Hi Folks,
I was intrigued by Grimm Audio's claims of the benefits of a wide baffle, so decided to explore how they affect perceived timbre, with some simulations, and start a new thread here. I've attached the complete pdf, but some high level observations are:
Grimm Design Philosophy: “What we can do is decide where to put the baffle-step frequency. Psychoacoustics tells us that below about 300Hz the ear will no longer clearly discern direct sound, first reflections or reverberant sound. Presumably the evolutionary background of this is that the vocal range starts here. Anyhow, we can live with a transition to omni below 300Hz. If we can get reasonable, constant directivity above that, we have a design spec!” “doing it right means: a single unit per “way”, spaced closely together on a baffle of at least 20” wide. Now, wide baffles have a reputation of less than stellar imaging. This is because diffraction artefacts become audible as distinct reflections whereas on narrow baffles they are perceived only as colouration. The solution is simple enough: generously round off the edges”
After modeling the diffraction of the Grimm LS-1 and comparing to a standard narrow baffle:
The LS-1 is a carefully designed system where the baffle width was chosen to counteract the specific driver’s change in response off axis out to 30 degrees, leading to greater flatness within a 30 degree listening window. Since the design is optimized for near field recording, this is staple requirement. The trade-offs are larger off axis nulls due to the larger woofer. However
• The wider baffle decreases the off axis response from 200 to 700 Hz by ~ 3 dB, relative to the standard narrow baffle. Stated another way, equalized to flat on axis, the LS-1 (or other large baffle designs) will have up to (the full spherical diffraction would need to be calculated to arrive at an exact value) 3dB less power response in this frequency range than a narrow monitor. A speaker with excess 600Hz can sound “hollow”, so this tuning will be the opposite of that (less hollow than pure accuracy?), but can also reduce the clarity of bass lines or thin out guitars.
• The similar is of course true for the narrow box. Equalized to flat on axis, the narrow monitor will have up to (same caveat for full spherical diffraction) 3dB less power response from 700Hz to 1.2 kHz than a wide baffle speaker. A speaker with less energy in this range can sound less “punchy”. The 500Hz to 1KHz region produces 35% of the intelligibility, while the range from 1 to 8KHz produces just 5% of the power but 60% of the intelligibility. So the wider speaker may offer enhanced intelligibility.
I think it is these last attributes: differences in power response; that probably most distinguishes the sound of the LS-1 relative to an equally well designed narrow monitor: it puts the power response “hole” in a different frequency range than a narrow monitor. This is akin to providing less “power” BDC while maintaining a flat on axis response.
I can see how this could be “interpreted” as more accurate. With less energy in the lower frequencies, there is less upward masking, therefore more perceived detailed. It’s impossible to ascertain if it’s truly more accurate, as true accuracy would involve a perfect engineer capturing the live event perfectly on his recording system, and a playback system with the same on axis and directivity as the recording system. So, is the LS-1 design or are other wide baffle systems more accurate than a narrow monitor? This is impossible to tell due to the circle of confusion. Is the LS-1 a very well-engineered loudspeaker with a relatively unique choice of trade-offs for the consumer? Definitely!
Epilogue: The LS-1 design brief on the Grimm Audio web site contains a comparison of LS-1 directivity to that of a narrow M-T-M. This is misleading as it’s an apples and oranges comparison. The negative effects of the MTM under discussion were due to two woofers increasing off axis comb filtering, along with the greater diffraction presented to the tweeter by the 2 woofers vs 1 woofer. Neither of these drawbacks will be worse relative to the LS-1 for a well-designed narrow M-T.
Full details attached. Comments welcome, looking forward to others' findings with wide baffles (especially measurements).
I was intrigued by Grimm Audio's claims of the benefits of a wide baffle, so decided to explore how they affect perceived timbre, with some simulations, and start a new thread here. I've attached the complete pdf, but some high level observations are:
Grimm Design Philosophy: “What we can do is decide where to put the baffle-step frequency. Psychoacoustics tells us that below about 300Hz the ear will no longer clearly discern direct sound, first reflections or reverberant sound. Presumably the evolutionary background of this is that the vocal range starts here. Anyhow, we can live with a transition to omni below 300Hz. If we can get reasonable, constant directivity above that, we have a design spec!” “doing it right means: a single unit per “way”, spaced closely together on a baffle of at least 20” wide. Now, wide baffles have a reputation of less than stellar imaging. This is because diffraction artefacts become audible as distinct reflections whereas on narrow baffles they are perceived only as colouration. The solution is simple enough: generously round off the edges”
After modeling the diffraction of the Grimm LS-1 and comparing to a standard narrow baffle:
The LS-1 is a carefully designed system where the baffle width was chosen to counteract the specific driver’s change in response off axis out to 30 degrees, leading to greater flatness within a 30 degree listening window. Since the design is optimized for near field recording, this is staple requirement. The trade-offs are larger off axis nulls due to the larger woofer. However
• The wider baffle decreases the off axis response from 200 to 700 Hz by ~ 3 dB, relative to the standard narrow baffle. Stated another way, equalized to flat on axis, the LS-1 (or other large baffle designs) will have up to (the full spherical diffraction would need to be calculated to arrive at an exact value) 3dB less power response in this frequency range than a narrow monitor. A speaker with excess 600Hz can sound “hollow”, so this tuning will be the opposite of that (less hollow than pure accuracy?), but can also reduce the clarity of bass lines or thin out guitars.
• The similar is of course true for the narrow box. Equalized to flat on axis, the narrow monitor will have up to (same caveat for full spherical diffraction) 3dB less power response from 700Hz to 1.2 kHz than a wide baffle speaker. A speaker with less energy in this range can sound less “punchy”. The 500Hz to 1KHz region produces 35% of the intelligibility, while the range from 1 to 8KHz produces just 5% of the power but 60% of the intelligibility. So the wider speaker may offer enhanced intelligibility.
I think it is these last attributes: differences in power response; that probably most distinguishes the sound of the LS-1 relative to an equally well designed narrow monitor: it puts the power response “hole” in a different frequency range than a narrow monitor. This is akin to providing less “power” BDC while maintaining a flat on axis response.
I can see how this could be “interpreted” as more accurate. With less energy in the lower frequencies, there is less upward masking, therefore more perceived detailed. It’s impossible to ascertain if it’s truly more accurate, as true accuracy would involve a perfect engineer capturing the live event perfectly on his recording system, and a playback system with the same on axis and directivity as the recording system. So, is the LS-1 design or are other wide baffle systems more accurate than a narrow monitor? This is impossible to tell due to the circle of confusion. Is the LS-1 a very well-engineered loudspeaker with a relatively unique choice of trade-offs for the consumer? Definitely!
Epilogue: The LS-1 design brief on the Grimm Audio web site contains a comparison of LS-1 directivity to that of a narrow M-T-M. This is misleading as it’s an apples and oranges comparison. The negative effects of the MTM under discussion were due to two woofers increasing off axis comb filtering, along with the greater diffraction presented to the tweeter by the 2 woofers vs 1 woofer. Neither of these drawbacks will be worse relative to the LS-1 for a well-designed narrow M-T.
Full details attached. Comments welcome, looking forward to others' findings with wide baffles (especially measurements).
Attachments
Last edited:
Hi All,
Based on email received, I want to clarify something in case it wasn't clear in the paper.
When the notch in the power response is moved to lower frequencies by wide baffles, it will tend to sound more like a frequency response change. We hear the integrated room response as timbre below ~ 500 Hz (Grimm quotes 300Hz, Toole 500Hz).
With a standard thin baffle, the power response is higher in frequency (~ 1kHz) and has less impact on the timbre as direct sound starts to dominate timbre. We integrate timbre over shorter time windows as you go up in frequency.
Of course, side wall reflections will still affect tonality at the higher frequencies if the side wall reflection is close enough (some quote 10 ms, others 15 ms) to the direct sound.
Comment still stands about the room modes having more affect on the power notch with a wide baffle.
This was a bit oh an "AHA" moment for me that I hadn't seen elsewhere, so thought I'd share.
Based on email received, I want to clarify something in case it wasn't clear in the paper.
When the notch in the power response is moved to lower frequencies by wide baffles, it will tend to sound more like a frequency response change. We hear the integrated room response as timbre below ~ 500 Hz (Grimm quotes 300Hz, Toole 500Hz).
With a standard thin baffle, the power response is higher in frequency (~ 1kHz) and has less impact on the timbre as direct sound starts to dominate timbre. We integrate timbre over shorter time windows as you go up in frequency.
Of course, side wall reflections will still affect tonality at the higher frequencies if the side wall reflection is close enough (some quote 10 ms, others 15 ms) to the direct sound.
Comment still stands about the room modes having more affect on the power notch with a wide baffle.
This was a bit oh an "AHA" moment for me that I hadn't seen elsewhere, so thought I'd share.
It's an intestesting topic. My experience was that narrow baffles provided a noticeably better channel separation and central image. I also found that the narrow baffle designs often gave the impression that the sound, or timbre, was more to the effect of the driver's direct radiation and less of that which had been reflected by the baffle.
A number of authors were at odds about wide versus narrow baffle width, and this seems ever-present in today's audio markets and DIY alike.
During the design of the Bowers & Wilkins' 800 series floor standing speakers, the designers stated they selected the midrange enclosure's 11" diameter as a balance between the desired cavity volume and imaging capability. Beyond 11", the imaging suffered, so it was said.
One speaker with a remarkable polar response was the NS-1000 Monitor. The radiation pattern was very wide, allowing mid frequencies through treble to be heard at virtually any angle. One would perhaps think the pattern would have been narower. While perhaps a bit off topic, I ended up simulating the baffle and the ripple was very minimal.
A number of authors were at odds about wide versus narrow baffle width, and this seems ever-present in today's audio markets and DIY alike.
During the design of the Bowers & Wilkins' 800 series floor standing speakers, the designers stated they selected the midrange enclosure's 11" diameter as a balance between the desired cavity volume and imaging capability. Beyond 11", the imaging suffered, so it was said.
One speaker with a remarkable polar response was the NS-1000 Monitor. The radiation pattern was very wide, allowing mid frequencies through treble to be heard at virtually any angle. One would perhaps think the pattern would have been narower. While perhaps a bit off topic, I ended up simulating the baffle and the ripple was very minimal.
Thanks kouiky
Large baffles increase the difference in time between the diffraction radiation off the edge, and the direct sound. Longer it is, the harder it is for the ear to fuse it to the direct sound and not pull the image.
Grimm claims that the large round overs resolve that drawback. A large radius will spread out the time of flight of the diffraction while lowering the level from any one part of the edge, so it's hard to imagine how the large radius' can eradicate this effect more than just change it
Large baffles increase the difference in time between the diffraction radiation off the edge, and the direct sound. Longer it is, the harder it is for the ear to fuse it to the direct sound and not pull the image.
Grimm claims that the large round overs resolve that drawback. A large radius will spread out the time of flight of the diffraction while lowering the level from any one part of the edge, so it's hard to imagine how the large radius' can eradicate this effect more than just change it
I put the question to the cognoscienti once in the form of a horn wall rolling onto a parallel room wall. Broadly speaking I was thinking toward guiding the waves as far as possible before the inevitable reflection, but I was advised (depending on the situation) to consider using the available space to diffract as much as possible before the wall.
I use a 'wedge' of absorption to mop up the higher frequency diffraction products without exacerbating that produced by the radius which matches the wall spacing and works with the mouth size to pass only lower frequencies to the wall.
I use a 'wedge' of absorption to mop up the higher frequency diffraction products without exacerbating that produced by the radius which matches the wall spacing and works with the mouth size to pass only lower frequencies to the wall.
It's been suggested that, like this, a spherical enclosure is ideal.. or maybe a decreasing radius with distance depending who you ask. It's a conclusion I've also arrived at time and time again. It would seem to be the basis for the LeCleach horn profile.
A word of encouragement here: It's not that hard, really. A year ago I had no idea how to do it, and today I'm doing it just fine - while keeping up with hectic professional and personal life, of course.
All you need:
- a computer for playback a DSP. Mine is super optimized, but almost anything will work.
- Acourate software: for digital crossovers, digital room correction, driver linearization, time alignment, etc.
- A multichannel DAC. Mine is a Lynx Hilo, but cheaper options available to try it out.
- a couple stereo amps, or as many monoblocks as you desire. I had a stereo tubed amp and powered subwoofers, and built Hypex UcD monos (very easy to do - my first builds).
- your existing passively xo speakers, remove the xo connectors, plug in connections from your amps.
And this is a more advanced one for actively crossing over with it: Computer Audiophile - Advanced Acourate Digital XO Time Alignment Driver Linearization Walkthrough
For me it has been a turning point in my audio journey.
I hope this helps!
You don't need to be a specialist. The software takes care of it and it can be as simple or as involved as you desire. It is quite simple, really. The articles should give you a good idea. From the "Advanced" article I only took away the concept - the actual mechanics were beyond me until I had the setup ready for implementation.
My two cents, of course.
There's also the Bodzio Ultimate Equalizer. For $150 vs $400 (Accourate), it provides dsp speaker crossover support, but no room correction. Unfortunately I've never heard it described as easy to use. In fact, it took me 10 minutes of searching just to find how much it costs and where to order. But its another option!
Last edited:
Our early experiments with wide baffles had the sound more audiably sticking to the speaker, unlike our full-on miniOnken which have the ability to disappear.
This speaker, with a wide curve off the baffle also pulls that off.
dave
These are a bit reminiscent of a PSB design from the mid 80s, where barton had a large baffle with similar angles. I remember liking it when I heard it at the time.
Be interesting to determine if adding larger rounds to the baffle edges allowed images to better free up from the speakers
You underestimate my software stupidity.
Nonetheless, you've encouraged me to at least try before doing an Elvis on my computer screen.How does it work when you are using a bass section and a separate midrange/tweeter section on top, such as Vapor Audio Joule, Wilsons, etc?
Bass sections seem to be mostly flat baffles with rounded edges. Say a 400mm wide baffle would have a baffle step F3 of 290Hz. If the xo was 400Hz then the woofers would need to compensate for the baffle step and by 400Hz the effect of baffle step will be gone. A flat baffle with 2" radiused edges seems ok at these frequencies?
If a midrange section was sitting above, say with a baffle width of 260mm (about 10 inches), the baffle step F3 would be at 460Hz...leaving the midrange to compensate for it. Increasing the midrange baffle width would lower the F3 making life easier for the midrange driver, but would go beyond the 11" quoted as the threshold claimed by B&W for intelligibility (mentioned earlier in this thread). What would be best practice for solving this?
Bass sections seem to be mostly flat baffles with rounded edges. Say a 400mm wide baffle would have a baffle step F3 of 290Hz. If the xo was 400Hz then the woofers would need to compensate for the baffle step and by 400Hz the effect of baffle step will be gone. A flat baffle with 2" radiused edges seems ok at these frequencies?
If a midrange section was sitting above, say with a baffle width of 260mm (about 10 inches), the baffle step F3 would be at 460Hz...leaving the midrange to compensate for it. Increasing the midrange baffle width would lower the F3 making life easier for the midrange driver, but would go beyond the 11" quoted as the threshold claimed by B&W for intelligibility (mentioned earlier in this thread). What would be best practice for solving this?
You underestimate my software stupidity.
Believe me, I'm far from an expert. Very far!
The articles I linked to are step by step instructions with Acourate. I'll be happy to help - IF I can...of course
There is also an Acourate yahoo group with very helpful folks, including the man behind Acourate. I'm not affiliated in any way, BTW, just a satisfied customer.
I could not calculate a passive xo. Since discovering this I've gone active and going into speaker building mode. A lot of fun!
It is an interesting study. Thanks for sharing.
Multiway crossover, delay, EQ can be had for free. You can use DRC, which is free, for auto correction, or use RePhase for manual correction. Then use Foobar or JRiver to put convolution into the audio path. It is not hard to do. There is a great guide for DRC in the full range forum. Otherwise, DRC is a bit harder to work with. RePhase is super easy to use. Having a good understanding of speaker measurements is the hard part.
Multiway crossover, delay, EQ can be had for free. You can use DRC, which is free, for auto correction, or use RePhase for manual correction. Then use Foobar or JRiver to put convolution into the audio path. It is not hard to do. There is a great guide for DRC in the full range forum. Otherwise, DRC is a bit harder to work with. RePhase is super easy to use. Having a good understanding of speaker measurements is the hard part.
Grimm chose to limit the maximum baffle size for the sake of appearance.
I simmed up the on axis diffraction if they extended the baffle to the floor (floor notch/reinforcement not included). Attached the graph. Also included sims with the Edge as a data check.
Extending baffle to the floor adds about a 1 dB boost on axis below ~ 200Hz. Since Grimm equalizes for flat on axis anechoic, that means the large baffle would lean out the perceived response by 1 dB below 200 Hz, given that it's the room power response that dominates timbre at those lower frequencies.
I simmed up the on axis diffraction if they extended the baffle to the floor (floor notch/reinforcement not included). Attached the graph. Also included sims with the Edge as a data check.
Extending baffle to the floor adds about a 1 dB boost on axis below ~ 200Hz. Since Grimm equalizes for flat on axis anechoic, that means the large baffle would lean out the perceived response by 1 dB below 200 Hz, given that it's the room power response that dominates timbre at those lower frequencies.
Attachments
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.