Leap calculates the total driver power with the driver TSP, enclosure volume, speaker system and the radiation impedance, complicated stuff, it is described in the Leap manuals. The total power is independent of environment (full space, half space,…), independent of enclosure shape, waveguides etc. These last parameters determine how the power is distributed in space and is translated to SPL values at different positions, but do not affect the total power.
To measure the total power exactly, you need to measure the SPL curves in the complete horizontal and vertical orbits around the cabinet. Like Floyd Toole describes in its analysis, very complicate to do it ourselves. Therefore the Leap calculation of the total power is not bad. The Vituix power calculation is only done with some frontal off axis measurements and of course is also an indication for the total power, but its absolute power value isn’t correct.
Of course the waveguide has large benefits, I completely agree with that. I can understand that making SPL flat with the tweeter waveguide does sound too bright, the midrange power becomes too high above 1 kHz w.r.t. the tweeter power (look to my EL3 design in previous post)
I have some problems with a SPL on axis that has to fall off for high frequencies to make the power response smooth. Then it is more recommended to use a 7 to 8 inch midrange and combine it with a tweeter with waveguide. For that case the on axis SPL can be made flat, combined with a flat smooth power curve. Maybe it is more recommended also to evaluate the speaker timbre with a stereo set.
It is a design choice, but my own recent experience with speaker design has learned me that the total power behavior is very important to realize a good sounding speaker in a room. Too low power for high frequencies tends to sound too absent, poor energy, a somewhat dull scene ...and that isn’t my taste for sound.
IMO it is important to understand the SPL-power stuff in the best way to realize your own taste of speaker sound.
For the Scanspeak tweeter SPL without waveguide simulation, I used a transducer model using the measured impedance and the datasheet TSP, placed in the cabinet in full space. I think it is close to a correct value.
To measure the total power exactly, you need to measure the SPL curves in the complete horizontal and vertical orbits around the cabinet. Like Floyd Toole describes in its analysis, very complicate to do it ourselves. Therefore the Leap calculation of the total power is not bad. The Vituix power calculation is only done with some frontal off axis measurements and of course is also an indication for the total power, but its absolute power value isn’t correct.
Of course the waveguide has large benefits, I completely agree with that. I can understand that making SPL flat with the tweeter waveguide does sound too bright, the midrange power becomes too high above 1 kHz w.r.t. the tweeter power (look to my EL3 design in previous post)
I have some problems with a SPL on axis that has to fall off for high frequencies to make the power response smooth. Then it is more recommended to use a 7 to 8 inch midrange and combine it with a tweeter with waveguide. For that case the on axis SPL can be made flat, combined with a flat smooth power curve. Maybe it is more recommended also to evaluate the speaker timbre with a stereo set.
It is a design choice, but my own recent experience with speaker design has learned me that the total power behavior is very important to realize a good sounding speaker in a room. Too low power for high frequencies tends to sound too absent, poor energy, a somewhat dull scene ...and that isn’t my taste for sound.
IMO it is important to understand the SPL-power stuff in the best way to realize your own taste of speaker sound.
For the Scanspeak tweeter SPL without waveguide simulation, I used a transducer model using the measured impedance and the datasheet TSP, placed in the cabinet in full space. I think it is close to a correct value.
Just wondering, out of curiosity, I am not very good with this... how does LEAP treat the radiation impedance of waveguides/horns and of ring radiators?
I have now ordered x-over parts to build a second stereo speaker. I decided that I will take the risk and assume the current x-over design is (very close) to the final design, so I did not ask for another round of prototyping money; I will pay this of my own pockets.
Don't hold your breath. It will be a while until I get the second speaker built, because life has some other things coming up, which will need some attention
What the ...? Of course waveguides, horns etc. directivity components increase DI and that way power after axial response is equalized to same as without directivity component. No sense to compare with different axial SPL.
What the ...? VCAD calculates power and DI estimations to full space by radial measurements in two planes either 0...180 or -180...+180 deg. But you should have adequate data for it. Not my bad if half space 0...90 deg or -90...+90 (or less) data is measured while target is to simulate power & DI to full space. Never given permission for that kind of laziness which saves total measurement time couple of minutes.
Anyway, I'm not reading this topic but can answer in VCAD thread if something suspicious is written. I'm not trying to challenge LEAP, but looks that interpretations about facts are somehow biased here. No way that I would estimate power & DI with some simplified simulation instead of actual measurement data - regardless of limitations due to two measured planes only.
Let's try to do a constructive discussion about this...
1. The power calculation method in Vituix is correct, indeed, if you have the complete set of measurements of the horizontal and vertical orbits (cfr. Floyd Toole). That has not been done for the Monkey Box. That is my remark to the Vituix power calculation for the Monkey Box, not more than that.
2. Applying a waveguide to a tweeter doesn't change the total radiated power, a waveguide only affects the horizontal and vertical radiation patterns, in a way SPL changes at different positions, more SPL on axis, less SPL off axis (DI is changing).
Briefly: using a waveguide, DI and SPL are changing; not the total radiated power!
That is how I understand it now.
1. The power calculation method in Vituix is correct, indeed, if you have the complete set of measurements of the horizontal and vertical orbits (cfr. Floyd Toole). That has not been done for the Monkey Box. That is my remark to the Vituix power calculation for the Monkey Box, not more than that.
2. Applying a waveguide to a tweeter doesn't change the total radiated power, a waveguide only affects the horizontal and vertical radiation patterns, in a way SPL changes at different positions, more SPL on axis, less SPL off axis (DI is changing).
Briefly: using a waveguide, DI and SPL are changing; not the total radiated power!
That is how I understand it now.
Last edited:
Yes.
However, an accurate estimate of the power response would require SPL curves measured with the microphone positioned at many evenly distributed points on a full sphere around the speaker.
I don't have a fixture that would allow mounting the speaker and rotating it both horizontally and vertically. Even if I had one, positioning the speaker at all possible combinations of, say, 15° intervals both vertically and horizontally and measuring the SPL response curves at these positions would require a humungeous amount of work. Angling the speaker exactly by hand at more than 500 positions and measuring the SPL curves at these positions is a bit more than "just a few more minutes of work". The fixture would need to have (stepper?) motors to move the speaker to the right positions, and they would need to be controlled by the measurement software to do automated measurements. I don't have such equipment available to me, so I simply can't do it.
Given the limitations of realistically achievable data, I do see some value in modelling the power response, even if it's just a model that will never be accurate.
As it is now, we only have horizontal off-axis measurements. Therefore, Vituix can only do a 2 dimensional analysis. I don't know what Vituix does to integrate over the full 3 dimensional sphere -- maybe by analysing the SPL responses projected onto a vertical cylinder around the box? Does Vituix also use some kind of model to extrapolate 2D data to 3D in order to model the vertical dispersion on top of the horizontal measurements?
I also tried to better understand how much "off" the Vituix estimation of the power response is due to currently missing measurement data at 90°...180° (i.e., the SPL behind the speaker). To this end, I estimated the horizontal dispersion at different angles beyond the 90° off axis level using the diffraction calculator. From my eyeballing, I'd estimate the mean SPL at 90°...180° to be about 15 dB lower than from 0°...90° (at 1 kHz). Since the diffraction calculator neglects the beaming of the waveguides, this difference can be expected to be quite a bit more due to the waveguides.
If in reality the mean SPL were 20 dB lower in the 90°...180° half space than in the 0°...90° range, the summed total from 0°...180° would be higher by about 0.8 dB than the sum from 0°...90°. While this is just a rough estimate, it shows that the effect of neglecting the 90°...180° data is not huge. In fact, it might be quite a bit less than the difference of the true, full 3D result one would get from spherically distributed measurements (or an accurate model, although that might not exist).
I might be in the wrong boat with all this, but overall I believe that getting a significantly better power response curve than what we already have would mean a lot more effort, and I am not able to implement this with what I have available.
I thought that a waveguide / horn would act as an impedance transform, allowing a better coupling of the dome (or ring radiator) to the air. This would allow the driver to transfer more power into the air. Am I mixing up something?
Yes correct, if we assume this transform function is improved on axis and worse off axis with a waveguide, it can be understood. I try to find some reference docs on this item, but still haven't found yet. I say now that the waveguide doesn't change the total power, but some confirmation by others would be nice.
In the past, I have compared in Leap its calculated power response with the power calculated out of complete horizontal and vertical orbit measurements in 15 degrees steps, also in Leap. The two results were quite the same. So I have some trust in the Leap power calculation.
Maybe this is useful. The first page (and the beginning of the second page) say something about impedance matching and loading the driver.
This is good to know. Is it possible to somehow trick LEAP to restrict the calculation of the power response using the front half-space only, ignoring the sound emitted to the rear of the speaker? This might help a bit more to understand how much the power responses calculated in full-space vs. front-space only compare to each other (at least for non-waveguide systems).
I also found seemingly contrasting definitions of how to calculate the power response.
Some definitions just say that the power response is the integral of the sound power radiated over a sphere around the speaker (here or here ("the Harman approach)). I guess this is what Paul uses, and what I always implicitly assumed without knowing it might also be something else.
Other definitions show a weighed integral with a weight factor of the form "sin(theta)" (where theta is the horizontal off-axis angle), giving most weight to the SPL curves near theta = 90° (here, and possibly also the Vituix manual on page 26, although it does not say what theta_n is). Interestingly, the vertical off-axis angle does not show up anywhere in these definitions. I don't understand the purpose of the weighting and where it comes from. What's up with this?
It seems the two types of power responses would need to be interpreted in different ways.
That's what I am trying to figure out. To this end I need to understand how much "better" the result will be if I'd add a small number of measurements that I can actually do with what I have available.
Last edited:
Matthias, I used this article of Tylka to calculate the directivity index and also the power, using the horizontal and vertical orbit measurements. I used it to compare the Leap power calculation with the power out of the orbit measurements. There is a good explanation of the weighting you need.
I recalculated the weighting factors to 15 degree steps instead of 5 degree steps like the example in the article to limit the number of simulations.
https://www.princeton.edu/3D3A/Publications/Tylka_3D3A_DICalculation.pdf
I can have a look in Leap of the impact of only using measurements at the front side for the power.
I am afraid there will be an error in the power response vs. frequency, because the speaker is omnidirectional at low frequencies and unidirectional for high frequencies.
I recalculated the weighting factors to 15 degree steps instead of 5 degree steps like the example in the article to limit the number of simulations.
https://www.princeton.edu/3D3A/Publications/Tylka_3D3A_DICalculation.pdf
I can have a look in Leap of the impact of only using measurements at the front side for the power.
I am afraid there will be an error in the power response vs. frequency, because the speaker is omnidirectional at low frequencies and unidirectional for high frequencies.
Last edited:
Yes that is correct. Using a waveguide, the axial SPL increases for the same power, the axial response can be attenuated more in a way the power decreases. Remark that the waveguide doesn't affect the power. Due to a higher DI by the waveguide, the axial SPL becomes higher, it can be attenauated more and so also the power can be decreased.
For example a 8 inch midrange and a tweeter, the midrange power is lower than the tweeter power for the same SPL at the x-over point, you place a waveguide on the tweeter, the power of the tweeter can be decreased and the SPL and power of midrange and tweeter (with waveguide) become equal, perfect.
But, in the Monkey Box the 3 inch midrange and the tweeter (without waveguide) have about the same power for the same SPL around the X-over point. Using a tweeter waveguide, the midrange power becomes higher than the tweeter power for the same SPL on the x-over point. To make the power response transition more smooth, some midrange and global SPL roll off is needed. That is what I tried to make clear for Matthias, that no waveguide would be better maybe for that project.
Power & DI estimation with dual plane measurement sequence is fully accepted method i.e. error is decent - also with typical multi-way speakers. Error is even less when we measure just single drivers for speaker simulation because lobing at XO range does not play any role.
Most of the error occurs while multi-way simulation based on dual plane driver measurements if drivers are not close to vertical or horizontal line relative to wage length at XO frequency. That is possible to verify with some constructions by rotating speaker temporarily in the simulation while power response is monitored. Unfortunately this feature is easy in VCAD only for myself in debug mode.
More important for capturing correct tilt of power and DI responses is that we measure also rear sector and angle step is not too long. That is not a problem with turning table. Rear sector takes few extra minutes which is nothing compared to whole measurement setup arrangements. Simple manual turning table does the job.
Anyway, instructions are written to follow. I'm not so pleased to follow discussion about suspicious or wrong results if something relevant was ignored.
Last edited:
I have implemented power & DI calculation by measurements to VCAD to support all possible speaker constructions; radiator combinations, leaking boxes etc.
Simulation of power & DI spectrum by some parameters such as driver and box dimensions is nice and available as long as calculation formulas are known. But that's not always the case.
Another challenge for simulated power is XO ranges; how to sum calculated power of different drivers/ways in order to get accurate result close to XO frequency. That depends on mechanical construction (distances), directivity of radiators and some incidental features such as phase match and diffraction. In addition, if/when we want to monitor individual off-axis responses while designing crossover network, simulated power turns even more just extra work for the designer due to needed additional parameters.
Sorry if this post looks defensive. Not possible to use several months for research and programming power calculation for different radiator types to save users time few minutes/project. Accurate result is still not guaranteed due to my limited brain
Last edited:
Alright, I am trying to better understand the background of the power response. What does power response really mean in mathematical-physical terms? How do the various measurement methods that can be implemented using relatively simple methods compromise the result as compared to a complete analysis? Once I understand these things I should be in a position to suggest a useful approach to improve the currently available measurement data for the Monkey Coffin, if necessary.
I had another look at the Tylka paper. It defines the "average power spectrum" (or "sound power spectrum") as the average of computed from the measured power over many directions. It does not give a formula, and so the "sin(theta)" weighting term does not show up in the "average power spectrum" or "sound power spectrum". The sin(theta) term only shows up in the definition of the directivity index (DI).
So, if I get you right, the definition of the "power response" does include the sin(theta) term. Hmm, that leaves me wondering what the definition of the "power response" really is. It does seem to be something else than "power spectrum" or "sound power spectrum" as defined in the Tylka paper.
I need to better understand where the sin(theta) term comes from:
I'd like to understand this better. Why is it fully accepted? Is it because doing full spherical measurements is just out of question for many loudspeaker designers, so people just use the dual plane approach in the assumption (or hope?) they are not misguided by the (unknown?) compromises they made with this assumption? Or is it because the dual-plane measurement have been studied and shown to really give the same result, or at lease the deviation between dual plane and full spherical analysis is well understood and quantified?
Or, in other words, what does "the error is decent" really mean? It would be useful to see some data or charts comparing the results obtained from the two methods. kimmosto, since you worked on this a lot, I'd guess you also needed to know and understand these things. Can you provide me with some comparison data for some insight?
Thinking about this a bit more, the sin(theta) term gives the most weight to the sound radiated to the sides and the top/bottom of the speaker, whereas the forward and backwards sound is
Again, I'd like to understand this better. All I have until now is my super-crude estimate of 0.8 dB error if the rear sector is missing in the summation (post 845). kimmosto, can you provide me with some comparison data for some insight?
As you may have guessed by now, I am not the type who follows instructions without understanding what I am doing. To me, instructions often just exist to delegate responsibility for a failure to the people who did not follow the instructions exactly. With only a few exceptions (traffic rules come close), I believe it's vital that people always think about what they are doing, so they understand what's going on.
Sorry for nagging, and for getting a bit off topic, but I am trying to learn and understand how to interpret the power response, and how it is important.
I had another look at the Tylka paper. It defines the "average power spectrum" (or "sound power spectrum") as the average of computed from the measured power over many directions. It does not give a formula, and so the "sin(theta)" weighting term does not show up in the "average power spectrum" or "sound power spectrum". The sin(theta) term only shows up in the definition of the directivity index (DI).
So, if I get you right, the definition of the "power response" does include the sin(theta) term. Hmm, that leaves me wondering what the definition of the "power response" really is. It does seem to be something else than "power spectrum" or "sound power spectrum" as defined in the Tylka paper.
I need to better understand where the sin(theta) term comes from:
- What is the purpose of the weighting by the horizontal angle?
- Why is there a weighting by the horizontal angle, but not by the vertical angle?
I'd like to understand this better. Why is it fully accepted? Is it because doing full spherical measurements is just out of question for many loudspeaker designers, so people just use the dual plane approach in the assumption (or hope?) they are not misguided by the (unknown?) compromises they made with this assumption? Or is it because the dual-plane measurement have been studied and shown to really give the same result, or at lease the deviation between dual plane and full spherical analysis is well understood and quantified?
Or, in other words, what does "the error is decent" really mean? It would be useful to see some data or charts comparing the results obtained from the two methods. kimmosto, since you worked on this a lot, I'd guess you also needed to know and understand these things. Can you provide me with some comparison data for some insight?
Thinking about this a bit more, the sin(theta) term gives the most weight to the sound radiated to the sides and the top/bottom of the speaker, whereas the forward and backwards sound is
Again, I'd like to understand this better. All I have until now is my super-crude estimate of 0.8 dB error if the rear sector is missing in the summation (post 845). kimmosto, can you provide me with some comparison data for some insight?
As you may have guessed by now, I am not the type who follows instructions without understanding what I am doing. To me, instructions often just exist to delegate responsibility for a failure to the people who did not follow the instructions exactly. With only a few exceptions (traffic rules come close), I believe it's vital that people always think about what they are doing, so they understand what's going on.
Sorry for nagging, and for getting a bit off topic, but I am trying to learn and understand how to interpret the power response, and how it is important.
Long story…
Difficult for me to understand at which point there is not a good understanding w.r.t. power response. Some general stuff below, maybe this will help a little.
The acoustic energy emitted by the driver can be calculated as a function of the applied force F, moving mass, acoustic impedance and frequency w (see Small articles). In Leap the power of a driver is calculated like that.
This energy is emitted to the sphere around the speaker with a certain reference efficiency of the driver, efficiency as a function of fs, Vas, sound speed c, Qes, Re (see Small articles)
If you know the radiation pattern of the driver in all directions, the acoustic intensity at each point in the sphere around the speaker can be calculated. Calculating the integral over the sphere gives the emitted power, that has to be the same as the power calculated out of the system parameters like mentioned above. This integral calculation is possible with an omnidirectional sound source, where the radiation pattern is uniform in all directions. For a driver with finite dimensions it is more complicated of course. Therefore we need a measurement like described in the Tylka document.
To understand the weighting, you have to read chapter 4 of Tylka very attentive. There are made steps along the z-axis with a perpendicular plane moving along the z-axis in discrete steps, a plane which cuts the unit sphere around the speaker. The steps are defined as theta angle steps along the horizontal and vertical orbits, each step corresponds with a surface strip of the unit sphere around the speaker. Each surface strip has a different value, the closer theta to 90 degrees (upper sphere side), the larger the surface area of the strip and the more impact on the directivity (power) calculation. I hope this is clear ??
Remark that the Tylka document describes a directivity measurement, but the power can be find in the numerator of the DI definition.
Some remark concerning the power calculated in Leap. The power is expressed as an omnidirectional point source with a SPL value, a source which power is equal to the calculated power of the driver. This means that the power sum of 2 drivers is just the SPL sum of the 2 point sources at the same point. It means that impact of driver positions and radiation angles are not included. IMO it is already good information to evaluate the separate power levels of the filtered drivers to get a rather good idea of the power sum. To calculate the complete behavior, the Tylka method has to be done within Leap. I have tried out once with a 3-way with X-over and it gave good results. You could see the power dips on the x-over points due to driver positions and the x-over impact. For design evaluation the separate filtered driver powers are enough, the x-over behavior and driver position impact is predictable.
Power response vs. frequency is important to evaluate accurately, because it tells you how the speaker will sound like. As an example, with a 8 inch midrange and a 1 inch dome, designed with a flat SPL and a smooth power response, you will see that the high frequency power will become about 8 dB lower at high frequencies. On the contrary, a 3 inch midrange and a 1 inch dome, also a flat SPL and a smooth power, the high frequency power can become about 6 lower. Both speakers will sound completely different. It is a choice that has to be made as a function of the room, the application of the speaker, home, PA,… personal taste of sound,…
PS. I have more detailed theoretical background of all this stuff, but not available in one document (not yet).
Difficult for me to understand at which point there is not a good understanding w.r.t. power response. Some general stuff below, maybe this will help a little.
The acoustic energy emitted by the driver can be calculated as a function of the applied force F, moving mass, acoustic impedance and frequency w (see Small articles). In Leap the power of a driver is calculated like that.
This energy is emitted to the sphere around the speaker with a certain reference efficiency of the driver, efficiency as a function of fs, Vas, sound speed c, Qes, Re (see Small articles)
If you know the radiation pattern of the driver in all directions, the acoustic intensity at each point in the sphere around the speaker can be calculated. Calculating the integral over the sphere gives the emitted power, that has to be the same as the power calculated out of the system parameters like mentioned above. This integral calculation is possible with an omnidirectional sound source, where the radiation pattern is uniform in all directions. For a driver with finite dimensions it is more complicated of course. Therefore we need a measurement like described in the Tylka document.
To understand the weighting, you have to read chapter 4 of Tylka very attentive. There are made steps along the z-axis with a perpendicular plane moving along the z-axis in discrete steps, a plane which cuts the unit sphere around the speaker. The steps are defined as theta angle steps along the horizontal and vertical orbits, each step corresponds with a surface strip of the unit sphere around the speaker. Each surface strip has a different value, the closer theta to 90 degrees (upper sphere side), the larger the surface area of the strip and the more impact on the directivity (power) calculation. I hope this is clear ??
Remark that the Tylka document describes a directivity measurement, but the power can be find in the numerator of the DI definition.
Some remark concerning the power calculated in Leap. The power is expressed as an omnidirectional point source with a SPL value, a source which power is equal to the calculated power of the driver. This means that the power sum of 2 drivers is just the SPL sum of the 2 point sources at the same point. It means that impact of driver positions and radiation angles are not included. IMO it is already good information to evaluate the separate power levels of the filtered drivers to get a rather good idea of the power sum. To calculate the complete behavior, the Tylka method has to be done within Leap. I have tried out once with a 3-way with X-over and it gave good results. You could see the power dips on the x-over points due to driver positions and the x-over impact. For design evaluation the separate filtered driver powers are enough, the x-over behavior and driver position impact is predictable.
Power response vs. frequency is important to evaluate accurately, because it tells you how the speaker will sound like. As an example, with a 8 inch midrange and a 1 inch dome, designed with a flat SPL and a smooth power response, you will see that the high frequency power will become about 8 dB lower at high frequencies. On the contrary, a 3 inch midrange and a 1 inch dome, also a flat SPL and a smooth power, the high frequency power can become about 6 lower. Both speakers will sound completely different. It is a choice that has to be made as a function of the room, the application of the speaker, home, PA,… personal taste of sound,…
PS. I have more detailed theoretical background of all this stuff, but not available in one document (not yet).
Last edited:
sin(theta) weighting converts radial measurements in plane(s) to total radiated power on spherical surface i.e. to compatible with multiple measurements all around the speaker. I guess Tylka paper includes graphical explanation why sin(theta) is needed for that.
This is practical method especially for diy because rotation of speaker or mic around speaker so that each measurement would cover the same area on spherical surface around the speaker is mechanically challenging without multi-axis rotation table or combination of speaker and mic rotation and some math for position setpoint calculation. Measurement in few planes (with constant angle step) is easy, fast and mechanically compact.
sin(theta) weighting is needed for all planes - not just for horizontal.
Each measurement needs also weighting by coverage angle if angle step varies within 0-180 deg sequence. This is included also in VCAD main program (but not in Calculator tool).
Nominator in Tylka (1) represent total radiated power and numerator is axial power. So Q(f) formula for DI include also total radiated power -> measurements all around speaker is not mandatory.
We know that it's not fully accurate especially with asymmetrical complete multi-way constructions where drivers are located diagonally. But it's good enough method for textbooks e.g. Acoustics Sound Fields and Transducers by Beranec & Mellow or Tylka's paper. We have quite good reason to trust that it's also good enough for diy. I would say it's state of the art in diy-scene and better to take advantage of it.
About rear sector. Significance depends on speaker type (mono, dipole, ...) so universal method & instruction is to measure 90-180 deg without questioning. Rear measurement gives also final power tilt to 4pi which enables using empirical data what tilt range is okay for decent acoustics.
I have compared effect of angle step. I don't have access to acoustic lab with anechoic where "all around the speaker" could show some actual advantages over dual plane.
This depends on what point "understanding" is reached. Before that moment it's better to follow than violate if you are forced to do some actions in early phase.
Hi Kimmosto,
Isn't it possible to add the individual radiated power values of the separate drivers in VCAD? A calculation as a function of the system parameters, like described by R.Small.
IMO very useful for concept studies, before having done any measurements.
It is implemented in Leap, but it should be nice to have available in VCAD also.
Just a question/idea...
Isn't it possible to add the individual radiated power values of the separate drivers in VCAD? A calculation as a function of the system parameters, like described by R.Small.
IMO very useful for concept studies, before having done any measurements.
It is implemented in Leap, but it should be nice to have available in VCAD also.
Just a question/idea...