|
|
Post by audiobill on Apr 12, 2016 16:50:44 GMT -5
OK, so Dirac is a fancy tone control, as suspected.....
|
|
bootman
Emo VIPs 
Typing useless posts on internet forums....
Posts: 9,358
|
Post by bootman on Apr 12, 2016 18:40:28 GMT -5
Luckily, now we have parametric equalizers (although it's nice to have something like Dirac do it all for us automatically).  It would be nice if Emotiva came up with a dedicated Dirac processor box that could be connected to a stereo/AV premap. Price it at $399.95 and I'll be your first customer... It may not be up to Emotiva to do that. |
|
|
|
Post by Cogito on Apr 12, 2016 18:42:14 GMT -5
It would be nice if Emotiva came up with a dedicated Dirac processor box that could be connected to a stereo/AV premap. Price it at $399.95 and I'll be your first customer... It may not be up to Emotiva to do that. Well, it has to be Emotiva. If it doesn't have those blinding laser Emotiva LEDs, it just won't do! |
|
bootman
Emo VIPs
Typing useless posts on internet forums....
Posts: 9,358
|
Post by bootman on Apr 12, 2016 20:07:34 GMT -5
It may not be up to Emotiva to do that. Well, it has to be Emotiva. If it doesn't have those blinding laser Emotiva LEDs, it just won't do! Well I'm sure Dirac also has a say in it. |
|
|
|
Post by garbulky on Apr 12, 2016 21:19:56 GMT -5
OK, so Dirac is a fancy tone control, as suspected..... That's like saying a tweeter is a fancy subwoofer. Or a set of electrostatic panel speakers is a fancy tin can and a wire. Sure you could say that, but it ignores a more accurate reality which is more than one facet. A tone control and DIRAC both change the sound. But the tone control tends to change the sound across a centered point. So if you want a bass boost, then you've got an overall bass boost around say 70 hz solely. It's activated solely depndingo n how the person feels about it. If he wants more bass he gets more bass etc. DIRAC and tone correction are similar because that they both do change the sound. Other things that change the sound tubes, Solid state, wires, the room, your couch, microphones, pcs, ears, the producer's decision, ambient temperature. Everything changes the sound. DIRAC is room correction. It measures the sound IN THE ROOM. Then it basically finds out how deviated it is from a straight wire with gain and then attempts to change the sound back to a straight wire with gain. Basically it's supposed to replicate the sound the way it's supposed to sound if the sound wasn't distorted by the room. Also DIRAC isn't just an eq where it simply lowers a certain frequency range to compensate for peaks and dips. It is more complicated than that. However...it obviously cannot do anywhere near a perfect job. Sound is too complex for that. And imo we are a long way from effective room correction where you get a straight wire with gain. |
|
klinemj
Emo VIPs
Official Emofest Scribe
Posts: 15,101
|
Post by klinemj on Apr 13, 2016 2:44:38 GMT -5
OK, so Dirac is a fancy tone control, as suspected..... I could see why you say that, and in one sense it is. But...it also takes impulse response issues...which no tone control ever did. So, in that light, it is far more than a fancy tone control. Mark |
|
|
|
Post by audiobill on Apr 13, 2016 4:05:21 GMT -5
Thanks, guys. You've both described how it works, but at its core still manipulating frequency response, no?
And Gar, Dirac isn't correcting the room, it's adjusting the frequency response of the overall system to compensate for room anomalies.
May be a good thing sonically , but still "distortion" as Keith uses that term IMO.
|
|
|
|
Post by Cogito on Apr 13, 2016 6:29:43 GMT -5
Well, it has to be Emotiva. If it doesn't have those blinding laser Emotiva LEDs, it just won't do! Well I'm sure Dirac also has a say in it. All they are going to want is their cut for licensing. Nothing unusual. |
|
|
|
Post by vneal on Apr 13, 2016 6:31:47 GMT -5
In the old days mid end receivers came with traditional tone controls. If you wanted something better you bought separates which had no tone controls
|
|
|
|
Post by audiobill on Apr 13, 2016 7:46:07 GMT -5
Everyone sweating .0000001% distortion and ruler flat frequency response from 1 to 100khz - then overlay with +/- 12db "correction" !!
Again, I trust my ears.
|
|
|
|
Post by garbulky on Apr 13, 2016 8:59:09 GMT -5
Thanks, guys. You've both described how it works, but at its core still manipulating frequency response, no? And Gar, Dirac isn't correcting the room, it's adjusting the frequency response of the overall system to compensate for room anomalies. May be a good thing sonically , but still "distortion" as Keith uses that term IMO. Yes it is manipulating frequency response no matter how you look at it. |
|
KeithL
Administrator 
Posts: 10,276
|
Post by KeithL on Apr 13, 2016 9:55:34 GMT -5
Actually, it's doing a lot more than just adjusting frequency response... (or you could say that it's doing more stuff more carefully). In reality, any practical filter circuit designed to alter frequency response also alters phase response (which is one way of looking at timing; phase describes aspects of time and frequency) - and that's absolutely true of tone controls. It is basically impossible to alter frequency response without introducing phase shift - and old style analog tone controls in specific introduce significant phase shift. Now, as long as we're only talking about stereo, and we make the same adjustment to both channels, the phase shift at a given frequency will be equal for each. And, while we humans are sensitive to differences in phase between channels, we aren't especially sensitive to different phase shifts at different frequencies in the same channel. (So, if you have a mono recording, and you turn up the treble, and, as a side effect, the treble control introduces a +90 degree phase shift at 5 kHz relative to 50 Hz, you probably won't notice it. And, if you have a stereo recording, and you turn up the treble, and it adds an EQUAL +90 degree phase shift at 5 kHz to both channels, you probably won't notice it either.) This is what would happen, for example, if you turn up the treble to correct for slightly dull speakers, or a dead room, or a dull-ish recording. HOWEVER, now let's look at room correction. What happens if we make careful measurements and determine that one speaker needs to be boosted at 5 kHz, while the other doesn't? If we simply turn the treble up in that channel, then our adjustment might introduce a +90 degree phase shift at 5 kHz ONLY IN THAT CHANNEL. As it so happens, relative phase information between our two ears is one of the most important ways in which our brains determine where something is located. So, assuming we have introduced this anomaly by our correction, here's what will happen. Any sound that appears in only one channel will be mostly unaffected. However, for any instrument that has harmonics above 5 kHz, and that appears in both channels (like ones located in the center stage), there will be a frequency-dependent "shift" in location. Even worse, since the phase shift varies with frequency, any instrument that has various harmonics will seem to be smeared from left to right. (Remember that this is happening specifically because, since we've applied the tone adjustment differently in the two channels, we've also got a different frequency-dependent phase shift in the two channels.) Now imagine all this with five channels, or with a graphic equalizer..... Since we're talking about room and speaker correction rather than just frequency response, we also have to deal with more complex issues. The sound you hear is a combination of direct and reflected sound; and, since reflected sound has traveled further than direct sound, it is also phase-shifted, often in complex and interesting ways (you can have reflections from different surfaces, at different distances, all interacting with the direct sound). And speakers themselves also introduce phase shift - often several hundred degrees of it across the audible spectrum. What Dirac basically does is to analyze all this, and do its best to create custom filters that correct both frequency response and phase issues. Dirac can do this because, unlike simple tone controls, or even most parametric equalizers, it gets to choose between different filters to pick ones that have the best frequency and phase characteristics to correct a particular situation. (It's not simply a matter of "doing EQ without the bad side effects". Dirac actually measures both frequency and phase, and does its best to correct both of them - at the same time. This means that Dirac can in fact correct imaging issues related to time delay or phase - even if the frequency response is already just fine and doesn't need correcting. And it can correct frequency response issues without introducing the annoying imaging anomalies you would get by making different adjustments to each channel with tone controls. And it can do both at the same time.) (In many specific models, including many simple ones used for modelling subs in rooms, the frequency response and phase are closely linked in a relatively simple relationship, and correcting the frequency response often also gives you a "reasonable" phase response. However, when you're talking about a real room, and multiple speakers, with a wide frequency range, and multiple drivers, and complex crossovers, it's often far more complex.... at which point you have to address both frequency response and phase separately - and avoid solutions that correct one at the expense of the other. This is what Dirac excels at... and does far better than simpler "automatic EQ only" solutions.) Thanks, guys. You've both described how it works, but at its core still manipulating frequency response, no? And Gar, Dirac isn't correcting the room, it's adjusting the frequency response of the overall system to compensate for room anomalies. May be a good thing sonically , but still "distortion" as Keith uses that term IMO. |
|
|
|
Post by audiobill on Apr 13, 2016 10:14:13 GMT -5
Interesting, thanks.
So with all this data manipulation at the output, seems like there is little need to be concerned about frequency or phase response of components in the chain - since it will be "fixed in the mix".
I don't really believe it, just toying with ideas!
The best answer is good components in a properly treated room, IMO.
And as little manipulation of the music as humanly possible.
|
|
KeithL
Administrator
Posts: 10,276
|
Post by KeithL on Apr 13, 2016 11:03:48 GMT -5
But some of them do.... The XMC-1 has Dirac, AND parametric EQs, AND Tone Controls (with fully configurable turnover points). And the XSP-1 has "Tone Trims" - which are really just tone controls with a very limited range. Are you really asking why current Emotiva products have no classic tone controls? |
|
KeithL
Administrator
Posts: 10,276
|
Post by KeithL on Apr 13, 2016 11:16:03 GMT -5
You got it. The other thing to remember is that nothing is perfect. There absolutely ARE situations that even Dirac can't correct. (It can't turn an echo chamber into an anechoic chamber, or vice versa, or enable a 4" woofer with 1/4" of travel to deliver room-shaking bass; and it can't remove THD or IM distortion from the signal, or prevent your speakers or other equipment from producing them.) There are also limitations in terms of processing power. Simply put, the fewer problems that need to be corrected, the more likely it is that Dirac will be able to do a better job of correcting them, and getting the corrections exactly right. So, the better your starting point, the better your results will usually be. Think of it like using Photoshop to fix a picture: Photoshop can make a huge improvement in most pictures; and someone who is skilled with it can do a better job than an amateur; but there are still limits as to what it can do; and, the better the picture you start out with, the better your end product will be for a given level of skill on the part of the operator. Dirac is a very GOOD room correction system, which means that it can produce better results than most others, but it is still limited by what it has to work with. Interesting, thanks. So with all this data manipulation at the output, seems like there is little need to be concerned about frequency or phase response of components in the chain - since it will be "fixed in the mix". I don't really believe it, just toying with ideas! The best answer is good components in a properly treated room, IMO. And as little manipulation of the music as humanly possible. |
|
|
|
Post by yves on Apr 14, 2016 7:13:17 GMT -5
Interesting, thanks. So with all this data manipulation at the output, seems like there is little need to be concerned about frequency or phase response of components in the chain - since it will be "fixed in the mix". I don't really believe it, just toying with ideas! The best answer is good components in a properly treated room, IMO. And as little manipulation of the music as humanly possible. Sometimes when boosting high frequencies with an equalizer, you'll hear the hollow sound of a phase shifter effect. But what you're hearing is comb filtering already present in the source that's now brought out by the EQ. You're not hearing the phase shift itself. As long as the amount of phase shift can be kept sufficiently low, the phase shift itself only becomes audible if it is changing, fairly rapidly — albeit a filter with a high Q (narrow peak) will typically introduce stronger unwanted artifacts so in practice, the correction desired from the EQ curve being applied must carefully be weighed against the detrimental audible side effects it may cause. The moral of the story IMO is treat your room first, apply EQ only if still needed after that. |
|
KeithL
Administrator
Posts: 10,276
|
Post by KeithL on Apr 14, 2016 9:35:52 GMT -5
I would like to add one thing to what you said (I don't disagree with any of it). The part about phase shift not being very audible unless it's varying is true - but only as long as the phase shift is equal in both/all channels. If a given sound appears in both channels (in stereo, if something is "in the center", it's really equal in both left and right), and you shift the phase in both of them equally, you probably won't notice it. However, if you do something that results in an UNEQUAL phase shift in the two channels over some range of frequencies, then the location in the sound stage of anything that's affected will be wildly altered. (A small relative shift will move the location to the left or right; a large shift will often result in a weird effect where you can hear the instrument or voice, but it seems to be coming from someplace entirely outside the sound stage. And, if the shift affects only some of the harmonics of a given instrument, it will make that instrument seem smeared oddly around the sound stage.) This is the sort of situation that is unlikely to occur with "tone controls" that apply to both channels equally, more likely if you have independent tone controls and set them differently, and most common with independent EQ for each channel - like you find in room correction. It's also true that, as you say, an EQ filter with a high Q, which you might prefer because it will have very little effect outside the range of frequencies it's set to affect, will be the most likely to introduce significant phase shift and other artifacts. Anyone familiar with old style "electronic" equalizers should also be aware that, while some of these effects cannot be entirely eliminated, modern DSP-based EQ solutions are usually programmed to at least minimize unwanted side effects - and are much less prone to really serious ones. Interesting, thanks. So with all this data manipulation at the output, seems like there is little need to be concerned about frequency or phase response of components in the chain - since it will be "fixed in the mix". I don't really believe it, just toying with ideas! The best answer is good components in a properly treated room, IMO. And as little manipulation of the music as humanly possible. Sometimes when boosting high frequencies with an equalizer, you'll hear the hollow sound of a phase shifter effect. But what you're hearing is comb filtering already present in the source that's now brought out by the EQ. You're not hearing the phase shift itself. As long as the amount of phase shift can be kept sufficiently low, the phase shift itself only becomes audible if it is changing, fairly rapidly — albeit a filter with a high Q (narrow peak) will typically introduce stronger unwanted artifacts so in practice, the correction desired from the EQ curve being applied must carefully be weighed against the detrimental audible side effects it may cause. The moral of the story IMO is treat your room first, apply EQ only if still needed after that. |
|
|
|
Post by 509Paul on Apr 14, 2016 9:40:19 GMT -5
Is there a way to measure or calculate phase shift?
|
|
|
|
Post by garbulky on Apr 14, 2016 10:02:11 GMT -5
What is phase shift? What does it sound like? Anybody have real world examples or an audio file where it happens?
|
|
|
|
Post by audiobill on Apr 14, 2016 10:09:14 GMT -5
Here we go!!
|
|
