Home Theater Forum and Systems banner
21 - 40 of 42 Posts

·
Banned
Joined
·
4,838 Posts
Discussion Starter · #21 ·
It has been strongly hinted that my views are primarily based on personal experience in disregard of the scientific method and properly conducted studies.

I will state simply:
  • I have high regard for the hard-working researchers in the audio industry and their work to further our understanding of acoustics and psychoacoustics for our benefit.
  • I appreciate the importance of basing my own work and experiments on well-founded principles.
  • Few have the luxury of conducting double-blind studies to prove every statement we make. Yet we are fortunate to live in a time and part of the world where phenomenal resources are within reach of anyone who wishes to contribute to the stretching of the limits of our understanding of the state of the art.
  • It is all about sound. In the end, our ears are an important instrument which must be included in the experiment.
  • Ears are fallible. Imaginations are powerful. Any individually-derived conclusion must be scrutinized relative to proven facts.
  • I am not afraid to disagree with an industry expert if I believe he has missed an important point in his research or in the advice given based on that research. There is as much danger (that is a strong word in this context) in over-agreeing with and over-generalizing the results of a study as there is in ignoring them. I am making no personal accusation here, simply stating a trend I have observed over time, and my reason for responding to Toole's paper in the first place.
  • There is very little (none?) of the detail in Toole's paper that I disagree with. Only that his hoped-for "calibration methodology that could be applied throughout the audio industry" (abstract) appears to disregard a critical slice of the high-end listener pie.
 

·
Premium Member
Joined
·
1,562 Posts
jtalden: Thank you for an excellent summary. I thought I read a little bit more into the part about room EQ:

  • If desired objectives are not achieved because of poorly controlled speaker directivity, get better speakers. "Equalization cannot change loudspeaker directivity; the remedy is a better loudspeaker."(p516)
  • Room EQ is of limited value, is generally useless in correcting for speakers with poorly controlled directivity. "Equalization is very limited in what it can “correct,” yet the notion that changing the signal supplied to a sound system consisting of an unknown loudspeaker in an unknown room can “equalize” or “calibrate” a system is widespread."(p517)
Points well made. I think we are in basic agreement.
I am use to hearing those 2 bullets as basic rote from what seems like most all speaker manufacturers. I am not sure why they stress those points so hard with all the evidence that full range EQ can be very helpful in many cases. Harmon developed and sells full range EQ devices (I think) so that makes such firm statements puzzling.

In my personal case I EQ full range manually with marked success. I have backed off to achieve flat direct speaker output similar to the guidelines in the paper several times in the past to try to find the improved sound quality that is promised; no success so far. My preferred overall steady-state house curve though is very close to the curve that the experienced listeners prefer in the paper. That target works well for me. The response above maybe 800Hz is fine when the speakers are EQed flat so using direct sound speaker output as flat there is no issue for me assuming some additional EQ roll-off can be applied if needed to meet the LP steady-state house curve. My inclination is though that the target LP response is more important than a flat speaker output in the midrange.

My issue is in the midrange 300 to 800Hz. The most notable issue in that area is that both the FL and FR LP responses have a significant sag around 350-500Hz which is a room effect and not the result of floor or ceiling reflection or any other reflection point I have been able to identify. If I don't boost this range I loose significant information that reduces sound quality. The phone ringing in; 'Why don't you call me any more' by Alicia Keys, is almost completely lost as one obvious example. The boost brings up the SPL, but does it at a 20 ms delay. I hear no ill effect from this late arriving sound, but maybe 'experienced' listeners would on some material.

As a result of my limited experience; my take would be that; after taking all the measures possible within the imposed constraints regarding the room size/shape, speaker and LP location, and room treatments then it is worth full range EQ experimentation. It may not help, but there are a huge number of testimonials on the web from users of commercial automated EQ systems that suggest that, if done well, it can significantly help in many cases. I would agree that the sound quality may be better (at least theoretically) if the setup was adjusted better to avoid the root cause of issues if possible. Many of us have significant setup and room constraints that limit what we can do in that regard however. Possibly the opponents of full range EQ just lump all these situations into a basket labeled 'bad setup' and discount them? Probably not, but I wish I understood the real situation in that regard.

Regarding this paper I would be tempted to attribute at least some the difference to the narrower and more controlled dispersion of the horn based speakers commonly used. I think Toole did mention however that similar results are achieved using typical domes in home theaters so…

Below is a FL wavelet using my preferred EQ with a filter providing; +5.5dB, 401Hz, 4.5Q. The direct sound is boosted that amount in that range. The result is that the SPL dip at the LP is filled 20ms late. My FR EQ is identical.

Colorfulness Graphic design Graphics software Art Graphics
 

·
Registered
Joined
·
21 Posts
Possibly the opponents of full range EQ just lump all these situations into a basket labeled 'bad setup' and discount them?
My understanding is that Mr. Toole targets an ideal world. This seems ideal since content creators and cinemas should have the funds available to reach this ideal world.

He brushes over home theaters, and I agree with his sentiment here also. Those of us who have been "playing" DSP for years, still don't have a clear understanding of all of the variables. I personally have spent a significant amount of time on research and practice, and I still learn new things all the time (part of the joy for me). And yet, we expect those with little understanding of any of the variables to jump in head first into DSP. Sure, as our collective knowledge increases, and with the help of commercial solutions, the situation probably isn't anywhere near as dire as it once was, but still, if Mr. Toole is nudging people (those I guess who otherwise don't know or don't care) away from DSP, and towards improved playback equipment, and/or room design, surely this is a good thing.

On the subject of "SS&I", there are so many variables, is it really any wonder why no one has attempted significant research?

For best SS&I, your speakers should have this directivity pattern, with this amount of toe-in, placed this distance from side walls, that have this amount of absorption. If the absorption of the walls is less, you should change toe-in by this amount. If the speakers are placed further from the walls, you should change toe-in by this amount, etc, etc (so many variables). Do we really want to go there? Still, to this day, the industry can't even decide on a correct calibration level ffs.
 

·
Banned
Joined
·
4,838 Posts
Discussion Starter · #24 ·
Harmon developed and sells full range EQ devices (I think) so that makes such firm statements puzzling.
I forgot about that. Yes, JBL's ARCOS system is full-range, based on PEQ, not IR filters. Their justification is probably that it is part of the JBL Synthesis pro line, and the speakers in that line are all controlled-directivity (horn) designs, so there is some degree of alignment with Toole's recommendations.

In my personal case I EQ full range manually with marked success.
My current setup includes Dirac Live, and I have done a fair amount of reflection control and setup optimization beforehand, so DL does not have to correct a huge amount. Switching it in/out, the difference is noticeable but not strong. I think of it as more of a polishing tool - FL & FR frequency response matching is tightened so imaging sharpens nicely.

I have hand-tuned full range for years, always finding the result to be a great improvement. Again, the amount of change was never much, usually cuts only.

Any auto EQ I have heard in fairly well-treated rooms has been an improvement.

My preferred overall steady-state house curve though is very close to the curve that the experienced listeners prefer in the paper.
As does mine. I end up with about 2 dB of boost from 120 to 30 Hz, about 2 dB of rolloff from 5 kHz to 10 kHz, and I throw in a 1 dB presence peak - or to be more accurate, I allow that peak, which is already there in the speakers - at around 2 kHz.

I have backed off to achieve flat direct speaker output similar to the guidelines in the paper several times in the past to try to find the improved sound quality that is promised; no success so far...

My inclination is though that the target LP response is more important than a flat speaker output in the midrange...

As a result of my limited experience; my take would be that; after taking all the measures possible within the imposed constraints regarding the room size/shape, speaker and LP location, and room treatments then it is worth full range EQ experimentation. It may not help, but there are a huge number of testimonials on the web from users of commercial automated EQ systems that suggest that, if done well, it can significantly help in many cases. I would agree that the sound quality may be better (at least theoretically) if the setup was adjusted better to avoid the root cause of issues if possible. Many of us have significant setup and room constraints that limit what we can do in that regard however. Possibly the opponents of full range EQ just lump all these situations into a basket labeled 'bad setup' and discount them? Probably not, but I wish I understood the real situation in that regard.

Regarding this paper I would be tempted to attribute at least some the difference to the narrower and more controlled dispersion of the horn based speakers commonly used. I think Toole did mention however that similar results are achieved using typical domes in home theaters so…
In a bad room with bad (poorly controlled directivity) speakers, it seems logical that the promised improvement of not EQing might be more obvious. Most who are interested in discussing such matters already have pretty good rooms and speakers, so they end up falling into the exception category. I think this is worth noting, that on a relativity scale there are no doubt clear examples that would illustrate the bad effects from room EQ that Toole suggests, but they are probably in situations horrific enough that a more experienced ear would turn and run away rather than try to EQ. The traveling front-of-house sound reinforcement mixer probably has to deal with this more, although FOH arrays tend to be designed for some degree of controlled directivity, as you suggest, so once again the technology leans in the direction of falling into the exception bucket where a smidge of EQ helps more than it hurts. I regularly visit a number of clubs in Lincoln/Omaha to hear live bands, and all have remarkably good sound, most with small vertical arrays. The trend for decent in-house sound in clubs over the last 20 years is definitely good news for touring bands and those who go to hear them (although I have heard a few mind-boggling exceptions, too).

Below is a FL wavelet using my preferred EQ with a filter providing; +5.5dB, 401Hz, 4.5Q. The direct sound is boosted that amount in that range. The result is that the SPL dip at the LP is filled 20ms late. My FR EQ is identical.

View attachment 100194
Pardon my dense question, what is causing the 20 mS delay? Are you using linear-phase filtering? That seems like a lot, although I can think of no particular reason for concern. Just curious.

On the subject of "SS&I", there are so many variables, is it really any wonder why no one has attempted significant research?
Sadly, I can only agree with you. Like art, I know it when I see it - as Leonard said so well, "It is simply an attempt to find the placement that gives the overall best impression of creating a sound space in which performers are playing music." But getting there with a given set of speakers in a given room is very much trial and error, with certain trends to be followed from experience.

I would be tickled pink if there was even a clear-cut set of acoustical measurements that could tell you when you were "there" or close to it. I have attempted this with FR and phase measurements and have so far been disappointed. A setup with good left/right FR and phase matching can have unclear, imprecise imaging, and when sharp imaging is achieved, FR and phase matching can look way off comparatively.:huh:

Adding in delayed reflection info and ratio of direct-to-reflected sound might help, but the latter is not easily measurable. It all gets nightmarish pretty fast.

As AJ has inferred, a good set of definitions might be a good starting place.
 

·
Premium Member
Joined
·
1,562 Posts
...what is causing the 20 mS delay? Are you using linear-phase filtering? That seems like a lot, although I can think of no particular reason for concern.
This measurement is with IIR PEQ filters using my DCX units. It also includes an FIR phase correction filter designed in rePhase to just remove the direct sound phase rotation; no impact to SPL. The FIR filter has no impact on sound quality that I can hear. It is only there as it helps cleanup the Phase/GD/SR charts. The FIR filter is only active on Stereo Music as it is being implemented using a Foobar2000 music server.

Yes, the 20ms delay for SPL fill seem large to me also, but I haven't heard any ill effects; only a significant improvement in tonality. The cause appears to be a reflection null causing the initial hole in the response at 0ms. I haven't been able to identify the source of that reflection. That is followed by 3 other nulls near the same freq that prevent earlier SPL fill-in. My back wall is 9' from the LP so a reflection from there would arrive about 16ms later. Possibly that is the primary source of the fill-in? Possibly it is more a combination of several late arriving reflections. I really am really challenged to attribute cause to this type of acoustic issue so these are just guesses.
 

·
Super Moderator
Joined
·
1,674 Posts
I am use to hearing those 2 bullets as basic rote from what seems like most all speaker manufacturers. I am not sure why they stress those points so hard with all the evidence that full range EQ can be very helpful in many cases. Harmon developed and sells full range EQ devices (I think) so that makes such firm statements puzzling.
Toole is of course not a manufacturer, so his statements are based merely his own research, while Harman is a manufacturer and free to sell what they think consumers may want. I don't recall him saying full range EQ can't make speakers with poor on/off axis, etc. sound better. If he did then I would have to disagree.

The response above maybe 800Hz is fine when the speakers are EQed flat so using direct sound speaker output as flat there is no issue for me assuming some additional EQ roll-off can be applied if needed to meet the LP steady-state house curve. My inclination is though that the target LP response is more important than a flat speaker output in the midrange.
So your speakers are axially flat nearfield above 800hz, but not below? Are they passive or active/DSP? Do you have measurements?

My issue is in the midrange 300 to 800Hz. The most notable issue in that area is that both the FL and FR LP responses have a significant sag around 350-500Hz which is a room effect and not the result of floor or ceiling reflection or any other reflection point I have been able to identify.
I have yet to see a room where monopole speakers (99.9% of market) don't show effects well above Schroeder, as can be clearly seen in Tooles measurements as well.
Have you tried others speakers, or better yet, speakers with different polar response that yours, in the same position?

Possibly the opponents of full range EQ just lump all these situations into a basket labeled 'bad setup' and discount them? Probably not, but I wish I understood the real situation in that regard.
Not sure who these "opponents" are, but Tooles data certainly indicates that full ranged EQ shouldn't be needed in the majority of rooms with well designed speakers, but that some EQ, perhaps staring as high as 500-800hz, where room modal effects are strongest, will inevitably be beneficial.
That is not the same as saying that some full range EQ can't make things sound better, given that a lot of speakers on the market won't fit Tooles "well designed" category.

cheers,
 

·
Premium Member
Joined
·
1,562 Posts
Toole is of course not a manufacturer, so his statements are based merely his own research, while Harman is a manufacturer and free to sell what they think consumers may want. I don't recall him saying full range EQ can't make speakers with poor on/off axis, etc. sound better. If he did then I would have to disagree.
Yes, I understand and agree with this comment and the impact of it. In Harmon's case it makes the understanding of what is done to boost sales and what is real engineering conviction difficult to decide. In Toole's case I am sure it all makes sense within the scope he covers. My only issues why it doesn't follow for me? Usually things we expect to be true bear out in listening sessions whether or not they are real. In this case the clear preference is counter to the expectation.

So your speakers are axially flat nearfield above 800hz, but not below? Are they passive or active/DSP? Do you have measurements?
My comparison is between near field axially flat speaker output from 300-20kHz Vs adding the indicated filter that fills a dip around 400Hz that otherwise occurs when measuring at the LP. These are DIY SEAS H1456/H1212 per the 'Equipment List' under my name. DCX2496 provides XO, delay and EQ as needed. The 16L Box is 2nd order/sealed, Qb=~0.5 the poly fiber fill is, if anything, possibly a little over filled from typical. So, there is not much that can go wrong with the MR response from 100-2k range. The speaker is EQ corrected for baffle step, the overall rising response of that driver and for a small resonance peak around 800Hz similar to the response shown in the SEAS driver spec sheet. I have all sorts of measurements and can show some that support this assertion and some that don't. Indoor measurements are somewhat very dependent on window setting and positioning trying to prove it is flat 100-500 is problematic. so we better not go there. I don't have outdoor measurement, but the preponderance of my work suggest the near-field response of the speaker is basically flat with the base EQ until I add this 400Hz filter and one or 2 more below 300Hz to address other modes below the Schroeder.

I have yet to see a room where monopole speakers (99.9% of market) don't show effects well above Schroeder, as can be clearly seen in Tooles measurements as well.
Room effects? Yes, mine have serious room effects that appear in the midrange up to 600Hz. The question is, it better to EQ those even though it is above the cited upper end of the Schroeder transition frequency or not. My experience is that it is clearly demonstrable that it is better to EQ out my dip even that happens with 20ms delayed.

Have you tried others speakers, or better yet, speakers with different polar response that yours, in the same position?
No, I haven't changed speakers. I have nice reasonably uniform horizontal dispersion as shown below. I have close phase tracking through the XO so the central lobe is stable. The vertical dispersion is as expected for a 7" MW and 1" dome; fairly narrow and asymmetric (not shown). It's unavoidable with this configuration.

Not sure who these "opponents" are, but Tooles data certainly indicates that full ranged EQ shouldn't be needed in the majority of rooms with well designed speakers, but that some EQ, perhaps staring as high as 500-800hz, where room modal effects are strongest, will inevitably be beneficial.
That is not the same as saying that some full range EQ can't make things sound better, given that a lot of speakers on the market won't fit Tooles "well designed" category.
If it is recognized that in some cases EQ departing from flat axial speaker output may provide some benefit as high as 300-600Hz then that accommodates my situation. It may be that simple.

Possibly Toole's guideline is intended to provide a very good reliable result that is acceptable for all practical applications and I am just fussing over the last bit of improvement possible beyond that? At least that is one possibility.

It's not like the sound is poor without the filter it is just noticeably better on some material with it. I am very pleased with most all aspects of my current setup. It measures well and has very good SS&I. My interest is as a hobbyist who wants to learn and experiment to develop better understanding and thus drag out the last bit of sound quality that this setup can muster - just for fun. Since I have no source of comparison to other well calibrated systems or to experienced listeners, there is no telling how my setup really sizes-up by comparison. If you or Wayne want to stop by and critique the sound quality, please do. Unfortunately that is probably not going to appear high on your to-do list however. :(

Thanks to all for all the comments. Every perspective helps me better understand this.

30" Unnormalized Horizontal Dispersion 0-90°:
 

Attachments

·
Banned
Joined
·
4,838 Posts
Discussion Starter · #28 ·
Not sure if this is exactly the statement being referred to:

"Equalization cannot change loudspeaker directivity; the remedy is a better loudspeaker."(p516)

Granted it does not say directly that EQ can/will not sound better, but it seems to be implied pretty strongly.
 

·
Super Moderator
Joined
·
1,674 Posts
Yes, I understand and agree with this comment and the impact of it. In Harmon's case it makes the understanding of what is done to boost sales and what is real engineering conviction difficult to decide. In Toole's case I am sure it all makes sense within the scope he covers. My only issues why it doesn't follow for me? Usually things we expect to be true bear out in listening sessions whether or not they are real. In this case the clear preference is counter to the expectation.
Regarding Harmans products, I believe on their most sophisticated stereo integrated, there was an option for EQ only<500hz iirc, or the option to EQ full range.
I'm still a bit confused by your case. Aren't you saying no EQ is used/required above 800hz?

My comparison is between near field axially flat speaker output from 300-20kHz Vs adding the indicated filter that fills a dip around 400Hz that otherwise occurs when measuring at the LP. These are DIY SEAS H1456/H1212 per the 'Equipment List' under my name. DCX2496 provides XO, delay and EQ as needed. The 16L Box is 2nd order/sealed, Qb=~0.5 the poly fiber fill is, if anything, possibly a little over filled from typical. So, there is not much that can go wrong with the MR response from 100-2k range. The speaker is EQ corrected for baffle step, the overall rising response of that driver and for a small resonance peak around 800Hz similar to the response shown in the SEAS driver spec sheet. I have all sorts of measurements and can show some that support this assertion and some that don't. Indoor measurements are somewhat very dependent on window setting and positioning trying to prove it is flat 100-500 is problematic. so we better not go there. I don't have outdoor measurement, but the preponderance of my work suggest the near-field response of the speaker is basically flat with the base EQ until I add this 400Hz filter and one or 2 more below 300Hz to address other modes below the Schroeder.
Yes, I saw your speakers were DIY and figured you would have some form of diffraction loss filtering (how much was added is unclear). Yes, best measured free space, but regardless, it seems you've accounted for that. So obviously more of a speaker/room modal issue.
One of the main reasons I got in the business. A 6.5" midbass is going to be largely omnidirectional to around 800hz, I assume your baffle isn't much wider than 8 or so inch, so you will get a lot of front wall reflection behind the speaker, combined of course, with all other directions as it radiated 3D, so its never quite that simple. Solution there by my way of thinking, is cardioid radiation, something quite doable with DSP. Then EQ if needed. "Fix" the source first.

Room effects? Yes, mine have serious room effects that appear in the midrange up to 600Hz. The question is, it better to EQ those even though it is above the cited upper end of the Schroeder transition frequency or not. My experience is that it is clearly demonstrable that it is better to EQ out my dip even that happens with 20ms delayed.
It's always source/room effects. Not separable. Put a cardioid or dipole where a monopole was, same room, very different result at the LP. Can't be viewed independent of one another.
Yes, the upper range recommended by the paper may need revisiting and I have always contended with Olive and Geddes et al, that EQ alone is not enough. That the radiation characteristics of the source can go beyond what EQ can address. Then make so called passive/active "treatments" a final step, if needed. Clearly you are hearing audible benefit from EQing higher than recommended.


No, I haven't changed speakers. I have nice reasonably uniform horizontal dispersion as shown below. I have close phase tracking through the XO so the central lobe is stable. The vertical dispersion is as expected for a 7" MW and 1" dome; fairly narrow and asymmetric (not shown). It's unavoidable with this configuration.
Yes, that is very good response. Perhaps some diffraction effects (not sure of your baffle edge shaping), but very good overall. I imagine a lowish XO. By "different", I meant to see if the 300-600hz dip was consistent.

If it is recognized that in some cases EQ departing from flat axial speaker output may provide some benefit as high as 300-600Hz then that accommodates my situation. It may be that simple.

Possibly Toole's guideline is intended to provide a very good reliable result that is acceptable for all practical applications and I am just fussing over the last bit of improvement possible beyond that? At least that is one possibility.

It's not like the sound is poor without the filter it is just noticeably better on some material with it. I am very pleased with most all aspects of my current setup. It measures well and has very good SS&I. My interest is as a hobbyist who wants to learn and experiment to develop better understanding and thus drag out the last bit of sound quality that this setup can muster - just for fun. Since I have no source of comparison to other well calibrated systems or to experienced listeners, there is no telling how my setup really sizes-up by comparison. If you or Wayne want to stop by and critique the sound quality, please do. Unfortunately that is probably not going to appear high on your to-do list however. :(
Nothing to disagree with there and if I'm ever up that way, thanks for the invite. Closest I usually get is Axpona CHI.

cheers,
 

·
Premium Member
Joined
·
1,562 Posts

I'm still a bit confused by your case. Aren't you saying no EQ is used/required above 800hz?
Sorry for the confusion. It is probably due to my way of thinking of the overall EQ situation. I was trying to differentiate between the 'base EQ' and the 'room EQ' portions of my total EQ settings. I'm calling 'base EQ' the filters that are needed to create a flat near-field speaker response. This would be a part of a speaker design that is intended to be sold commercially. It may include compensation for driver response, baffle step and possibly even diffraction effects. The 'room EQ' portion are the filters needed to achieve the desired house curve at the LP. The point was that above 800Hz the filters needed to create a flat speaker response work reasonably well to also achieve the preferred house curve at the LP. Only possibly a some minor roll-off adjustment would be needed for my LP distance. I am agreeing with the paper that having a smooth direct response of the speaker would work fine above 800Hz in my case.

It was my preference for a filter at 400 Hzthat I was confused about. I, perhaps mistakenly, understood that it would be better not to use that 'room filter' as it is above the modal range.

Some charts may help the understanding. Below is; 'base EQ' for the FL and FR speakers, LP measurement that include the total EQ filters (The 4 shown above and 3 lower ones including the 400Hz filter), 16" measurement with the 'base EQ' (The 4 filters shown). While my total EQ was actually determined at the LP. The measurement at 16" shows that the direct response above 800Hz is relatively flat hence the observation that there is good EQ agreement above 800Hz for these 2 methods in this case.

Other info relative to your comments:
As you noted, I have tried to keep the XO on the low side to help keep a smoother horizontal dispersion handoff between drivers. The XO approximates an acoustical LR-24 at about 2.2kHz. The front wall is 20 feet behind the speakers so no early reflections there. There are the diffraction effects of the box design and possibly that is at least partially why the higher frequency filters are needed in the base EQ to get a flat direct response. I also attached a measurement file of the 3 front speakers as the LP in case other charts of the overall setup are of interest.

Text Line Font Parallel Design

Text Green Line Blue Font

Text Blue Line Plot Yellow

View attachment LP DLNA-2-3.mdat
 

·
Super Moderator
Joined
·
1,674 Posts
Thanks and sorry for late reply, got busy with some orders. Looks good.
I think we are in agreement that no or minimal EQ should be needed >800hz and I'm not quite sure if we disagree with Toole, as his own data show issues to 500+ Hz in most rooms, so maybe the "room" to "speaker" transition region is well above 200-300hz, possibly as high as 800hz.
20' from front wall? Lucky you.

cheers,

AJ
 

·
Premium Member
Joined
·
1,562 Posts
Hi Raimonds,
Thanks for posting that.
I have read Dr. Toole's popular book, his recent paper and seen your comments. As you point out, there is little in specific and practical guidelines for studio/home EQ setup. I find it impossible to understand Dr. Toole's research and the citations well enough draw any confident conclusions of my own on how to proceed with EQ. I am looking forward to more detail coming now that both Dr. Toole and David Smith mentioned that something is in the works. Possibly that will be more helpful. My background in product development and application engineering includes some time on an SAE standards committee. There is lots of 'corporate interest' lobbying in that type of process that impedes development of clear standards. No company wants to see their products/practices fall into a disadvantageous market position. It seems to me the music creation industry is all over the place in terms of studio EQ practices. Hopefully the charts I've seen showing gross variability are dated and things have improved. Possibly this new work will push it along. I am not too confident however as the market forces are strong and entrenched.

I am not familiar with your products and practices and I found it difficult to follow your thoughts on how EQ is best done.

> Sound intensity in a window of some size at the LP?
> Localized Power response in a window of some size at the LP via either 'moving mic method' (MMM) or individual mic positions?
> Should there be a defined house curve for a defined setup or are there too many parameters to make that practical?
> I understood that you advise linear phase response at the LP through the full range. Correct? Some hobbyists indicate linear phase in bass range is not as important and others do not hear any improvement. Also I am understanding that linear phase at the LP would be in reference to the direct sound and not in trying to chase the phase rotations due to room influences (reflections and modes).
> Do you agree that a single measurement at the LP position is adequate for bass range EQ?
> Do you agree that in most situations the EQ result above maybe 800Hz will be reasonably similar whether the measurements are done via quasi-anechoic at say 1m Vs the LP if an MMM method is used given that there may be appropriate HF roll-off/absorption due to distance?

If my situation is typical, EQ of the bass and HF range practices seem relatively clear compared to the 300-800Hz midrange. Asymmetric room setup and limited acoustic treatments as I have make the EQ of room effects in that range difficult to address. I would agree that it is impossible for EQ to correct the issues, but find that it can help mitigate the severity of them. A professional studio is better able to address the source of the issues through room design and acoustic treatments. I have self-imposed constraints in that area and instead just want a clearer understanding of a process or guidelines to mitigate the damage.

I am not expecting a detailed response to any of this long post, but If you would care to cite a reference to your thoughts on any of these areas, or provide thoughts on any of this here, I would appreciate that. I am mostly just trying to understand how best to deal with the midrange EQ in a nonideal room situation. I like my current midrange EQ setup far better than avoiding EQ in that range. Many experienced hobbyest have advised that any EQ there to address room effects is not advised. I am more inclined to suggest to other hobbyists that they can improve the sound quality via EQ in this range if reasonable methods are used.
 

·
Registered
Joined
·
1,295 Posts
My uncle has told me that FR is hugely not important... as long as your within a +/- 3 or 5db then your good.

He does praise proper toe of the speakers.... position within the room... and exactly symmetrical layout and having the speakers elevation, plumb and level exact to to an accuracy of .05°. He said this is vital for pinpoint sound-stage.

I've listened and his SSI is amazing even though there are a few faults in the FR (particularly the highs falloff too fast imho).
 

·
Premium Member
Joined
·
1,562 Posts
Thanks Talley,
I take that to mean that the anechoic axial response of the speaker should be flat and room effects should not be addressed. Recently, a smooth off-axis response is also often stated as an important factor. This apparently holds up so long as the room is symmetrical and the setup is ideal for speaker, listening positions and appropriate acoustical treatments. I have seen that often stated before and it is rational to me that very good results could be expected. Most, Including Dr. Toole seem to agree that the bass range still benefits from EQ and I assume your uncle does also. But... Does that mean high end studios do not ever EQ above the Schroeder frequency to achieve improved results? Is 1/3 octave RTA, DRC, Audyssey, YPAO, Dirac, Acourate, et.al., all marketing schemes with no real value? I really don't know what EQ practices are common there, but suspect some EQ is applied. If not, then does this change in the case of non ideal setups as mine? Possibly EQ is more valuable in that case. I would think that it might be even more effective there.

I was mostly fishing for more practical guidance relating to what best practices are. From my experience it seem very beneficial to EQ the midrange and to shape the HF. I am a hobbyist having only experience with my room setup, so I am interested as to how my methods compare to more professional ones.
 

·
Registered
Joined
·
1,295 Posts
Thanks Talley,
I take that to mean that the anechoic axial response of the speaker should be flat and room effects should not be addressed. Recently, a smooth off-axis response is also often stated as an important factor. This apparently holds up so long as the room is symmetrical and the setup is ideal for speaker, listening positions and appropriate acoustical treatments. I have seen that often stated before and it is rational to me that very good results could be expected. Most, Including Dr. Toole seem to agree that the bass range still benefits from EQ and I assume your uncle does also. But... Does that mean high end studios do not ever EQ above the Schroeder frequency to achieve improved results? Is 1/3 octave RTA, DRC, Audyssey, YPAO, Dirac, Acourate, et.al., all marketing schemes with no real value? I really don't know what EQ practices are common there, but suspect some EQ is applied. If not, then does this change in the case of non ideal setups as mine? Possibly EQ is more valuable in that case. I would think that it might be even more effective there.

I was mostly fishing for more practical guidance relating to what best practices are. From my experience it seem very beneficial to EQ the midrange and to shape the HF. I am a hobbyist having only experience with my room setup, so I am interested as to how my methods compare to more professional ones.
I see value in EQing... but my uncle is a die hard. His system is pure analog. source, pre, amp... speaker done. I think he would value from it but nothing I can do to change his 45yrs of hi-fi.
 

·
Registered
Joined
·
12 Posts
Thanks, guys, for this really interesting discussion (& you especially Wayne, for getting it going)! My knowledge is not sophisticated enough to understand all that has been said, but I do get (and appreciate) some of it. Thanks again.

Jack Brent
 

·
Premium Member
Joined
·
9 Posts
Hi Raimonds,
Thanks for posting that.
I have read Dr. Toole's popular book, his recent paper and seen your comments. As you point out, there is little in specific and practical guidelines for studio/home EQ setup. I find it impossible to understand Dr. Toole's research and the citations well enough draw any confident conclusions of my own on how to proceed with EQ. I am looking forward to more detail coming now that both Dr. Toole and David Smith mentioned that something is in the works. Possibly that will be more helpful. My background in product development and application engineering includes some time on an SAE standards committee. There is lots of 'corporate interest' lobbying in that type of process that impedes development of clear standards. No company wants to see their products/practices fall into a disadvantageous market position. It seems to me the music creation industry is all over the place in terms of studio EQ practices. Hopefully the charts I've seen showing gross variability are dated and things have improved. Possibly this new work will push it along. I am not too confident however as the market forces are strong and entrenched.

I am not familiar with your products and practices and I found it difficult to follow your thoughts on how EQ is best done.

> Sound intensity in a window of some size at the LP?
> Localized Power response in a window of some size at the LP via either 'moving mic method' (MMM) or individual mic positions?
> Should there be a defined house curve for a defined setup or are there too many parameters to make that practical?
> I understood that you advise linear phase response at the LP through the full range. Correct? Some hobbyists indicate linear phase in bass range is not as important and others do not hear any improvement. Also I am understanding that linear phase at the LP would be in reference to the direct sound and not in trying to chase the phase rotations due to room influences (reflections and modes).
> Do you agree that a single measurement at the LP position is adequate for bass range EQ?
> Do you agree that in most situations the EQ result above maybe 800Hz will be reasonably similar whether the measurements are done via quasi-anechoic at say 1m Vs the LP if an MMM method is used given that there may be appropriate HF roll-off/absorption due to distance?

If my situation is typical, EQ of the bass and HF range practices seem relatively clear compared to the 300-800Hz midrange. Asymmetric room setup and limited acoustic treatments as I have make the EQ of room effects in that range difficult to address. I would agree that it is impossible for EQ to correct the issues, but find that it can help mitigate the severity of them. A professional studio is better able to address the source of the issues through room design and acoustic treatments. I have self-imposed constraints in that area and instead just want a clearer understanding of a process or guidelines to mitigate the damage.

I am not expecting a detailed response to any of this long post, but If you would care to cite a reference to your thoughts on any of these areas, or provide thoughts on any of this here, I would appreciate that. I am mostly just trying to understand how best to deal with the midrange EQ in a nonideal room situation. I like my current midrange EQ setup far better than avoiding EQ in that range. Many experienced hobbyest have advised that any EQ there to address room effects is not advised. I am more inclined to suggest to other hobbyists that they can improve the sound quality via EQ in this range if reasonable methods are used.
Thank you for your valued opinion and questions!
I will try to answer.

*) As I am pointing to focus on loudspeaker performance I do not use such term as Listening Point. Even very early users of my works were referencing lack of any sweet spot and any connection to some listening point when the correction was introduced. Of course, there are some special cases when loudspeakers have very narrow directivity. Special measures are taking place in such cases.
*) The loudspeaker’s Sound Power response (for far field applications) is base.
*) The MMM is a trial to reinvent some part of this work, published in 2005
http://www.google.com/patents/US8121302

*) I did not see any benefit of use of Linear Phase filters in any application. I am using Minimum Phase correction filters from very start in 2002. Because of 1) no latency 2) the minimum phase problems of loudspeaker are corrected ideally- as from amplitude as from phase prospective. The beauty of that you can see in LF correction when large GD caused by main resonance of loudspeaker, is removed.

*) I do not understand any attempts to do some correction from one point measurement (or few points) that are connected to some LP. Because of fact that any trial to fill up some dip on FR (caused by room modes or interference – non minimum phase behavior) is requesting narrow resonant peak on equalizer’s FR. That two „processes” are not with opposite sign – they do not compensate each other as it is with basic assumption of the use of correction that - predistortions, which are introduced by equalizer, will be compensated (neutralized) by distortions of system that we are truing to equalize.

If we are looking for exception quality in LF we should turn to near field setup with appropriate EQ. The TDA EQ is for that.

Our, human, listening perception is build such way that it extracts the „color” of sound source even in case of far field. We will percept any predistortions as distortions of sound source if such predistortions are not closely connected to real distortions of source – loudspeaker. This is true especially up from midrange.


... as long as your within a +/- 3 or 5db then your good.
This is true for listening for enjoyment and when „problems” are wide band.

The job of recording/mix engineer is requesting decision making within accuracy of 1...2 dB and the monitor system with respective accuracy is requested.

If you are willing to catch (listen for) nuances of different „handwriting style” of each engineer/producer/label, you should have a loudspeaker system with appropriate accuracy.

...the anechoic axial response of the speaker should be flat...
It is impossible to find true axes even for loudspeakers with smooth and wide directivity...


But, if little bit touch the room, the most degradating factor of room are its windows.
Nothing can be more problematic than a real large window next to yours speakers.

BR,
Raimonds
 
21 - 40 of 42 Posts
Top