At one time there was a mention of an ambient plugin for miniDSP for the Car crowd, based on L-R and R-L, band passed and enough delay. No idea if it was ever released.
http://www.diymobileaudio.com/forum/technical-advanced-car-audio-discussion/81935-minidsp-intelligent-rear-fill-plugin.html
http://www.diymobileaudio.com/forum/technical-advanced-car-audio-discussion/81935-minidsp-intelligent-rear-fill-plugin.html
Last edited:
Here is a question. Although I recently said my ambient project isn't a 5.1 clone, what does 5.1 synthesized from 2 ch really give me? Dolby Pro Logic 2? Is it simply a R-L, L-R scheme? Is it something more? If its purpose would be to provide ambience, and my current project gives me all the ambience I need, then wouldn't 5.1 be redundant?
No, just way more experienced at this than me at least, and quite a few of those reading with interest I'm sure. The comment was not meant to help you, just to add something to the discussion.
So, the JRiver 5.1 upmix does not add delay, does it?
I will add that I agree with Mike that your ultimate goal of trying to improve the realism of the presentation is the same as that of the 5.1 channel delivery system. And that the 5.1 delivery method coupled with 5.1 recordings is the right way to go. But there may be other ways to achieve the same goals, maybe simpler ways, with fewer channels and customized to the rooms you have.
So, the JRiver 5.1 upmix does not add delay, does it?
I will add that I agree with Mike that your ultimate goal of trying to improve the realism of the presentation is the same as that of the 5.1 channel delivery system. And that the 5.1 delivery method coupled with 5.1 recordings is the right way to go. But there may be other ways to achieve the same goals, maybe simpler ways, with fewer channels and customized to the rooms you have.
Last edited:
I don't think it does, but it's one of those secret sauce recipe's again. I can't find too much on what it does. Only that the mains signal in Audio seems to be left intact. In Video they try and get all phantom info to the center channel. But I couldn't find any info on what they do for the surrounds. The info I describe here is from 2012 or so. They may have put in changes. I've listened to its separate channels, rather easy to do with Metaplugin to route the output on headphones. Maybe one day I'll try and figure out what they are doing. Up-mixing a couple of test tracks should show some logic I guess.
But it would be easy enough to add your own delay if needed. Check it with a loop measurement for instance.
But it would be easy enough to add your own delay if needed. Check it with a loop measurement for instance.
You did see the big grin, right?
I guess what i really want to know is what 5.1 does in the rear channels. For me, I dont have a single 5.1 recording and dont do movies through my main audio system. So therefore, my interest is what 5.1 does when fed 2.0. And even more specifically, when there is no center channel.
One thing seemingly we all agree on is whats best can be very room specific. Whatever it is 5.1 does to a 2.0 input, its not going to adjust itself to my rooms specific acoustics.
I've asked myself this question. I know Logic 7 changes the signal with steering. Not what I wanted. I don't know enough about Logic II yet to answer that with confidence. Some people like the upmix, but more seem to enjoy Logic 7. I'm liking my own experiments enough to not worry about what I might be missing. I try to stick to simulating a better room. I do not want to steer sound. I can play 5.1 content over my 4 speakers with a phantom center and plenty of bottom end (both fronts receive a full range signal).
I like surround in movies though. Fly-by's are fun and entertaining. Due to lacking the bottom end in the surrounds I can't enjoy quad recordings to the fullest.
I have yet to try/hear a mesmerizing 5.1 audio recording. Any recommendations? Preferably something that was meant to be 5.1 from the start. I don't look forward to hearing tricked out audio with drums from the back etc. I'd just want to hear more of the venue, like being there. As that's my goal in 2 channel audio as well.
For the most part, I am with you (what I bolded). But I am curious what other types of induced rear/lateral returns are doing.
What is steering?
I have never heard a 5.1 setup of any kind, so I couldn't recommend anything.
Logic 7 sends some sound to the surrounds based on content and phase and might even remove it from the mains (at least that's what I'm reading in the Logic papers). That's what I refer to as "steering the sound". That's really altering the stereo content. I don't believe Dolby's variant does that. I'm only borrowing the reverb ideas from Logic 7. I want my mains do the initial wave front.
I can believe surround has a better shot at a believable experience, if the content was recorded as such. I'd love to try some of that type recordings in my 5.1 to 4.0 mode.
I can believe surround has a better shot at a believable experience, if the content was recorded as such. I'd love to try some of that type recordings in my 5.1 to 4.0 mode.
Last edited:
For you and me at least, we are leaving out a source. Our kickers. So we could define sources as:
mains
delayed and directional (120 degrees +/-) rear (kickers)
delayed ambient (90 degrees in my case)
sub-woofers
In my case, the answer would be 2.2.4.3
This is starting to look like a barker sequence
In any event, my point is that regardless of the scheme name (Logic 7, Dolby, 5.1), beyond the mains and subs, its about late arriving energy. On top of this, is the room contribution (what energy from previously described sources that bounces around after the first wavefront from each source).
Now, this is starting to really task my brain in its complexity, so for sanity sake I am going to make it simple for myself and instead of thinking of all these things separately, think of them as one thing. That one thing being the synthesis of all the above. And in reality, that is what we hear. Not 6 or 8 or 10 separate things, but one thing. That said, how good its going to sound is really only about two things.
1) Initial goals and design
2) How well the different parts integrate
So let us put the components in some sort of order. Lets do this temporally.
Mains & Subs (first wavefront) (t = 0)
Kicker (second wavefront) (t = 25ms)
Ambient sources (third wavefront) (t > 25ms)
So now what I am saying is that if integration is key, and our components are working at different temporal points, we could start to think about this like a crossover. Get each part doing what it does best, but give special attention to the XO points. In this case, instead of dividing spectra, we are dividing time. And therefore the points we need to give special attention to are where they intersect temporally.
This is how it works at my house. Probably some similarity to yours.
mains
delayed and directional (120 degrees +/-) rear (kickers)
delayed ambient (90 degrees in my case)
sub-woofers
In my case, the answer would be 2.2.4.3
This is starting to look like a barker sequence
In any event, my point is that regardless of the scheme name (Logic 7, Dolby, 5.1), beyond the mains and subs, its about late arriving energy. On top of this, is the room contribution (what energy from previously described sources that bounces around after the first wavefront from each source).
Now, this is starting to really task my brain in its complexity, so for sanity sake I am going to make it simple for myself and instead of thinking of all these things separately, think of them as one thing. That one thing being the synthesis of all the above. And in reality, that is what we hear. Not 6 or 8 or 10 separate things, but one thing. That said, how good its going to sound is really only about two things.
1) Initial goals and design
2) How well the different parts integrate
So let us put the components in some sort of order. Lets do this temporally.
Mains & Subs (first wavefront) (t = 0)
Kicker (second wavefront) (t = 25ms)
Ambient sources (third wavefront) (t > 25ms)
So now what I am saying is that if integration is key, and our components are working at different temporal points, we could start to think about this like a crossover. Get each part doing what it does best, but give special attention to the XO points. In this case, instead of dividing spectra, we are dividing time. And therefore the points we need to give special attention to are where they intersect temporally.
This is how it works at my house. Probably some similarity to yours.
Last edited:
Side wall reflections will create a sense of spaciousness, but it's always the same, regardless of program material. It can mask spacial cues in a recording. For this reason (and the fact that I often engage a holographic generator circuit (inter-aural cancellation), I prefer to set up my speakers along the longer wall of the rectangular shaped living/listening room. The back wall gets closer to my chair, but it's effect is somewhat similar to both channels, so doesn't damage the stereo effect as much as side wall reflections which can dominate stereo effect spacial cues in the program content. The hologram circuit works much better if the side walls are further away.
Side and rear speakers fed the right signals can make for a much improved listening experience in many cases, but, according to papers by David Griesinger, it's of primary importance that they don't roll off in the lower mid and bass, if you want the best of envelopment, Most surround speakers seem to be small, so they can find a place to be in the typical living room. I guess this is why I'm just trying to do the best version of a 2 to 3 channel "extractor", and keep all the speakers on one side of the room. The best side and/or rear channel speakers might well be just woofers. Multiple woofers appear to work substantially better with typical room acoustic effects.
As for room treatment, I've found that putting absorbtive material in all 2 and 3 surface corners kills room ringing much better than large absorbers on wall surfaces away from corners, like I've seen so many do. I nailed a 2 inch cotton rope in all my corners except at the floor, and it made a huge difference using the hand clap test. The rest of the reflections are arguably useful, in that they are many and random, so tend to fill in each others comb filter effect cancellations (above about 300HZ anyway - below that is another story).
Side and rear speakers fed the right signals can make for a much improved listening experience in many cases, but, according to papers by David Griesinger, it's of primary importance that they don't roll off in the lower mid and bass, if you want the best of envelopment, Most surround speakers seem to be small, so they can find a place to be in the typical living room. I guess this is why I'm just trying to do the best version of a 2 to 3 channel "extractor", and keep all the speakers on one side of the room. The best side and/or rear channel speakers might well be just woofers. Multiple woofers appear to work substantially better with typical room acoustic effects.
As for room treatment, I've found that putting absorbtive material in all 2 and 3 surface corners kills room ringing much better than large absorbers on wall surfaces away from corners, like I've seen so many do. I nailed a 2 inch cotton rope in all my corners except at the floor, and it made a huge difference using the hand clap test. The rest of the reflections are arguably useful, in that they are many and random, so tend to fill in each others comb filter effect cancellations (above about 300HZ anyway - below that is another story).
I'll go on a tangent for a while discussing all of my mayor steps in processing (leaving changes in DRC settings out of this for now, as those are more specific to my type of speaker)
I'm all for avoiding early side wall reflections. I started out in an untreated room, with one prominent side wall due to an asymmetric placement of speakers. My first step was to damp that wall (into the corner) with a pretty huge damping panel, slightly off the wall so each wave hitting that one would travel trough it 2 times.
That changed the soundstage width on that side. Making it way more inline with the other side that didn't have such a boundary. My next step was the corner and wall on the opposite side. At that point my mayor reflections were all coming from the wall behind the listening position. Due to the small room they were about 7 ms behind the mains. Yet they did give a immersive effect to the sound heard. Like the music was all around me. Not the phantom center, as that was in front of me, the sides were wrapping around my listening area.
When I placed damping on that back wall, the imaging improved, everything got more focused and clear. But gone was that immersive (too strong actually) sense of the sides.
But part of me liked that feel, even though it was too early and wrapped the stage around me. Not on all recordings I must say. Some recordings remained wide and in front.
After living with that for a while I started with the kicker. Band passed, attenuated and delayed about 20 ms. It gave me some of that sense of those early reflections without the extreme wrap around. It also separated the front stage more. Somehow it gave more body to the separate sounds in the stage.
In the mean time I played with an "S" curve in mid/side processing to get the tonal balance the same between the center and sides. At first with help of the JRiver effect "Surround Field" as mentioned before on this thread. Gradually this surround field effect was turned off completely. The "S" curve remained. It was based on info in this thread: http://www.diyaudio.com/forums/multi-way/277519-fixing-stereo-phantom-center.html
I continued with some cross talk experiments. I noticed the right settings made my phantom center more clear and coherent. But the sides were all messed up. I've always noticed at least 2 different types of phantom content. Some songs had a very coherent voice from the start. Other recordings, still with the voice prominent in the center were harder to follow. It's those voices that became more coherent with cross talk cancelation. I tried to apply it only to the center channel (using my mid/side processing chain). I never got it to work without messing up parts of the stage. At one point I had great sides, great center but the content panned somewhere in between got a soft focus.
I recorded the cross talk cancelation signal with pink noise and compared it to the signal without cross talk cancelation. I noticed 2 mayor dips in the cross talk cancelation signal. After that I removed the cross talk cancelation DSP and made 2 small EQ cuts similar to the mentioned dips above. Low and behold the better coherency of the phantom center remained. Like I said, only on some songs the phantom was less clear than on other songs. Simply a bit harder to understand the words. While most material never had that problem. I tried to focus on songs from the same album that had both types of phantom center.
For those willing to experiment, I used Christina Aguilera's album "Stripped" for this. Not because I'm a huge fan or anything, just because of the two distinctly different vocals in the center. Plus that "Stripped" CD has a lot of fun stuff in regards of stage manipulation which is why it made it to regular use as reference.
The song: "Impossible" has always had a crystal clear vocal in the phantom center.
In comparison the song: "Underappreciated" the center vocal was more difficult to follow at times. Less focused or coherent. It's this one that benefitted from those two EQ cuts.
The cuts: 3700 Hz, Q= ~2.7 Gain = -1 dB and 7270 Hz, Q=2.7 Gain = -0.7 dB. Only applied to the mid part of my mid/side processing. (Very specific to my setup I might add as it has to do with my listening distance and speaker placement)
Higher gain cuts screwed up the tonal balance of vocals that were good to begin with. I have no idea why there is such a big difference in both these vocal parts. One always being crystal clear and the other more dependent on some processing. I bet the processing is different as I can hear the difference even on headphones.
For me the above 2 recordings were the most clear examples on the same album. I did find one other song with both types of coherency in the same song. The party track: "Fireball" exhibits both types in one song. The main vocal part has always been clear and focused. But the second voice, slightly behind the first one at about 1 minute into the song is less focused (for me). Again the 2 minor cuts helped to clear up the coherency of those lines.
Fast forward to ambient again. As I have said, my ambient info, at that time more kicker than anything else was mainly L-R and L-R with the phantom mixed in at a lower level and slightly different timing (3 ms difference). With the center band passed at 3.5 KHz, 12 dB I lost coherency again in those more difficult to follow vocal parts. I went back to pure L-R and R-L and the phantom improved, but with a penalty of tonal balance.
At that point I added (late) reverb to the ambient mix. First without any phantom center in the ambience setup, at a later point I mixed it back in but this time cut off at 2 KHz, 12 dB. That's where I am today.
Slight changes were made to keep the tonal balance in check, straightening out the targets of the mains and adjusting the ambience to compensate.
My ambience channels don't go low. Cut off on the bottom is about 200 Hz. Going lower might add more envelopment, but they simply cannot reach that low. No multiple subs or other parts are involved. Yet I get a very satisfactory enhancement out of my ambience.
With the added reverb I get more usable depth than I had before. I run a 2 pole all-pass filter at 500 Hz on the ambient channels to delay the lower end of it. Subjectively this helped create more space.
So that's my entire trip in one post. Basically one thing led to another and in between I revisited prior setting to confirm my preference. Meaning I try with and without all of the single steps I have done to see if I still like it.
I'd love to get more input on cross talk cancelation improving the coherency of the phantom center, any others that experienced it like that? Bob?
While I'll keep tinkering and revisiting old ideas I'm pretty pleased so far. The ambience experiments have been one of the most satisfying additions for me. Trying hard to limit it to a minimum but keeping the enhancement I feel it brings.
Bob, did you ever measure at the listening position specifically to see early reflection data?
I'm all for avoiding early side wall reflections. I started out in an untreated room, with one prominent side wall due to an asymmetric placement of speakers. My first step was to damp that wall (into the corner) with a pretty huge damping panel, slightly off the wall so each wave hitting that one would travel trough it 2 times.
That changed the soundstage width on that side. Making it way more inline with the other side that didn't have such a boundary. My next step was the corner and wall on the opposite side. At that point my mayor reflections were all coming from the wall behind the listening position. Due to the small room they were about 7 ms behind the mains. Yet they did give a immersive effect to the sound heard. Like the music was all around me. Not the phantom center, as that was in front of me, the sides were wrapping around my listening area.
When I placed damping on that back wall, the imaging improved, everything got more focused and clear. But gone was that immersive (too strong actually) sense of the sides.
But part of me liked that feel, even though it was too early and wrapped the stage around me. Not on all recordings I must say. Some recordings remained wide and in front.
After living with that for a while I started with the kicker. Band passed, attenuated and delayed about 20 ms. It gave me some of that sense of those early reflections without the extreme wrap around. It also separated the front stage more. Somehow it gave more body to the separate sounds in the stage.
In the mean time I played with an "S" curve in mid/side processing to get the tonal balance the same between the center and sides. At first with help of the JRiver effect "Surround Field" as mentioned before on this thread. Gradually this surround field effect was turned off completely. The "S" curve remained. It was based on info in this thread: http://www.diyaudio.com/forums/multi-way/277519-fixing-stereo-phantom-center.html
I continued with some cross talk experiments. I noticed the right settings made my phantom center more clear and coherent. But the sides were all messed up. I've always noticed at least 2 different types of phantom content. Some songs had a very coherent voice from the start. Other recordings, still with the voice prominent in the center were harder to follow. It's those voices that became more coherent with cross talk cancelation. I tried to apply it only to the center channel (using my mid/side processing chain). I never got it to work without messing up parts of the stage. At one point I had great sides, great center but the content panned somewhere in between got a soft focus.
I recorded the cross talk cancelation signal with pink noise and compared it to the signal without cross talk cancelation. I noticed 2 mayor dips in the cross talk cancelation signal. After that I removed the cross talk cancelation DSP and made 2 small EQ cuts similar to the mentioned dips above. Low and behold the better coherency of the phantom center remained. Like I said, only on some songs the phantom was less clear than on other songs. Simply a bit harder to understand the words. While most material never had that problem. I tried to focus on songs from the same album that had both types of phantom center.
For those willing to experiment, I used Christina Aguilera's album "Stripped" for this. Not because I'm a huge fan or anything, just because of the two distinctly different vocals in the center. Plus that "Stripped" CD has a lot of fun stuff in regards of stage manipulation which is why it made it to regular use as reference.
The song: "Impossible" has always had a crystal clear vocal in the phantom center.
In comparison the song: "Underappreciated" the center vocal was more difficult to follow at times. Less focused or coherent. It's this one that benefitted from those two EQ cuts.
The cuts: 3700 Hz, Q= ~2.7 Gain = -1 dB and 7270 Hz, Q=2.7 Gain = -0.7 dB. Only applied to the mid part of my mid/side processing. (Very specific to my setup I might add as it has to do with my listening distance and speaker placement)
Higher gain cuts screwed up the tonal balance of vocals that were good to begin with. I have no idea why there is such a big difference in both these vocal parts. One always being crystal clear and the other more dependent on some processing. I bet the processing is different as I can hear the difference even on headphones.
For me the above 2 recordings were the most clear examples on the same album. I did find one other song with both types of coherency in the same song. The party track: "Fireball" exhibits both types in one song. The main vocal part has always been clear and focused. But the second voice, slightly behind the first one at about 1 minute into the song is less focused (for me). Again the 2 minor cuts helped to clear up the coherency of those lines.
Fast forward to ambient again. As I have said, my ambient info, at that time more kicker than anything else was mainly L-R and L-R with the phantom mixed in at a lower level and slightly different timing (3 ms difference). With the center band passed at 3.5 KHz, 12 dB I lost coherency again in those more difficult to follow vocal parts. I went back to pure L-R and R-L and the phantom improved, but with a penalty of tonal balance.
At that point I added (late) reverb to the ambient mix. First without any phantom center in the ambience setup, at a later point I mixed it back in but this time cut off at 2 KHz, 12 dB. That's where I am today.
Slight changes were made to keep the tonal balance in check, straightening out the targets of the mains and adjusting the ambience to compensate.
My ambience channels don't go low. Cut off on the bottom is about 200 Hz. Going lower might add more envelopment, but they simply cannot reach that low. No multiple subs or other parts are involved. Yet I get a very satisfactory enhancement out of my ambience.
With the added reverb I get more usable depth than I had before. I run a 2 pole all-pass filter at 500 Hz on the ambient channels to delay the lower end of it. Subjectively this helped create more space.
So that's my entire trip in one post. Basically one thing led to another and in between I revisited prior setting to confirm my preference. Meaning I try with and without all of the single steps I have done to see if I still like it.
I'd love to get more input on cross talk cancelation improving the coherency of the phantom center, any others that experienced it like that? Bob?
While I'll keep tinkering and revisiting old ideas I'm pretty pleased so far. The ambience experiments have been one of the most satisfying additions for me. Trying hard to limit it to a minimum but keeping the enhancement I feel it brings.
Bob, did you ever measure at the listening position specifically to see early reflection data?
Last edited:
Hmm, Toole's studies indicate that strong first reflections from the sidewalls can be "listened through" and do not mask program spatial cues. (I'm not 100% convinced, but thought I should cite that.) But, IACC works best in anechoic conditions, and anytime you are concerned with delaying / reducing first reflections, the "long wall" speaker setup makes sense:
Loudspeaker Placement in Small Rooms | Richard's Stuff
I must point out though, 2" rope in the corners cannot be an effective room treatment. While it is known that bass builds in corners, the rope would have no effect at low frequencies (too small). It is reflective at high frequencies (dense), and has not enough surface area to affect mid or high frequencies (not an effective placement to boot). Unless your walls / floor / ceiling are exceptionally diffuse, the reflections are far from random and can be predicted with ray tracing. The most effective absorption locations are at first reflection points, and behind the listener.
That said Bob I would also be interested in your IACC experiences.. do you have a thread centered on the subject? Does it have anything to do with ambiophonics?
If the rooms I've seen from JBL are any indication I'm guessing his early reflections would be considered kinda late for us less fortunate people
.
If the rooms I've seen from JBL are any indication I'm guessing his early reflections would be considered kinda late for us less fortunate people
Haha.. yeah. That isn't the basis for these studies though. I think they were performed anechoically at the NRC, using a second speaker to produce the "first reflection". After flipping through the book I can't find any recommendation for "minimum delay to first reflection".. going to have to read it all again :$