You were gonna post a link? You said "check this".
-
WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!
Audyssey Room EQ Review
- Thread starter amirm
- Start date
Sorry got busy and forgot.
https://www.andyc.diy-audio-engineering.org/mso/html/index.html
There’s this too:
Last edited:
What do you mean manual approach?
The payment is more for the hardware than the software, as you need something to apply filters in real time to EQ... Audyssey is one option (but it always EQs them as one sub after setting delays and levels), minidsp is another (can EQ any way you want, although the "normal" way seems to be setting delays manually based on REW measurements and then applying filters to both subs equally), and some use both, but not sure how well that works, though both seem to be pretty decent at EQ of subwoofers according to the reviews I've seen, and in any case if one applies enough filters to flatten the response, theoretically the other will have not much left to do.
Audyssey's appoach of time aligning and level matching individually and then to EQ the pair as a whole is the only way that make sense to me. If you EQ'd them individually there's no way of knowing what the response would be when the two were sum'd in the room as a whole, you'd then have to do the sum over again to get the desired response. I've been using Audyssey with the dual sub feature for a few years now and get excellent results.
Which is why we have computer software to try millions of combinations in a matter of minutes. If you tried to do this manually you’d spend your entire life doing it.
It seems to me you could do the same thing using REW. If I got some subs (two) I'd initially place them equidistant to my listening position as the speakers and then measure them using REW & UMIK microphone at the listening position when combined with the output from my speakers (crossed over at a sensible crossover frequency, I'd probably choose 100Hz or a bit less). I'd use the measurement to set a Gain Level for the subwoofers so that their response flowed evenly into the rest of the frequency response, then I'd use REW to create some filters to remove peaks, with the view to end up with a smooth as possible bass frequency response at listening position that flows evenly into the frequency response of my JBL 308p Mkii speakers.
In terms of phase integration, I'd have to look some more into that because I'm currently unclear how I capture that measurement (I think it's included in REW), and I understand that phase would be related to equidistant physical positioning of the subwoofer with the speakers (I think), and I'm also thinking I could use delays to account for any necessary distant differences between subwoofers & speakers in order to phase align.
Onto the actual physical connection scheme I'd use to hook up the two subwoofers. I know this is gonna influence how much control re EQ & delays I have over the subwoofers, but I've thought of a few options.
Option 1 could be simply connecting each subwoofer in-line with each respective left/right speaker and relying on the subwoofers pass through and the subwoofers built in crossover setting options. So in this option I understand depending on subwoofer I can set a crossover point and it will do the necessary Low Pass for the subwoofer and then apply a corresponding High Pass to the signal continuing on to the speakers. In this instance I could still use my laptop (music source) to do bass & speaker EQ using EqualiserAPO, because the bass EQ would still be utilised by the receiving subwoofer during its Low Pass and the speaker EQ would still be passed on through the subwoofer's High Pass - the only problem I can see is that I can't apply any delay in the signal in relation to the subwoofer because the EQ is all bundled as one channel complete EQ (for both speaker & sub).
Option 2 could be using my miniDSP 2x4 (non HD) to manage the subwoofers and crossover. I've got one spare left/right output on my miniDSP which I would connect up to the two subwoofers using a Y-splitter RCA cable (the other one would be used by left & right JBL speakers). I'd use the EQ banks within the miniDSP to apply a Low Pass Filter and any bass EQ as well as delays for time alignment (although these EQ's & delays would have to apply equally for each sub due to using Y-Splitter, but I guess that's ok as long as subs are equidistant to listening position and speakers). Another thought, re miniDSP setup, I could just route left channel to left subwoofer & right to right subwoofer, that way I'd have individual control of EQ & delay to each sub via the miniDSP, and I'm thinking that still might give me some more flexibility of sub placement as long as I set the crossover low enough so I can't localise the subs during listening (100Hz or below should be fine for this right?).
I kind of prefer Option 1 above, if it gives me enough flexibility to end up with a good result, because my miniDSP 2x4 (non HD) is not of the highest spec in terms of SINAD, etc, I think it's rated at around 90dB in the miniDSP specs (actually I'm wrong, the specs listed are "ADC dynamic range, un-weighted > 98 dB" and also "DAC dynamic range, un-weighted > 98 dB"), so I'd prefer to keep it out of my audiochain for music listening and instead use EqualiserAPO on my laptop....however I would continue to use the miniDSP in the connection chain from my TV for movie & TV watching - currently I use a manual switch box to simply switch inputs from TV to laptop and the miniDSP just sits in the TV chain before the manual switch. I'd prefer to keep it that way for audio fidelity reasons which leads me to prefer Option 1 above, although it is perhaps providing less control over the subs, most notably in delay (I'd have to use any delay functions included in the subs themselves, if they offer such features, I don't know).
Yeah, I've not thought a great deal about it, right now is the most in-depth I've considered the option of buying & integrating subs. Not totally sure if I've considered everything correctly here or if there's some things I've missed, but welcome to hear from people to confirm my thoughts or to offer improvements or insight to problems associated with my different ideas.
Hi Galz, as above.
Last edited:
Hi. I agree in some ways with you on doing option 1. Also in other ways for example if your room is 60dB ambient then 90dB up is like 150dB. So maybe my understanding is wrong but then it makes the 90 SINAD at least worth trying also. The other thing I noticed it may support fewer parametric eq points as the HD?
Anyway it also is something I wouldn’t really worry about. You get the biggest gains in the first few peq points.
The SINAD of the miniDSP 2x4 is unknown as far as I can find out, in their official specs they talk of dynamic range, which I think is just referring to SNR, so they're advertising >98dB SNR....so this doesn't include any distortion element, so SINAD is unknown. It could be worth trying for music listening though, but one drawback is that the miniDSP can only output a max 0.9V, so it means upping the Gain on the speakers (& subs) and subjectively I feel that my 308p speakers perform better at a lower gain but I've not been rigorous about this. The good thing about the miniDSP 2x4 is that there is a "hidden" way of activating more available PEQ bands. In the crossover section of the miniDSP software you can change the configuration from Basic to Advanced, and with Advanced that means you have to enter filters as biquads, and when entering filters as biquads into that section you can apply 8 different PEQ bands within that space (they don't have to be anything to do with the crossover, can be any filters you like). So this means you have the 8 PEQ bands there, plus 2 more banks of EQ that contain 6 filters each, so that's 8+6+6 = 20 bands of PEQ filters available for each channel, albeit the crossover bank that I mentioned where you can put the 8 PEQ bands - there are only two of these, one for left & right channels before they get further broken down into four outputs (x2 Left & Right), so those 8 PEQ bands by definition have to be applicable for both Left Channel outputs, and the other Crossover bank by definition has to be applicable for both Right Channel Outputs. So this would mean I'd have to use the Crossover 8 PEQ bands for Speaker EQ rather than for crossover High Pass / Low Pass and instead I'd place the High Pass & Low Pass filters within the EQ section for each output.
About Option 1 I listed in my previous post, where I don't use the miniDSP and instead rely on the crossover built into the sub, the good thing is I'm thinking I can fine tune some of the response at the crossover point using filters in Equaliser APO - for instance if the crossover built into the sub is not very configurable and causes a dip in the response in that area then I can use EQ filters to boost that general area before the sub applies it's crossover. So I wouldn't be totally at the behest of the subwoofer crossover, but I would be regarding the steepness of crossover. It's true though I won't have delay options to allow for different sub placement positions, unless there are delay options built into some subs, where I'd be hoping you could set the delay of the sub independently of it's pass through to the main speaker (thereby allowing different delays for the speaker & sub)? I'd have to research what configuration options different subs include, I've not looked into it at all yet.
Thank you in advance for any guidance. I am a first-time room EQ utilizer and have been playing with Odyssey EQ the mult xt32 version that comes with the Denon 3700h.
I think my main questions are as follows:
1. When using odyssey, the software shows very specific places between the two front speakers that it wants you to measure. The software also says don't measure more than 2 ft away from the initial measurement place. So if you have a second sofa or love seat off to the side and outside of the sweet spot, you just don't measure over there?
2. The software recommends setting the subwoofers at 12:00 but that's giving me a negative 6.5 correction on my two subwoofers. I have read in other places that you want to get that negative 6.5 down to about two or three would you agree with that? If so then I will rerun the measurements with the subwoofer set more like 10:00 or 9:30.
3. I saw amir's post about using the curve editor but what a cluster. No matter how I try to raise the 20 bar the slope always goes up another 5 DB before it goes sloping back down and nothing I can think of will create a smooth looking slope like a Amir shows in his picture. For now I have settled on keeping 20 about zero and that gives me the 5 DB bass boost that I can then slope down to 200. (Bonus question, I have the revel M16 which have a bit of a bass boost at 100 would I be better off sloping down to 100 instead of 200?)
4. I also noticed that Adyssey set my LFE crossover to 120. That should be changed to down to around 80 shouldn't it? It also crossed my revel c25 at 40 and my m16s at 60. I am under the impression that I want my crossovers for the front and center at 80 and my surround left and right at 100. Does that sound right?
5. I also noticed that in the $20 mobile app that each of the manual curve editor results are speaker set by speaker set... You have to choose either front or Center or subwoofer or surrounds. Not globally? Are we only interested in changing the slope for the fronts?
Appreciate any feedback.
I think my main questions are as follows:
1. When using odyssey, the software shows very specific places between the two front speakers that it wants you to measure. The software also says don't measure more than 2 ft away from the initial measurement place. So if you have a second sofa or love seat off to the side and outside of the sweet spot, you just don't measure over there?
2. The software recommends setting the subwoofers at 12:00 but that's giving me a negative 6.5 correction on my two subwoofers. I have read in other places that you want to get that negative 6.5 down to about two or three would you agree with that? If so then I will rerun the measurements with the subwoofer set more like 10:00 or 9:30.
3. I saw amir's post about using the curve editor but what a cluster. No matter how I try to raise the 20 bar the slope always goes up another 5 DB before it goes sloping back down and nothing I can think of will create a smooth looking slope like a Amir shows in his picture. For now I have settled on keeping 20 about zero and that gives me the 5 DB bass boost that I can then slope down to 200. (Bonus question, I have the revel M16 which have a bit of a bass boost at 100 would I be better off sloping down to 100 instead of 200?)
4. I also noticed that Adyssey set my LFE crossover to 120. That should be changed to down to around 80 shouldn't it? It also crossed my revel c25 at 40 and my m16s at 60. I am under the impression that I want my crossovers for the front and center at 80 and my surround left and right at 100. Does that sound right?
5. I also noticed that in the $20 mobile app that each of the manual curve editor results are speaker set by speaker set... You have to choose either front or Center or subwoofer or surrounds. Not globally? Are we only interested in changing the slope for the fronts?
Appreciate any feedback.
1. It is not recommended to go farther than 2ft from the first measurement to ensure good results at the main listening position. If you wanted, you could save 2 Audyssey presets: one focused on the MLP and one including the off-axis seats. Switch between them based on if you have off-axis listeners.
2. -6.5dB is fine.
3. The curve editor is certainly clunky, but this behavior is because the M16s are correctly set to "Small". Remember there is a crossover so the subwoofer takes over at 60Hz with your current settings. You are defining the target curve not the filter response, so you don't need to over compensate for the M16 100Hz anechoic boost.
4. The LFE setting is a low pass filter, not a crossover. You can leave it at 120Hz. The one-size-fits-all recommendation for crossovers is 80Hz. That value works in many cases, but is not necessarily optimal depending on your specific in room response. If you don't have the means to measure the result, I'd probably default to 80Hz.
5. Unfortunately, you have to manually change the curve for each speaker set. Many people have requested the feature to be able to copy a curve to another speaker set. When you upload settings from the app to the AVR, all of the edited curves will be applied.
Thank you very much for your responses! All of those were very helpful. One note, Audessey flagged the M16s as Large. I had to change to small manually. I suppose I'll try both 80 and 60 for the crossovers and see if I have a preference. I found a youtube video which recommended turning off the mid bass compensation which I did. I was surprised to see that dip at 2k. I also turned off Dynamic EQ. Audessey is really interesting to play with! Again, many thanks for your response.
Edit - One other recommendation was to cut off the Audessey EQ at 700. I know the M16s already have some roll off at 20 so I tried the 700 cut.
Last edited:
Disabling midrange compensation is a good step. There are many speaker designs where an anomaly in the crossover leads to a dip in this region. Audyssey seems to think this is common enough that they put a dip in the default target curve meant to avoid overcorrecting and messing up the direct sound. Even the Revel M16 has a slight deviation from its otherwise smooth PIR from 2-3kHz. It's not enough to warrant keeping a dip in the correction target though. I'd actually suggest you try limiting the Audessey correction to below ~500Hz to maintain the M16 natural midrange and treble response.
Dynamic EQ is a loudness equalization that introduces bass and treble boost at lower volumes. It also boosts the surround channels at lower volume. I find it useful with the reference offset set to maximum to maintain correct tonality at low volumes. I do wish the surround boost was separately adjustable though.
What are your thoughts on the seemingly mandatory High Frequency Roll Off #1 or #2?
Well we know that some amount of high frequency rolloff is natural for an anechoically flat speaker placed in a normally reflective room. It's a moot point for me because I either edit in my own target curve or limit the correction range to the modal region.
Right, of course. Doesn't matter.
Perhaps already mentioned, but a note from me.
I had Audyssey activated for subs and for main monitors <300 Hz and a sloping target curve. Happened to measure the system at LP with Audyssey (reference curve) vs Audyssey (L/R bypass) and noticed that I got more energy with the bypass (red curve) than the reference curve. So I am using Audyssey for room correction only <80 Hz at the moment. Sounds better as well.
I had Audyssey activated for subs and for main monitors <300 Hz and a sloping target curve. Happened to measure the system at LP with Audyssey (reference curve) vs Audyssey (L/R bypass) and noticed that I got more energy with the bypass (red curve) than the reference curve. So I am using Audyssey for room correction only <80 Hz at the moment. Sounds better as well.
If adjusting to the Harman Curve via the app per Amir's recommendation after running Audessey, wouldn't one need to turn down the subwoofer levels since the added bass now coming from the speakers was never accounted for in the original measurements?
What's the point to use that curve if you turn the subwoofer down? The whole point of that curve is to increase the bass. Though make sure to adjust the subwoofer and other speaker curves to be the same as the speakers, I think Amir's test included only 2 speakers. I personally prefer to increase the bass with dynamic EQ, but you need to know how to mess with the reference level offset (mostly set 0 for movies, and play with it for non-movie content for best result, although I find -10 works well for most non-movie stuff).
Dynamic EQ does work well and I used it on and off as well. My only complaint is it messes with the souround speaker levels depending on the volume just like bass. I found it raised the levels quite a bit too much when listening at lower volumes. Some may like that part of DEQ but ultimately it caused me to not use it.
I have Dynamic EQ on...seems to work well on lower volumes. As for the curve, I adjusted all of the speakers LRC and Surrounds to the same curve but I didn't touch the sub....I left it flat and I have more bass than I can handle without adding more there. In fact, my question is whether the sub should be turned down a bit as the EQ for the sub expects a flat FR and now all of the speakers have added bass. Seems in my mind that the sub should go down because more bass is coming through the other speakers. Am I not thinking about it correctly?
Last edited:
Similar threads
