The starting baseline is a speaker with a flat gated on-axis response above 200-300Hz and a smooth roughly linear directivity vs frequency curve. Then EQ the continuous pink noise response below 200-300Hz at the listening position to be roughly flat or slightly tilted down -1dB per octave. Then adjust the bass level up or down to taste.
Incidentally, I think this is consistent with Olive's findings in David S's post quoted by dantheman above, although I don't think the third factor should be called flatness, in case readers mistake it for levelness. Everyone, Olive included, knows that a system that produces the same SPL in the bass and the treble when excited by continuous pink noise (i.e. levelness) will sound much too sharp in the treble and lacking in the bass. The third factor, "wideband in-room response smoothness", actually refers to a smoothly falling pink noise response curve, in fact linearly falling when measuring dB against log frequency.
Last edited:
Why is this though ? Has anyone looked seriously into why this is the case, or do we all just take it as a given ?
I've noticed that whenever I measure the in room response at the listening position of a speaker where the bass is subjectively well balanced with the rest of the range, the entire bass response (averaging out peaks and dips) measures roughly 6dB-8dB higher than the rest of the spectrum, even with only one speaker playing during measurement.
The transition range between this "boosted" bass and the rest of the range seems to be from about 100-400Hz, from 400Hz up to ~10Khz the response is essentially flat, maybe falling very gradually.
(This is consistent with what would be expected from a speaker of a typical baffle size which measures flat on-axis in an anechoic chamber, then placed into a room)
This is an un-gated measurement made with an omni-directional microphone relatively far from the speakers, and it doesn't matter whether it's based on un-gated noise techniques or a sine wave sweep.
Trying to equalize the bass down to a "flat" response measured in this fashion results in a sound that is extremely thin, unpleasant and lacking in bass and warmth.
But why ? If the room boosts the gain at low frequencies so much compared to an anechoic response, why should this sound right, and trying to equalize it out sounds wrong ? Equalizing out the individual room node peaks provides an improvement, but trying to remove the overall ~6dB gain across the entire bass region makes things worse.
Surely if you reproduce the tonal balance of an original recording, including allowing for room gain in the bass, it should sound right ? (Yet it doesn't)
I'm wondering whether it's due to the whole circular reference problem that we have in audio - where music is mixed to sound right on monitor speakers, and other available speakers, and speakers are designed to measure flat in an environment (anechoic) in which we never listen to them.
If we start off with the arguably arbitrary premise that a speaker should be designed to measure flat in a truly anechoic environment, somewhere a speaker will never be used, then such a speaker introduced into a room will have the bass boosted because a room, even well away from most boundaries is still much closer to 2pi than it is to 4pi, thanks primarily to the nearest boundary - usually the floor in a home or the wall in a mixing studio.
If such a speaker is then used for mixing music such as in a recording studio, this low frequency gain will then be "allowed for" by the engineer during the mixing process, whether consciously or not, resulting in a mix where the bass is less than it otherwise would have been. Most engineers also "check" the mix on a few other speakers, such as home systems, thus reinforcing this circular reference issue.
Does a flat in room response sound thin because the "reference" to which recordings are produced are speakers which measure anechoically flat but have increased bass installed in real listening and mixing rooms ?
Another possible factor is that speakers are invariably designed and measured as single speakers, anechoic chamber or otherwise, and a stereo pair will introduce an additional skew in the balance between bass and the rest of the frequency range in a reflective room.
In a far field (in room) listening position approximately eqi-distant from a left and right speaker, the bass from two speakers will add coherently and give an overall 6dB increase over a single speaker, (not allowing for differences in excitation of the lateral room node between a single speaker and two speakers) however, midrange and treble due to their short wavelengths and random phase relationships in the reverberant/reflective field will add non-coherently, resulting in a shift in the power response of the room from a single speaker to a stereo pair which effectively gives an extra 3dB of gain to the bass in a stereo pair referenced to a single speaker, with a transition corner frequency somewhere around 300Hz - the frequency at which most rooms go from nodal to non-nodal.
So are we in this predicament precisely because the reference for measuring speakers is so different from where speakers are generally used, at least for mixing music ?
What does music sound like played in an anechoic chamber on a speaker that measures flat anechoically ? Ignoring the loss of reflections and reverberation, is there still a perceived ~6dB lack of bass on typical recordings ?
Last edited:
Hi,
Thanks for posting that link.
I think the most interesting aspects of it are the statement that stereophonic phantom imaging is unnatural phenomenon (page 5), and the remedy to it is cross talk cancellation (pages 6-8), and his first hand cure to the problem is to use anti-HRTF filtering (pages 10-12).
Maybe in the future we can see stereo-dipole like implementations more common.
- Elias
Thanks for posting that link.
I think the most interesting aspects of it are the statement that stereophonic phantom imaging is unnatural phenomenon (page 5), and the remedy to it is cross talk cancellation (pages 6-8), and his first hand cure to the problem is to use anti-HRTF filtering (pages 10-12).
Maybe in the future we can see stereo-dipole like implementations more common.
- Elias
Originally Posted by tnargs
Everyone, Olive included, knows that a system that produces the same SPL in the bass and the treble when excited by continuous pink noise (i.e. levelness) will sound much too sharp in the treble and lacking in the bass.
Everyone, Olive included, knows that a system that produces the same SPL in the bass and the treble when excited by continuous pink noise (i.e. levelness) will sound much too sharp in the treble and lacking in the bass.
Sure, it's well understood. But I can't do it justice, bit rushed now. Part answer lies in how the room affects the direct/reflected sound proportions at different frequencies, and how the ear distinguishes between direct and reflected sound at different frequencies.
Except you shouldn't be EQing the in-room response to a -1dB per Octave curve. The flat anechoic response above 200 will transfer over to a flat direct or early soundfield in the listenng room. That is what you want.
Room curves are a roundabout way of getting to a flat direct field without measuring it directly, yet no generalized curve will work for all rooms or all speaker directivities.
Everything I read points in one direction as a universal spec:
Design the speaker to be very flat and smooth on axis under anechoic conditions. Minimize the off axis variations to cover the likely listening window.
Give flat and extended bass response but design with regard to the room boundaries. That is, below 200 Hz design around in-room response. Allow for some loudness (compensation) related bass rise.
Sound power should be free from peaks but may have holes.
David S.
Thanks guys, some good thoughts and info. Much of what I was looking for.
DBMandrake is going where I've been thinking. What is typical in-room response of the mixing and mastering suite? Especially mastering, where the final balance is made. I wonder if there has been a survey made of mastering suites and their measurements? Of course for me, that would have to include many now long gone. Since the mastering suite is where the music is tweaked to "sound good" - it would be interesting to know what those rooms sound like.
Recordings have gotten heavier in bass over the years, but its that just a matter of taste or is it related to mastering suites?
DBMandrake is going where I've been thinking. What is typical in-room response of the mixing and mastering suite? Especially mastering, where the final balance is made. I wonder if there has been a survey made of mastering suites and their measurements? Of course for me, that would have to include many now long gone. Since the mastering suite is where the music is tweaked to "sound good" - it would be interesting to know what those rooms sound like.
Recordings have gotten heavier in bass over the years, but its that just a matter of taste or is it related to mastering suites?
Pano-Funny you ask. I had the same question a couple years back and contacted my local mastering studio. He was VERY happy to have someone come listen that "would appreciate the sound."
Studio B Mastering It sounded wonderful from what I remember. He keeps telling me that he's coming to listen to my stereo...
While this may be the case in practice, how much of this being "correct" as a spec is simply the circularity of the whole situation, and not some fundamental truth?
Eg, Step 1 - decree the spec to be flat anechoic response based on a theoretical ideal, Step 2 - design a speaker that matches this requirement, Step 3 - place it in a mixing room that causes the bass below ~200Hz to inevitably be boosted somewhat, Step 4 - master recordings to sound balanced with said speakers, and now these recordings will (generally) sound best with a speaker following the same flat anechoic but in-room bass boosted response.
Step 5 - build consumer speakers to match this decreed standard, and the end result is overall balanced, when used in rooms, provided that the recording engineer took some effort to balance the response to sound good in their mixing.
Step 6 - recording engineers are forced to adhere to this circular situation, because otherwise if they were to equalize the bass flat in their monitoring booth and balance the recording based on that, the bass will be excessive in normal listening environments with un-EQ'ed speakers that are used in rooms but designed based on anechoic response.
Step 7 - return to Step 2 😉
However, what happens if you were to try to single (stereo) mic record a performance and play it back on such a system with no EQ whatsoever, what will it sound like ?
Many mic's used for recording roll off below 100Hz or so, but I've noticed if I make recordings of every day sounds using a measurement microphone which is flat to below 30Hz, and play it back without alteration on a system that is balanced to sound good on commercial recordings, (eg an in-room bass response that is raised roughly 6dB over midrange/treble) the bass does indeed sound excessive.
I remember recording the sound of a fairly air tight door closing normally from about 3 metres away with a measurement mic, playing it back, and the low frequency (below 50Hz) "whump" from the door closing was quite startling and out of proportion to the higher frequency components of the sound, when compared to the live event. Also the low frequency drone from vehicles passing by outdoors that was hardly noticeable when heard during the recording process, dominated the higher frequency components of the car sound when played back.
This suggests to me that in the current situation we are relying on EQ during the mastering process to get a pleasing over all result.
All good advice, and so long as recording practice is based on these standards, speaker designers will need to follow them to get good sounding results, even if the in-room response of the speakers is technically incorrect...
In general recordings have got a LOT more heavy on bass over the years, but I'm not sure there is one single reason for it.
For example play some late 70's / early 80's Alan parsons project CD's and on many songs you'll find some very thin sounding stuff with anaemic bass, even on a speaker that measures +6dB in-room in the bass, as we're discussing.
The same albums on vinyl were much warmer sounding, (although still a bit short on bass by today's standards) so I think some recording engineers went through a "brightness craze" when CD's first came out in the early 80's, where the response seem to be tilted up towards the treble.
Maybe they were trying to show off the high frequency response of CD's - who knows. The end result was a pretty unsatisfactory balance, at least to me. Some CD's of that period sound far better with a bit of variable slope bass boost.
On the other hand some modern recordings have so much bass it's crazy.
One thing I have wondered is whether some modern studios are trying to EQ the bass response of their monitor speakers "flat" in-room. If they are, it will cause them to mix recordings that have even MORE bass in them than before. Perhaps this is happening in some cases ?
I think it's largely a matter of changing tastes though. Loads of bass and treble and dynamic range compression seems to be the norm for many music genres today unfortunately. Too much bass and treble can be compensated for in the playback chain to some extent but there's nothing we can do about the compression 🙁
This relates to a point I made a long way back in this thread I believe. That is, the premise that flat is not correct ignores the roll of the recording engineer. If the recording engineer monitors the recording on a system with flat axial response then he is going to make the adjustments for HRT and all other types of effects while he sits there and monitors playback. Now if you play back that recording on a speaker with flat response and it doesn't sound right then the issue should not be the axial response. It has to be one or more of the following: room effects, power/polar response, or the recoding engineer has a tin ear.
The obvious contradiction here is that if I happen to listen using the same speakers as the recording engineer in a room with very similar characteristics, the "flat is not correct concept" would say that I need to roll off the highs of the same speaker used to monitor the recording in the first place.
It comes down to simply that if a well recorded piece of music doesn't sound correct on your playback system it is because there is a substantial deviation between the characteristics of you room and speakers compared to those used in producing the recording in the first place.
In the exchange I had with Davey yesterday about transfer functions for the Orion one point was missed that I realized over night. Those modified transfer functions were developed to yield flat on axis response. After that the tweeter level was adjusted subjectively and after that the shelf filter was added.
No, those transfer functions already include the tweeter level adjustment and the shelf filter. The final result is a non-flat, on-axis, free-field, acoustic response.....although not shelved down nearly as much as -3.2db. 🙂
Cheers,
Dave.
Last edited:
That may be true for "engineered" recordings. However . . . if you record an orchestra with a pair of NT1000 ORTF (low noise, nicely "flat", reasonably constant polar, ie. a good mic for the purpose) and play the recording in a "normal" living room with anechoically "flat on axis" uniform polar response loudspeakers it will sound too bright. That's with *no* equalization by the recording or mastering engineer. Why is that?
Of course when the playback system is balanced to make that recording sound right "engineered" recordings are all over the ballpark. At least we know why that is . . .
I'm not as worried as others are about the "circle of confusion". The real issue is what gives us consistency of perception.
Much of this discussion has been about the relative merits of measuring steady state response in a room vs. measuring the direct response or anechoic response. People that measure the room curve generally accept that a flat target will sound too bright, but don't seem too concerned about why? Clearly room size also plays a big factor and the larger the room the more you need to roll off the room curve (this by convention). The most extreme case is the Cinema "X" curve with 10 or more dB of rolloff. The modern X curve has a set count related rolloff. Why?
Now the only way this makes sense to me is that the room curve is a very poor measure of perceived balance. We aren't really measuring the right thing. Since most of these room curves can be met with a fairly flat direct field you have to wonder if that isn't the real key?
There is a lack of an absolute standard, yet if we can get most recordings and most speakers to be a good fit then do we really care? Lets say we woke up tomorrow to find that every B&K measuring microphone had a previously unknown 5 dB mound at 3k. Every speaker had been designed with a complimentary divot (so that it measured flat in an anechoic chamber) and most every recording engineer ended up goosing the 3k region to counteract the dished out speakers. Would we care? It's a bit as if RIAA EQ happened by accident rather than on purpose. As long as playback matched recordings then we would still be okay.
Studios can be inappropriatly EQed but they also tend to sample a lot of external recordings and also to send their recordings from studio to studio. If they are the "odd man out" it is generally revealed over time.
David S.
The microphones lacking any sort of human-like HRTF characteristic ?
An orchestra in a concert hall is something of a special case in that it's a large very reverberant environment, so there is a lot of high frequency energy coming at you from many different angles.
Human head and ears will attenuate high frequencies from many of the different incident angles far more than the simple cardioid patterns from a pair of microphones like those.
The net result is that for a given reverberant field an actual pair of ears on a head will be collecting less high frequency energy than a couple of cardioid microphones in a "stereo" configuration.
Have you tried making a true binaural recording with a dummy head in the same environment ? Does it also have a perceived excess of high frequencies ?
Hi Dave,
Yes I realize they include the shelf. My point is the how the shelf came about, and from the history on SL's page and Don's discussion it certainly lookls like 3.2 dB to me.
The transfer functions were first modified to bring the response to flat according to SL's page,
Then a subjective over all tweeter level adjustment was made, again according to SL pages,
They apparently were not happy with the result of simply reducing the gain on the tweeter so,
What ever the final free field response ended up as I don't know. It appears that for the V 3.0 it was -1.8 dB according to SL's site and Don's write up of the progression of the mods. With V 3.2 it appears to be -3.2, again, both by the plots on SL's site and as discussed in the revision history by Don,
It certainly seems pretty clear but there are no acoustic measurements I can find that show the actual final result.
All that is fine. The speaker sounds better, great. There is absolutely noting wrong with that.
The issue I have has nothing to do with any of that. The issue is simple; that flat is not correct for stereo implies that recording engineers mix recordings too hot. Whether it be 1.8dB or 3.2 dB, or anything else, if the recording engineers were "rolling off the highs" appropriately in the recording process then flat would be correct. So the blame lies with the recording industry? Are all recording engineers are turning up the treble? Do the all suffer from hearing deficiencies? Does this permeate the entire music industry be it rock, jazz, classical, what every? Are there no quality recordings? Let's use a little common sense here. No, I think it is necessary to ask where is the problem, in the recording studio or in the home? It should be apparent that the problem is not at the recording end but rather that the speakers and environment used for play back are significantly different that those used for monitoring during production. What do you suppose would happen if SL were sitting in his listening room with a pair of (flat) Orions without the shelving filter mastering a recording. After he was done would he then have to insert the shelf for playback? I don't think so.
So now, as usual everybody can jump on my back and say what an arrogant SOB I am and how I am besmirching SL and all that. I have heard it all before. So please, don't kill the messenger. I'm just repeating what is presented and making a simple observation. I'm not asking any body to start a revolution, I 'm not even asking anyone to agree with me. (I really hesitate to post this. )
No rule or principle according to me.
I can still hear above 20 khz, so if I may, I will consider myself a reference listener. Almost 29 years old to be specific.
When I designed the crossover of my so far best speakers, I switched several HF attenuation resistors and tried several crossover slopes and ended with 3rd order electrical crossover and 1 ohm series resistor.
Below you'll see a FR taken in laboratory conditions (the certificate that comes with the driver), unsmoothed off axis response taken in small room (16-17m2) and smoothed off axis FR taken in the same room.
Note that on the smoothed and unsmoothed FR's there are two HF drivers working together (back then I wanted a 300+ w rms system with near field of at least 4m so i seeked for some sort of directivity.. - that's another story).
So back to topic... I listened to the slightly padded ribbon tweeter for several months like that - mean sensitivity of the system above 150 hz is about 89-91 db/w/1m all the way to 40 khz... and it's OB up to 10 khz.
What I noticed is that it is a matter of habit and a matter of record, but it's much more a matter of habit! After several months I EQ'd + 3-4db at 16 khz and little less at 10 khz.
Then I switched back to linear with only 4-6 db at 20-30 hz, but not always...
Every time I changed the settings I liked the new ones more over the old ones. But I switched between mainly two settings all the time.
Then I got a turntable and stopped using the EQ with it! The RIAA of the preamp was pretty enough for everything... only +3db @ 50 hz sometimes... and + 1-2 db@ 100 hz from the power amp... But linear is as good as it gets!
Then again i noticed that at 0.8 grams or pressure I need more gain from the preamp and at 1.5 grams the highs begin to sound a little congested as from a overdriven compression driver...
Probably my biggest breakthrough was limiting the listening of CD's and switching to vinyl, flack and other lossless formats... - and that's very serious!
Before that I had remarks on both my main systems and speakers. I had plans for upgrading one of them and replacing the other. Now I'm happy with whatever is recorded.
In the time of listening to CD's I was alway in some sort of searching and redoing and tweaking something. Crossover this, crossover that, not enough highs, too much highs, not enough bass, unconvincing bass, distortion and etc...
I have investigated what is a "Reference Listener" - this should be male person about 40-45 years old and preferably a sound engineer. At ages below 40 people hear higher sounds stronger and the low sounds weaker, above 40-45 it is the opposite. So a 40 year old male sound engineer with healthy and well trained hearing should be the reference listener. - And that's important. Not just everybody can lay creditable opinion about sound.
Another very strong consideration is the fact that distortion can make us reduce the volume. If it is recorded distortion or noise we'll do exactly as described in this topic - reduce the volume of the range that carries distortion and noise. When we reduce the volume the unpleasant sounds fall beneath our hearing threshold and beneath the HF driver response threshold - you know that distortion and noise are always quieter than the main signal. - Having in mind the above it is obvious that the better the tweeter the bigger the problem is! And the biggest problem is with ribbon tweeters.
In my opinion this topic and this particular reply of mine have much to do with the effects observed with the EnABL process... in a way...
I can still hear above 20 khz, so if I may, I will consider myself a reference listener. Almost 29 years old to be specific.
When I designed the crossover of my so far best speakers, I switched several HF attenuation resistors and tried several crossover slopes and ended with 3rd order electrical crossover and 1 ohm series resistor.
Below you'll see a FR taken in laboratory conditions (the certificate that comes with the driver), unsmoothed off axis response taken in small room (16-17m2) and smoothed off axis FR taken in the same room.
Note that on the smoothed and unsmoothed FR's there are two HF drivers working together (back then I wanted a 300+ w rms system with near field of at least 4m so i seeked for some sort of directivity.. - that's another story).
So back to topic... I listened to the slightly padded ribbon tweeter for several months like that - mean sensitivity of the system above 150 hz is about 89-91 db/w/1m all the way to 40 khz... and it's OB up to 10 khz.
What I noticed is that it is a matter of habit and a matter of record, but it's much more a matter of habit! After several months I EQ'd + 3-4db at 16 khz and little less at 10 khz.
Then I switched back to linear with only 4-6 db at 20-30 hz, but not always...
Every time I changed the settings I liked the new ones more over the old ones. But I switched between mainly two settings all the time.
Then I got a turntable and stopped using the EQ with it! The RIAA of the preamp was pretty enough for everything... only +3db @ 50 hz sometimes... and + 1-2 db@ 100 hz from the power amp... But linear is as good as it gets!
Then again i noticed that at 0.8 grams or pressure I need more gain from the preamp and at 1.5 grams the highs begin to sound a little congested as from a overdriven compression driver...
Probably my biggest breakthrough was limiting the listening of CD's and switching to vinyl, flack and other lossless formats... - and that's very serious!
Before that I had remarks on both my main systems and speakers. I had plans for upgrading one of them and replacing the other. Now I'm happy with whatever is recorded.
In the time of listening to CD's I was alway in some sort of searching and redoing and tweaking something. Crossover this, crossover that, not enough highs, too much highs, not enough bass, unconvincing bass, distortion and etc...
I have investigated what is a "Reference Listener" - this should be male person about 40-45 years old and preferably a sound engineer. At ages below 40 people hear higher sounds stronger and the low sounds weaker, above 40-45 it is the opposite. So a 40 year old male sound engineer with healthy and well trained hearing should be the reference listener. - And that's important. Not just everybody can lay creditable opinion about sound.
Another very strong consideration is the fact that distortion can make us reduce the volume. If it is recorded distortion or noise we'll do exactly as described in this topic - reduce the volume of the range that carries distortion and noise. When we reduce the volume the unpleasant sounds fall beneath our hearing threshold and beneath the HF driver response threshold - you know that distortion and noise are always quieter than the main signal. - Having in mind the above it is obvious that the better the tweeter the bigger the problem is! And the biggest problem is with ribbon tweeters.
In my opinion this topic and this particular reply of mine have much to do with the effects observed with the EnABL process... in a way...
Attachments
Well count me in with the concerned. 🙂
Although the basis behind a gently falling power response in the treble range being desirable for most speakers makes sense (EQ'ing the room response flat in the treble with most speakers would tilt up the on-axis response undesirably) the large discrepancy between a measured "flat" in room steady state response from bass through to midrange, and the considerably elevated bass response actually required to provide a balanced sound is something that I find hard to explain other than being an effect of the "circle of confusion".
If nothing else, it's something that people need to be aware of when making measurements. Many attempts by people to EQ their speakers / listening environment fail miserably when they naively accept flat steady state frequency response measured at the listening position with an omni-directional microphone as the target. Even if they allow for a small roll off in the power response of the tweeter it doesn't take care of the elephant in the room, which is the bass response.
Many naive auto-eq systems make the same mistake, both with the bass and the treble, and also attempt to make narrow band corrections in the midrange and treble based on single point measurements. (That's a whole different can of worms though...)
You're right that the room curve or steady state room measurement is a very poor correlation with perceived balance. I've always been disappointed by just how poorly it matches up with perceived balance, to the point that it's almost useless. (Well, maybe an exaggeration, but it seems that way sometimes)
It must come down to the difference between on-axis response vs power response reflected back by the room, however there is some debate as to the relative importance of the two. Some say the on axis response is the predominate factor, while others believe the power response is just as important.
Probably the relative importance depends on whether you listen in the near-field where the direct signal is stronger, or far away where you're in the diffuse field of the room.
My gut feeling is that the power response isn't as important as many people suggest, and provided the speakers are far enough away from the nearest walls to delay early reflections enough (more than about ~0.7m) and you're close enough to be listening in the near-field, (which you usually are to get good imaging) that the on axis response by far and away predominates the perceived balance, unless there are huge variations in power response. Small dips in power response seem to go unnoticed unless you're vastly off axis or a long way from the speakers. If I had to guess I'd put the relative importance of on-axis response to power response around 80/20.
If the ear is predominately judging the balance from 200Hz up based on a time windowed on-axis response, yet steady state room measurements are largely affected by the speaker/room power response, it's easy to see why steady state room response doesn't have a good correlation with perceived balance, especially across different speaker types, or rooms for that matter.
All of which is a long roundabout way of agreeing that provided that the power response is not dramatically varying with frequency (such as in the case of a single large full range driver where the treble is tightly beamed) then flat on-axis anechoic response above 200Hz is going to sound a lot closer to balanced than EQ'ing based on a room measurement.
What's going on below 200Hz in regards to balance I'm a bit less convinced about though, it may well be the circle of confusion reigning here, but provided it's known about and accounted for I don't really mind 😉
No, it does not. "Flat" is a fine standard for signal capture and delivery (any standard would be fine were it consistently applied). Different rerpoduction devices in different listening environments (where the variation occurs) would then be tuned so that "flat" input sounded right on those devices in those environments. This is pretty much what THX does for the film environment . . . establish a delivery standard, and separate "tuning" standards for soundstage, studio, theater and home environments so that the common signal "sounds the same" in all of them. For "audio only" you could extend the listening environment to headphones and car, which require their own specific equalizations.
You should be able to switch between headphones and loudspeaker in your listening room and hear the same balance. To do so will require different equalization for the two reproduction devices. And ideally both will then match the balance of the original performance as heard "live" (assuming such a thing ever happened . . . "studio" recordings being adjusted to their own "reality"). As it stands, though, there is no "delivery standard", and no consistency between the equalizations used by different engineers and for different music genra, and we might as well be arguing about fixed or variable inflection points and the best boost/cut slopes for our tone controls (which would itself be better than going on and on about "SL this and SL that" . . .).
And in Conclusion...
Wow... I think the original question that started this discussion was something like, "Is flat where it's at or should I have tone controls in my playback preamp?" I think it's pretty darn clear to me anyway that the answer is yes. Definitely have tone controls in your preamp (yet so few high-end preamps have them...). There are so many issues and variables...
Especially technical people seem to get all caught up in technical accuracy as they understand it. Yet I've never met anyone who knew it all. Acoustics are very complex. The recording process is riddled with all the same acoustic related problems as the playback situation. My head mic recordings sound so balanced that I have to wonder what goes on in so many recording studios. I think that is a big variable. It amazes me how much most of us know about all of this, and yet we still don't have a clear answer as to why we need to correct the final result more often than not, with tone control, to have a really great sound.
Bottom Line: Flat seems to me to be a good starting point, and then from there adjust tone controls to your individual liking. Use your own judgment as to when it sounds right.
The best tone controls in my opinion, would be a four section Baxandall (or James, the passive version) variable slope, designed such that you could create, among other things, your own variation of a reverse Fletcher-Munson "Loudness compensation" curve. I'm in the process of designing and building such a preamp right now. This discussion has confirmed my thoughts and goal. 😎
Wow... I think the original question that started this discussion was something like, "Is flat where it's at or should I have tone controls in my playback preamp?" I think it's pretty darn clear to me anyway that the answer is yes. Definitely have tone controls in your preamp (yet so few high-end preamps have them...). There are so many issues and variables...
Especially technical people seem to get all caught up in technical accuracy as they understand it. Yet I've never met anyone who knew it all. Acoustics are very complex. The recording process is riddled with all the same acoustic related problems as the playback situation. My head mic recordings sound so balanced that I have to wonder what goes on in so many recording studios. I think that is a big variable. It amazes me how much most of us know about all of this, and yet we still don't have a clear answer as to why we need to correct the final result more often than not, with tone control, to have a really great sound.
Bottom Line: Flat seems to me to be a good starting point, and then from there adjust tone controls to your individual liking. Use your own judgment as to when it sounds right.
The best tone controls in my opinion, would be a four section Baxandall (or James, the passive version) variable slope, designed such that you could create, among other things, your own variation of a reverse Fletcher-Munson "Loudness compensation" curve. I'm in the process of designing and building such a preamp right now. This discussion has confirmed my thoughts and goal. 😎
- Status
- Not open for further replies.