Off-topic.
As I understand it, there are no threads on this forum regarding multi-channel Class D amplification with a fully digital input for building active speakers, and discussing certain issues, for example, related to the absence of analog input, has no designated place. I believe I need to create a new thread and try to discuss the questions that interest me.
Could you please advise on the appropriate section of the forum to do this?
As I understand it, there are no threads on this forum regarding multi-channel Class D amplification with a fully digital input for building active speakers, and discussing certain issues, for example, related to the absence of analog input, has no designated place. I believe I need to create a new thread and try to discuss the questions that interest me.
Could you please advise on the appropriate section of the forum to do this?
By the way, if you need an affordable Class D amplifier with a power of 100W, featuring excellent output characteristics, I know someone in Ukraine who recently designed and assembled boards for such a Class D amplifier. He has created a very good amplifier with feedback after the choke. The amplifier itself has quite high characteristics at its output and is implemented on a power chip with a bridge output, significantly reducing the requirements for the quality of the single-polarity power supply. In other words, the amplifier is not demanding on the power supply and shows high measured performance levels when using the most common power supplies. Additionally, this amplifier has already undergone listening tests among audio enthusiasts in Ukraine, and the subjective impression of the sound is at a very high level.
If anyone is interested in a more detailed overview of this amplifier, I can ask the developer to create a thread on this forum where he will describe all its parameters and provide measurement results.
Off-topic.
Created a thread about modules for active three-way speakers without analog input. Anyone interested in discussing the need for such a solution, please join the discussion.
https://www.diyaudio.com/community/...52-for-implementation-active-speakers.408225/
Created a thread about modules for active three-way speakers without analog input. Anyone interested in discussing the need for such a solution, please join the discussion.
https://www.diyaudio.com/community/...52-for-implementation-active-speakers.408225/
I must agree with @Azrael and @drewmc. l believe that a loudspeaker’s EQ should not be based on individual hearing capability.
If you assemble a group of guitarists, and let them listen blind to several different acoustic guitars played by the same player, they will have no difficulty identifying each specific guitar. Each of the listeners has a different head shape, different ear shape, and different level of hearing loss, but each brain sorts out the peculiarities, and identifies each guitar.
Now if we played a recording of those same guitars to the listeners, using a high quality playback system, the result will be the same. The playback system has a “flat response”, although this simple term usually encompasses a great deal of complexity because a “flat response” means more than a simple flat on-axis response… it normally takes into account the sound power response and the early reflection response, among other things.
However, if the playback system has an altered from flat, I think the listeners will have some trouble identifying the guitars, because the playback will have a different spectral balance than the live instruments did.
If you assemble a group of guitarists, and let them listen blind to several different acoustic guitars played by the same player, they will have no difficulty identifying each specific guitar. Each of the listeners has a different head shape, different ear shape, and different level of hearing loss, but each brain sorts out the peculiarities, and identifies each guitar.
Now if we played a recording of those same guitars to the listeners, using a high quality playback system, the result will be the same. The playback system has a “flat response”, although this simple term usually encompasses a great deal of complexity because a “flat response” means more than a simple flat on-axis response… it normally takes into account the sound power response and the early reflection response, among other things.
However, if the playback system has an altered from flat, I think the listeners will have some trouble identifying the guitars, because the playback will have a different spectral balance than the live instruments did.
I'm curious. How many people acknowledge that individual hearing characteristics, for example, influence the perception of the tonal balance in a composition? If a composition is crafted with consideration for tonal balance, a person whose frequency response differs from that of the performer will experience a different structure of the tonal balance, meaning they will receive altered information. In other words, if two people have different frequency responses in their hearing, they will have different opinions about the sufficiency or insufficiency of the tonal balance in the same composition. This phenomenon is widespread when several individuals listen to the same speaker system and express their opinions about what they feel is lacking or excessive in the speakers.
2-cents: Different seats at a live performance hear very different tonalities, and none are the same as any of the performers' perception. Yet, our brain will auto-EQ and they all sound "real enough" to us, because this type of "harmonic distortion" is a NATURAL PHENOMENON we've biologically evolved and culturally adapted to compensate for, subconsciously. However, the ARTIFICIAL distortions of our audio equipment still shout "I'm a recording, I'm not real".
I'm not saying individualized flat response is right. But typical non-flat FR certainly doesn't sound right. And I think the brain will quickly readjust to a flat-ish response and appreciate it, given the opportunity -- like sitting front-and-center at the live concert.
I'm not saying individualized flat response is right. But typical non-flat FR certainly doesn't sound right. And I think the brain will quickly readjust to a flat-ish response and appreciate it, given the opportunity -- like sitting front-and-center at the live concert.
I generally agree. This phenomenon shouldn't be exaggerated, but excluding it entirely is not advisable either. Personally, I don't mind what frequency response I get from my speakers on their axis; what matters to me is that the end result is satisfying. I also recommend this approach to others— a flat frequency response on your speakers is not the ultimate truth
But the whole point is that the general preference is remarkably similar among the audience. So, yes you do, although you may think otherwise.
The metric for good speakers is well established by Toole and Olive. It has nothing to do with art or voodoo.
And, related: Hifijims creations here on diya meet the target criteria admirably.
Individual hearing characteristics are a fact of life. And also apply to every live sound we hear. So using some correction in the reproduction chain seems pointless to me, apart from taste. Nobody changes the sound of a live instrument or voice because of individual hearing characteristics. So even from that perspective it seems moot to me.
I've been away from my PC and only watched this thread from my phone with out being able to log in and respond. I suspected the low cost class D amplifiers causing the problem. I believe someone already mentioned that the ICE power amps have the same problem, maybe not as severe, not sure if they have some newer models that are more linear or if Parts Express sells them. On the other hand I would be OK if they caused a roll off on the the high end, I would voice my speakers that way anyhow.
Edit
I just checked Parts Express, they do have the new model Ice Power amplifiers and the specs of the ICEpower 250ASX2 looks like it has solved the problem, but it isn't cheap. Distortion looks good if you stay under 100W.
Last edited:
I mentioned the class D amps offered by @Erica.C to @hifijim before. Eric just confirmed:
Eric's test data shows about a 0.7 dB rise at 20 kHz (~0.25+ dB @10kHz) at both 4 & 8 ohms. Distortion before clipping is typically well <0.1% -- probably an order of magnitude higher than the Hypex, but I don't perceive the difference. Noise is very low, low enough not to be any issue with either Hypex or Eric's amps. For the $, they are amazingly good. The only feature I'd love to see is an input-sensing power on/off, which is so useful for multi-ch active systems.
I've been using a couple of the FFA001V3 (150w/ch stereo amp module) in my audio lab for a few months wherever needed in various active systems I have going. Not sure that I'd call it "critical listening" but in general I have heard no appreciable difference between these and Hypex NC250MP for bass (up to ~120Hz) and mid (100~1000Hz or 600~6000Hz) amps. Also used as a stereo amp for the passive loudspeakers that come through. Haven't tried it as a treble amp yet.
Eric's test data shows about a 0.7 dB rise at 20 kHz (~0.25+ dB @10kHz) at both 4 & 8 ohms. Distortion before clipping is typically well <0.1% -- probably an order of magnitude higher than the Hypex, but I don't perceive the difference. Noise is very low, low enough not to be any issue with either Hypex or Eric's amps. For the $, they are amazingly good. The only feature I'd love to see is an input-sensing power on/off, which is so useful for multi-ch active systems.
Thanks to everyone for continuing the discussion about class D amps, and the various options. My current amps are discussed in post 13, although at the time I did not know that these would be the amps I would choose.
https://www.diyaudio.com/community/threads/compact-low-cost-active-3-way-speaker.402812/post-7441164
Given that I have already characterized the frequency response of the amps when connected to my speakers, there is really no reason for me to change amps at this point.
For a person looking to replicate this project, I might recommend a different amp. The @Erica.C amps look like an excellent value. Many people have a collection of older class AB amps laying about, and these are great candidates.
If someone wants to use one of the self-contained class D amps sold on Parts Express (like I did), I would recommend that they confirm they have an acceptable frequency response above 10k. This would be an easy measurement of the final speaker system. They would need a single FR scan on-axis with the tweeter. A simple USB mic would be fine, and the gate window could be as short as 2 ms, which means it could be done in almost any environment. This scan would confirm if any high freqency adjustment is necessary in the DSP filter.
j.
https://www.diyaudio.com/community/threads/compact-low-cost-active-3-way-speaker.402812/post-7441164
Given that I have already characterized the frequency response of the amps when connected to my speakers, there is really no reason for me to change amps at this point.
For a person looking to replicate this project, I might recommend a different amp. The @Erica.C amps look like an excellent value. Many people have a collection of older class AB amps laying about, and these are great candidates.
If someone wants to use one of the self-contained class D amps sold on Parts Express (like I did), I would recommend that they confirm they have an acceptable frequency response above 10k. This would be an easy measurement of the final speaker system. They would need a single FR scan on-axis with the tweeter. A simple USB mic would be fine, and the gate window could be as short as 2 ms, which means it could be done in almost any environment. This scan would confirm if any high freqency adjustment is necessary in the DSP filter.
j.
Search the forum. Individual hearing preference and ability is acknowledged in tens of thousands of threads on diyaudio.
You are misguided. Flat frequency response is not just a reference. Its a very GOOD reference. Also it is one of the most important indications of quality.
When its degree is established, you can work with what you've got towards your own personal preference, consumerpanel derived preferences as B&O did extensively or you can choose one of the Harman curves.
Cheers!