@ oon,
If the speaker is meant to provide the raw material and facilitate the added ambience then you will need one speaker array for each instrument, otherwise placement is going to depend on a stereo image to some degree, or at least some kind of omnidirectional stereo arrangement 😕.
If the speaker is meant to provide the raw material and facilitate the added ambience then you will need one speaker array for each instrument, otherwise placement is going to depend on a stereo image to some degree, or at least some kind of omnidirectional stereo arrangement 😕.
I would have thought the engineer intended them to be played in a room and would have made some allowance for that?
It depends on the frequency, how "bendy" it is. 😉
>I would have thought the engineer intended them to be played in a room and would have made some allowance for that?
I really have no idea about what allowances they could make - I assume they mix to sound good over a variety of speakers; near field, farfield, headphones, car stereo. But it's all for "sounding good" over whatever it is they anticipate their best customers will be listening through.
I think our room just plops over the top of that sound, like a giant convolution. This is where I just begin to understand all the room treatments; to take away some of the degree of that convolution and get more towards the sound the recording engineer originally heard.
Which of course assumes your speakers sound enough like whatever it was they were listening to, when they dialed things into what they anticipated their customers would like and pay for.
I have to wonder if some of these engineers would say "It's actually not how I'd like it personally, but I have to mix toward what's going to sell, otherwise I'd be doing the artist a grave dis-service". Who knows...
I really have no idea about what allowances they could make - I assume they mix to sound good over a variety of speakers; near field, farfield, headphones, car stereo. But it's all for "sounding good" over whatever it is they anticipate their best customers will be listening through.
I think our room just plops over the top of that sound, like a giant convolution. This is where I just begin to understand all the room treatments; to take away some of the degree of that convolution and get more towards the sound the recording engineer originally heard.
Which of course assumes your speakers sound enough like whatever it was they were listening to, when they dialed things into what they anticipated their customers would like and pay for.
I have to wonder if some of these engineers would say "It's actually not how I'd like it personally, but I have to mix toward what's going to sell, otherwise I'd be doing the artist a grave dis-service". Who knows...
Last edited:
No we usually don't mix for something. We mix for it to sound 'good' whatever it means.
We usually works on known room or if we don't know it, we have nearfield 'reference' loudspeakers ( which can be horrible things like NS10 imho but for mixing you don't really need super accurate things) AND recordings we know sounds goods (as references).
Each rooms is different, this is true for domestic and control rooms. But in CR we ask to have the minimum possible Early Reflections and some coherency about reverb ( it must be decorelated).
Search the forum about LEDE principle. Mitchba posted the AES paper about LEDE where you'll find most target of this principle ( others exists which are valid too, i repeatidly posted about CID concept from Bob Walker (BBC) which can be found on their R&D archive, or T.Hidley zero environnement rooms).
Here again there is differences between each stage of production and mastering rooms are differents than recording/mixing rooms, as well as loudspeakers.
We usually works on known room or if we don't know it, we have nearfield 'reference' loudspeakers ( which can be horrible things like NS10 imho but for mixing you don't really need super accurate things) AND recordings we know sounds goods (as references).
Each rooms is different, this is true for domestic and control rooms. But in CR we ask to have the minimum possible Early Reflections and some coherency about reverb ( it must be decorelated).
Search the forum about LEDE principle. Mitchba posted the AES paper about LEDE where you'll find most target of this principle ( others exists which are valid too, i repeatidly posted about CID concept from Bob Walker (BBC) which can be found on their R&D archive, or T.Hidley zero environnement rooms).
Here again there is differences between each stage of production and mastering rooms are differents than recording/mixing rooms, as well as loudspeakers.
Last edited:
5.5 to 11 feet. So half that from the wall.
Likely a compromise, 20-30 ms is likely more appropriate (and where we got the best performance, but still had noticeable time smear)
dave
6mS is mentioned by Linkwitz,
"If the speaker is at least 1 m (3 ft) in front of the wall, then the reflected sound is delayed by about 6 ms, which is sufficient not to be perceptually summed with the direct sound. The cognitive process disassociates it from the speaker and attributes it to the room with a sense of space." Frontiers
Is it his own observation or has it been established elsewhere?
What does he mean by "summed"?
"If the speaker is at least 1 m (3 ft) in front of the wall, then the reflected sound is delayed by about 6 ms, which is sufficient not to be perceptually summed with the direct sound. The cognitive process disassociates it from the speaker and attributes it to the room with a sense of space." Frontiers
Is it his own observation or has it been established elsewhere?
What does he mean by "summed"?
The estimates of how sepearpated in time th edelayed signal needs to be is larger than that from what i recall in Toole.
dave
dave
Needs to be for what though? There is the precedence effect, but I don't think he's referring to that?
Ah, see ambiance extraction here Precedence effect - Wikipedia
Ah, see ambiance extraction here Precedence effect - Wikipedia
Last edited:
Reverb is ideally a diffuse echo. 20-30ms is better, that way you don't have to think to separate them and it isn't fatiguing.
The summing is related to Haas effect. Under 6ms both direct and ER merge in our brain ( but comb filtering severely corrupt the rendering).
Imho, under 10ms/-15db is too much a compromise for my own pref.
Imho, under 10ms/-15db is too much a compromise for my own pref.
Mostly talked about is that there's a difference for what type of music one is enjoying.
The classical music/concert listeners seem to like (or benefit more from) earlier reflections (say 6 to 10ms) than the more studio album oriented listeners. At least that's what I've seen mentioned in many prior discussions. That might explain what Dr. Geddes advocates (no early reflections) vs Toole (advocating the benefit of early reflections). Toole likes the classical genre and Geddes is more of a studio recording listener.
A Haas kicker like I use would come in somewhere between 15 to 25 ms after the direct sound. The focus for me, personally was to have a reflection free zone (or rather damped/absorbed reflections in a real room) of about 20 ms. Granted, I like to listen to more studio oriented or originated works. The longer reflection free (or absorbed) zone does make a large difference for imaging queues.
The Haas limit is often quoted to be as long as 35 to 40 ms but in reality it largely depends on the frequency content as it will differ and be shorter the higher one goes in frequency.
Looking at the IR recorded at the listening position before the haas kicker:
With the added Haas kicker it looks like this:
The hole between the first sound and the Haas kicker is the 'reflection free zone' I mentioned. These are actual in room measurements at the sweet spot.
The classical music/concert listeners seem to like (or benefit more from) earlier reflections (say 6 to 10ms) than the more studio album oriented listeners. At least that's what I've seen mentioned in many prior discussions. That might explain what Dr. Geddes advocates (no early reflections) vs Toole (advocating the benefit of early reflections). Toole likes the classical genre and Geddes is more of a studio recording listener.
A Haas kicker like I use would come in somewhere between 15 to 25 ms after the direct sound. The focus for me, personally was to have a reflection free zone (or rather damped/absorbed reflections in a real room) of about 20 ms. Granted, I like to listen to more studio oriented or originated works. The longer reflection free (or absorbed) zone does make a large difference for imaging queues.
The Haas limit is often quoted to be as long as 35 to 40 ms but in reality it largely depends on the frequency content as it will differ and be shorter the higher one goes in frequency.
Looking at the IR recorded at the listening position before the haas kicker:
With the added Haas kicker it looks like this:
The hole between the first sound and the Haas kicker is the 'reflection free zone' I mentioned. These are actual in room measurements at the sweet spot.
Attachments
Last edited:
Too short for my liking, but getting it even shorter can only confuse the actual imaging queues, yes.
One thing that has not been mentioned so far is the on-axis and off-axis response relationship. A smooth transition from one driver to the next without affecting the off-axis response is the PRIMARY requirement for a wide, deep, and stable soundstage. This is a pre-requisite without which there is only limited benefit to the other tricks.
In traditional cone and dome two-ways, this is very hard to achieve because the tweeter comes in with very wide directivity whereas the woofer is beaming at the crossover frequency. I believe that the ear can use the difference between the direct sound and reflected sound to tell the location of the speaker. To hide the location, the reflection must be a copy of the on-axis sound. A waveguided tweeter can solve this problem. See Heissman Acoustics for examples.
For the OP's specific case, making sure the speakers are as far away from the front wall as is practical and generating strong side-wall reflections will help with spaciousness.
In traditional cone and dome two-ways, this is very hard to achieve because the tweeter comes in with very wide directivity whereas the woofer is beaming at the crossover frequency. I believe that the ear can use the difference between the direct sound and reflected sound to tell the location of the speaker. To hide the location, the reflection must be a copy of the on-axis sound. A waveguided tweeter can solve this problem. See Heissman Acoustics for examples.
For the OP's specific case, making sure the speakers are as far away from the front wall as is practical and generating strong side-wall reflections will help with spaciousness.
Ironically though, a resulting dip from off-axis (horizontal axis) listening in a typical 2-way loudspeaker is usually around 1-3 kHz:
-which tends to enhance subjective depth.
-which tends to enhance subjective depth.
Agreed, but how real is that depth? On systems with good soundstaging, instruments can appear both in front and behind the speakers. If you get depth on every recording, which is what a power response hole in 1-3 kHz might do, then that's not right either.
It is all illusion.
With good recordings the best of the speakers we have built, the back wall will disappear and be replaced by the artists. But only on good material.
dave
With good recordings the best of the speakers we have built, the back wall will disappear and be replaced by the artists. But only on good material.
dave