Jump to content

charbuggie

Members
  • Posts

    49
  • Joined

  • Last visited

Recent Profile Visitors

2135 profile views

charbuggie's Achievements

Advanced Member

Advanced Member (3/9)

25

Reputation

  1. If anyone is curious I do have REW files I don't mind sharing.
  2. I have had some time to spend listening and I am very happy. My room made life difficult and as you know I went the DSP route. This is one of many ways to skin this cat and honestly it was easier for me because I really did not know what I was doing. Mitch has been great to work with along with some others that have helped me. Mitch was very patient as well and the service he offered fit me perfectly. Bottom line now things are much clearer and it does not sound like a blob of sounds. Crisp and accurate is how I would describe it. In the bottom end it is unreal what clean bass sounds like. I asked Mitch to give me a synopsis of what we did and here it is. Like I said there are other ways to go about this but this route was the easiest for me and I am done. Time to enjoys it. Rob asked me to write up a bit about the before and after DSP measurements of his Klipsch Jubilee with KPT-1802-HLS sub. A little bit of background before we get into the measurements to assist in understanding what the DSP is doing and why. Software DSP running on a computer is incredibly powerful and transparent. In the pro audio industry, virtually every piece of analog hardware has been modelled in DSP to precisely and accurately replicate the sound for any particular h/w device. These are routinely purchased as a “plugins” and used in Digital Audio Workstations (DAW’s) in the creation and production of music. There are literally thousands of plugins today for music creators to choose from. Some examples here: https://www.uaudio.com/uad-plugins.html We can also use software DSP to model the “ideal” loudspeaker with perfect frequency and timing response. Loudspeakers are typically minimum phase devices, but adding crossovers and placing them in rooms alter both the frequency and timing response of the sound arriving at our ears. Linear phase digital crossovers are one of the few types of crossovers that sum properly in the frequency and time domain. Some good reference material here and here. We used linear phase digital XO in Rob’s system with XO points at 80 Hz and 450 Hz. Placing speakers in rooms activate room modes and can deviate from minimum phase response. John Mulcahy, author of REW, has a technical paper on minimum phase and explanation on a common cause of non-minimum phase behavior in rooms, even at low frequencies. The DSP we use can correct for non-minimum phase behavior in Rob’s room. We want the sound arriving at our ears to be the ideal minimum phase response removing the effects of non-aligned drivers, loudspeaker crossovers and non-minimum phase room responses. We want to hear the sound that is on the recording to arrive at our ears with no frequency or timing distortions so we hear the music as the artist, producers, and engineers intended. Here is the frequency response of Rob’s system before DSP: We have a smooth response from 15 Hz to 20 kHz to well within the studio spec of +- 3dB tolerance. Let’s look at the step or timing response: The after DSP shows that the direct sound from all speakers are arriving at the ears at the same time. In addition to the precision driver time alignment, the rooms non-minimum phase behavior is gone. This is the “ideal” minimum phase response. One should note that both the left and right channels are virtually identical both in the frequency and time domain. This is what accounts for a perfect stereo image with a rock solid phantom center image and increased depth of field. Another common phrase to describe this is that the speakers have disappeared. It has been my experience that these two measurements, frequency response and timing (step) response cover the main aspects of showing a system that has been restored to its ideal minimum phase response. Of course in REW, we can look at numerous other charts that have different view, but these two are the main ones. My book on, “Accurate Sound Reproduction using DSP” goes into great detail with all of the different chart views. I also show that the corrected frequency and timing response is just not for one location, but across a wide sweet spot covering a 6’ x 2’ grid with little deviation in frequency and timing response. A note on DSP/DRC software programs. There are quite a few to choose from and I have tested most of them. Audiolense, Acourate, and Focus Fidelity Designer are the top 3 commercial DSP programs available today. With the right know how, they are able to accurately and precisely restore the ideal minimum phase response arriving at one’s ears. Almost all other DSP/DRC software fail in one way or another to reach that goal.
  3. Jriver is running on a windows computer. I was wondering could I use an IPAD and get the Jriver remote for it and would it interface with the windows computer. Right now the 2020 Apple iPad (10.2-inch, Wi-Fi, 32GB) is a great deal on Amazon and everyone says the IPAD is the way to go but I dont need compatibility issues with the windows 10 computer running Jriver and controlling my system. Or I can get a cheap Samsung A7?
  4. I have finally figured out how to loopback in the Lynx Hilo and will get some REW data up here soon. I was looking for some input on how to handle the CD from 10K to 20K. Right now it drops off about 12DB in that range with a constant slope. I do have the ability to get Mitch to implement a High Shelf and help it some. It is the nature of these drivers and curious what others have done?? Maybe just leave it be or boost it some.
  5. I was wondering if Running an application like Team viewer would work?
  6. I have read Jremote2 and Gizmo are the two options. Has anyone tried Gizmo?
  7. Mike, You are so correct, small movements of speakers or listening position make a huge difference. I wish I had endless funds and could build the ideal room for the jubes and the 1802 sub but I am a working stiff and I have to make due with what I have, I will be honest There is something about digital I do not like but the world of digital allows for corrections I would never be able to do in the analogue world. Since you were up here Mike, I got rid of the RME interface and went with the Lynx Hilo. The Hilo is a much better product and so much easier to implement. Also you hit the nail on the head, Once I got the Hilo working I managed to get the CD Player to work using the Jriver Live in and the quality of a CD sounds so much better. It is amazing now, as soon as you play something off my hard drive that has about 2 TB of music, you know real quick if it is descent quality or not. Also when you were here Mike I had not received anything from Mitch and the Configuration we were listening to was something a great guy, Brad, off the Avnirvana forum had helped me put together. Mitch Emailed me the other night and said Klipsch Engineered the Jubes and the 1802 sub to work together perfectly. He said my setup can follow the MIN Phase textbook target response just about perfectly. He also said out of all the systems he has dialed in my setup is king top to bottom, even the $65K to $100K Wilson speakers can not reproduce as well. So that Speaks Highly of the Chief Engineer that designed and built these speakers. So I owe MikeTN a big one because he is the one that got me moving to Buy the Jubes and go the route I did. I do plan to get some REW info on this thread when I get time and work with Mitch on some final tweaks. I am a very happy camper for the most part. Only thing I want to do is figure out a way to run Jriver from a tablet at my listening position.
  8. The room is what it is and from playing around over the last couple years moving things just inches makes a huge difference. I wish I had a better room but it is what I have. To answer some of the questions I had to ask Mitch. I am not an expert here and all I am trying to do here is share my experience and provide people good information so they can decide for themselves on what is best for them. In the past Chris has been very helpful and no one knows these speakers better them him and Mike B. I wish I had time to learn more and get into the bug guts of the physics but I dont. I know enough to be dangerous is all I can say. The effect of flattening the phase of the Jubilees as much as possible (in your case, using pretty significant FIR filters with a lot of delay) will not only make the bass so much better, it will also change the "listenability" of the loudspeaker quite significantly. See the following thread for a discussion of that phenomenon: Yes, we are indeed flattening the phase at low frequencies. This is what the excess phase correction does. One of the issues that I've got with taking measurements at the listening position is that you're baking in those nearfield reflections in the FIR filters, which is why I asked about microphone location. I've found that the effects of using Dirac Live (Full) to be problematic in that it doesn't allow the user to take "minimum phase measurements" in-room. Instead, it forces the user to move the microphone back too far (at the LPs), then doesn't do a very good job in excluding the nearfield reflections, particularly in the 100-300 Hz band with the Jubs (which don't have directivity issues at this band--like almost all other loudspeaker do) Nope this is the difference between Dirac and Audiolense. Audiolense uses frequency dependent windowing so that we take care of the room modes and room reflections up to 300 Hz or more, and then we are only adjusting for the direct sound after that. So we are not baking in any reflections after that. That’s the key. We are indeed restoring full minimum phase response which is what the article on ‘What is Accurate Sound” all about. If using TAD TD-4002s, that "big boost" above 10 kHz isn't really an issue. It isn't really a "big boost" since the raw response of the TADs on K-402s actually rises above 13 kHz, without chattering, so that roll off shown is actually "baked in" via EQ (assuming you're using TADs). The TADs don't chatter. I would bet that Mitch probably hasn't heard them--since I believe he's been using JBL 2" compression drivers, and all of them do chatter above 10-13 kHz, so perhaps he's designed in something that actually isn't desirable. I'd recommend talking to him about putting that response back (unless you're no longer using TADs). If those compression drivers are Celestion Axi2050s (as the throat of the K-402s appear to be from your small picture you posted above), then I'd say that allowing the axial response above 10 kHz of them to roll off takes away significant amounts of "sparkle". At least, that was my experience with the Axi2050--which literally comes alive when that top octave is EQed flat to about 18-19 kHz on-axis. I recommend revisiting that. I am using Radian PB 950 Beryllium drivers on the K402's and I have attached the data sheet that came with one of the drivers. I am open to any suggestions on these drivers as I finalize the corrections.
  9. 1. Can you tell me where the microphone is for these measurements, i.e., how far from the front baffle? The mic is at the listening position as we are applying “room correction”. 15’ 6” away from the baffle on each speaker If you're taking measurements at the listening position(s)--more than a metre away--can you post the phase response plot and a spectrogram plot (i.e., before FIR filtering)? I am not sure what value these will be as at 15.5ft there will be so many room reflections, so the phase and spectrogram will be a mess. Maybe I am missing something? What order of crossover filters are you using? Second order, fourth order, etc. Linear phase FIR filters, 2nd order. They are steep slope. Based on my limited listening they really blend in great. Once I figure out how to loop back to Jriver through the Hilo, I will run some REW sweeps. I have attached a PIC to give you an idea how my crossovers are setup. Are you still using REW to take measurements, or something else? Audiolense that sweeps each individual driver and then REW can be used to get the corrected 2 channel response. From reading the forums many people have validated that Audiolense simulations are within a .25 dB of the real measurements. I am looking forward to validating that with REW JRIver (running on a PC) typically requires a pretty hefty PC (in terms of its computing horsepower) to apply the FIR filters--can you identify what PC you're using? JRiver also requires that your PC is running all the time and is your digital preamplifier. How do you handle analog sources, such as a turntable? Do you ever plan to run more than just stereo with a subwoofer, i.e., how would you convert to a 5.1? How would you expand to 6 channels using JRiver? My computer is a newer I-5 with 16MB of RAM and has no issues. I spent about $600 on hardware building it a while back. One of those budget builds but I had case and power supply already. I know a guy from a forum that is running a stereo triamped systems using JRiver on an i3 2 GHz with 8 gig RAM using only 10 to 30% of the CPU. Can you tell me how many FIR taps are being used in each channel (bass bin, K-402/TAD)? I had to ask Mitch this one and he said 65,536 FIR taps per channel. The Lynx Hilo apparently goes for $2300 (USD), the Earthworks microphone goes for $600, and the microphone preamp goes for ~$800 (i.e., the total microphone cost is $1400). Audiolense XO costs ~$460. Is that the product you're using from Audiolense? Any other add-on software that's required? I bought some nice hardware to do this and I can assure you that you can do this at about 40% of the cost I spent. Motu and RME makes some great products that are much cheaper and I sold some stuff laying around and had the funds. Heck you can use a UMIK-1 MIC and save a ton. Can you give us a ballpark on the service costs using Mitch's service? $500 for up to 6 DSP FIR filters, but it is a process where the art is how to use the science to achieve the result. I am very inexperienced and time is precious for me. IMO Mitch offers a great service and so far I am very happy. I do have to dial somethings in because I just received my first filter and I need to listen and figure out what I want. But let me say this the BASS will hit you in the chest like a train and the low end is so clear now. I was listening to Phil Collins in the Air tonight and when he the bass drum, damn… I still have to optimize settings in Jriver and the Hilo which I will work towards. All of this is very new to me because I was a plug and play guy all my life… When I contacted Mitch asking questions, here are two articles he recommended I read before making my decision Understanding Digital Room Correction For Audiophiles Article By Mitch Barnett Of Accurate Sound (enjoythemusic.com) What is Accurate Sound? - Bits and Bytes - Audiophile Style
  10. I wanted to share some info on something I am trying and update everyone as it progresses. I am Jube owner and I have the KPT-1802 owner. Over the last year or so I have been trying to get the system dialed in. Unfortunately I am not a technical expert on audio and I have struggled getting things dialed in. My room is a killer. You move the speakers or listening position a couple inches and things really change. So one day I stumbled across a service offered by Mitch Barnett from accurate sound. He offers a DSP Calibration service: Not all DSP/DRC software are created equal. In our testing there are significant sound quality differences between DSP software products. Our DSP calibration service only use these state-of-the-art DSP/DRC software packages: Acourate, Audiolense and Focus Fidelity. Unlike other DSP correction services, we use advanced mathematical analysis of your speakers and room ratio to calculate the optimum FIR correction (eq) filter based on psychoacoustics. Our filters are designed to work in both the frequency and time domain to restore accurate sound arriving at your ears. Not only direct sound, but sound over time, specifically at low frequencies. The proof is in listening to superior 3D imaging with smooth and clear bass. Our DSP calibration service is designed to restore accurate sound at your listening position. We offer up to 6 FIR correction filters for comparison and fine tuning the result to your personal preference while achieving accurate sound. So I sent him and email and decided to give it a try. First of all I had to buy some hardware. I ended up using a Lynx Hilo as my audio interface and it works great. I tried RME products first but their ASIO drivers did not work well with the ASIO drivers in Jriver. I have to use Jriver because I am using the convolution engine in Jriver. I also ended up buying a Earthworks M23R Mic and a Grace m101 PRE for the Mic. I am using a pass labs X250 on the bass bins, a pass labs XA25 on the horns and an old Crown Studio reference II on the SUB. Here is some more info on the 1st set of filters. The -3 dB point is 15 Hz. You are getting +- 2.5 dB variation across the entire frequency response, well within the +-3 dB response at the studio. While there is some high shelf boost as you can see in the “correction filter on top” I don’t think it is excessive and probably don’t want much more, but that is for your ears to decide. I don’t boost my cd and waveguide combo too much above 10 kHz either as I find it sounds much more natural without the big boost in the 10 to 20 kHz range. I only used 4 dB of max correction in the Correction Procedure Designer. The default is 6 dB and I have used up to 12 dB of max corr with really bad rooms. So 4 dB is enough to totally smooth out the response but not too bad in the headroom department. I am listening now and the Bass response is incredible. I will try ands answer questions but I am not the technical guy. I will be working with Mitch to dial in a few things and finally enjoy the setup. I plan on getting REW setup to loopback through REW and get some real time measurements.
  11. I'm using these crossovers to dial in room, then I will get the minidsp SHD out and figure it out. Today I spent alot of time moving the speakers around in the corner I'm in trying to kill that node around 60hz. Tonight I'm having a friend come over that can help me move them to the other side of room. I obviously have some boundary interference. Don't let anyone tell you different, the room is the most important variable in my mind. Using REW for the first time has made it so apparent.
×
×
  • Create New...