- Joined
- Jun 5, 2020
- Messages
- 4,804
- Likes
- 3,747
In my experience, Audyssey detects and does not attempt to boost null regions of the response. It will fill in the edges and make it narrower, which helps make it less audible.
-
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's Next Generation of Room Correction (MultEQ-X)
- Thread starter sweetchaos
- Start date
There were both peaks and dips, as you would normally see in the bass region. The average level of the bass region seemed to be around 3db above the baseline, so setting the curve to +3db in this region should have resulted with less correction, rather than more. In any case, as the boost is broadband, I would expect to see the overall level rise by 3db anyway. Especially when it already seemed to be at that +3db before EQ, and the EQ took it down.
I'm a bit lost on your description now.
One thing that may be relevant is that the EQ filters can't change the overall gain of the subwoofer. If your curve change made the whole frequency range of the subwoofer 3dB higher, nothing much happens. You only get to change the shape of the curve within its range. (This is my experience with older Audyssey, but I don't think this has changed).
So imagine you changed the target curve from totally flat (ew) to sloping down 8dB from 20Hz to 20kHz (yay). The sub should now be louder than the mains. But that won't happen. Both the sub and the mains will have downward-sloping response, but the sub's slope will be going from +2dB to -2dB (say) rather than +4dB to 0dB.
The 0dB point for the sub filter is chosen as the average of the sub's limited-bandwidth response, not the average of the system's full-bandwidth response. At least in Audyssey Pro, this is visible in the "after" graphs, and may be so in the current apps. The 100Hz response shown for a speaker might be +2dB, while for the sub it might be 0dB.
So if you have higher bass overall in your target curve, you have to adjust the sub master trim to get the sub's response to stitch together correctly with the mains.
(I stand to be corrected if any of this has changed - maybe in MultEQ-X?)
One thing that may be relevant is that the EQ filters can't change the overall gain of the subwoofer. If your curve change made the whole frequency range of the subwoofer 3dB higher, nothing much happens. You only get to change the shape of the curve within its range. (This is my experience with older Audyssey, but I don't think this has changed).
So imagine you changed the target curve from totally flat (ew) to sloping down 8dB from 20Hz to 20kHz (yay). The sub should now be louder than the mains. But that won't happen. Both the sub and the mains will have downward-sloping response, but the sub's slope will be going from +2dB to -2dB (say) rather than +4dB to 0dB.
The 0dB point for the sub filter is chosen as the average of the sub's limited-bandwidth response, not the average of the system's full-bandwidth response. At least in Audyssey Pro, this is visible in the "after" graphs, and may be so in the current apps. The 100Hz response shown for a speaker might be +2dB, while for the sub it might be 0dB.
So if you have higher bass overall in your target curve, you have to adjust the sub master trim to get the sub's response to stitch together correctly with the mains.
(I stand to be corrected if any of this has changed - maybe in MultEQ-X?)
Last edited:
It was for speakers, not subwoofer. However, it's possible that as I limited the EQ frequency to 300, and set the filter center frequency at 200, then from the software's point of view I was boosting the entire frequency range, although from my point of view I was boosting only the the bass. Although if that's really the case, this makes bass boosting impossible without also EQing the high frequencies, which seems wrong? Is that actually what was happening? Is this unique to Audyssey? Are there any feasible workarounds, so that I can EQ just the bass but keep the overall room gain or even apply a bass boost? How does it work in Arc/Dirac when limiting EQ to just bass?
I didn't even get to the sub, obviously with how things work one should just boost the overall subwoofer level so that it matches with the speaker levels at the crossover region. But first one should actually succeed in boosting the bass of the speakers...
I didn't even get to the sub, obviously with how things work one should just boost the overall subwoofer level so that it matches with the speaker levels at the crossover region. But first one should actually succeed in boosting the bass of the speakers...
Last edited:
Hi
The people at Audyssey do not seem to understand the value of customer retention. I was all ready to buy the software and start using it.
Well this is the response I got from them when I inquire about the transfer to a new model in case of my old (X-3400H) having to be replaced ..
This is my correspondence with them... NOT pleased . Sour about Audyssey whose most interesting feature is the DEQ for me. If I can find a serious alternative to DEQ, , I will bolt out...
Peace.
The people at Audyssey do not seem to understand the value of customer retention. I was all ready to buy the software and start using it.
Well this is the response I got from them when I inquire about the transfer to a new model in case of my old (X-3400H) having to be replaced ..
This is my correspondence with them... NOT pleased . Sour about Audyssey whose most interesting feature is the DEQ for me. If I can find a serious alternative to DEQ, , I will bolt out...
| |
|
Peace.
This is literally the only reason I haven't purchased yet. I've read elsewhere there is an exception to be made for AVRs that fail, but I can't find any such verbiage on Audyssey's site and now you have this reply straight from support. I'd love to try it but this is just too limiting given the price
While this is probably not a great way of selling software, we can't say that this wasn't communicated very early on, and can't really say nothing wrong about the customer support doing anything wrong here. You can vote with your wallet, but consider that any alternatives that can EQ bass properly are going to cost much more than an X3700h with MultEQ-X.
Honestly, considering the price of alternatives, the 200$ is not what would make me buy or avoid an Audyssey device. For example, if the issue I mentioned in #944 is real with no workaround, I find that a much greater reason to pay a lot more than extra 200$ for an alternative, if I ever decide to increase the budget by that much.
Hmm, yes, it's plausible to me at least that you might see the same "average of filtered region is 0dB" effect in a limited EQ of the mains as you do with a sub. Would explain what you're seeing.
(What you were saying was so reminiscent of my own sub experiences I just assumed we were talking about a sub).
Hi
It is not a matter of how much it cost. It is of how it is sold,
Could be personal: I find it a underhanded: Sell the option wit the AVR. Done. Don't sell a stand alone software and then attached it to a gear. I don't see this as a customer-friendly strategy. They may get away with it for a while. In the big schem of things $200.oo is not that much, it remains that such their response is lame: "Wait until your receiver crap out , then get our software, which is by the way, not transferable to your next purchase" ?
Let's admit there are better ways to make money...
It is not a matter of how much it cost. It is of how it is sold,
Could be personal: I find it a underhanded: Sell the option wit the AVR. Done. Don't sell a stand alone software and then attached it to a gear. I don't see this as a customer-friendly strategy. They may get away with it for a while. In the big schem of things $200.oo is not that much, it remains that such their response is lame: "Wait until your receiver crap out , then get our software, which is by the way, not transferable to your next purchase" ?
Let's admit there are better ways to make money...
If I understood him right, he just wanted to restore some room gain that XT32 presumably must have cut, if that's the case it should work, except if in that case there happened to have too many nulls that you mentioned within the range he tried to low shelf it up. Also if that's case, then instead of using what he called a low shelf type filter, but use REW to identify where the original room gain frequency points were, and boost those only, then XT32 (via the App) will likely comply. At least it worked for me that way in the past.
Edit: Posted too soon, missed a few..
Two cents… I tried Room EQ and other forms of DSP including Audyssey extensively in two or three homes on different setups and I must say it superficially sounds better (smooth) but overall results in an anemic less natural sound to my ears. My opinion is this - if your speakers are of reasonable high quality then any anomalies will be room related - it is much better to either acoustically treat the room or just accept that the room casts some anomalies (which are real) and live with it. RT60 Measurements will help define what treatments are needed.
The problem (two cents) with DSP is that it affects the direct sound (60% of what you hear) as well as the reverberant sound (40% of what you hear). So in order to flatten out an anomaly in the room you are applying a correction to all the sound when it is actually a problem from only 40% of sound field, if you have a reasonably good speaker. This affects the balance of sound in the range of the adjustment done by DSP and sounds unnatural because of the change in direct sound level. Direct sound level is really important to stereo imaging and integrity of the stereo illusion. A vocalist or instrument covering a range of frequencies appears to come convincingly from a point source if the sound field is consistent across the frequency range.
That said, the non directional sound field below 150Hz should benefit from DSP without affecting soundstage but once again I find acoustic treatment sounds better than DSP - this is puzzling since our lower frequency hearing is not supposed to be broken into direct or reflected but it is what I have observed. I also observe that large 15 or 18 inch woofers sound more like the real thing (a real kick drum head) than small 6 or 8 inch multiple bass woofers (equivalent area) - another puzzle which is hard to explain - something to do with transient response and damping on large drivers I suspect.
The problem (two cents) with DSP is that it affects the direct sound (60% of what you hear) as well as the reverberant sound (40% of what you hear). So in order to flatten out an anomaly in the room you are applying a correction to all the sound when it is actually a problem from only 40% of sound field, if you have a reasonably good speaker. This affects the balance of sound in the range of the adjustment done by DSP and sounds unnatural because of the change in direct sound level. Direct sound level is really important to stereo imaging and integrity of the stereo illusion. A vocalist or instrument covering a range of frequencies appears to come convincingly from a point source if the sound field is consistent across the frequency range.
That said, the non directional sound field below 150Hz should benefit from DSP without affecting soundstage but once again I find acoustic treatment sounds better than DSP - this is puzzling since our lower frequency hearing is not supposed to be broken into direct or reflected but it is what I have observed. I also observe that large 15 or 18 inch woofers sound more like the real thing (a real kick drum head) than small 6 or 8 inch multiple bass woofers (equivalent area) - another puzzle which is hard to explain - something to do with transient response and damping on large drivers I suspect.
Last edited:
Room treatment does more than just flatten the response, it also reduces decay times of room modes, which room EQ can't do. However, even after room treatment, you are very likely to still have peaks and nulls due to room modes, especially in the lower bass frequencies, so you still need room EQ for bass even in a very well-treated room.
As for Audyssey and shelf filters, if it really "sets the average level" only within the equalized frequency range, then that's completely broken from my point of view and something they must fix to be competitive. I sent them a question about it and hope maybe others here can test to confirm/deny it, as I don't have direct access to the system on which I tried this (was done remotely, and I haven't purchased the app myself).
As for the economical model, while I also don't like this kind of approach, again I think people should vote with their wallets, and even if it wasn't marketed that way, in the end it's 200$ extra per receiver if you intend to use the receiver until it dies, and if you consider the software necessary, this is how you should count it when comparing to the alternatives (which are mostly much more than 200$ extra, and we just shouldn't be ignoring that either!).
As for Audyssey and shelf filters, if it really "sets the average level" only within the equalized frequency range, then that's completely broken from my point of view and something they must fix to be competitive. I sent them a question about it and hope maybe others here can test to confirm/deny it, as I don't have direct access to the system on which I tried this (was done remotely, and I haven't purchased the app myself).
As for the economical model, while I also don't like this kind of approach, again I think people should vote with their wallets, and even if it wasn't marketed that way, in the end it's 200$ extra per receiver if you intend to use the receiver until it dies, and if you consider the software necessary, this is how you should count it when comparing to the alternatives (which are mostly much more than 200$ extra, and we just shouldn't be ignoring that either!).
I have only used my $20 app not the $200 one, based on my experience I would be in the "deny.." group
. In that regard, it seems to work like the Mult EQ X app, that you can enter the boosts/cuts and the App see to take the entry points and do its FIR thing. Again, cuts work well almost as expected whereas boost only work reasonably as expected as long as the boosted frequency points are not the room mode/nulls. Please do share the response from Audyssey.But what if you boost (or cut) uniformly up to 300Hz and also limit EQ to 300Hz? Will you get any boosting/cutting? Seems like you don't, which is a serious issue.I have only used my $20 app not the $200 one, based on my experience I would be in the "deny.." group
Going by memory, pretty sure it would, except as I emphasized, at the frequency points where the room mode dips were. This is going by memory and I have too many REW plots all over the place to find them. I should do a new one for may be just 5 mic positions to save time, but it will have to be next week. If I have time I may do one with Dirac too just to compare.
No need to recalibrate, it would suffice to just change the curve of current calibration and make 2 measurements - one with eq limited to 300Hz and flat curve, and another with same limit but 5db boost up to 300Hz and flat from there. Check the average measured SPL below 300Hz between the 2 curves, if it is closer to 0db or 5db.
Last edited:
tjcinnamon
Addicted to Fun and Learning
- Joined
- Mar 20, 2021
- Messages
- 541
- Likes
- 221
I've calibrated with MultEQ-X and I'm checking my results. Obvisously the scale of the REW Y axis gives a skewed sense of "flatness". A big Y axis makes it seem flat; zooming in makes it look 'worse'.
Whats an accurate scale to use to evaluate 'flatness'?
Whats an accurate scale to use to evaluate 'flatness'?
Just look at rhe scale, how high/low are you peaks/dips - numerically.
Good point, the flat curve is not affected by changes previously made withthe App. Thank you for saving me time.
- Joined
- Jun 5, 2020
- Messages
- 4,804
- Likes
- 3,747
I just turn on DEQ and that is usually enough. I have never felt like I needed to add a bass boost to speakers on top of DEQ, but if I did, there are tone controls that affect the front L & R speakers. I know it's not exactly what he wants to do, but it should work.
Last edited: