I can adjust timing on all virtually created reflections. Diffusion would be like manny smaller queues, but at a fixed timing due to geometry. It adds ambient sound but works rather different from hard reflections.
Personally, I went for a 'reflection free' zone of about 17 ms. Thats when the ambient queues kick in.
The 'reflection free' zone isn't completely free of course in a reasonably standard living room with only 3 big panels. The graph I posted shows these smaller reflections.... The Haas kicker was off for that measurement.
Personally, I went for a 'reflection free' zone of about 17 ms. Thats when the ambient queues kick in.
The 'reflection free' zone isn't completely free of course in a reasonably standard living room with only 3 big panels. The graph I posted shows these smaller reflections.... The Haas kicker was off for that measurement.
You have one larger source, I have 25 smaller sources all with slightly different timing
. The sum of those smaller reflections get smeared out in time. Lower in level than one source would show, but spread out over a larger time frame. The higher frequencies are more directional, the curtain is accoustically invissible at lower frequencies. Speaker is at a 15 degree toe-in.Before placing the damping panel it (the side wall) was quite obvious in measurements.
The stage was artificially wide on the left side on all songs at that time.
Last edited:
As always , I find that I am in total agreement with you. It was actually after reading the long WE thread, I got an urge to find a way to describe where I feel the controversy originates from.@koldby, I see your point there, once we find out where the theory does not match the reality with science, we often hunt down the why, don't we.
I have tried to write down a thoughtful answer, but I can't find any. I was bothered by the view on the WE thread that the objective crew seemed to be convinced we have booked a lot of progress trough the years. I can't always see it like that. Living here in Europe I still prefer a building or house that was made in the ~ 30's to the way it is made today. We can do it faster and cheaper today, but will it last as long?
Is that even real progress? The market has miniaturized our audio equipment, at least the main stream commercial crap shows us that trend. Well, that simply doesn't sound the same to my ears.
For a few years I was actually worried mp3 would take over as the main stream media. The 'fast food' equivalent of the music industry, instant gratification guaranteed. Who needs dynamic range anyway.
Throw away those big old recording studios while you're at it, you can do the same with a simple laptop in your home!
See why I shouldn't respond today?
I think that as long as we don't forget to keep listening, really listen to music and still get enjoyment from that we'll probably be alright.
It doesn't matter what your personal preference is, as long as you find something enjoyable in it. For me it's the hunt for that illusion that sends shivers down my spine. The one that brings tears to my eyes from pure emotion.
(that has happened, more than once, over the past few years)
As long as I still get that, from time to time, I'll be continuing this hunt.
@Joseph Crowe, I don't disagree with your view of that speaker/curtain position. However I tested this extensively with measurements and listening tests. The curtain is not a heavy thick fabric, so that helps.
Part of getting rid of the obvious first reflections included many measurements and listening tests, toe in and out of the speakers etc. In this process keeping the spouse happy as well.
Most of the results are somewhere in this thread complete with graphs. The goal has always been to find a good balance between the left and right side at the listening spot due to having to live with this asymmetric setup. This asymmetry actually helps at bass frequencies due to having that left speaker tucked so close into that corner. It's responsible for the lowest octaves, something the right array couldn't do as well. It does have it's own trouble spot around 60-70 Hz, where I shift the energy to the right side. That's one thing I'll try to balance with the subs.
I might be able to pull it out just a little when the subs are ready, but it will be a few cm at most.
Like most things in Audio it is finding the compromise you can live with. However the measurements look very clean (for a listening seat).
Part of getting rid of the obvious first reflections included many measurements and listening tests, toe in and out of the speakers etc. In this process keeping the spouse happy as well
.Most of the results are somewhere in this thread complete with graphs. The goal has always been to find a good balance between the left and right side at the listening spot due to having to live with this asymmetric setup. This asymmetry actually helps at bass frequencies due to having that left speaker tucked so close into that corner. It's responsible for the lowest octaves, something the right array couldn't do as well. It does have it's own trouble spot around 60-70 Hz, where I shift the energy to the right side. That's one thing I'll try to balance with the subs.
I might be able to pull it out just a little when the subs are ready, but it will be a few cm at most.
Like most things in Audio it is finding the compromise you can live with. However the measurements look very clean (for a listening seat).
I think I must have typed in at least 3 replies, the one that stuck was the most gentle/polite one I could come up with
.
Due to an error I've heard this same setup within Home Theatre, except I forgot to invert the 0.27 ms delayed signal. I was pleasantly surprised with the results.
I'll have to run many more experiments but tried this with 0.27 ms delay, a 1 ms delay and a 3 ms delay, all using the same signal. Definite differences could be heard, the 1 ms delay sounding the least favourable of all. Getting a more fuzzy image. Time was up before I could pick a winner between 0.27 and 3 ms, or try even larger numbers.
But I didn't expect this trick to sound as well as it did. The level of the cross talk part mixed back in is 18 dB down compared to main, as said, a little goes a long way.
Opening up REW while comparing, subtracting and adding plots a 0.54 ms delay should sound even sweeter than the 0.27 ms delay. I'll try that next.
Yes it does. Even with eyes open but it works best with eyes closed.
It differs from song to song, it can be a very convincing illusion.
When I still had early reflections, all the songs seemed to image in the same way. That got old fast.
Some songs are small and intimate, other project a wide stage. There's a lot of difference in depth perception as well between songs.
The sound goes in front of and past the speakers, beyond to the sides as well, depending on the recording. I can play with the properties of it by changing my settings. Goal is to keep it as natural as possible, easier said than done. Even perceived height of individual sound sources seem to vary.
Still working on it thought, as said; if I can (am able to) combine everything I've heard so far is will be one big trip!
It differs from song to song, it can be a very convincing illusion.
When I still had early reflections, all the songs seemed to image in the same way. That got old fast.
Some songs are small and intimate, other project a wide stage. There's a lot of difference in depth perception as well between songs.
The sound goes in front of and past the speakers, beyond to the sides as well, depending on the recording. I can play with the properties of it by changing my settings. Goal is to keep it as natural as possible, easier said than done. Even perceived height of individual sound sources seem to vary.
Still working on it thought, as said; if I can (am able to) combine everything I've heard so far is will be one big trip!
Last edited:
Hmm, still feels strange quoting myself but it saves me some typing
.I predicted sweeter sound... after hearing it I would call it smoother. But not necessarily in every good way. Don't get me wrong, it was pretty good, though a little too smooth perhaps? Polite sounding? It made everything a sound bit smoother, but didn't have the "hook" in me. Tonal balance was pretty good, but it didn't capture me.
I've had a long session, playing with levels and kept the timing the same throughout. I think we need either a very short delay or a longer one to not loose the detail.
More will follow, hopefully soon...
Another fun experiment, shortened the cross talk signal back to 0.272 ms again, better!
But in light of the magic WE 15a horn puzzle I tried something new as well. I added another signal, delayed and attenuated, also band passed. Much like what the ambient back channel gets, but coming from the mains themselves.
My simple theory: a horn with an abrupt ending like the WE 15a is bound to have a wave traveling back trough the horn and cause a delayed "copy" as a 'late return'.
Much like what we saw in that document by JMLC I linked to earlier, but due to the horn size it would be delayed much longer than any of the horns presented in that paper. If the 15a has a 15 feet horn path, it would result in a delay of about ~28 ms.
Armed with that simple idea I decided to try it out. First feeding it the (L+R) signal only. I still had it high passed at about ~600 Hz.
Much to my surprise it caused an auditory scene with seemingly unending depth... Though it did sound awful, the overall tonal balance was horrid.
I re-arranged (once again) my routing within JRiver to accommodate all these ideas. A picture says more than lots of words, it looks like this:
How on earth is that going to sound good.
Well it does, and doesn't. It has potential, but as can be seen has almost unlimited possibilities in adjustments. The 'doesn't' part is that it needs fine tuning to really make it shine. Phantom vocal is just too large right now. I've heard it do better.
It looks way more wild than it is though. There's a main signal, which gets split up in mid and side (the regular scheduled program). A copy of mid and side get's rerouted and adjusted back into the main signal, with a short (0.27 ms) delay. It is a high-passed, attenuated signal, but kept pretty simple. Acting as a (anti) cross talk helper. All of the phantom material gets a copy of this CT helper, while on the pure side signal gets deducted. (warning this still leaves a (lower in level) copy in the opposite channel).
I'm too pleased with the side tonal balance to have it in there.
This part was tested by itself before the madness as can be read below started.
The mid most chain is the 'late return', the 'fake horn reflection' (if I should even dare call it that) delayed 25 ms. On the right most channels is the usual ambient mix with a delay set to 28 ms, making it arrive a little later than the 'late return'.
Don't get scared, I haven't lost it completely. I'm just curious enough to fool around with ideas here. I hate to admit that it is much easier to create a (very cool) sense of depth with this 'late return' coming from the fronts.
The ambient channels arriving laterally are better at creating envelopment. Still it seems to work real well together like this. And it does make it sound real pleasant for a lot of very differing material being played back.
An example song where it works well:
Kimbra - Plain Gold Ring. A song full of effects, and it really creates a beautiful "space" with the floating sounds all around.
It's working very well on Ayreon - Into the Electric Castle. I played two songs, Valley of the Queens and The Castle Hall. Probably a few more tomorrow, this was fun!
A track that always gets played to check its results: The Doors - Ghost Song...
Well this song does not need much to sound real cool. It still works well though. What I think is it has enough "room" mixed in it to not need much help. Jim sounded a bit bloated in the lower registries. This is a song where they took a poem recited by Jim and arranged the music around him after he was gone. I think it works well, sometimes scary even.
I played many more songs, I couldn't tell you if I'll keep this or not. Basically it's curiosity driven good fun, but that's where the mid/side EQ started too! Some things work way too good to let go.
I think I can get about 90% of the effect the 'late return' has with/from my ambient speakers alone. If it turns out it is worth keeping, I'd want to have dedicated speakers (which I would need to hide from sight very well) instead of mixing it back into the mains.
For those interested in more WE 15a stories: https://www.lencoheaven.net/forum/index.php?topic=7616.0
Reading that thread made me feel the whole WECO scene almost seems like a religion. I didn't get much from that thread though, I still think the bigger part of their magic lies in that horn. But in that thread you wouldn't be taken seriously unless you have the whole shebang, with all the right parts.
It describes how that demo in 2012 Munich went where other WE owners were present. All I hoped to find was perhaps a set of measurements, like an IR pulse, the phase/FR plot and a wavelet. That's probably not going to happen, ever.
But in light of the magic WE 15a horn puzzle I tried something new as well. I added another signal, delayed and attenuated, also band passed. Much like what the ambient back channel gets, but coming from the mains themselves.
My simple theory: a horn with an abrupt ending like the WE 15a is bound to have a wave traveling back trough the horn and cause a delayed "copy" as a 'late return'.
Much like what we saw in that document by JMLC I linked to earlier, but due to the horn size it would be delayed much longer than any of the horns presented in that paper. If the 15a has a 15 feet horn path, it would result in a delay of about ~28 ms.
Armed with that simple idea I decided to try it out. First feeding it the (L+R) signal only. I still had it high passed at about ~600 Hz.
Much to my surprise it caused an auditory scene with seemingly unending depth... Though it did sound awful, the overall tonal balance was horrid.
I re-arranged (once again) my routing within JRiver to accommodate all these ideas. A picture says more than lots of words, it looks like this:
How on earth is that going to sound good
.Well it does, and doesn't. It has potential, but as can be seen has almost unlimited possibilities in adjustments. The 'doesn't' part is that it needs fine tuning to really make it shine. Phantom vocal is just too large right now. I've heard it do better.
It looks way more wild than it is though. There's a main signal, which gets split up in mid and side (the regular scheduled program). A copy of mid and side get's rerouted and adjusted back into the main signal, with a short (0.27 ms) delay. It is a high-passed, attenuated signal, but kept pretty simple. Acting as a (anti) cross talk helper. All of the phantom material gets a copy of this CT helper, while on the pure side signal gets deducted. (warning this still leaves a (lower in level) copy in the opposite channel).
I'm too pleased with the side tonal balance to have it in there.
This part was tested by itself before the madness as can be read below started.
The mid most chain is the 'late return', the 'fake horn reflection' (if I should even dare call it that) delayed 25 ms. On the right most channels is the usual ambient mix with a delay set to 28 ms, making it arrive a little later than the 'late return'.
Don't get scared, I haven't lost it completely. I'm just curious enough to fool around with ideas here. I hate to admit that it is much easier to create a (very cool) sense of depth with this 'late return' coming from the fronts.
The ambient channels arriving laterally are better at creating envelopment. Still it seems to work real well together like this. And it does make it sound real pleasant for a lot of very differing material being played back.
An example song where it works well:
Kimbra - Plain Gold Ring. A song full of effects, and it really creates a beautiful "space" with the floating sounds all around.
It's working very well on Ayreon - Into the Electric Castle. I played two songs, Valley of the Queens and The Castle Hall. Probably a few more tomorrow, this was fun!
A track that always gets played to check its results: The Doors - Ghost Song...
Well this song does not need much to sound real cool. It still works well though. What I think is it has enough "room" mixed in it to not need much help. Jim sounded a bit bloated in the lower registries. This is a song where they took a poem recited by Jim and arranged the music around him after he was gone. I think it works well, sometimes scary even.
I played many more songs, I couldn't tell you if I'll keep this or not. Basically it's curiosity driven good fun, but that's where the mid/side EQ started too!
Some things work way too good to let go.I think I can get about 90% of the effect the 'late return' has with/from my ambient speakers alone. If it turns out it is worth keeping, I'd want to have dedicated speakers (which I would need to hide from sight very well) instead of mixing it back into the mains.
For those interested in more WE 15a stories: https://www.lencoheaven.net/forum/index.php?topic=7616.0
Reading that thread made me feel the whole WECO scene almost seems like a religion. I didn't get much from that thread though, I still think the bigger part of their magic lies in that horn. But in that thread you wouldn't be taken seriously unless you have the whole shebang, with all the right parts.
It describes how that demo in 2012 Munich went where other WE owners were present. All I hoped to find was perhaps a set of measurements, like an IR pulse, the phase/FR plot and a wavelet. That's probably not going to happen, ever.
Attachments
I'll add a separate disclaimer... By now a few readers might think I've lost it completely . Well, I do a lot of these experiments and only time will tell if it's worth keeping (even parts of) it.
By no means would I think this would get me (close to) the WE 15a sound. Luckily I'm not that delusional. I'm merely interested in how and why the room or other parts of playback changes our perception so much.
Usually a "Haas Kicker" is down by about 12 dB compared to the mains. It's delay can vary from about 15 to 25 ms on average. This late addition is down even further, and still has quite an effect on perception, it probably wouldn't work as well if I still had higher levels of early reflections.
Think of OB speakers, that back wave is going to make it back to the listening position as well. But only placement and room treatment could determine when and how much. Compared to that, these experiments are quick, adjustable and easy to leave behind if it doesn't work out.
More than half, way more actually, of my experiments don't work out for the long run. Some of them do though, and if I didn't try anything I would never know.
Hope I don't bore you guys to much though.
. Well, I do a lot of these experiments and only time will tell if it's worth keeping (even parts of) it.By no means would I think this would get me (close to) the WE 15a sound. Luckily I'm not that delusional. I'm merely interested in how and why the room or other parts of playback changes our perception so much.
Usually a "Haas Kicker" is down by about 12 dB compared to the mains. It's delay can vary from about 15 to 25 ms on average. This late addition is down even further, and still has quite an effect on perception, it probably wouldn't work as well if I still had higher levels of early reflections.
Think of OB speakers, that back wave is going to make it back to the listening position as well. But only placement and room treatment could determine when and how much. Compared to that, these experiments are quick, adjustable and easy to leave behind if it doesn't work out.
More than half, way more actually, of my experiments don't work out for the long run. Some of them do though, and if I didn't try anything I would never know.
Hope I don't bore you guys to much though
.
Cool experiments. It will all take some time to digest.
Agree ...
...Hope I don't bore you guys to much though
Indeed you don't bore us and with reference of below kolby post remind me how shall we ever get to core of stuff if we didn't research all possibilityes that these modern DSP engines actual can replicate if we willing to live with the relative small delay it takes to process audio stream to be repaired to the better for actual situation, so thanks please keep on rocking the research.
By the way think had one big subjective discovery for some time myself that need advanced FIR filter in DSP engine to be executed but need some more time to investigate if my ears are fooled or the big grin will continue , never the less its fun and educating having the possibilityes and power of modern DSP engines
.
Just had another listen and changed a few parameters, to see what it does.
I also let my son listen, and played a track from Ayreon. The start of that song had spacey effects but the rest had a more normal stage. He was disappointed not everything wrapped around him like the start.
I played some Steely Dan to show him it was still there, if the track had it...
Don't tease us, please share!
I also let my son listen, and played a track from Ayreon. The start of that song had spacey effects but the rest had a more normal stage. He was disappointed not everything wrapped around him like the start.
I played some Steely Dan to show him it was still there, if the track had it...
By the way think had one big subjective discovery for some time myself that need advanced FIR filter in DSP engine to be executed but need some more time to investigate if my ears are fooled or the big grin will continue , never the less its fun and educating having the possibilityes and power of modern DSP engines
Don't tease us, please share!
Last edited:
...Don't tease us, please share!
Yes of course but i have to make shure in the end i not make myself look foolish in judge of what i hear
its because in objective domain this filter turns some stuff upside down and logical in first place a really crazy idea to ever try out, i made many crazy investigations down the road because we can with these DSP engines and they mostly fall apart which is good if they really is stupid idea and breaks obejtive domain rules, but this time sound happen is a big smile so far. Well having this conversation reminds me you a good judge and have good ears so will try convolute a track or two over the weekend and share via PM to get your opinion on sound and then lets see if filter holds water for a technical reveal and public sharing
Last edited: