Close. As volume increases low and high frequencies become more pronounced (not just low frequencies), the inverse as volume decreases. What this does is attempt to make it so that things don't sound different at different volumes as it compensates for how our sensitivity to different frequencies changes with volume. Does that make sense?

@@Elliott-Designs So it's better than loudness eq? I'm a bit confused; explain it if possible in like movie/game sounds if possible

No worries!

No problem!

Be careful with too many taps at low frequencies as that actually could introduce some noticeable time domain artifacts if corrections are too sharp, especially boosts. I would recommend going with something like Multi Sub Optimiser instead of you have more than one, as that actually uses the phase change of minimum phase filters to control how to subs interact to give a smoother distribution of bass over the space, should be a lot easier to implement too!

Thanks

I'd say so, as unlike the windows feature you're able to adjust it for different content (as different contents have different targets even if -14dbFS is meant to be the standard for streaming, it's not always followed, particularly for Netflix and Disney plus I believe. Don't hold me to this, but I believe I may remember reading a few years ago that all the Windows "loudness equalisation" tries to do is make it so that volume differences between applications isn't as prominent. I believe it must act as some kind of normaliser, unlike this method which makes the audio sound the same (bass, mid, high balancing) at different volume levels given that the way our hearing works is different at different volume levels.

Hello, improve phase in terms of what exactly sorry? In terms of correcting a passive speaker's phase from its crossover components, or are you referring to something else?

Hello there. How many different audio inputs and outputs do you need on the device with the filters?

@@Elliott-Designs i need actually just 2.1 , but the problem is , regular 2.1 sound card is recgonized by windows only as 2 chanells .

I see the issue, those cards have a fake subwoofer channel. What you want to look for is a 5.1 or 7.1 card with multiple analogue outputs, you can then choose to only use a portion of those channels, they're quite inexpensive too, all you'll then require is an amplifier for each of those outputs, one one that covers all that you need. Hope this helps 🤞

I'm afraid there is not, unless it is done between your final output and your computer in terms of the audio chain.

So glad to hear it worked so well for you! Sorry, I don't use IPhones, on android you can sometimes find some systemwide EQ apps that support convolutions, particularly when rooted. You may find there's some IOS apps that can also do this for you, but unfortunately I couldn't name any specifics. Best of luck!

I'm not actually sure. I have a feeling it may just be JSON formatting. If you create some filters using the GUI then save that config, then you'll be able to see how it formats it.

No worries, as long as you don't mind the extra delay you can configure your playback in your DAW to different output types, DirectSound will likely work, but I'm pretty sure the standards for such interfaces like ASIO are not compatible since it's a different audio chain that operates independently. Other than that, make sure to try different methods of applying equaliser apo to your interface via the configurator and then restart the computer before attempting to use it, because it will need that in order to apply to configuration. Hope that helps. You also mentioned you had to figure a lot out on your own, any chance you could give some input on what I wasn't clear on so I could perhaps make a follow up in the future for others that were also struggling? Thank you.

before I saw your video I had downloaded the app to tune my speakers. It took me sometime to understand what I was doing and though there were some tutorials on REW and APO there seems to be certain functions and features that are not well explained in both programs. For instance in APO the signal chain was not so evident. The way you explained it was fine and understandable but it took me about 3 hours of messing around to understand it. I did not have your video at the time for reference. Right now I have been searching for a way to tune the out put channels of my DAW with AP0 for ATMOS. It seems not possible however maybe you know of a way. Thanks for your time. @@Elliott-Designs

@@metalplatedfacejob Ah, yes, unfortunately not something that can be done with ATMOS currently as it's a tightly controlled ecosystem, that and windows audio only supports up to 8 channels as standard.

yeah it seems like it should be easy however it is not.@@Elliott-Designs

Sorry to hear that. If it's not for critical mixing purposes you could always just adjust the value inside of the equaliser apo program until you're happy with how it sounds.

@@Elliott-Designs Is there somewhere that explains what the 3 dials mean? I couldn't find docs or anything, so I'm just turning knobs without even understanding what they do.

@@rofii_vt Reference level and reference offset are do the same thing, but in the opposite direction. Let's say your program is mixed to be listened to at 75dB in the listening position (when playing pink noise at -14dB (14LUFS), so you do your test with pink noise at -14dB with windows volume itself at max, that tells you should set reference level to 0. Now let's say it measures 78dB instead, that's 3dB too loud, so what do you do? You set reference level at -3dB, because your reference volume is 3dB quieter than your max volume. It's the same for if you're below your target e.g. 72dB, then at max volume you're 3dB too quiet, so you set reference level at 3dB, because the reference level is 3dB louder than your maximum. The second knob, the offset, is purely for convenience. Let's say we already have reference level set up so we never have to touch it again, but this other program material has it's loudness described as 26LUFS at 75dB (75dB when playing pink noise at -26dB), we originally calibrated to 14LUFS, so if we do 26-14=12. That tells us this new programme material has 12dB more difference between our maximum/peak volume and our average/talking volume. This means the target has moved so that the calibrated listening level is 12dB quieter for the same setting. So we can adjust the reference offset to -12dB to temporarily compensate. Attenuation is simply a factor that describes how much loudness correction you want to apply. If at 1, it applies all of it, the closer to 0, the less the effect is applied.

Thank you! I'd love to do a follow up at some point! Currently trying to complete a long term project first- my concrete casted counterpart to Nick (from ToidsDIYAudio) and I's RuleBreaker speaker. Unfortunately due to long setting times and the many iterations required to get it right, it has been taking far too long to complete. I'd also love to hear what you want from a new series if you have any ideas? I have plenty of my own, but it's nice to see what my audience wants too 😁.

@@Elliott-Designs Wow that sounds like the dream team of DIY audio on RU-vid! If that's the case, I would love to see y'all build a DIY speaker at some certain price point start to finish (sounds like you're doing that already) and then use these correction tools on that same speaker to produce a great house curve. It would be awesome to see that whole process in a digestible format, like your videos are. Also to expand on that, I would love to know if it is possible to use this convolution process in a Digital Audio Workstation, for mixing and mastering purposes.

@@dedik8SKB I'm so going to make a video on that. There's plenty of free tools available. Thank you

Sorry it took me so long to get back to you. First of all, just so you know, parametric EQ will likely only be minimum phase (but in your options, if you see somewhere that says minimum phase, click it anyways since it may be a drop-down to select linear or mixed phase - if that's what you want that is). Other than that, unfortunately I am not experienced with Hypex amps, but what should always work is entering the values manually. If you go into REW EQ and choose generic instead of RePhase, you can then go into the EQ filter banks at the top and depending on how many bands of EQ you have (8, 10, 12, 16 are common values but you may have a different amount available) you can set more by selecting them from None to Peaking, or set less by selecting Peaking to None. Then carry on generating the EQ filters as per usual, go back into the filter banks and then you can use those values to fill out the parametric EQ banks in your Hypex amp manually. If using minimum phase EQ, I recommend taking an RMS average of all your speakers and applying the same EQ to all of them so that when they play at the same time you don't have phasing problems.

Still active. Just unfortunately very busy and current projects are taking far too long. Nonetheless, still here to help. What convolver are you using?

@Elliott-Designs booger by voxengo, I've fixed the problem with loading it, but now I'm trying to find a way to cut down on latency while keeping the fluctuations negligible :(

@@FriendlyAudio unfortunately that's not really a thing because the way waves work, the most you can do is use the optimisation filter or lose that high resolution variance in those low frequencies. However, if you use "Psychoacoustic" smoothing in REQ (room EQ wizard) you'll see how little definition we can actually detect in such low frequencies anyways. You may find that it's enough 🤞.

@@Elliott-Designs I haven't messed with rew that much, I need to watch more videos on how to use it

I need to create a follow-up. There's a reason for manual methods... Also, feel free to just tweak until it sounds right to you. That's totally valid. It's just not necessarily accurate. 👍

I wanted to add, you're probably right in terms of adjustments as an end-user/consumer of the content, this is because unfortunately the mixing/mastering isn't always done properly and doesn't always follow these loudness guidelines, because of this you may just be better off doing it by ear, unless you know it was properly produced or don't mind a little variation. However for mixing and mastering, it's important to have loudness set up correctly for the platform you are distributing to so that you can more accurately mix and master your content for your user and how they'll be hearing it.

how did you do it on peace?

Hello 👋. The EQ process itself should be perfectly fine for audio quality if you stick to these rules: if the total gain from the sum of your filters is positive (red gain value when viewing the summed output at the bottom), you'll want to apply a negative preamp on all your outputs (on all of them so the levels don't become mismatched). No positive gain peaking filters (only cuts), and no too high of a Q (sharpness of the dips, etc). Both of these are to avoid noticeable time domain ringing (think echo). Although if you follow those rules the EQ process itself doesn't hurt sound quality, the way you apply the filters might. You may want to have a look at some of my videos on room correction, or there are plenty others that have done it too which might have some simpler methods if you don't mind changing how the speakers sum (if you want to know what I mean by speakers summing differently from typical methods, I have a video on linear phase filtering explaining that). Is there anything else I can try and help you with 🙂?

@@Elliott-Designs For now I am absolutly satisfied with your unsver and before asking I have to deep in a bit. Thank you very much :)

@@janulik9535 No worries, and whenever you have any more questions, please don't hesitate to ask 🙂

@@Elliott-Designs Thank you

If you're talking about what I believe you're talking about, what you're doing is effectively doubling the colouration of your speakers and your room, that initial convolution is of your original measurements and effectively "describes your room", try enabling the convolution whilst listening through headphones, you'll see what I mean. If I'm understanding you correctly that is. You might want to have a look at my series on linear phase correction, to see how to correct for your room and speakers. That's if you want to try and be faithful to the what the post-production audio engineers intended anyways, if you prefer the colouration of your room then keep it 🙂, after all you're listening to it back, it's not like it's critical as if you were a pre/post-production engineer 👍. Oh, as for the long delay, that's because it's quite literally doubled the decay time of your room.

Not that I can find I'm afraid. That's the thing about MacOS unfortunately. Not as many things are developed for it. Unfortunately there's not much I can do about that

@@Elliott-Designs Thank you for your response!

That's odd, they should be one and the same, the reason that dip occurs is because that is where the port "takes over" in terms of loading. If you go onto Nick's website ToidsDIYAudio, you can start a thread and tag me in if you'd like, that way I can see your measurements and try and help a bit better.

@@Elliott-Designs yeah, the ideal condition would be that. But the box tuning frequency and port tuning frequency can be different sometimes. My box for woofer is a bit small for its VAS

@@user-dh7lt4we2t In that case I'd recommend applying an IR window to "remove your room" as much as you can from your nearfield measurement and then analysing your measured phase response manually to determine how you want to phase correct your port region.

@@user-dh7lt4we2t You could also just try both and then compare the phase responses of the measurements with either correction applied, without an anechoic chamber or at least a portable amplifier and a large field to act as a ground plane, these iterative methods are sometimes the best you can do.

Oh, that doesn't sound great 😕. Unfortunately I don't have a Mac myself. Sounds like an LTSpice problem more than the add-on itself SpicyTL. Might be worth reaching out on an LTSpice thread you can find on electronics forums. Sorry I couldn't be more help

@@Elliott-Designs oh... Don't worry! I've mentioned it to let your viewers know about the issue.

@@wrongIQ Thank you for letting everyone know! I've been sure to pin your comment. If you find a solution at some point, it would be great if you could share it 👍

Thank you very much for such high praise! Anything in particular you might like to see from me?

@@Elliott-Designs Thank you for the offer. I don't even know what to ask for yet... Your phase correction and frequency response technology was the most understandable among others. I was able to repeat this on two different pairs of my speakers (there are two more pairs of speakers from my collection left without correction, they are waiting for their finest hour) In all cases, I was very pleased with the result. I heard a lot of new sounds in my favorite music recordings, I was able to hear a very clear transparent space between the instruments, and even sounds from behind... and in general, any music has become more pleasant to sound!

@@syelchin so glad it's been useful to you 🙂

Today I came up with the idea of whether it is possible to get rid of thd harmonic distortion of the speaker in a similar way? The speaker introduces parasitic harmonics in the sound, but if you generate reverse (antiphase parasitic harmonics) that will correct the distortion at the output? Do you have any guesses how to do this using the same programs?

@@syelchin Very interesting idea, theoretically this should be possible, however the very high Q peaks that represent the harmonics will likely cause noticeable pre-ringing, however this is something you would have to test. Even so, you would need accurate measurements of the driver distortion so your results aren't skewed. One way I'd imagine doing it if you don't have an anechoic chamber would be to make a low distortion battery powered amplifier, a laptop, and a calibrated measurement microphone all out to the middle of a field, place the back of the speaker on the ground and rig the microphone to take vector averaged measurements of the driver. You could try other methods but I wouldn't be so sure of the success rate of those. I'd very much be interested in hearing how your testing goes. If you wanted to you could post about it on ToidsDIYAudio.com under the DSP section 👍

I did not make this add-on for LTSpice, nor make LTSpice itself, but you can fairly easily approximate a room's low frequency gain, much easier than having to put it into the program too. Take the longest dimension of the room and multiply it by two. This is the wavelength of the frequency where room gain will start (approximately). E.g. Starting frequency = speed of sound / (2*L). Where L is length in metres. For a perfectly sealed room this will continue with a trend of 12dB/Oct all the way down to 0 Hz. It is likely the slope will be less though, since this is under ideal circumstances. You also then have boundary gain. Have it next to one wall? Assume +3dB. Have it next to a wall and the floor, +6dB. In the corner of a room on the floor, +9dB. Again, these are all approximations. But I hope they help.

Thanks!

Totally, in fact, a small room can help get more (although less accurate) bass. This is called room gain, it's the same with car audio, but there it is called cabin gain. I run down to 30Hz flat in my UK bedroom which is much smaller than your average living room. If I had a larger driver in my subwoofer, or more subwoofers then I'd be able to get even lower.

@@Elliott-Designs I guess I worded it wrong. I meant with room gain do we need speakers that can output much lower than 50hz? Do things get messy without room correction that most people don't do? Love what you're doing here. I'm a carpenter looking to start building speakers. I'm trying to learn all I can about designing speakers but I may need to get help from a designer in the end.

@@bellrobert1978 Not at all, in fact produced music is designed to be played back with room gain. If it's too much bass for you, you can always build a sealed subwoofer (no port, closed box) and utilise its shallow roll off. My content doesn't go into much of this stuff yet, so if you want to know more of things like this, I highly recommend my friend Nick's channel. Toid's DIY Audio. Also, thank you for the high praise! The best place to start in speaker design/building would be to start with a kit/build plan, it's a great introduction into the hobby that way.

Hey Matt 👋. Thank you!

Glad you enjoyed it!

Thank you!

The simplified description would be a mass-spring damper system if comparing it to mechanical engineering, but yes, loudspeaker ports are definitely a planned topic!

And yes, resistances etc would be discussed. Essentially I just wanted to explain particles and the sound transmission process as a whole, give some fundamental information for some more interesting upcoming videos I have planned

Thanks Nick! I wanted to lay some fundamentals down before getting into some more interesting, and speaker centric topics.

But yeah, it's definitely been a while 😅

@@Elliott-Designs sounds like a great idea

miniDSP HD is one, but it's still rather limited in terms of the amount of taps/samples you can use which severely limits low frequency performance with convolutions. Anything else (that I'm aware of) would be in pro audio, and therefore definitely not cheap. It's why I cover it in software, not many good implementations on hardware based DSPs at the moment unfortunately.

@@Elliott-Designs too bad.. I already have t rack dsp, and i dont think this would be upgrade of mine.. I did stumbled on ir loader from tc eletronics, but i dont know if that would work the same since it has impedance made for guitar

@@ivanmaskov Sorry, I'm afraid I have no experience with either of these things and therefore can't help you with this issue. Hopefully you can find someone that can help you. Best of luck 🤞

I'm glad you liked it! Also, thank you for your kind words, I certainly tried my best. Wave physics can be rather difficult to convey at times

What is it you are using it on? I might be able to help. There are some systems though that seem to just be completely incompatible. Does it say anything in particular when applying a filter in the configurator? Have you tried the 32bit version (or 64bit if you were running 32bit)?

Double blind AB tests are of course the best way to go about this, but I'm not in a situation to perform them myself. But with enough taps the pre-ringing will actually be below the noise floor (physically inaudible - all from what I've learned anyways), which is what I propose, although I do understand most hardware DSPs aren't capable of such high taps. I'm not sure what you mean by pre-attack hints, since that isn't what pre-ringing does (perhaps you are thinking of a different phenomenon). Anyways, I'm happy you're happy with what you have. I'm just putting the information out there for people to make their own informed decisions. I'm glad you've made yours!

It might also be that you prefer the time smearing that minimum phase gives you, and that's totally valid too!

I'm going to do more testing. I can set taps ( quality in my case) to very high and yes CPU is the only limit as always. I'm planning highest linear vs minimum phase. I have eq at every speaker driver and on the final LR bus so setting globally to high might blow up my system lol. Just for info I use VST pluggins in a DAW for my active system.

@@sqcaraudio sounds good! Just be sure to have all drivers be set to the same amount of taps, otherwise you'll have the audio being fed into each at different points in time! 👍

Yeah just did a quick test. All chans. Sounded different, less grain? All songs had a tad deeper depth but..... Then an isolated kick and snare intro.. Slurp slurp thwack . Back to minimum phase. To be fair I know the more gain manipulation requested in a linear phase filter the more pre ringing happens and if the source is an isolated transient etc.

If it's not working for you, likelihood is something went wrong with the initial install. You can head onto toidsdiyaudio.com, there's a forum on there where I'd be happy to help, but there's absolutely no need for such language, I'm making informational videos to help people. If my video is too quiet for you that's because I use a wide dynamic range for better audio quality. Perhaps I shall compress my audio in future videos for those that don't have the volume capability. Thanks for pointing that out (although you could have done it a little more politely). Not sure what you're referencing with the Realtek stuff, perhaps if you could rephrase that part? This program is one of the best for software loudness equalisation and EQ. As for not equalising RU-vid sound, it won't be equalising any sound (not just RU-vid). The EQ functions systemwide. Hope this helps 👍

@@Elliott-Designs It's not the APO EQ that doesn't work, it's just the equalization module. The equalization function works perfectly in the realtech HD menu! Equalizes all sounds, including your video. BUT. If I activate the APO software, the equalizer check box disappears from the realtech HD menu. The APO software blocks the realtech sound effect modification menu. The consequence: I can't adjust compensation either in Windows or in APO. It is not enough for me to only be able to set EQ schemes. Due to videos with different volumes, an equalizer must be turned on. But if I want to listen to music, it must be turned off. That's my problem.

@@retrobudapest Ok, I see now. Yes, with EQAPO you will not have access to the realtek EQ as well as EqualiserAPO's EQ, this is because EQAPO embeds itself around the realtek driver. Luckily, you'll have the same equalisation abilities inside of EQ APO (parametric and graphical EQ). Actually, you mention videos with different volumes, is perhaps the feature you are missing called compression or normalisation? Either way, all videos on RU-vid should be fairly similar in volume level. They are all equalised to a loudness of -14LUFS. Going from program to program is a different issue of course. That requires several different preamp sections in EQAPO that you have to turn on or off depending on what platform you are using. I have a good write-up of integrated loudness of different video and audio sources on toidsdiyaudio.com if you wanted to implement this. You mention it not being enough to just be able to "set EQ schemes", what do you mean by this, what is missing? Hopefully I can try and help solve this for you.

@@retrobudapest KP, you are misunderstanding what the "loudness correction" feature in Equalizer APO is for. It is not like ReplayGain which normalizes loudness between tracks, albums, etc. I don't know of any software that would do that *between* Windows applications. I'm amazed your Realtek software even had an option for that. You might have a tough time finding something that works. You probably just need to ride the volume knob all day like the rest of us! hahaha

@@mjodr A program called Wale will do it.

No, I've not seen that issue before. Have you tried reinstalling it? The next step I'd go for after that is trying the 32bit version or the 64bit version (whichever one is different to the one you were trying to use). It might be that the driver isn't linked in properly with the configurator, so you could try changing which device it's applied to, or try the experimental settings option shown in the video. Give these a go, I'd love to hear if it worked or not, if it didn't, we can find an easier way to communicate, e.g. email or toidsdiyaudio.com, and I can try and help from there

@@Elliott-Designs oh and also when i try to open configurator editor it doesnt open do yk why?

@@flyinq_baconz9695 That sound like a bad install/initial setup. Try reinstalling

I know, and you can follow the built in correction playback if you want that, it's just not as accurate, that's all.

@@Elliott-Designs That's the reason you've got so many dislikes.

​@@kalumanderI'm ok with that. People can still use the standard method if they'd like. But given the fact that based on what the audio source is, netflix, RU-vid, Blu-ray, changes what the reference volume is, going through REW just let's the user actually know what they are targeting. Thanks for the feedback though 🙂

Thank you, and very true about the delay. You can get a lot done with 10000-11000 taps still, and that delay isn't too bad when the rest of the chain has very little delay. Also, you can lip sync correct either in a video player like VLC, some TVs have it, also most AVRs. You can add delay to display output essentially to resync. Wikipedia says audio can be delayed up to about 140ms before it becomes noticeable for lip sync mismatch, those numbers above should give that. ^those numbers work really well for low frequency corrections, but only really when optimisation is used like in rePhase to get more out of some taps. Also, doesn't matter for music, so you can go nuts with that if you have the processing power

I know, it was an output error from my video editor output, I wasn't aware it fed mono audio into stereo single channel. So sorry about that, glad you liked it otherwise. In all videos since this one I have fixed the issue

Do you mean you want to apply it to an individual audio file (one with music on for example)?

If so, you can use a VST plug-in like the one on convolver.sourceforge.net/vst.html, and use it in a DAW, or since it's a VST2 type plugin, you can also use it in Davinci Resolve which is a free video editor, that can also just be used for audio editing. Hope this helps 👍

Processing power is only in comparison to audio processors like minidsp or other typical hardware based processors. Any slightly modern PC should be able to do just fine. A raspberry pi for example can handle processing 8 channels, each with more taps than you'll ever use.

As for the 192khz issue, head over to toidsdiyaudio.com, there's a forum section called DSP. I'll make a topic for EqualiserAPO, if you could post a few screenshots of it all there, I might be able to help 🙂👍

@@Elliott-Designs thank you for your reply. I uninstalled apo then changed backed to 192 np, I don't remember changing it to 48 so was just curious if thats the max using app. I'll check out that link as well and hopefully get informed by the time I'm ready to apply all 3 of these, again thank you very much. I'm so tech challenged, sometimes I think I was born in the wrong era and would have faired better when the most advanced tech was the telegraph... 😂

@@phillipmorris9847 Nope, it doesn't max out, there's something strange going here, as I said, I'd be happy to help on the forum toidsdiyaudio.com/forums/discussion/equalizer-apo-troubleshooting/ There's even people on forum pages (not the Toid's one) that have used up to 384khz/32bit no issue Likelihood is that it's a communication error between the driver and eqAPO, there's a few things we can try to fix that!

Per channel, but because I use a spatial average type measurement (MMM-moving mic measurement), it uses periodic pink noise, I'll probably make a tutorial at some point. I digress... Because I use that method, it means descrepencies between left and right channels are less. When doing individual measurements, or just a few averaged, differences in the responses are typically exaggerated between the channels. But yes, I have a different filter for each of the speakers.

Typically the lifelessness comes from the EQ itself not being done right. Filters from FIRs compared to just PEQ aren't actually any different when they are both minimum phase, assuming you have enough taps applied of course. Also, I recommend checking out my other tutorial on how to do the loudness correction properly, the audio out built into the program for the calibration isn't the right volume. Also, if you aren't already using loudness correction, I highly recommend it.

It certainly is! This one's based on the old Fletcher-munson curves though, not the new ISO standard. Honestly though, there's very little difference, to my ears it works just as well 👍

@@Elliott-Designs that is awesome ! Thank you for passing that information 😍.

@@EngineeringEssentials No worries! That's why I'm here!