The No.1 Website for Pro Audio
 All  This Thread  Reviews  Gear Database  Gear for sale     Latest  Trending
Loudspeakers and rooms - a scientific approach
Old 3 weeks ago
  #61
Lives for gear
 
Brian M. Boykin's Avatar
Quote:
Originally Posted by Fhorn88 View Post
One of the problems is of course, the lack of neutral speakers. The point of this thread is to make people look beyond sales talks about loudspeaker and into comprehensive(!) anechoic data.
And to make things worse, an active set of monitors could have EQ networks within the amp that corrects for the drivers making anechoic FR flat when in actuality your 1v 1kHz tone coming off your console is being EQ’d. I’ve spent hours doing sweeps through my gear learning what it’s doing to frequencies. Sometimes EQ may be required within the network to compensate for a driver that bumps or cuts a frequency. Tontest this you’d have to have sweeps on both sides of the active network and then anechoic numbers to compare to. I can see why companies would not want to share this information. Same reason major studios don’t want to share their room FR.

Statistics are like a bikini. What they reveal is suggestive, but what they conceal is vital.
Old 3 weeks ago
  #62
Gear Nut
 

Some anechoic measurements in the ANSI-2034A standard

A neutral 3-way cone/dome loudspeaker ($2000 /piece)



A not-so-neutral 3-way cone/dome loudspeaker ($5000 /piece)




A neutral PA line array module ($4000 /piece)




A neutral coaxial design ($5000 /piece)




A crime against humanity ($5000 /piece)




A small two way cone/dome ($100 /piece)

Old 3 weeks ago
  #63
Lives for gear
 
Brian M. Boykin's Avatar
Quote:
Originally Posted by deedeeyeah View Post
one more thing on dsp though: pls don't talk about 'room correction' when using a few filters to adjust fr! no filter or dsp can correct the room, you need mechanical devices such absorbers, diffusors, bass traps etc. to do this!
(i know the term is getting used a lot - i find it plain silly!)
I don’t believe in room correction through EQ. I was trying to say, that’s what I thought DSP was. I had apparently fallen victim to that belief. Hopefully I’ll be able to grasp DSP and apply it properly.

Last edited by Brian M. Boykin; 3 weeks ago at 09:28 PM..
Old 3 weeks ago
  #64
Gear Guru
 

Quote:
Originally Posted by Fhorn88 View Post
Let me put it in simpler terms then. If it measures flat anechoically (neutral) with well behaved directivity, we will perceive that loudspeaker, even when put in a room, as neutral - even if the steady state measurements say otherwise. The exception being below the transition frequency of the room, where the wavelengths are too long and we can no longer seperate the direct from the reflected sound.
Ok, but how does that relate to your objection? All you're still saying, seemingly, is that the measurement doesn't reflect perception, correct?

But aren't we adjusting the speaker to get the preferred perceived tonal quality?
Old 3 weeks ago
  #65
Gear Nut
 

Quote:
Originally Posted by mattiasnyc View Post
Ok, but how does that relate to your objection? All you're still saying, seemingly, is that the measurement doesn't reflect perception, correct?

But aren't we adjusting the speaker to get the preferred perceived tonal quality?

That is the point - if one starts with a neutral loudspeaker, set up properly - there is literally, above the transition, nothing to adjust. You know it to be calibrated at the factory, to a standard - if they share those measurements that is. If the desired balance is off - then you know exactly what needs to be corrected in the recording, since your playback system is truthful.

If you start with a lesser or unkown loudspeaker, and have to adjust it in-situ.. Well then.. Since we know in-situ full range measurements are not nearly as reliable as full anechoic measurements, it's a literally a guessing game. You'll be tweaking a target curve, while listening to recordings, made through other (unkown) loudspeakers, which you will in turn use to judge new recordings.. It's the circle of confusion. And the inverse of whatever the true response is of your new system, will likely get imprinted on whatever you are producing in terms of material.

Granted, for many this is a new approach to things - all posted with the best of intentions, to up the standard for all.

It's all further explained on page one.

Last edited by Fhorn88; 3 weeks ago at 06:45 AM.. Reason: typo
Old 3 weeks ago
  #66
Gear Nut
 

Some anechoic measurements as portrayed by Soundstage network, made in the anechoic chamber of the NRCC.

Listening window and horizontal off-axis response of a two-way monitor of a three letter brand that markets to both studio and home users.






Listening window and horizontal off-axis repsonse of a better designed loudspeaker around the same price point



Old 3 weeks ago
  #67
Gear Guru
 

Quote:
Originally Posted by Fhorn88 View Post
That is the point - if one starts with a neutral loudspeaker, set up properly - there is literally, above the transition, nothing to adjust. You know it to be calibrated at the factory, to a standard - if they share those measurements that is. If the desired balance is off - then you know exactly what needs to be corrected in the recording, since your playback system is truthful.
- Not everyone wants a tonal balance that's "neutral". Examples have already been given.

- Not everyone ends up with a "neutral" speaker, for several reasons (though see below).

What do you recommend in those cases?

Quote:
Originally Posted by Fhorn88 View Post
If you start with a lesser
Just because a speaker is not "neutral" doesn't mean it's "lesser" in my opinion.

Also, having looked at the charts you posted how do you define "neutral"? That first 3-way speaker you showed a graph of has a difference of about 8dB between about 3k and 4.5k. The coax looks 'better' with only 5dB swings as far as I can see.

Quote:
Originally Posted by Fhorn88 View Post
or unkown loudspeakers, and have to adjust them in-situ.. Well then.. Since we know in-situ full range measurements are not nearly as reliable as full anechoic measurements, it's a literally a guessing game. You'll be tweaking a target curve, while listening to recordings, made through other (unkown) loudspeakers, which you will in turn use to judge new recordings.. It's the circle of confusion.
I have to question this though. Those of us who work as engineers for a living aren't just paid to push buttons but to use our ears, and to reference previous material. There's a reason why engineers take content of certain types and reference that when mixing. It goes beyond "I wonder what these speakers really are like" to "I want to match x, y, z aesthetically". We don't spend our times listening to just one set of speakers in one environment.

And again; see above regarding what "neutral" really is...

Quote:
Originally Posted by Fhorn88 View Post
And the inverse of whatever the true response is of your new system, will likely get imprinted on whatever you are producing in terms of material.
So I end up with the inverse of that 8dB difference I see in the first chart, correct? Unless I 'correct' it, right?

Quote:
Originally Posted by Fhorn88 View Post
Granted, for many this is a new approach to things - all posted with the best of intentions, to up the standard for all.

It's all further explained on page one.
You shouldn't have put this in the newbie section in my opinion. It goes far beyond what many newbies can absorb and is really neither "short and simple". Seems more appropriate in other sections..
Old 3 weeks ago
  #68
Gear Guru
 

Quote:
Originally Posted by Fhorn88 View Post
Some anechoic measurements as portrayed by Soundstage network, made in the anechoic chamber of the NRCC.
What's your point? Some speakers are flatter than others?
Old 3 weeks ago
  #69
Gear Nut
 

Quote:
Originally Posted by mattiasnyc View Post
- Not everyone wants a tonal balance that's "neutral". Examples have already been given.

- Not everyone ends up with a "neutral" speaker, for several reasons (though see below).

What do you recommend in those cases?
Quite simply, if you feel that way, this thread isn't for you. I'm not here to tell you what to do, I am presenting objective information, for those that are interested in it.


Quote:
Originally Posted by mattiasnyc View Post
Just because a speaker is not "neutral" doesn't mean it's "lesser" in my opinion.
I feel different. This is also not an issue in any other aspect of artistic creation and production. Video editing wouldn't be performed on an uncalibrated system - ever - with good reason. They also wouldn't try to wing it by 'adjusting the picture until it looks right'

Quote:
Originally Posted by mattiasnyc View Post
Also, having looked at the charts you posted how do you define "neutral"? That first 3-way speaker you showed a graph of has a difference of about 8dB between about 3k and 4.5k. The coax looks 'better' with only 5dB swings as far as I can see.
I see +/-3dB swings on the first one - on-axis only, which, if you would have read page one, is not the primary target due to small interference effects. The green line, aka the listening window closely resembles what we perceive as the direct sound field, this as determined by correlating technical measurements to blind listening experiences (not only by me, by research groups with serious funding behind it).

They are both essentially neutral, just different shades of it. Featuring slightly different responses, with different directivity properties - both will provide a similar, though not identical listening experience - they will be however both miles ahead of some other loudspeakers.

To find out which one works the better in your own environment - the blind test procedure is outlined on page one. This isn't a brand or type endorsment thread either - it's about standards and procedures.


Quote:
Originally Posted by mattiasnyc View Post
I have to question this though. Those of us who work as engineers for a living aren't just paid to push buttons but to use our ears, and to reference previous material. There's a reason why engineers take content of certain types and reference that when mixing. It goes beyond "I wonder what these speakers really are like" to "I want to match x, y, z aesthetically". We don't spend our times listening to just one set of speakers in one environment.

And again; see above regarding what "neutral" really is...


So I end up with the inverse of that 8dB difference I see in the first chart, correct? Unless I 'correct' it, right?
If you are aiming for a neutral, natural end result, on a non-neutral monitoring system - you may very well end up with the inverse. Unless you feel you can compensate for that - which may or may not be the case. In any case seems like more chance of error and more trouble than it needs to be?

But, I can understand what you say. My personal feeling is that one should have at least one system that represents a neutral sound field, a standard metric. This does not mean you have to agree with it, or apply it. As I said, this thread is meant for those that want to achieve this, but are struggling.




Quote:
Originally Posted by mattiasnyc View Post
You shouldn't have put this in the newbie section in my opinion. It goes far beyond what many newbies can absorb and is really neither "short and simple". Seems more appropriate in other sections..
Perhaps - on the other hand, it is best to spark an interest early on in my opinion before getting set in a fixed way of doing things. It doesn't mean one has to understand the matter entirely from the start. A simple curiousity is often enough to get people to start thinking more critically. There's nothing more difficult during an install then arguing with seasoned professionals. That's when proposing blind testing comes in handy - some simply flat out refuse to participate, but most can be persuaded that way.
Old 3 weeks ago
  #70
Gear Nut
 

Quote:
Originally Posted by mattiasnyc View Post
What's your point? Some speakers are flatter than others?
Yep - and that the price tag doesn't necessarily reflect that. In fact, nor do most manufacturers. There is a serious lack of objective information in the monitoring/loudspeaker world.
Old 3 weeks ago
  #71
Lives for gear
 

Quote:
Originally Posted by Fhorn88 View Post
(...) if one starts with a neutral loudspeaker, set up properly - there is literally, above the transition, nothing to adjust (...)
this assumption imo is fundamentally flawed (or at the very least, it's wrong to put it in absolute terms).

could you please give us a link to the studies you are referring to?
Old 3 weeks ago
  #72
Gear Nut
 

Quote:
Originally Posted by deedeeyeah View Post
this assumption imo is fundamentally flawed (or at the very least, it's wrong to put it in absolute terms).

could you please give us a link to the studies you are referring to?
They are not free - however if you are an AES member you may want to look into these

http://www.aes.org/e-lib/browse.cfm?elib=5276
http://www.aes.org/e-lib/browse.cfm?elib=5270
http://www.aes.org/e-lib/browse.cfm?elib=12794
https://secure.aes.org/forum/pubs/co...ns/?elib=12847

Furthermore these principles have been a standard in the European Broadcasting Union for over 20 years - it puzzles me why people treat it with such indignation.
Old 3 weeks ago
  #73
Gear Nut
 

A free paper (from Harman - so there is an obvious brand link - but true nonetheless)

https://www.harman.com/sites/default...oScience_0.pdf

Quote:
There are some things equalization can and cannot do:
1. It can help with some, but not all, loudspeaker problems. With comprehensive laboratory measurements (like Figs. 9–12) to work from, equalization can be used to make a
good speaker sound better. Once a loudspeaker is in a normal room, we lose the ability to
measure it in ways that allow us to be totally analytical. Without knowing what is wrong,
we don’t know whether equalization is the right solution for the problem. For example,
equalization can change frequency response, but it cannot change directivity, yet together
they determine the basic shape of a room curve. Poor directivity control, as a function of
frequency, can only be cured by using a better loudspeaker. In general, if the speaker has
been competently designed, it should probably be left alone at frequencies above about 300
to 500 Hz, whatever the room-curves look like
From the European Broadcasting Union tech documents (1998) - Listening conditions for the assessment of sound programme material: monophonic and two–channel stereophonic

https://tech.ebu.ch/docs/tech/tech3276.pdf

Quote:
2.1. Direct sound
The direct sound is defined as the sound field which would be measured, using the same loudspeakers,
under anechoic conditions, i.e.: without the early reflections and the reverberation caused by the listening room
(see Sections 2.2. and 2.3.).
Quote:
2.4. Operational room response curve
The operational room response curve is an important criterion in the evaluation of the mutual influence of
the loudspeaker and the listening room, and hence for the assessment of the listening conditions. It corresponds
well with the subjective assessment of reproduced sound.

...To avoid degrading the quality of reproduction, electrical equalization should be used carefully. It is advisable
to make the corrections in the low–frequency range
(f < 300 Hz) only.
Old 3 weeks ago
  #74
Gear Nut
 

Free AES paper

http://www.aes.org/tmpFiles/elib/20190920/17839.pdf

"The Measurement and Calibration of Sound
Reproducing Systems"
Old 3 weeks ago
  #75
Lives for gear
 

Quote:
Originally Posted by Fhorn88 View Post
They are not free - however if you are an AES member you may want to look into these

http://www.aes.org/e-lib/browse.cfm?elib=5276
http://www.aes.org/e-lib/browse.cfm?elib=5270
http://www.aes.org/e-lib/browse.cfm?elib=12794
https://secure.aes.org/forum/pubs/co...ns/?elib=12847

Furthermore these principles have been a standard in the European Broadcasting Union for over 20 years - it puzzles me why people treat it with such indignation.
thx (for the link)!

regarding standards: every standard serves a purpose and there is nothing wrong questioning a standard from time to time, especially since there were advantages in related fields (such as audiology in the last 20 years), when there is a significant amount of findings (maybe seemingly) contradicting earlier findings which initially led to setting the standards or just by accepting that there are other ways to see the world (or, may i add, that there are other bearers of the same message with a different attitude and tone: no need to lecture me and others on the topic)...

Last edited by deedeeyeah; 3 weeks ago at 09:16 AM.. Reason: first sentence edited
Old 3 weeks ago
  #76
Gear Nut
 

Quote:
Originally Posted by deedeeyeah View Post
thx!

regarding standards: every standard serves a purpose and there is nothing wrong questioning a standard from time to time, especially since there were advantages in related fields (such as audiology in the last 20 years), when there are significant amount of findings (maybe seemingly) contradicting earlier findings which initially led to setting the standards or just by accepting that there are other ways to see the world (or, may i add, that there are other bearers of the same message with a different attitude and tone: no need to lecture me and others on the topic - could even be that someone around here has equal or even superior knowledge and experience in the same field as you!)
Correct - and there are some things in the 1998 paper that already could be improved upon. Overall it still holds up very well. The free AES paper from 2015 would elaborate on it I suppose. This thread wasn't meant to lecture anyone, I hope that is not how your perceive it. I did expect some discussion of course, so I feel the need to respond. Not to lecture, but to answer the questions asked and/or to substantiate my claims. Sceptisism is healthy, and indeed evidence should be provided whenever claims are made. It is in a way, why I started this thread afterall. Again not to lecture, or to impose my views, but to present results of research performed, to those that relate to my writings.
Old 3 weeks ago
  #77
Lives for gear
 
Brian M. Boykin's Avatar
So I’m little Joe Schmo down here that got into recording roughly 15 years ago. I’ve been building up my studio ever since. I consider myself part of the 1st generation of the home studio DAW craze. I’m currently treating a room for tracking and mixing. I have an measurement mic and REW to flash my room and track my progress. I’m using monitors from the 1990’s which have served me well but I’ve been tossing around the idea of upgrading. This is all to place perspective on me.

Now, I spent a day flashing my room roughly a month ago. The FR is pretty straight forward. The waterfall took a little learning and I’m not totally there. I’ve been reading on the Acoustical forum since January relearning list info and getting up to date on new info. I just read a large portion of Floyd Toole on a trip to Seattle. I’ve read almost all of Rod Gervais’s book Building like the Pro’s. I’be exhausted Ethan’s website, GIK’s website, and countless others. His is more perspective. I’m currently dropping $500 to $700 a month on materials to build my own broadband absorbers.

So here’s the poop. I get everything done and flash my room and I have that 200Hz dip in my FR that is shown in Fhorn’s first graph. I have zero data on my current monitors to tell me if that’s my room or the monitors themselves. I now go post my results on the Acoustical forum and plead for help. What do you think I’m gonna be told?

Get rid of that console. Get a smaller desk. Your cloud is not thick enough. Your back wall needs more treatment. Yada yada yada.

And so I start throwing money at a problem that is in the monitor and not the room.

Thank you Fhorn. I spent a good part of my day shopping for new monitors yesterday. I absolutely have to have those FR numbers to know how my room is progressing. Otherwise it’s a crap shoot.

Deedee, most all of the monitors I looked at yesterday had DSP filters built in. I know you said your beyond simple filters so I think a dedicated thread to DSP would be a benefit.
Old 3 weeks ago
  #78
Gear Nut
 

Quote:
Originally Posted by Brian M. Boykin View Post
So I’m little Joe Schmo down here that got into recording roughly 15 years ago. I’ve been building up my studio ever since. I consider myself part of the 1st generation of the home studio DAW craze. I’m currently treating a room for tracking and mixing. I have an measurement mic and REW to flash my room and track my progress. I’m using monitors from the 1990’s which have served me well but I’ve been tossing around the idea of upgrading. This is all to place perspective on me.

Now, I spent a day flashing my room roughly a month ago. The FR is pretty straight forward. The waterfall took a little learning and I’m not totally there. I’ve been reading on the Acoustical forum since January relearning list info and getting up to date on new info. I just read a large portion of Floyd Toole on a trip to Seattle. I’ve read almost all of Rod Gervais’s book Building like the Pro’s. I’be exhausted Ethan’s website, GIK’s website, and countless others. His is more perspective. I’m currently dropping $500 to $700 a month on materials to build my own broadband absorbers.

So here’s the poop. I get everything done and flash my room and I have that 200Hz dip in my FR that is shown in Fhorn’s first graph. I have zero data on my current monitors to tell me if that’s my room or the monitors themselves. I now go post my results on the Acoustical forum and plead for help. What do you think I’m gonna be told?

Get rid of that console. Get a smaller desk. Your cloud is not thick enough. Your back wall needs more treatment. Yada yada yada.

And so I start throwing money at a problem that is in the monitor and not the room.

Thank you Fhorn. I spent a good part of my day shopping for new monitors yesterday. I absolutely have to have those FR numbers to know how my room is progressing. Otherwise it’s a crap shoot.

Deedee, most all of the monitors I looked at yesterday had DSP filters built in. I know you said your beyond simple filters so I think a dedicated thread to DSP would be a benefit.
The 200Hz dip in one of the speakers is on account of it being a two-way transmission line. If I saw correctly yours are a sealed box or bass-reflex design and I don't think the cause of the dip is loudspeaker related.

I know you have a vaulted ceiling, but it may be worth to at least try and simulate your situation. Here for example http://tripp.com.au/sbir.htm

If you would be shopping for loudspeakers, feel free to PM me for any advice in your situation. I'd be happy to provide a shortlist based on your environment and overal goals.
Old 3 weeks ago
  #79
Lives for gear
 
Brian M. Boykin's Avatar
Quote:
Originally Posted by Fhorn88 View Post
The 200Hz dip in one of the speakers is on account of it being a two-way transmission line. If I saw correctly yours are a sealed box or bass-reflex design and I don't think the cause of the dip is loudspeaker related.

I know you have a vaulted ceiling, but it may be worth to at least try and simulate your situation. Here for example http://tripp.com.au/sbir.htm

If you would be shopping for loudspeakers, feel free to PM me for any advice in your situation. I'd be happy to provide a shortlist based on your environment and overal goals.
I was giving a hypothetical. I’m the case of my monitors I don’t have published FR of the anechoic numbers so IDK the origin of any dip in my system. Wether it be monitor related or room related. Isn’t that a partial point of your discussion? I’m spending money to create a nice smooth FR response with nice decay times and I don’t know what I’m starting with. Even if I’m going for a house curve that’s not neutral/flat I have to know what I’m starting with.
Old 3 weeks ago
  #80
Gear Nut
 

Quote:
Originally Posted by Brian M. Boykin View Post
I was giving a hypothetical. I’m the case of my monitors I don’t have published FR of the anechoic numbers so IDK the origin of any dip in my system. Wether it be monitor related or room related. Isn’t that a partial point of your discussion? I’m spending money to create a nice smooth FR response with nice decay times and I don’t know what I’m starting with. Even if I’m going for a house curve that’s not neutral/flat I have to know what I’m starting with.
True - however for a basic sealed or bass reflex design, such a dip would have to be electronically implemented by the manufacturer for it to occur, which I've never seen done :-)

But it is always better to know what you're starting with, indeed!
Old 3 weeks ago
  #81
Gear Guru
 

Quote:
Originally Posted by Fhorn88 View Post
Quite simply, if you feel that way, this thread isn't for you. I'm not here to tell you what to do, I am presenting objective information, for those that are interested in it.
When you state that there's little to do as long as the speaker is neutral and call that objective information then it's not really up to me. You're at that point either wrong or right and if you're right we should all apply whatever it is that you're recommending.

One difficulty I have with a fair amount of what you've written so far is that you're avoiding to answer things head-on. I asked specifically what to do in the cases I mentioned and your previous replies rather than answer specifically you simply gave a bunch of wider information. This time you completely avoid a simple response. I actually don't really see what's difficult in just replying (though I have a hunch).

Quote:
Originally Posted by Fhorn88 View Post
I feel different. This is also not an issue in any other aspect of artistic creation and production. Video editing wouldn't be performed on an uncalibrated system - ever - with good reason. They also wouldn't try to wing it by 'adjusting the picture until it looks right'
Video editing absolutely is done on "uncalibrated" systems, like, all the time. If you're talking about coloring then that's a different matter, but it also actually then becomes a context which is relevant as a comparison for us, and not necessarily in favor of what you imply.

Quote:
Originally Posted by Fhorn88 View Post
I see +/-3dB swings on the first one - on-axis only, which, if you would have read page one, is not the primary target due to small interference effects.
You're right. I read it as 10dB between every two lines, not 5dB.

Quote:
Originally Posted by Fhorn88 View Post
If you are aiming for a neutral, natural end result, on a non-neutral monitoring system - you may very well end up with the inverse. Unless you feel you can compensate for that - which may or may not be the case. In any case seems like more chance of error and more trouble than it needs to be?
I'm actually saying that many users won't end up with a speaker as flat as the ones you point out, and that they do what they can with their acoustics, and then all that's left is any other means (i.e. short of buying new speakers or spending more on treatment). As long as it's within reason it seems to me entirely reasonable to adjust the speaker output to accomodate the subjective wishes of the engineer - which in this case could mean striving towards a 'neutral' perceived tonal balance,

but I'm also saying that some as explained earlier don't want an entirely 'flat' system, simply because working on one doesn't actually lead to the desired results. Someone with hearing that is oversensitive in a particular range might intuitively mix softer in that range which then as you point out leads to an inverse effect in the actual rendered mix. To adjust for that preference a curve can surely be applied to counteract that.

I feel compelled to ask if your background is in studio design (including acoustics and/or monitoring) and/or engineering... (?)

Quote:
Originally Posted by Fhorn88 View Post
Quote:
Originally Posted by mattiasnyc

You shouldn't have put this in the newbie section in my opinion. It goes far beyond what many newbies can absorb and is really neither "short and simple". Seems more appropriate in other sections..
it is best to spark an interest early on in my opinion before getting set in a fixed way of doing things. It doesn't mean one has to understand the matter entirely from the start. A simple curiousity is often enough to get people to start thinking more critically. There's nothing more difficult during an install then arguing with seasoned professionals. That's when proposing blind testing comes in handy - some simply flat out refuse to participate, but most can be persuaded that way.
Well the topic relied heavily on acoustics, film mix stages and cinema playback systems etc so either the acoustics and/or the post production sections might have been good targets for the thread.

Quote:
Originally Posted by Fhorn88 View Post
Quote:
Originally Posted by mattiasnyc

What's your point? Some speakers are flatter than others?
Yep - and that the price tag doesn't necessarily reflect that. In fact, nor do most manufacturers. There is a serious lack of objective information in the monitoring/loudspeaker world.
I think it's pretty widely known that some speakers are flatter than others, and that there's often a set of graphs available from a manufacturer and that we can look at those to make at least some judgements.
Old 3 weeks ago
  #82
Gear Nut
 

Quote:
Originally Posted by mattiasnyc View Post
When you state that there's little to do as long as the speaker is neutral and call that objective information then it's not really up to me. You're at that point either wrong or right and if you're right we should all apply whatever it is that you're recommending.
Hi Mattias. That's the idea of introducing standards. I'll try to elaborate further.

Quote:
Originally Posted by mattiasnyc View Post
One difficulty I have with a fair amount of what you've written so far is that you're avoiding to answer things head-on. I asked specifically what to do in the cases I mentioned and your previous replies rather than answer specifically you simply gave a bunch of wider information. This time you completely avoid a simple response. I actually don't really see what's difficult in just replying (though I have a hunch).
Ok - I'll address them head on.

Quote:
Not everyone wants a tonal balance that's "neutral". Examples have already been given.
Relating to hearing problems: obivously these very specific issued should be dealt with in the monitoring chain, as long as the end user knows and compensates for this in a proper manner. Since these issues can vary wildly and are user specific, I dare say that in some cases it would be a job better suited to people with normal hearing. Not everyone with hearing issues is well aware of them.

Quote:
But aren't we adjusting the speaker to get the preferred perceived tonal quality?
To give a proper answer I'll put the conclusion into perspective:

Over 50 years ago, Floyd Toole started with a simple series of loudspeaker blind testing, to see what people actually preferred. It turned out many different people, preferred the same loudspeakers. Investigation through the means of anechoic measurements revealed that the loudspeakers that were preferred, were the ones with the least issues. These issues being resonances, uneven directivity, and a lack smoothness in terms of frequency response. These tests were done with many participants, and in various locations. Loudspeakers with a lack of these issues were always rated the highest in blind testing. While the overall scores went up or down, depending on the qualities of the room, the preference order did not change. Meaning our pereception of loudspeakers remains virtually the same, independent of room influences.These tests continue up to this day, on a worldwide scale. The procedures have been much refined since the early days, and there is now a model that can predidct loudspeaker preference in blind testing, with quite amazing accuracy. Research groups have shown their results in the Audio Engineering Society and it formed the basis for the ANSI (American National Standards Institute) 2034A standard.

More recent testing has shown that the better loudspeakers often end up in statistical ties, with certain program material often giving the edge, if it allows for certain differences to be revealed. This is subject of another test, on program material. The conlusion is that is in fact 1)The program material is the limiting factor in many instances. The standard needs to be raised in general 2) That a lack of resonances and smooth spatial distribution are always preferred, but that spectral balance (severe frequency response irregularities aside) are often recording dependent. The technically 'best' loudspeaker because of this, cannot sound the best all the time.

When you talk about using recordings, that have been finalised using other (perhaps unknown) loudspeaker, adjusted in a perhaps unknown way, to 'calibrate' new systems - which in turn will be perhaps used to produce new recordings, which in turn may be used to 'voice' other systems.. It's a neverending cycle. So the answer to your question is the following: The perceived tonal quality, when using a high-quality, neutral loudspeaker, must be adjusted in the recording, not in the monitor chain.


Quote:
Originally Posted by mattiasnyc View Post
Video editing absolutely is done on "uncalibrated" systems, like, all the time. If you're talking about coloring then that's a different matter, but it also actually then becomes a context which is relevant as a comparison for us, and not necessarily in favor of what you imply.
This would relate then to the subject of spectral balance. Video is not my area of expertise, yet in my experience in working with others in the field, calibration is an important part of the chain.





Quote:
Originally Posted by mattiasnyc View Post
I'm actually saying that many users won't end up with a speaker as flat as the ones you point out, and that they do what they can with their acoustics, and then all that's left is any other means (i.e. short of buying new speakers or spending more on treatment). As long as it's within reason it seems to me entirely reasonable to adjust the speaker output to accomodate the subjective wishes of the engineer - which in this case could mean striving towards a 'neutral' perceived tonal balance
For a non-neutral system to start with, yes - there is no other way than to try and replicate those measurements yourself, and adjust as needed. The key point is in how those measurements are acquired, above the transition frequency.

Quote:
Originally Posted by mattiasnyc View Post
but I'm also saying that some as explained earlier don't want an entirely 'flat' system, simply because working on one doesn't actually lead to the desired results. Someone with hearing that is oversensitive in a particular range might intuitively mix softer in that range which then as you point out leads to an inverse effect in the actual rendered mix. To adjust for that preference a curve can surely be applied to counteract that.
I agree - in which case it would be necessary to assess your hearing and learn your own faults, so to say, and how it correlates to your working environment. Having your hearing checked once everywhile is never a bad thing in my opinion, especially when it's an integral part of your work.

Quote:
Originally Posted by mattiasnyc View Post
I feel compelled to ask if your background is in studio design (including acoustics and/or monitoring) and/or engineering... (?)
Civil Engineering - Applied Physics - and Music (Bachelor - performing arts) with a special interest in psychoacoustics since I am involved in both technical and artistic lines of work.



Quote:
Originally Posted by mattiasnyc View Post
Well the topic relied heavily on acoustics, film mix stages and cinema playback systems etc so either the acoustics and/or the post production sections might have been good targets for the thread.
I'm relatively new to gearslutz and don't tend to post in audio related forums really. I hoped by posting in this particular subsection it would attract a broader audience so to say, since the question of monitor choice often pops up at first sight, but perhaps you are right and the discussion has become technical to the extent that it is difficult to follow.



Quote:
Originally Posted by mattiasnyc View Post
I think it's pretty widely known that some speak ers are flatter than others, and that there's often a set of graphs available from a manufacturer and that we can look at those to make at least some judgements.
That is the idea! In combination with a blind test, according the procedure outlined on the first page it should provide the basis for a good monitoring environment.

Please also understand I'm not trying to lecture you, or anyone - a healthy discussion can be good. I hope that it gives some a new perspective on things, and for myself, it is good to learn the train of thought of those that have experience in either mixing/mastering/production/..

Last edited by Fhorn88; 3 weeks ago at 09:04 AM.. Reason: typo
Old 3 weeks ago
  #83
Gear Nut
 

Since it's always interesting to hear it "from the horse's mouth":

This 9 minute video is very much worth a watch

Old 3 weeks ago
  #84
Gear Guru
 

Quote:
Originally Posted by Fhorn88 View Post
Relating to hearing problems: obivously these very specific issued should be dealt with in the monitoring chain, as long as the end user knows and compensates for this in a proper manner. Since these issues can vary wildly and are user specific, I dare say that in some cases it would be a job better suited to people with normal hearing. Not everyone with hearing issues is well aware of them.
It honestly doesn't take much though, that's the thing. If you're oversensitive to content in the 2-5kHz range and consistently end up with too little energy in that range then with a flat system it is what it is. You either suffer through the discomfort (I wouldn't) or you adjust your monitoring to deemphasize that region. By letting your monitoring do that you are now able to push more energy into that area without feeling uncomfortable.

I was also pointing out that someone who has exhausted a situation, i.e. has invested in speakers and room treatment and is "stuck" with what it is, can deemphasize a range of frequencies to counteract such a situation hearing issue or not. You may not like it, but if it works it works.

Quote:
Originally Posted by Fhorn88 View Post
To give a proper answer I'll put the conclusion into perspective:

Over 50 years ago, Floyd Toole started with a simple series of loudspeaker blind testing, to see what people actually preferred. It turned out many different people, preferred the same loudspeakers. Investigation through the means of anechoic measurements revealed that the loudspeakers that were preferred, were the ones with the least issues. These issues being resonances, uneven directivity, and a lack smoothness in terms of frequency response.
Sure, yet curiously, if I read the paper correctly, the paper you linked to earlier that discussed mix stages for cinema had results among "uneducated" listeners that were pretty different from that of others. Obviously we're supposed to be professionals, but I'm just saying that there's going to be a range at play. Now, it actually doesn't take that much to prefer one tonal balance over another if we assume that the engineer has decent hearing. I incorrectly stated the peak/dip as being a 10dB delta earlier, but even a 3dB discrepancy can be enough to end up with mixes one doesn't like as much as with that range rebalanced.

Then again, that's why we have mastering. And on that note I'll just I suppose both support that "neutral" is almost always good enough despite personal preference and that tweaking the tonal balance is fine since a final adjustment will be made in the mastering stage anyway (this excludes mixing for the cinema/tv of course).

Quote:
Originally Posted by Fhorn88 View Post
When you talk about using recordings, that have been finalised using other (perhaps unknown) loudspeaker, adjusted in a perhaps unknown way, to 'calibrate' new systems - which in turn will be perhaps used to produce new recordings, which in turn may be used to 'voice' other systems.. It's a neverending cycle. So the answer to your question is the following: The perceived tonal quality, when using a high-quality, neutral loudspeaker, must be adjusted in the recording, not in the monitor chain.
I'll just reiterate what I implied earlier: Working sound engineers have experience that is of great value, and what you imply above simply isn't always an issue, exactly because of that experience.

This is why I asked you what your background is. Crudely put: I don't expect you to understand or agree if you lack the direct experience.

-- We should note here that if part of this "neutral is good" evidence is professionals having a subjective opinion about playback systems then you can probably bet a good dollar on those professionals being the very same that could correctly adjust to a system that sounds "different" using reference mixes/masters. I don't think it's a coincidence that some recordings/masters stand above others in terms of being references... (for example "Gaucho" for us old farts)

Quote:
Originally Posted by Fhorn88 View Post
This would relate then to the subject of spectral balance. Video is not my area of expertise, yet in my experience in working with others in the field, calibration is an important part of the chain.
Not for editing. Editing is done on anything from calibrated systems to laptops. Final color is hopefully done in a neutral environment.

Quote:
Originally Posted by Fhorn88 View Post
For a non-neutral system to start with, yes - there is no other way than to try and replicate those measurements yourself, and adjust as needed. The key point is in how those measurements are acquired, above the transition frequency.
Right, and to many experienced engineers having decades of work under their belt playing different pieces of music that are known entities help define how the system behaves, if it is "neutral"..
Old 3 weeks ago
  #85
Gear Nut
 

Quote:
Originally Posted by mattiasnyc View Post
It honestly doesn't take much though, that's the thing. If you're oversensitive to content in the 2-5kHz range and consistently end up with too little energy in that range then with a flat system it is what it is. You either suffer through the discomfort (I wouldn't) or you adjust your monitoring to deemphasize that region. By letting your monitoring do that you are now able to push more energy into that area without feeling uncomfortable.
I was also pointing out that someone who has exhausted a situation, i.e. has invested in speakers and room treatment and is "stuck" with what it is, can deemphasize a range of frequencies to counteract such a situation hearing issue or not. You may not like it, but if it works it works.
I don't disagree with that - knowing your issues and adjusting your playback chain is not a bad thing per se. The important point being that when you start from a known factor (a monitor on which there is comprehensive data available) - it is in fact easier to make this type of correction. If this type of data is unknown, it could be a fault in the monitor, for all you know. Sure, experienced people will have a way of learning this information - but many may not.


Quote:
Originally Posted by mattiasnyc View Post
Sure, yet curiously, if I read the paper correctly, the paper you linked to earlier that discussed mix stages for cinema had results among "uneducated" listeners that were pretty different from that of others. Obviously we're supposed to be professionals, but I'm just saying that there's going to be a range at play. Now, it actually doesn't take that much to prefer one tonal balance over another if we assume that the engineer has decent hearing. I incorrectly stated the peak/dip as being a 10dB delta earlier, but even a 3dB discrepancy can be enough to end up with mixes one doesn't like as much as with that range rebalanced.
The document in question was from Harman - they 'train' listeners (it's a free and unbiased process http://harmanhowtolisten.blogspot.com/) as to get quicker and more accurate feedback on products. This takes longer with regular listeners, although the results in broad terms do correlate. If I may quote
Quote:
Pulling all of this together is an important study [6] in which the opinions of 12 selected and trained listeners are compared to those of 256 listeners from various backgrounds. The relative ratings of the products were essentially the same for small groups of listeners extracted from each population. The consequential difference was in the statistical confidence one could place in the opinions. The selected and trained listeners were much more reliable in their ratings, meaning that trustworthy results could be obtained in much less time. The trained listeners also provided comments that were easily interpreted by design engineers to help them focus on aspects of performance that needed working on, while other listeners tended to use less technically descriptive terms.
Obviously this is indeed less of a problem for most professionals. And indeed, I've seen people accurately point out 0.25dB low Q differences in a blind test. That's amazing if you think about it. Which is exactly why this anechoic information is so valuable. It's difficult for most, and downright impossible for part of the spectrum to accurately assess those kind of measurements in-room. Since we can, and do, pick out small differences - it's important to know if you're adjusting the source material, or compensating for a potential discrepancy in the anechoic loudspeaker response.


Quote:
Originally Posted by mattiasnyc View Post
Then again, that's why we have mastering. And on that note I'll just I suppose both support that "neutral" is almost always good enough despite personal preference and that tweaking the tonal balance is fine since a final adjustment will be made in the mastering stage anyway (this excludes mixing for the cinema/tv of course).

I'll just reiterate what I implied earlier: Working sound engineers have experience that is of great value, and what you imply above simply isn't always an issue, exactly because of that experience.

This is why I asked you what your background is. Crudely put: I don't expect you to understand or agree if you lack the direct experience.

-- We should note here that if part of this "neutral is good" evidence is professionals having a subjective opinion about playback systems then you can probably bet a good dollar on those professionals being the very same that could correctly adjust to a system that sounds "different" using reference mixes/masters. I don't think it's a coincidence that some recordings/masters stand above others in terms of being references... (for example "Gaucho" for us old farts)
Sure, this must be true - otherwise we wouldn't have any great recordings out there. But again, the process I outline is also meant to be a stepping stone for those that don't have that insight yet, or may never achieve it otherwise. Hell, I've come accross more than one professional that still believes NS10s are an example of a neutral monitor. I also don't believe my starting point (a neutral, well behaved monitor loudspeaker) would in any way negatively affect a seasoned professional's capabilities. So while experience is a factor, it is simultaneously not a quantifiable metric. The process is certainly not meant to limit anyone's artistic freedom, but merely to provide a more consistent working environment so that a monitoring system is less likely to be a limiting factor. While there are good recordings, great recordings even, out there - there are also a ton of average and bad ones. While the artistic process is entirely subjective, the least we can do is raise the bar in general for monitoring and playback systems. As for my background/experience. Installing and calibrating systems is exactly what I do - even designing and building monitors from scratch has been part of this. So while -I- don't have to work on them, I do provide clients with neutral and reliable systems so that they can work on them with confidence.




Quote:
Originally Posted by mattiasnyc View Post
Right, and to many experienced engineers having decades of work under their belt playing different pieces of music that are known entities help define how the system behaves, if it is "neutral"..
I think that's where there might be a slight misunderstanding. As I say on page one "I hope my post will spark some interest, introduce some polite discussion - and hopefully be of help for those of you confused and struggling - for those of you trapped in that circle of confusion." I'm not here to lecture anyone - only to inform those who're looking into a different, more objective way of doing things.
Old 3 weeks ago
  #86
Lives for gear
The frequency response is one part of the neutrality.

You and Toole forgot: the distortions, the resonnance box, the membrane deformation...
Old 3 weeks ago
  #87
Gear Guru
 

Quote:
Originally Posted by Fhorn88 View Post
while experience is a factor, it is simultaneously not a quantifiable metric.
Fair enough, though it sure seems that something quantifiable was extracted out of it in the research you quoted. It just all reads to me as you having your cake and eating it too.

Quote:
Originally Posted by Fhorn88 View Post
While there are good recordings, great recordings even, out there - there are also a ton of average and bad ones. While the artistic process is entirely subjective, the least we can do is raise the bar in general for monitoring and playback systems.
Nobody is disputing that.

You made an objection and I didn't (and still don't) understand really what you were objecting to and why. Something about adjusting/correcting a speaker's output yet sometimes it's ok...

I'll leave this alone. Sufficed to say I think few true newbies (to this) will make it through all the text even on the first page and properly digest it.
Old 3 weeks ago
  #88
Gear Nut
 

Quote:
Originally Posted by dinococcus View Post
The frequency response is one part of the neutrality.

You and Toole forgot: the distortions, the resonnance box, the membrane deformation...
Flattered to be put in the same senctence, but I had nothing to do with the research itself I'm afraid.

Ask yourself the following: Is it likely that these research groups, with some of the best and brightest minds in their respective fields, backed by millions of dollars, have simply not taken the time and effort to investigate these matters?

The truth about distortions is that it is a general term, but in reality a quite complex matter. Studies have been performed on the audibility and auditory perception of distortions, and the truth is that there is no exact correlation between listening preferences and standard THD and IMD levels. This is one of those things where additional research is required. From a design perspective the only safety is 'as little as possible'.

If you are interested, here are two interesting AES papers:
http://www.gedlee.com/Papers/Distortion_AES_I.pdf
http://www.gedlee.com/Papers/Distortion_AES_II.pdf

As for the resonance box? the membrane deformation? If you mean what I think you mean, both very basic matters to assess in the design phase. This thread is not a loudspeaker design class so sorry to not go into this any further.
Old 3 weeks ago
  #89
Gear Nut
 

Quote:
Originally Posted by mattiasnyc View Post
Fair enough, though it sure seems that something quantifiable was extracted out of it in the research you quoted. It just all reads to me as you having your cake and eating it too.
The proof of the pudding.. So yes, listening is the critical test, and absolutely necessary, but it must be unbiased - blind. Then we can start correlating these 'subjective measurements' to 'technical measurements'.


Quote:
Originally Posted by mattiasnyc View Post
You made an objection and I didn't (and still don't) understand really what you were objecting to and why. Something about adjusting/correcting a speaker's output yet sometimes it's ok...

I'll leave this alone. Sufficed to say I think few true newbies (to this) will make it through all the text even on the first page and properly digest it.
To rule out any misunderstandings: I believe it was the subjective full range tweaking/EQ of the loudspeaker from the listening position that I objected to. Not in the sense that it cannot lead to a better result than the original situation, but that if one used high quality, neutral, well behaved loudspeakers (with the anechoic data to back it up) in the first place, that, apart from the bass range, there would be nothing to correct.

Exception: If one can obtain those anechoic measurements, or have them made - this would be a good baseground to make corrections to smooth out the response if you have a decent knowledge of loudspeaker design, or compensate in the case of user specific auditory issues. The issue lies in seperating the room from the loudspeaker, which cannot be done with high enough accuracy from the listening position, and to ensure that any decisions that are made to material that you are working on are of merit, and not an unbeknown correction for flaws in your monitoring system. Hope this clarifies my standpoint.
Old 3 weeks ago
  #90
Gear Nut
 

A real world example

I was asked recently to look at a monitoring system. The client owns a set of main monitors of a reputable brand, and was having trouble 'calibrating' the system to his (treated) room. I was asked to take a look. For the purpose of this thread, we will be looking at what happens from around 1Khz on up.

A comment was that the system had a lack of clarity, a fogginess. It had been suggested to the owner that this was a quality of the polypropylene midrange driver and that he ought to upgrade to a system that feature a smaller mid-dome with 'superior clarity'.

Not yet ready to spend a lot of money on a new set of monitors, I was asked to come in and take a look.

Since it was unpractical to take the speakers outside for this initial consultation. A listening window average was made using nearfields measurements.



While the speaker is within its specifications (using a +/-3dB window), we can see there are definitely some things that could be improved upon - since without full anechoic measurements and a better microphone setup we could not exclude some potential local anomalies, only broader filters were employed.

A short amount of time later already we had a much improved listening window curve



In A/B checks with instant switching between filter/no filter the client could easily pick out the new corrected filter. In his (and my) opinion, it made all the difference. I have suggested if he should desire, detailed full anechoic measurements could be made and probably make some further improvements to his monitors.

Let us look now at what happens at the listening position



We can definitely see some differences as well. The remaining dip at 3kHz is situated around the crossover between midrange and tweeter and shows up because of vertical reflections which are a part of what we measure in a room (see posts on page 1 for detailed explanation).
Had we used a standard target curve, with measurements taken at the listening position, we would have 1) filled this dip completely and made the direct sound field worse. 2) not been able to achieve the needed high frequency balance as accurate, and as quick.
Post Reply

Welcome to the Gearslutz Pro Audio Community!

Registration benefits include:
  • The ability to reply to and create new discussions
  • Access to members-only giveaways & competitions
  • Interact with VIP industry experts in our guest Q&As
  • Access to members-only sub forum discussions
  • Access to members-only Chat Room
  • Get INSTANT ACCESS to the world's best private pro audio Classifieds for only USD $20/year
  • Promote your eBay auctions and Reverb.com listings for free
  • Remove this message!
You need an account to post a reply. Create a username and password below and an account will be created and your post entered.


 
 
Slide to join now Processing…
Thread Tools
Search this Thread
Search this Thread:

Advanced Search
Forum Jump
Forum Jump