The short answer is, it is because the mastering engineers are finalizing the products in the tilted response environment, because they expect the music will be played back in that environment. All explained in Dr. Toole's paper.
Actually, I found that CD 2 way horn has another issue with the commercial recordings. Many mastering speakers and commercial speakers with dome tweeters have recessed room response around 2-3K, so that frequency region tends to be overpowered with CD speakers. Many people claim horn speakers sound harsh, but it is mostly because of this, I guess. I believe we should use BBC dip EQ with CD horn for more natural playback. Check Stereophile, you will notice many modern commercial speakers have BBC dip.
Last edited:
Can we get to a consensus about terminology ?
Response can be flat (without large peaks or dips) but still be tilted. When you write response, what do you mean by that ? Power response, frequency response (if so, axis angle should be stated and if it was measured gated or not) or maybe it is in-room response?
Departure from flat on axis frequency response is one of the most common things in speaker building since many speakers do not use any mechanism to control diffraction. In classic two way (6"/1" and i mention this because it is probably most exploited concept of all times) it is very likely you will measure dip on axis between 2kHz and 4kHz. If you were to tailor the crossover for flat on axis frequency response it would make a very nice peak 3dB high at that same frequency range off axis.
On the other hand, if you measure drivers like Kimmo Saunisto suggests and tailor the crossover with that data (0-180deg horizontal in 10 deg increments, vertical may vary based on what you want to achieve) it will be painfully obvious that off axis frequency response dominates what we hear and that measurement at 0 deg means very little in itself.
Power response does much better job on predicting how a loudspeaker will sound in a room. We can debate on how much power response should be tilted in regard to room size and properties of the room but it has to be flat (without sudden peaks and dips).
Response can be flat (without large peaks or dips) but still be tilted. When you write response, what do you mean by that ? Power response, frequency response (if so, axis angle should be stated and if it was measured gated or not) or maybe it is in-room response?
Departure from flat on axis frequency response is one of the most common things in speaker building since many speakers do not use any mechanism to control diffraction. In classic two way (6"/1" and i mention this because it is probably most exploited concept of all times) it is very likely you will measure dip on axis between 2kHz and 4kHz. If you were to tailor the crossover for flat on axis frequency response it would make a very nice peak 3dB high at that same frequency range off axis.
On the other hand, if you measure drivers like Kimmo Saunisto suggests and tailor the crossover with that data (0-180deg horizontal in 10 deg increments, vertical may vary based on what you want to achieve) it will be painfully obvious that off axis frequency response dominates what we hear and that measurement at 0 deg means very little in itself.
Power response does much better job on predicting how a loudspeaker will sound in a room. We can debate on how much power response should be tilted in regard to room size and properties of the room but it has to be flat (without sudden peaks and dips).
Last edited:
@ mbrennwa I think it is the "circle of confusion" article, but I'm having trouble finding the one I've seen before. This one on Harmon's website mentions it, but It's not the one I'm thinking of https://www.harman.com/sites/default/files/white-paper/12/11/2015%20-%2006%3A20/files/LoudspeakersandRoomsPt2.pdf
Tony.
Tony.
Last edited:
I think I have downloaded several different JBL or AES versions, but can't find them now, sorry. One of them is what wintermute linked.
Thanks Zvu, I'd not seen this, Have only watched the first 27 minutes so far, but thought that from this point YouTube and specifically at around 22:30 is particularly relevant, to the argument of doing an-echoic as opposed to in room, measurements. The bit from 33:30 also is very relevant.
Tony.
Tony.
Last edited:
Guess you are going to have big trouble understanding how Fletcher-Munson-type curves relate to the last couple of posts about identifying the right way to set FR.
An analogy might be that you can't tell much about the state of charge of a a battery from its voltage.
I sometimes wonder about the hearing skills of folks who design a system to be anechoicly flat and then declare it sounds perfect in their room. If I have changed my system in some way, I do many REW sweeps and average my mic readings and THEN I listen to my old favourite test recordings that I've used for decades and my new cowbell recording. Not a chance that it will sound right just based on FRs: always needs some fine-tuning by ear.
Maybe some day soon we'll have smarter freq measurements. Thinking about the importance of recognizing factors of human perception as in the smarter THD test that Geddes and Lee created recognizing the influence of perceptual masking, if I recall correctly.
B.
Last edited:
I have no problem that you prefer flat response speaker. It is your personal preference. Historically, most of the commercial recordings has been finalized by non CD speakers, 2-3k recessed at the listening position. It is simply because almost of all domestic speakers have not been CD, unfortunately. Check the frequency response of today’s most popular mastering speakers, B&W or Dunlavy. Those speakers either deliberately recess on axis response around that area, or we can easily imagine recessed off axis response at 2-4K, crossover point. Mastering engineers still follow the same standard as before, whether you like it or not. It is not a science, it is a history.
Here's a link where you can download Toole's "The Measurement and Calibration of Sound Reproducing Systems" paper, i.e. the "Circle of Confusion" paper. LINK. It's too large to upload here. I also attached a summary from one of his Youtube presentations (sorry it's so fuzzy). Notice when he says "a flat on-axis response is preferred" that he's referring to an anechoic response.
Attachments
Of course, JBL says so. They have to advocate CD speakers, but we all know what happened to JBL. They lost in high end audio market, and B&W won, unfortunately.
Here is the frequency response of B&W. I personally dislike the sound of those speakers, but they are one of the most referenced speaker today, and many mastering studios use. Other than recess, I think 10K boost is the trick B&W uses to sell those DIAMOND oriented speakers to average music listeners, not us.
Bowers & Wilkins 802 D3 Diamond loudspeaker | Stereophile.com
B&W 800 Diamond loudspeaker Measurements | Stereophile.com
Bowers & Wilkins 802 D3 Diamond loudspeaker | Stereophile.com
B&W 800 Diamond loudspeaker Measurements | Stereophile.com
Last edited:
It's more in the listening window. Just on axis is a very limited measurement be it ancholic or gated. What's really important is the off axis vs on axis balance as they should be very close. Take a look at the in room average vs the on axis and the directivity compared to the BW. Completely different design philosophy.
Rob
JBL Synthesis 1400 Array BG loudspeaker Measurements | Stereophile.com
Everything is a preference - as is listening inaccurate loudspeakers because of junk recordings. I agree to the fullest. I on the other hand choose not to listen to junk recordings.
If you think that anyone serious is mastering their recordings on B&W or Dynlavy's you might want to think again. Last two sentences are a product of your imagination. Either that or you are hanging out with some really low end studio crew. For at 30+ years there are products that are made to pay atention on the directivity and flat power response department - home or pro. Genelec, Klein+Hummel (Neumann these days), JBL, Kef, Revel and many others.
B&W hardly won anything. But their marketing department would want you to think that way. Since Nautilus nothing really interesting came out of that house - and it makes sense since Laurence Dickie left and joined Vivid Audio.
I'm choosing not to be a part of inertia caused by circle of confusion Toole is talking about. I'm listening accurate loudspeakers and music that sounds good to me. It will all be remastered eventually and, luckily for me, i don't listen too much of older music and recordings. For guys who listen pop music 80's and 90's it is a drag i guess. As for me, i wouldn't know.
Last edited:
Zvu, I'm from music industry, and I have worked with top mastering people in NY.
BTW, Genelec are used for mixing not for mastering.
I named few that pay attention about directivity and power response. I never wrote what are mastering or mixing monitors, nor what are home loudspeakers.
Problem is that you guys, who work like that and add to the circle of confusion, consider yourselves professionals. I'd be ashamed to confess i'm one of those that Toole was talking about. Really
EDIT: But then again, i'd probably never do it that way so i'd never have a problem of that sort.
Last edited:
No, I have been trying to solve the mystery of the circle of confusion here. There is no confusion. I wish you can understand the dilemma that today's mastering engineers have. They have to follow the standard whatever it is, 44.1K, MP3, loudness war, etc. Recessed high mid is just one of them.
Toole tells himself
https://www.harman.com/sites/default/files/LoudspeakersandRoomsPt2_0.pdf
Earlier similar threads to this
Smooth (Flat) vs. Accurate (Hi-Fidelity)
Why the emphasis on flat speaker response?
https://www.harman.com/sites/default/files/LoudspeakersandRoomsPt2_0.pdf
Earlier similar threads to this
Smooth (Flat) vs. Accurate (Hi-Fidelity)
Why the emphasis on flat speaker response?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.