Announcement

Collapse

HUG - here for all audio enthusiasts

The Harbeth User Group is the primary channel for public communication with Harbeth's HQ. If you have a 'scientific mind' and are curious about how the ear works, how it can lead us to make the right - and wrong - audio equipment decisions, and about the technical ins and outs of audio equipment, how it's designed and what choices the designer makes, then the factual Science of Audio sub-forum area of HUG is your place. The objective methods of comparing audio equipment under controlled conditions has been thoroughly examined here on HUG and elsewhere and should be accessible to non-experts and able to be tried-out at home without deep technical knowledge. From a design perspective, today's award winning Harbeths could not have been designed any other way.

Alternatively, if you just like chatting about audio and subjectivity rules for you, then the Subjective Soundings area is you. If you are quite set in your subjectivity, then HUG is likely to be a bit too fact based for you, as many of the contributors have maximised their pleasure in home music reproduction by allowing their head to rule their heart. If upon examination we think that Posts are better suited to one sub-forum than than the other, they will be redirected during Moderation, which is applied throughout the site.

Questions and Posts about, for example, 'does amplifier A sounds better than amplifier B' or 'which speaker stands or cables are best' are suitable for the Subjective Soundings area only, although HUG is really not the best place to have these sort of purely subjective airings.

The Moderators' decision is final in all matters and Harbeth does not necessarily agree with the contents of any member contributions, especially in the Subjective Soundings area, and has no control over external content.

That's it! Enjoy!

{Updated Oct. 2017}
See more
See less

Historical - DSP pool of information for Harbeth users...

Collapse
X
  • Filter
  • Time
  • Show
Clear All
new posts

  • Historical - DSP pool of information for Harbeth users...

    Hi Alan

    Trust you are well ? I am going to be cheeky now but rest assured if you tell me no then I promise not to be offended. It would seem that you have and still are looking at the possibility of looking in to adding a DSP unit to gently assist your speakers and improve the sound quality. Now I may be wrong but it would seem that the production and sale of such an item is likely to be some way off so what I wanted to know is this .

    There are some quite complex and sophisticated DSP and room correction units still in the market that in some cases have a really powerful ability to edit any Sound Frequency graph that is needed. So have you any information or pictures of the graphs and alterations you have been working with ? If they could be copied even just in a fairly rough way as the start of the basic curve then this could be fine tuned to produce some thing now close to what you are working to. Then those with these sort of units could try out for themselves and comment on how well it does or does not work for them in their systems and in their rooms. If we all agreed to report back on the findings with any details you would consider relevant then this might add an additional pool of information data that would help .

  • #2
    Digitally Assisted Harbeth Listening Experience - DAHLE

    Originally posted by zenith View Post
    Hi Alan

    Trust you are well ? I am going to be cheeky now but rest assured if you tell me no then I promise not to be offended. It would seem that you have and still are looking at the possibility of looking in to adding a DSP unit to gently assist your speakers and improve the sound quality. Now I may be wrong but it would seem that the production and sale of such an item is likely to be some way off so what I wanted to know is this .

    There are some quite complex and sophisticated DSP and room correction units still in the market that in some cases have a really powerful ability to edit any Sound Frequency graph that is needed. So have you any information or pictures of the graphs and alterations you have been working with ? If they could be copied even just in a fairly rough way as the start of the basic curve then this could be fine tuned to produce some thing now close to what you are working to. Then those with these sort of units could try out for themselves and comment on how well it does or does not work for them in their systems and in their rooms. If we all agreed to report back on the findings with any details you would consider relevant then this might add an additional pool of information data that would help .
    Very well thank you for asking. And you? Glad that the sun is finally shining after four months of more or less constant rain, and just in time for the London Olympics 2012.

    That's a really very interesting question you ask. It's surely a great idea that we pool our knowledge and experiences with "room correction" or "DSP".

    What should we call this new electronic technology which could move our listening fidelity up a notch or two. I'm looking for a word that is not too frighteningly 'digital' like DSP, too specific like 'room correction'. Maybe "Digitally Assisted Harbeth Listening Experience" - DAHLE. Could you accept that?

    Then we need to be absolutely clear what the objective is. Frankly, this DAHLE concept is on the edge of my own technical knowledge, and I'm very much aware that I have to understand (to an adequate if not deep extent) what it does and how it works, how it can be used and abused before I can put myself up as an intermediary between the technology and the ordinary user. And I'm not really at that point yet. The DSP people talk a language which may well be appropriate to mathematicians, but means little to me. It's far too theoretical. I'm not much concerned how it works as lines of computer code; I'm really only interested in how we here can jointly apply it to a virtually foolproof system in the ordinary user's home. And we must assume that the ordinary user has zero technical knowledge (why would a music lover need any technical knowledge at all?) and that whatever fancy solution we endorse is going to work almost literally out of the box. That's not a trivial matter.

    You'll have seem my mental work list of the issues to cover in a previous post. Let's just look at one which you touch on: how to train the DAHLE system. There are many variables: which model should we/could we/can we apply DAHLE to? M40.1 or P3ESR? Any model? What about the influence of the room? What about the fact that no two mechanical systems are exactly alike in every degree down to the finest detail? We need to be realistic. One of the really big initial decisions has to be this I think ~~~ Does the user take a standard, idealised lab measurement of a reference speaker and programme that into his DAHLE box, or does he measure how his actual speakers behave in his actual room and use that as the starting point?
    Alan A. Shaw
    Designer, owner
    Harbeth Audio UK

    Comment


    • #3
      To clarify some basic doubts as to the concept:

      Is this seen to be a box of tricks that will sit in the signal chain, between the amp and the speakers? It could physically be a part of the amp or the speakers, I suppose, but it need not?

      Will the presence of this box add some delay to the signal getting to the speaker?

      Comment


      • #4
        Signal delay?

        Originally posted by Kumar Kane View Post
        Will the presence of this box add some delay to the signal getting to the speaker?
        I meant to ask - will it add some delay to hearing the sound from the speakers?

        {Moderator's comment: it must unless we have invented a time machine. The rule of causality is that the output must arrive after the input.}

        Comment


        • #5
          DSP = correct sound, but to whom? Relearning in later life...

          Originally posted by A.S. View Post
          k ~~~ Does the user take a standard, idealised lab measurement of a reference speaker and program that into his DAHLE box, or does he measure how his actual speakers behave in his actual room and use that as the starting point?
          This is a big hurdle, indeed. Does the user know what should be an accurate reproduction? As you have mentioned previously that most users' system sounded bit too bright for your listening preference. There are equally as many users who boost the bass with the aid of subwoofers to their preference and would argue that's the correct sound. Unlearning their preference so late in life is not an easy thing to do.

          Give the users a perfect recording in a perfect room and yet they would still tend to tweak here and there. While we are happy to use whatever DSP available in the car audio, mini HiFi system but the so called audiophiles avoid DSP like a plague.

          IMO, for many a reflective room with overlapping frequencies is good sound to them and a room with enough damping and clear sound is declared to be dull and sterile. Who is Correct here?

          Until we find the actual definition and examples of correct sound, a DSP will be just another variable that may be liked or disliked by them. I have even heard from hardcore audiophiles that full orchestra sound in a international hall to be lacking the so called pinpoint accuracy, lacking separation and imaging that they are so accustomed to with stereophonic listening at home.

          Comment


          • #6
            Master curves for speaker correction?

            Hi Alan , I am also in good spirits but for me it is the prospect of the returning football season not something that will bother too many on this forum. By the way did you enjoy the Cassandra Wilson CD I sent to you ? Often wondered if you thought it was any good.

            With regard to your DAHLE question sadly my technical knowledge is far less than yours and all I can relate is what I think makes sense . What I think is needed is the data of what you design in to your speakers and how you feel they should sound in a perfect room . After all we must like the Harbeth sound if there is such a thing because all of us paid out our own money to buy them and we still have and use them in the system .

            Once that is established as a base line then what I feel is needed is something that would measure each and every system in its users home through his equipment and from his (and as an option ) others listening position. Once done the object then would be to try and match the readings from the user room to the baseline ideal room . I would then think that a sensible option would be for the user to have easy control to tweak and change that curve / shape whatever it is called and possibly add a little extra bass or reduce the bass to his or her taste.

            I am sure that for the majority they would set and match the baseline curve then forget about it and get on with listening but there should always be scope for those people who want to change things a little . If the readings and curves are all that is shared then the details can be added or used with many of the current DSP units that already exist and have the ability to put in and use pre created curves and adjustments .

            I would like to hear a system adjusted for as near as possible flat response across the full audio spectrum that covers human hearing if nothing else but to see what it sounds like . I may not actually like it even though I will be able know that is a flat response. But until you experience this and then perhaps what is judged to be the best tailored response you will never know what was originally intended.

            Having read some of the current theory that is being accepted about our hearing it never ceases to amaze me that we can never agree . all of not only have ears and the mechanical / organic devices to hear that are all different but also the pathways and networks we use are all different so it does become impossible for anyone to really know what anyone else really hears and this means we all must become more assertive and better at saying what we do and do not like and going with our own judgement .

            Comment


            • #7
              Random thoughts

              Had a bit of a think about this last night and this morning and in no paticular order are some of the things that came to me and might be ideas/views that those with greater technical knowledge can comment and help with .

              Why DSP well for me the main attraction is this is an additive fix. In most cases it makes a series of measurements and then devises a solution to take it to a known base setting . But it only adds in the chain it does not change anything that is not easily reversable and in most cases you even have a by pass to let you hear what it adds.

              What are we trying to achieve ? Well for me it is to let the audio chain that I have out together do its job as easily as possible . This for me at least does not mean I want change and improve my Harbeth speakers with my own design but I want to let the Harbeths (whose general sound I must like or why did I buy them ?).

              What is DSP good at , well it seems that some view it a whole Audio spectrum magic bullet but it does not appear that it is anything of the sort. It can I think improve bass clarity and stop it booming it can make it sound clearer and more atriculate but unless abused it will not accurately amke your system deliver lower bass this is most likely already defined by the working parameters of your speakers and these need to be eased and assisted to work not pushed out of there designed sound coverage.

              I would not and do not consider changing components within my speakers cross over they were designed and tested to work to produce the sound quality I liked . I do not have the test equipment the knowledge or the skill to think I can do a better job if I did then I would start my own company and try ane compete I am never going to do this so I will be content to leave things alone. If Harbeth ever design and come out with a repalcement that they say is an improvement then I will look at that otherwise leave well alone.

              What am I really asking about here , well if I do finally go the DSP root I would look to offer a full picture of the before and after reading that I used . If everyone else did the same then we could get a pool of results and look for the most frequently used changes . Individual rooms and taste would no doubt have different things that would show up but I do feel there would some changes that would be common to all. These would then help for someone (and I know this is not me) with the real Acoustic Knowledge to put together a base or first point adjustment that anyone just starting out could program in to whichever unit they DSP they had and this would give them a place to start with the most likely changes that would assist the system. From there it is then up to them they could need additional changes for their rooms or their tastes whatever but if we all have a base start then we have a useful fist postion.

              One thing that is vital is this is just information to share and to help not an excuse to look at someone elses set up and tell them it is wrong or right just raw data to analyse and find the common points. Perhaps then it would be better if those submitted were just numbered and no other details given.

              Ah well sorry for rambling on I am having great problems getting my DSP unit to work at all and if I fail this weekend then it goes back. But I will look at another alternative and try that so I am not giving up on DSP but one the model I currently have .

              Comment


              • #8
                DSP, carefully applied can increase realism

                Have now had some time to spend with a DSP unit that I was allowed to purchase on a 30 day Sale or Return basis. This was the Anti - Mode 2.0 Digital Processor . Things that I liked about this unit were that it came with a supplied microphone and that all processing was done in the unit no need for connection to a PC .

                First thing I learned after doing a room calibration was just how badly my room measured. What should have been as near as possible a flat line in many ways more resembled a cross section of the Alps with a very large number if very large peaks at various points. I make no claims to have the full knowledge of what this means but I do know it not good.

                Next came the ability on the unit at the push of a button to have a corrected digital out put and a bypassed digital output. So you could hear the difference. Some comments I am sure it could be used this way but I would not recommend one of these units as lazy mans correct speaker placement. I am convinced that you should take all the time and care that you would expect to do now to get your system working and sounding as good as it possibly can without any DSP. Then and only then when you are happy should you apply and insert the DSP.

                What I noticed it did which was very satisfying is that it did not change the balance or overall sound of my system . When Harbeth speakers came in to my system it was the natural unforced nature of the final sound that captured me. It was not impressive in the fireworks or in your face way some sounds are rather it allowed you to relax and listen more in to the music and hear what was being played and how it was being played.

                The AM unit does this but just a little bit more it makes the bass seem a little less but then when you listen you realise that you can hear every note being played but it has placed it within the group and it has its own space and place in the music not emphasised but integral . Vocals also sound more like a real person singing with a mouth , chest and diaphragm . No it will never rival live sound because I do not think any amplified system ever could but it does get you a little bit closer to the real thing and makes the performance as a stand alone performance more enjoyable.

                I do not claim or suggest that DSP is the cure for all problems or even if it is necessary for everyone's room or system. It works for me and it is the added quality that makes things just a little bit better more natural and more believable. I have some pictures of the graphs generated by the AM unit with before and after readings if anyone is interested I could post them here for comments from those with considerably more knowledge of this area than I have.

                Comment


                • #9
                  Anti-mode DSP system

                  zenith,

                  I have learned that you've got both Tact and Anit-mode. so what are your comments on these two units?

                  Since Anti-mode only had optical input, did you need to convert the digital pathway from coaxial or AES to optical? If yes, what device you are using?

                  I am very glad to read your comment as I am interested in Anti-mode matching my Harbeth M40. it seems that you might be the first owner of Harbeth speakers using Anti-mode.

                  Ken Wong

                  Comment


                  • #10
                    Day 2 with DSP

                    Well, some more feedback for any of you interested in digital (room) correction systems.

                    I now have two days of experience under my belt. I'm not sure if what I have learned working with one (nameless) system is typical of all RC systems or not, but my guess is that it will be fairly typical. The problem is this: the maths and deep internal workings are horrendously complex and the need to market the system paramount. So understandably, considering that the typical user has negligible technical knowledge (if he was more grounded in science, most audiophile gadgets would disappear through lack of sales), the DSP designers have to automate many of the functions. That's a fine idea as it goes, because it does open the market to the average Joe.

                    Unfortunately, DSP designers are not speaker people. As we don't have skills in their high tech world, they can't be expected to have skills in our mundane world of critically evaluating lousdpeakers in ordinary rooms with music. The consequence is, based on what I've experienced, that the system has to make many assumptions. And the most critical is the overall energy versus frequency in a normal room across the audio band, in other words, what the ideal frequency response shape should be, as plotted on a graph. Not in an anechoic, reflection free test environment, but in a real, ordinary room. And it seems that the DSP gurus assume that the frequency response should be close to that of a flat line. That's a great starting point - put the DSP box into auto mode, let it make that assumption, and calculate the correction factors to flatten out the horrible and inevitable peaks and troughs in the room. The result: peaks flattened, troughs (somewhat) filled-in and high frequencies far too bright - the consequence of assuming that everything should be a flat line.

                    The art (if you can call it that, I hesitate to call it science) of speaker design is to know when you have to back-off from pure theory and bend the rules a little to give the user a satisfying, fatigue-free, life-like listening experience in a normal, real world reflective room on music, not test signals.

                    I wonder if anyone has had success with a DSP system and if the result (at least on auto) was a subjectively rather lean and irritatingly bright tilt upwards. Any better on manual mode where the (smart) user can make some skilful choices?
                    Alan A. Shaw
                    Designer, owner
                    Harbeth Audio UK

                    Comment


                    • #11
                      Just a glorified graphic EQ?

                      Is a DSP system in 'manual mode', just a glorified graphic equaliser ?

                      I have experienced two DSP systems in the past, both trial units from two highly regarded Japanese manufacturers. On both occasions where these were used, the results were obviously worse than with them out of the system and both of the design teams agreed. One was on trial by a noted UK speaker company. No names, no pack drill.

                      Of course, this was just the results in my demo room, with my equipment at that time and might not be representative.

                      Comment


                      • #12
                        Careful stepping through the process

                        Originally posted by A.S. View Post
                        Well, some more feedback for any of you interested in digital (room) correction systems. I wonder if anyone has had success with a DSP system and if the result (at least on auto) was a subjectively rather lean and irritatingly bright tilt upwards. Any better on manual mode where the (smart) user can make some skilful choices?
                        I've had great success with DSP/EQ systems. It's a very complex topic, so I'm not sure where to start. I'll try to explain my process of thought briefly and I can go into detail if you or someone is interested.

                        I will stick to a situating with two speakers for now,and one listening position, as that is the situation that will apply to most here. I must also say I've got some experience, so not everything will be clear for everyone, feel free to ask questions.

                        First of all, DSP/EQ is nothing magical and it will not turn an acoustically bad (lots of hard, reflective surfaces, a lot of reverb) room into a good one. At all times, keep this in mind.

                        There are many variations of EQ/DSP systems, so not every system will have the same options. Some are very basic (e.g. a simple 6 band graphic eq) and some are very advanced many options (time alignment, high pass and low pass filters, parametric equalizers, limiters etc.)

                        What I need in the first place is a calibrated microphone and software to make the necessary measurements. Once I've got that sorted on I can proceed, microphone in listening position, measurement software ready.

                        I first take a measurement of the separate speakers, first in the close range, then at the listening position. Then I do a combined measurement (both speakers) at the listening position. Normally all is normal, but sometimes due to the interaction of both speakers with the room and themselves (walls might not be parallel, or the speakers are not symmetrical in relation to the room), you can see a boost/cancellation in certain frequency regions. I keep this in mind for the adjustments to the separate speakers.

                        I can now begin adjusting frequency regions (according to whatever options the EQ system gives). I usually do not boost regions, this can be dangerous to the speakers in many cases and rarely works out in general. I aim for a line that is as flat as possible, with respect to the close range measurement and without using too many EQ bands. I rarely use more than 4.

                        I will post a picture in a minute from a different computer, of a measurement I made at a friends' house which uses 3 EQ bands (from a simple 6 band graphic EQ, with a more complex system better results could have been achieved) for the main speakers and 4 bands for the subwoofer. This brought the response from 10hz to 20khz down to +/- 6db over the entire bandwith instead of the original +/- 15db (of which I sadly do not have a picture I can find).

                        The result was a much improved listening experience, everything we played sounded better, we AB tested it because the remote of the receiver had an "EQ off/on" switch.
                        Last edited by TimVG; 13-09-2012, 08:50 AM. Reason: clearing up something that was a bit vague

                        Comment


                        • #13
                          Measurement

                          Click image for larger version

Name:	full range.jpg
Views:	1
Size:	61.8 KB
ID:	62341

                          Picture of the measurement

                          Comment


                          • #14
                            A sideline?

                            Originally posted by A.S. View Post
                            Unfortunately, DSP designers are not speaker people. ... The art (if you can call it that, I hesitate to call it science) of speaker design is to know when you have to back-off from pure theory and bend the rules a little to give the user a satisfying, fatigue-free, life-like listening experience in a normal, real world reflective room on music, not test signals.
                            Sounds like a business opportunity to me.

                            Comment


                            • #15
                              Split LF and HF adjustment?

                              Fully appreciate the need for room correction DSP to be applied easily by the uninitiated if it is to sell. I would have thought that to avoid the high-frequency "thinness" resulting from applying a target of a flat frequency response would be to address the lower frequencies first by having a "standard" low pass filter (say 400Hz??) on the treatment.

                              Any corrections for the room at higher frequencies could be addressed separately. I believe there are DSP room correction boxes on the market that allow this approach: the added complexity of the system ought to be manageable by most potential users.

                              Comment

                              Working...
                              X