  I will have a go and report back.


If I could even understand that I'd be a happy guy wink.gif

Simple version: I was thinking that you if look at where the 'dip' is, you'd apply the adjustment at the midpoint between the two anchors, or as close as you can get given the limits of the Curve Editor. In essence you're assuming a kind of symmetry (even width) to where the adjustment should be, and linear interpolation. If not, ideally I'm thinking that if there's some formula we could use, we could plug in values in Excel, and all you'd need to do is edit the Curve Editor's underlying .CSV files to specify where the anchors and tweaks actually should be. Not as simple as dragging and dropping anchors, but probably more accurate.

More gobbly-gook version:
OTOH, depending on how the algorithm actually works, linear interpolation might be an artificial assumption, if the underlying statistical model (some sort of piecework regression or polynomial regression model) isn't using the straight Hz in its calculations. In that case, the question is where can we do the adjustment where it would be 'optimal' for the channels involved. Put into English, if the range you're interested in is 400 to 800 hz (in my example), what's the optimal point of adjustment to get the impact you want (the 3 db boost in the range of your dip)? And what should the proper anchors be to maximize the tweak?

There was some discussion either here or the main Audyssey thread about how octaves are calculated, but I never quite understood if that can be hooked into the Curve Editor per se.

I'm no Feri, but one of these days I have to read the Audyssey technical paper. It can't be too much worse than some of the Journal of Marketing Research stuff I've read for professional education:eek:

 

Thanks Stuart. I consider myself enlightened. ;)  How about just doing it visually but putting handles at +3dB at, say, 500 and 700 (with anchors at 400 and 800). Not sure if the Curve Editor allows that fineness of resolution though. I'll be tweaking it later to see. Visually :)

post #3373 of 5281
^^
I'd probably try both the single and double handles at + 3 db and see what works the best in practice in giving you the flattest response. I've done the more midpoint approach, but it can't hurt to try both. The algorithm's enough of a 'black box' that I don't know if you'll run into either the law of unintended consequences or wind up with similar results either way.
post #3374 of 5281

@ Jerry - incidentally Jerry, engaging Pgm2 has also given me a slightly flatter curve than Pgm1 did (as expected).

 

 

1000

 

As opposed to:

 

 

1000

 

I did not make any changes to the sub distances when going to Pgm2 as I didn't think it would make any difference, but perhaps I should check.

 

You know, once upon a time, I used to watch movies! :)

post #3375 of 5281
Quote:
Originally Posted by sdrucker View Post

^^
I'd probably try both the single and double handles at + 3 db and see what works the best.

Thanks. Will do.

post #3376 of 5281
Wow, this is getting complicated.
post #3377 of 5281
Quote:
Originally Posted by AustinJerry View Post

Wow, this is getting complicated.

Now you know how I feel when I read about construction or placement of room treatments, or some of Mark Seaton or Craig John's stuff on gain matching or post-distance tweak adjustments:).

We all have different comfort levels - or pain thresholds:eek:.
post #3378 of 5281
Quote:
Originally Posted by sdrucker View Post

At least that's what my eyeball glance at the curve would make me think...I'm at work and don't have Pro handy.
More a question for Pepar/Jeff: do you eyeball differences, or apply an octave distance rule (that is, a calculated Hz distance from an origin, using either the raw Hz or transforming them to logs) when you edit the .CSV file to make tweaks?
The grips/control points can be spaced at a minimum of 1/3 octave. Working with the Curve Editor and a mouse, I just placed a grip at 1000Hz and then another at 1260Hz. (~26% is a 1/3 oct interval.) I then tried to manually change it to 1250Hz. It accepted my edit, but when I checked it again, it had moved it to 1259.9Hz. I am considering, for creating target curves with CSV in a spreadsheet, recommending 27% ... to be safe. I found that you can stick whatever numbers you want into the cells, but if you create a target curve that is out of bounds, it won't load.

My current target curve is applied to the mains only, i.e. not to the sub channel, and has a boost at the low end and another at the upper end to "brighten up" Audyssey Reference as I don't have a way to apply any other base curve.



The 20 and 24000 are "stock", the none-zero entries are my tweaks and the other "0" entries are "anchors" (to reference) that isolate my tweaks. Remember, this is not applied to the sub channel.
Edited by pepar - 11/26/12 at 11:44am
post #3379 of 5281
Quote:
Originally Posted by pepar View Post

The grips/control points can be spaced at a minimum of 1/3 octave. Working with the Curve Editor and a mouse, I just placed a grip at 1000Hz and then another at 1260Hz. (~26% is a 1/3 oct interval.) I then tried to manually change it to 1250Hz. It accepted my edit, but when I checked it again, it had moved it to 1259.9Hz. I am considering, for creating target curves with CSV in a spreadsheet, recommending 27% ... to be safe.
My current target curve is applied to the mains only, i.e. not to the sub channel, and has a boost at the low end and another at the upper end to "brighten up" Audyssey Reference as I don't have a way to apply any other base curve.

The 20 and 24000 are "stock", the none-zero entries are my tweaks and the other "0" entries are "anchors" (to reference) that isolate my tweaks. Remember, this is not applied to the sub channel.

Super. If you don't mind, could you give an example with Keith's situation in the 400 Hz to 1 kHz range? And what formula are you using to ID octave width prior to spacing?
post #3380 of 5281
Quote:
Originally Posted by sdrucker View Post

Super. If you don't mind, could you give an example with Keith's situation in the 400 Hz to 1 kHz range? And what formula are you using to ID octave width prior to spacing?
I am only calculating +27% and -27%.

If there is a grip at 1000Hz, then 1270Hz above and 730Hz below. Above 1270Hz would be 1613Hz. Below 730Hz would be 533Hz.

Jeff
post #3381 of 5281
Quote:
Originally Posted by sdrucker View Post

Super. If you don't mind, could you give an example with Keith's situation in the 400 Hz to 1 kHz range? And what formula are you using to ID octave width prior to spacing?

Anchor at 400Hz and another at 1000Hz. I might visually select a midpoint for a grip and dial in a +3dB boost. But then the curve would need to be loaded and measured for it's effect. There is "room" for two control points between 400 and 1000, should that be necessary.

Jeff
post #3382 of 5281
Quote:
Originally Posted by pepar View Post

Anchor at 400Hz and another at 1000Hz. I might visually select a midpoint for a grip and dial in a +3dB boost. But then the curve would need to be loaded and measured for it's effect. There is "room" for two control points between 400 and 1000, should that be necessary.
Jeff

This is definitely tutorial fodder:). So....if you wanted two control points, maybe 508 Hz (27% above 400 Hz) and 730 Hz (27% below 1000 Hz)?

Or just pick 700 Hz and be done with it:).
post #3383 of 5281
Quote:
Originally Posted by sdrucker View Post

This is definitely tutorial fodder:). So....if you wanted two control points, maybe 508 Hz (27% above 400 Hz) and 730 Hz (27% below 1000 Hz)?
Or just pick 700 Hz and be done with it:).

Math-wise, the closest a grip could be placed to 1000Hz is 730Hz. The closest one could go below that is 533Hz ... comfortably above the 508Hz interval needed above 400Hz. But that's what *can* be done. Whether one midpoint or the two more or less evenly spaced points are used would be entirely based on which produced the desired result.

Jeff
post #3384 of 5281
Quote:
Originally Posted by pepar View Post

Math-wise, the closest a grip could be placed to 1000Hz is 730Hz. The closest one could go below that is 533Hz ... comfortably above the 508Hz interval needed above 400Hz. But that's what *can* be done. Whether one midpoint or the two more or less evenly spaced points are used would be entirely based on which produced the desired result.
Jeff

Got it - I hadn't tried the dual control points this way, so I wasn't sure which 'interval' took precedent.

Thanks for your tips - back to work for some of us:(.
post #3385 of 5281
Quote:
Originally Posted by sdrucker View Post

Got it - I hadn't tried the dual control points this way, so I wasn't sure which 'interval' took precedent.
Thanks for your tips - back to work for some of us:(.

You wouldn't have to put a point at 533 if there was one at 730. It *could* go at 508. And if one were there, the next one up wouldn't actually have to go at 730 ... it could go as low as 645. The rule only dictates the closest two points can be.
post #3386 of 5281
Quote:
Originally Posted by pepar View Post

You wouldn't have to put a point at 533 if there was one at 730. It *could* go at 508. And if one were there, the next one up wouldn't actually have to go at 730 ... it could go as low as 645. The rule only dictates the closest two points can be.

That's a key difference - minimum distance vs. what we'd like to do. Thanks for clarifying.
post #3387 of 5281

OK - I have done more Curve Editing to try to smooth that dip between about 300Hz and 1kHz. This is in addition to my earlier Curve Editor tweak where I boosted 55Hz to 95Hz by 3dB.

 

Here are the graphs. I think this is as much as can be done with the Curve Editor here - I have used the max 3dB boost for both tweaks. On the latest, I used two boots of 3dB - one at roughly 500 Hz and one at roughly 700Hz - anchored at 350Hz-ish and 800Hz-ish. I say 'ish' because it is hard to drag the little handles precisely so my 800 may well have been 790 for example. I got it to 'look' right on the Curve Editor screen.

 

All graphs are 24th smoothing first, followed by the same graph at 6th smoothing.

 

 

1000

 

 

1000

 

 

1000

 

 

1000

 

 

1000

 

 

1000

 

Crossovers were set by Pro at 80Hz for FL and FR and 100Hz for Centre. RS/LS/FLH/FRH were all set at 110Hz. I manually changed the FL & FR XOs to 100Hz in the AVP after uploading the cal. One small difference this time was that I chose 100Hz in Pro for the centre channel and left it there - don't know if that makes any difference but it seemed best to use the Pro setting of 100Hz if I was also going to manually set the FL & FR at 100Hz. 

 

I think the graph is a little flatter after doing the 350-800Hz tweak in the Curve Editor. When I did my listening tests, I thought I could hear a tad more 'presence' in the 300-1000Hz region but this may just be wishful thinking. The good news is that it didn't make the sound any worse and I think it may have made it better. Bass is with the Submersives in Pgm2 mode, after calibrating with Audyssey in Pgm1 mode, per the Submersive guru's suggestion (and Mark's). This accounts for the lift from about 30Hz down. I am undecided as of yet about whether to engage DEQ with Pgm2 or not. I will do more listening tests. It is easy to go back to Pgm1 + DEQ if I want to, at the push of a couple of buttons.

 

I honestly think I am done now. I don't think I can do better with my skill level and my room. How does it sound? Just absolutely, totally, undeniably, indubitably freakin' AWESOME!  

 

Any comments gratefully received as ever. Thanks to all the guys in this amazing thread for the endless help and encouragement.

post #3388 of 5281
Quote:
Originally Posted by kbarnes701 View Post

OK - I have done more Curve Editing to try to smooth that dip between about 300Hz and 1kHz. This is in addition to my earlier Curve Editor tweak where I boosted 55Hz to 95Hz by 3dB.

Here are the graphs. I think this is as much as can be done with the Curve Editor here - I have used the max 3dB boost for both tweaks. On the latest, I used two boots of 3dB - one at roughly 500 Hz and one at roughly 700Hz - anchored at 350Hz-ish and 800Hz-ish. I say 'ish' because it is hard to drag the little handles precisely so my 800 may well have been 790 for example. I got it to 'look' right on the Curve Editor screen.

All graphs are 24th smoothing first, followed by the same graph at 6th smoothing.



















Crossovers were set by Pro at 80Hz for FL and FR and 100Hz for Centre. RS/LS/FLH/FRH were all set at 110Hz. I manually changed the FL & FR XOs to 100Hz in the AVP after uploading the cal. One small difference this time was that I chose 100Hz in Pro for the centre channel and left it there - don't know if that makes any difference but it seemed best to use the Pro setting of 100Hz if I was also going to manually set the FL & FR at 100Hz. 

I think the graph is a little flatter after doing the 350-800Hz tweak in the Curve Editor. When I did my listening tests, I thought I could hear a tad more 'presence' in the 300-1000Hz region but this may just be wishful thinking. The good news is that it didn't make the sound any worse and I think it may have made it better. Bass is with the Submersives in Pgm2 mode, after calibrating with Audyssey in Pgm1 mode, per the Submersive guru's suggestion (and Mark's). This accounts for the lift from about 30Hz down. I am undecided as of yet about whether to engage DEQ with Pgm2 or not. I will do more listening tests. It is easy to go back to Pgm1 + DEQ if I want to, at the push of a couple of buttons.

I honestly think I am done now. I don't think I can do better with my skill level and my room. How does it sound? Just absolutely, totally, undeniably, indubitably freakin' AWESOME!  

Any comments gratefully received as ever. Thanks to all the guys in this amazing thread for the endless help and encouragement.

That 350 to 800 Hz region looks smoother to me (but wasn't the other curve you plotted from 20 Hz to 20 kHz at 1/12th smoothing)? If anything you may be understating the improvement because you're plotting at 1/24th. You way want to read what Pepar and I were going back and forth on for more precise control points, but regardless it looks smoother in that region IMO.

And that's one smooth FR < 55 Hz.
post #3389 of 5281
Quote:
Originally Posted by sdrucker View Post

That 350 to 800 Hz region looks smoother to me (but wasn't the other curve you plotted from 20 Hz to 20 kHz at 1/12th smoothing)? If anything you may be understating the improvement because you're plotting at 1/24th. You way want to read what Pepar and I were going back and forth on for more precise control points, but regardless it looks smoother in that region IMO.

And that's one smooth FR < 55 Hz.

Thank Stuart. Yes, I’d forgotten about that pesky 1/12th smoothing - I don't even know how that happened - I only ever use 1/24th and/ir 1/6th for the printed graphs. I'm pretty happy with it. All I have to do now is decide whether I prefer Pgm2 and no DEQ or Pgm1 plus DEQ on the Submersives. I'll need to do some more listening tests to determine that.

post #3390 of 5281
Quote:
Originally Posted by kbarnes701 View Post

Thank Stuart. Yes, I’d forgotten about that pesky 1/12th smoothing - I don't even know how that happened - I only ever use 1/24th and/ir 1/6th for the printed graphs. I'm pretty happy with it. All I have to do now is decide whether I prefer Pgm2 and no DEQ or Pgm1 plus DEQ on the Submersives. I'll need to do some more listening tests to determine that.

OK, you're off the hook. You can go to sleep now:).
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Receivers, Amps, and Processors
AVS › AVS Forum › Audio › Receivers, Amps, and Processors › The Audyssey Pro Installer Kit Thread (FAQ in post #1)