Interview with Allex Mendis
Disclaimer: The opinions published here are not representative of the Sydney Audio Club Inc.
In our March Meeting, we had the wonderful opportunity to listen to one of our members Allex Mendis present the TechTalk segment. In it, he discussed the topic of loudspeaker measurements and how it relates to sound quality. It all started when Allex heard about the launch of Tag McLaren in 1998. He read up on the science behind the design and he was intrigued. The aim of the company was to make the best music reproduction system without compromise. When it was first released, the system was out of reach of most people’s budgets. Several years later, Allex was able to acquire different pieces on the used market, quite an achievement as the brand was relatively unknown in Australia. Around 2001, Allex obtained an SPL meter and this allowed him to make some crude measurements. Today, the equipment has improved and his knowledge on the topic has expanded. Here, we follow up the TechTalk with a few quick questions on the topic.
In our March Meeting, we had the wonderful opportunity to listen to one of our members Allex Mendis present the TechTalk segment. In it, he discussed the topic of loudspeaker measurements and how it relates to sound quality. It all started when Allex heard about the launch of Tag McLaren in 1998. He read up on the science behind the design and he was intrigued. The aim of the company was to make the best music reproduction system without compromise. When it was first released, the system was out of reach of most people’s budgets. Several years later, Allex was able to acquire different pieces on the used market, quite an achievement as the brand was relatively unknown in Australia. Around 2001, Allex obtained an SPL meter and this allowed him to make some crude measurements. Today, the equipment has improved and his knowledge on the topic has expanded. Here, we follow up the TechTalk with a few quick questions on the topic.
JK: The digital equaliser in the Tag McLaren AV Processor sounds very much like the equaliser feature built in to some playback software such as JRiver, MediaMonkey, etc? How do they differ exactly?
AM: The principle is the same. With both, a filter is applied at a particular frequency to reduce the peak amplitude. 20 years ago, Tag McLaren AV processors were the only ones I was aware of to have this feature and it was done in the device itself. JRiver and other playback software are similar in principle, but they apply this is in software. Applying it in the AV processor means any input connected will benefit from the EQ.
JK: For those who don’t know, can you please explain why one would want to achieve a flat frequency response (in the room) from their loudspeakers. What does that achieve sonically?
AM: A flat frequency response provides equal sound pressure levels to the listener. Equal levels at all frequencies means you will hear what is in the music, without your speaker or your room altering the sound. If the response isn’t flat, the sound will be altered and worse, the harmonics of dominant frequencies will overpower the incoming notes, meaning you will hear something different to the original recording.
JK: Suppose you’ve measured the frequency response and you found peaks or dips at certain frequencies. Can you explain what you would then do to try and correct / flatten these frequencies? Would you use room treatments, digital EQ or both?
AM: You can only apply correction (i.e. filters) to frequencies below 200-250 Hz and you can only reduce ‘peaks’. We want to avoid increasing the 'dips' because that can make the sound worse. For example, you may introduce distortions that do not exist in the music. Simply put, you can't hear any dips so it’s best to leave them untouched.
Above the transition frequency, the direct sound of the speakers dominates so you cannot successfully apply EQ most of the time. Any peaks or dips are usually a characteristic of the speaker and it could be due to any of the following: a bad design, a weak chassis, internal reflections, resonances or poor choice of drivers. The only solution then would be to change the speakers. You need to start with good speakers that have a flat frequency response, apply room treatment (acoustic absorbers), place the speakers in the best location of the room and put in multiple subwoofers and then apply EQ to the system.
AM: The principle is the same. With both, a filter is applied at a particular frequency to reduce the peak amplitude. 20 years ago, Tag McLaren AV processors were the only ones I was aware of to have this feature and it was done in the device itself. JRiver and other playback software are similar in principle, but they apply this is in software. Applying it in the AV processor means any input connected will benefit from the EQ.
JK: For those who don’t know, can you please explain why one would want to achieve a flat frequency response (in the room) from their loudspeakers. What does that achieve sonically?
AM: A flat frequency response provides equal sound pressure levels to the listener. Equal levels at all frequencies means you will hear what is in the music, without your speaker or your room altering the sound. If the response isn’t flat, the sound will be altered and worse, the harmonics of dominant frequencies will overpower the incoming notes, meaning you will hear something different to the original recording.
JK: Suppose you’ve measured the frequency response and you found peaks or dips at certain frequencies. Can you explain what you would then do to try and correct / flatten these frequencies? Would you use room treatments, digital EQ or both?
AM: You can only apply correction (i.e. filters) to frequencies below 200-250 Hz and you can only reduce ‘peaks’. We want to avoid increasing the 'dips' because that can make the sound worse. For example, you may introduce distortions that do not exist in the music. Simply put, you can't hear any dips so it’s best to leave them untouched.
Above the transition frequency, the direct sound of the speakers dominates so you cannot successfully apply EQ most of the time. Any peaks or dips are usually a characteristic of the speaker and it could be due to any of the following: a bad design, a weak chassis, internal reflections, resonances or poor choice of drivers. The only solution then would be to change the speakers. You need to start with good speakers that have a flat frequency response, apply room treatment (acoustic absorbers), place the speakers in the best location of the room and put in multiple subwoofers and then apply EQ to the system.
Picture: Frequency response graphs. The top diagram shows the response with the room untreated.
The bottom diagram shows the response after the room has been treated.
JK: Is it possible that a certain loudspeaker measures poorly in one room, but then measures significantly better in another room (with different dimensions / acoustic properties)?
AM: Above the transition frequency, every speaker will behave the same way in any room. A good speaker will be good and a bad speaker will be bad, regardless of the room. Below the transition frequency, it’s mostly the room that dominates. So, you need to 'tune' the system to get the best out of it.
The bottom diagram shows the response after the room has been treated.
JK: Is it possible that a certain loudspeaker measures poorly in one room, but then measures significantly better in another room (with different dimensions / acoustic properties)?
AM: Above the transition frequency, every speaker will behave the same way in any room. A good speaker will be good and a bad speaker will be bad, regardless of the room. Below the transition frequency, it’s mostly the room that dominates. So, you need to 'tune' the system to get the best out of it.
Picture: Frequency vs dB chart showing the transition frequency at 250 Hz
JK: If someone was experiencing issues with bass because their listening room is lacking in size, how can multiple subwoofers help alleviate the issue? How many subwoofers would you recommend and where would you position them?
AM: To achieve the best bass response, research suggests 2- 4 subwoofers ideally placed in all four corners of the room. They can be small subwoofers, but they will 'fill' the 'dips'. Then any 'peaks' can be reduced using EQ to provide a fairly flat response at multiple listening positions near the 'sweet spot'.
JK: Given your experiences so far, what advice would you offer to someone who’s just getting into the hobby and about to purchase their first system?
AM: Go online and search for actual acoustic measurement data for speakers provided by third-party websites. You cannot trust data provided by the manufacturers. This is not always possible as data is difficult to find. Ignore the prices and just look for well designed speakers and how they measure. Then try to listen to them first before purchasing them. Remember, you cannot 're-design' or 're-manufacture' faulty speakers. You can only tune your room below the transition frequency. Don't waste money by constantly upgrading equipment, thinking the next speaker, amplifier, source, cable, etc. will improve the sound. Hearing different sounds does not mean it is better. In the end, it has to work as a 'system' to reproduce the sound in the source as faithfully as possible.
Once you have a properly calibrated, measured and tuned system, you don’t have to worry about changing equipment anymore. Any differences or issues you can hear is almost always in the recording, and not in your system. Then you can just sit back and enjoy the music.
JK: What was the best sound system you have ever heard to date? Which speakers & components did it include, and what did you notice about the room itself?
AM: The best system I have ever heard, was built by the owner himself. He designed and built the speakers using concrete, following scientific principles and treated the room accordingly.
JK: If you were starting today and building your system from scratch, how would you go about it? What would you prioritise?
AM: Start with the best speakers for your listening space (e.g. floorstanding or standmounts) and then add other components to match the requirements of the speakers to create a system.
JK: And for your room specifically, what would they be?
AM: I was considering getting another pair of Tags but they’re very expensive to ship and they’re also hard to find (there’s only 76 of them in existence around the world). If not the Tags, I’d go for the Revel Salon 2 speakers paired with a few monoblock amplifiers.
JK: You mentioned that a good active loudspeaker system will beat any system with passive loudspeakers in it. Can you please elaborate on why you think this is the case?
AM: Passive loudspeakers has an internal crossover network which splits the incoming frequencies to match what each individual driver can handle. Active loudspeakers on the other hand has multiple amplifiers that receive different frequencies (already split by an active crossover). This means each of these amplifiers in an active loudspeaker only has to amplify a smaller frequency range, so it can be either more or less powerful, based on the driver it is driving. Also, there won’t be any crosstalk, more headroom and less distortion. So a well-designed active loudspeaker system will outperform nearly an passive loudspeaker system.
To simplify this, we can think of it this way: In a passive loudspeaker system, you’re attaching resistors, capacitors and inductors to the speaker cable coming from the amplifier to drive your speaker, while sending a full range signal. In an active loudspeaker system, the cable from the amp is connected directly to the speaker driver (without anything else) and only has to deal with a small frequency range. Therefore, active loudspeaker systems are more efficient and accurate in reproducing music.
An active loudspeaker system needs to be tuned properly, but the extra complexity of doing so can discourage people and they miss out from realising how much better active loudspeakers can sound.
Thank you Allex for taking the time to answer these questions. It was certainly very informative and we look forward to another one of your TechTalks in the near future.
JK: If someone was experiencing issues with bass because their listening room is lacking in size, how can multiple subwoofers help alleviate the issue? How many subwoofers would you recommend and where would you position them?
AM: To achieve the best bass response, research suggests 2- 4 subwoofers ideally placed in all four corners of the room. They can be small subwoofers, but they will 'fill' the 'dips'. Then any 'peaks' can be reduced using EQ to provide a fairly flat response at multiple listening positions near the 'sweet spot'.
JK: Given your experiences so far, what advice would you offer to someone who’s just getting into the hobby and about to purchase their first system?
AM: Go online and search for actual acoustic measurement data for speakers provided by third-party websites. You cannot trust data provided by the manufacturers. This is not always possible as data is difficult to find. Ignore the prices and just look for well designed speakers and how they measure. Then try to listen to them first before purchasing them. Remember, you cannot 're-design' or 're-manufacture' faulty speakers. You can only tune your room below the transition frequency. Don't waste money by constantly upgrading equipment, thinking the next speaker, amplifier, source, cable, etc. will improve the sound. Hearing different sounds does not mean it is better. In the end, it has to work as a 'system' to reproduce the sound in the source as faithfully as possible.
Once you have a properly calibrated, measured and tuned system, you don’t have to worry about changing equipment anymore. Any differences or issues you can hear is almost always in the recording, and not in your system. Then you can just sit back and enjoy the music.
JK: What was the best sound system you have ever heard to date? Which speakers & components did it include, and what did you notice about the room itself?
AM: The best system I have ever heard, was built by the owner himself. He designed and built the speakers using concrete, following scientific principles and treated the room accordingly.
JK: If you were starting today and building your system from scratch, how would you go about it? What would you prioritise?
AM: Start with the best speakers for your listening space (e.g. floorstanding or standmounts) and then add other components to match the requirements of the speakers to create a system.
JK: And for your room specifically, what would they be?
AM: I was considering getting another pair of Tags but they’re very expensive to ship and they’re also hard to find (there’s only 76 of them in existence around the world). If not the Tags, I’d go for the Revel Salon 2 speakers paired with a few monoblock amplifiers.
JK: You mentioned that a good active loudspeaker system will beat any system with passive loudspeakers in it. Can you please elaborate on why you think this is the case?
AM: Passive loudspeakers has an internal crossover network which splits the incoming frequencies to match what each individual driver can handle. Active loudspeakers on the other hand has multiple amplifiers that receive different frequencies (already split by an active crossover). This means each of these amplifiers in an active loudspeaker only has to amplify a smaller frequency range, so it can be either more or less powerful, based on the driver it is driving. Also, there won’t be any crosstalk, more headroom and less distortion. So a well-designed active loudspeaker system will outperform nearly an passive loudspeaker system.
To simplify this, we can think of it this way: In a passive loudspeaker system, you’re attaching resistors, capacitors and inductors to the speaker cable coming from the amplifier to drive your speaker, while sending a full range signal. In an active loudspeaker system, the cable from the amp is connected directly to the speaker driver (without anything else) and only has to deal with a small frequency range. Therefore, active loudspeaker systems are more efficient and accurate in reproducing music.
An active loudspeaker system needs to be tuned properly, but the extra complexity of doing so can discourage people and they miss out from realising how much better active loudspeakers can sound.
Thank you Allex for taking the time to answer these questions. It was certainly very informative and we look forward to another one of your TechTalks in the near future.