Do you really need Audyssey, Dirac or YPAO for the best Atmos experience?

You can calibrate your home cinema system better than any automated system by using the graphic equalizer and REW. This video introduces manual calibration basics and its advantages and compares with automated calibration systems. It will be followed up with a SOTA manual calibration tutorial.
Important addition to the video:
__________________________________________________________________________________________________
Full details of the Marantz / Denon graphic equalizer filters (Q factors of the last two EQ bands are different than shown in the video (1 octave) - it seems they have taken some shortcuts due to 32kHz internal processing)
Configurable PEQ
Enabled Type Frq(Hz) Q Factor
1 TRUE Auto PK 63 1.41
2 TRUE Auto PK 125 1.41
3 TRUE Auto PK 250 1.41
4 TRUE Auto PK 500 1.41
5 TRUE Auto PK 1000 1.41
6 TRUE Auto PK 2000 1.41
7 TRUE Auto PK 4000 1.41
8 TRUE Auto PK 8000 1.15
9 TRUE Auto PK 16000 0.53
__________________________________________________________________________________________________
ppm: part per miilion (not pulse per minute)
"home theater system"
"home theatre setup"
#rew #roomcorrection #subwoofer #hometheater #calibration #cinemasound #dolbyatmos #audiophile #hometheatre #bassboost #bass #alignment #aligning #bassmanagement #dirac #audyssey #ypao #diraclivebassmanagement #mcacc #pioneer #denon #marantz #onkyo #yamaha #diraclive #multeq #multeq-x #multeq-xt32 #multeq-xt #63Hz

[@AdryDoic]

This is pure Gold and must stay on top of all Audyssey videos on YT. i have upgrade my Receiver with top Denon and was fiddeling around with al this audyssey stuff within hours and days .. never sounded right .. i have some audiophile big Towers and center. this perfectly explains why to my ears nothing sounds better then simply turn off audyssey and have my front in Fullrange ! i do some little adjustmet with manual EQ and SVS manual PEQ and thats it ! THX great job how did you that RT60 thing for that reference curve ?

[@ocaudiophile]

Check out Audyssey ART video for great sounding results with Audyssey

[@X1238-u1v]

I hope the next video will come out as soon as possible, you are the only one who provides advanced content, thank you.

[@ocaudiophile]

Cheers!

[@johnaps]

Perfect videos! Nowhere else have i found this good information, i keenly await the todo video about manual calibration, time aligning and db vector averaging, and how we measure the atmos... No one else i think has ever spoken about all this... Keep up the good work and i hope you get thousand of subcribers!!!

[@ocaudiophile]

Thank you, I'm working on that video as we speak but it's taking time.

[@phred593]

Thank you very much for sharing this wonderful content. Very helpful!

[@ocaudiophile]

You're very welcome!

[@NeemGiri]

Excellent, I am trying to understand calibration of speakers by your video, Your voice is clear & sweet

[@ocaudiophile]

Thanks and welcome

[@ts6640]

Simply an incredible piece of work!

[@carltonmann3563]

I want to thank you for this fabulous technique. I've used Audyssey for years and it never sounded right for full range so I stopped it at 300 Hz. But your technique makes full-range sound remarkable. I used a previous measurement in the Audyssey app Multeq-x because it lets me import REW filters. I measured my system with both left and right speakers active, used IR windows set to 15 cycles, Psychoacoustic smoothing, and generated EQ filters in REW then imported them with Multeq-x. I've listened to the result for about a week and it sounds fantastic. However, I think that Audyssey did better on my subs than the manual method, so I left the bass to Audyssey. There was a dip in the bass response with the manual method, for some reason, and Audyssey was very flat. Based on my experience, I think a lot of Audyssey users could benefit from your method. Once again, thank you.

[@ocaudiophile]

Cheers!

[@baronbob6971]

This channel is pure gold! Thanks OCA!

[@ocaudiophile]

Much appreciated!

[@kewlbug]

Excuse my ignorance, Are you running everything through Windows? I've got things decent with my Yamaha A-2070.. mini DSP HD on the subs

[@ocaudiophile]

Yes all in Windows and applicable to all recieevrs but the EQ band gain limits are different in Yamaha (better!). Also watch the complementary video to this one.

[@HiFinest343]

Great series you’ve got running here, I predict very fast channel growth as this is a unique feature set on KZbin here!! So funny you just released this video as I recently just updated my speaker configuration on my 8805; I added a parallel connection to my A90J to act as a Front L/R height channel(s) for my pseudo 5.2.4 system. This can be seen on my Instagram. In making this change, Audyssey is void and needs to be rerun. I manually calibrated for distance and SPL at the MLP, and I’ve never listened at this position without Audyssey turned ON. I can’t believe how much better it sounds!! The only thing I have corrected is the bass region using PEQ on my SVS subs; I corrected two sizable peaks prior to my last Audyssey run, so those PEQ filters are still enabled. I can’t believe how much better it sounds, makes me sad I’ve been missing out this whole time! I may not rerun Audyssey at all lol!

[@ocaudiophile]

Welcome to the club 😂 You'll be missing dynamic volume when listening at low levels but in every other aspect manual calibration oblitrates Audyssey!

[@HiFinest343]

Nice! How about setting speakers to Large instead of Small? If a system is very capable (500W speakers with 500W amplifiers), why would you need to cut them off? Especially if you listen to 10 below reference, it will hardly use any power? Thank you!

[@ocaudiophile]

@@HiFinest343 you have to phase align all the active woofers in the room for flat bass response. When you set fronts to large, you will have to treat them as 2 more subwoofers. Watch my subwoofer alignment video, or use MSO.

[@HiFinest343]

@@ocaudiophile nice I will take a look! When phase aligning, do you fix the mains in place, then tweak the subwoofer distance and phase, or do you have to adjust all of them until it aligns? Thanks!

[@HiFinest343]

@@ocaudiophile oh and do you agree to set to large if you have equipment that supports it, or do you still prefer to set to small so you don’t have double duty frequencies betwixt multiple drivers?

[@TrueStereo-]

Thank you for the videos. Saving me a lot of time.

[@joek6207]

Looking forward for the next video!

[@ocaudiophile]

It was supposed to be included in this video but was gonna get too long so...

[@janrademacher3680]

A great overview - Thanks a lot for your efforts!

[@ocaudiophile]

Your welcome!

[@danielmarshall8255]

Really really enjoying this series and want to try it out but do I have to leave all the speakers on large? I have kef q950/650 and they can handle it but my surround is kef ci250 (which audyssey sometimes sets to large). I haven't noticed you setting any xover. Will setting to small and crossing over break the calibration? If not, where should xover be? I have Denon3500 doing all the work. Thx!

[@ocaudiophile]

Watch my supreme Audyssey video, crossover selection is towards the end. Dolby Atmos Music video also covers it.

[@joek6207]

I know it’s $200 but MultiEQ-X from Audyssey allows a lot more flexibility with the room correction. You can skip certain speakers, lower, or draw custom target curves. You can even upload REW target curves I’m a huge fan of DEQ because it gives a home theater feel without having to pump the volume to get subs/surrounds.

[@ocaudiophile]

It's $150 at the moment 😉

[@joek6207]

@@ocaudiophile nice! I bought it when it first came out. Definitely worth it! I was coming over from Dirac Live. It was much more flexible and I had more control.

[@ocaudiophile]

I am using a Minidsp for my subs, a "highly" customized Audyssey in Preset 1 and a manual Graphic EQ set up in Preset 2. Preset 2 sounds noticably better at 75dB but for lower volumes which I frequently end up having to listen at, dynamic volume has no substitute. So I switch to Preset 1. MultEQ-X is a nice tool and offers great customization for the average user. More importantly, it pushes a lot of people down the rabbit hole of perfect sound and they will all end up with REW and Umik-1 mics pretty soon :)

[@joek6207]

@@ocaudiophile I like manual calibration much better but I have 11 channels and not all can be minidsp’d. I wish MiniDSP would make an HDMI model. Maybe team up with the HDFury company

[@joek6207]

@@ocaudiophile also, I went from Audyssey to Dirac for a few years then back to Audyssey. Dynamic EQ is second to none for home theater. The aggressiveness of it can be tempered by up to 15dB. I listen to MV at -20 and it sounds good. If I go lower than that, then I should be setting DEQ to -5 or -10. I’ve got a bunch of room treatments and I’ve been experimenting with those. I’ll be doing that cross corr time alignment technique with my subs soon.

[@igornazarov8460]

If you don't mind, I have question regarding time alignment and vector averaging of a speaker measurements from different points. Why is it necessary? IMO, level averaging is the right way (also, time alignment is irrelevant for level averaging). Thease are not measurements from different speakers from the same point (which require alignment and vector averaging), but the opposite. For example, if mic records 75dB at 100Hz at one point and same 75dB at 2 meters from that point, we will still hear this level at these two points, even if phases are 180 degre shifted between these points. 75 Hz will not cancel out, becase they are in different locations.

[@ocaudiophile]

Ideally, they should be aligned by cross correlation first before vector averaging. This operation yields the closest representation of the speaker response in a wider area at the MLP that can be filtered.

[@igornazarov8460]

Thank you for such rapid response. I understand why you need a cross corellation for the vector averaging, but still can't understand why vector averaging produces better results for filtering. When we listen to the sound, we hear it from a single point at a time, and phase in other points should not matter. In your video, you show how vector averaging produces a dip at low frequency, when there was no such dip in all three measurement points. So if we compensate for this dip using a filter (which we shouldn't, but this is just an example), then we will measure (and I believe hear) a corresponding peak in these 3 positions. So how this is closes representation of a true response in MLP if we will measure a peak in MLP after the crrection?

[@miroslavpetkov9895]

Hello OCA, after some tests over the weekend, I have a question. When we create our target curve for the calibration, we don't check in target settings the add room curve box. I think this led to overall lower db levels in the speakers 20-200 Hz range, which led to higher Hz-levels crossovers overall (150 Hz front/60 Hz center/150 Hz surround/200 Hz FrontDolby/ 150 Hz BackDolby). Do you have any own thoughts on my observation? Do we need to add the room curve to Dr. Toole curve or not? Thank you for your advices.

[@ocaudiophile]

Your observation is accurate. If you apply frequency dependent windowing to the a speaker response (and you normally should before creating filters), then you may need to apply a tilt to the original target curve because fdw will bring HF down due to quicker termination of high frequencies in the room.

[@miroslavpetkov9895]

@@ocaudiophile My comment was ment for the Audyssey A.R.T.video, but somehow it landed here. So due to the windowing by creating the 500-20000 Hz filter there is a reduction of the lower frequencies of the speacers. You have not mentioned a tilt in the Audyssey A.R.T.video. Can you give me a hint, how to compensate for the loss and by how much? Thank you!

[@ocaudiophile]

The method in ART video takes care of room reflections with a better technique (right window trimming) than FDW. Default target curves can be used as is there. Ideally, you should end up applying no correction above 200Hz anyway.@@miroslavpetkov9895

[@youtube改名也太難了吧]

Thanks for your great video! It seems that REW has cross correlation in the beta version now. Did DIRAC mentioned their method anywhere? I'm currently using DIRAC but after watching your video I'm thinking to use REW+rephase for correction. Is there suggested software for whole windows to apply convolution filter? Thanks!

[@ocaudiophile]

Equalizer APO (and it's free)

[@christiandickey]

Quick question, once everything is corrected with audyssey off and speakers set to large. Do we then go back and set speakers to small and turn audyssey back on?

[@ocaudiophile]

I don't know which tutorial you are asking this question about. Manual tuning or Audyssey calibration. Regardless, the answer should defo be in the video.

[@nofx1st]

Great in-depth info. Thanks for sharing! I got 3 questions though: 1. Can you explain what Dirac actually does, is it timealigment in terms of fixing impuls response delays between the speakers + frequency correction via eq, or is there more? That cannot be easily done manually? 2. Can you comment on the new Yamaha avrs YPAO measurement system, especially the 3D measurement that should enable the avr to map speakers in 3D similar to Trinnovs speakers remapping. 3. Is that work in a useful way at all? is new YPAO more accurate than audyssey (eg. Using more tabs for measuring?) Thanks for shedding some light on those complex topics Stefan

[@ocaudiophile]

1. There's more. Whan Dirac means by filters in the time domain is actually frequency band specific phase inversion filters. 2. I haven't actually tested the new YPAO system but for manual calibration ie without using any automated system like Audyssey or YPAO, Yamaha has the best pool of settings among brands. I don't think it's a direct match for Trinnov because Trinnov uses a special 3D microphone for measurements. 3. As mentioned I haven't tested the new YPAO but it's not hard to beat Audyssey which is not being updated since 20 years.

[@nofx1st]

@@ocaudiophile thanks for the quick responses. Do you know if there's a way to do the more complex corrections like phase inversion filters or impulse response correction via yamaha on board features? Or is it always necessary to use devices like minidsp or the like? So I guess my question is: What is the best set of corrections you could do with yamaha on board features?

[@ocaudiophile]

@@nofx1st You can use IIR filters with Yamaha (PEQ equalizer) which are known to be the best for bass frequencies regardless of what method you use. It will not do phase inversion or any othewr FIR filter because it does not allow users to use convolution filters like every other vendor. For that you will need a HTPC because even MiniDSP is limited in the number of taps for its FIR filter capacity. Trinnov for example is in fact a HTPC.

[@HarryM2022]

Very good and informative video. Many thanks for the effort to put this together. Assuming that someone wants to use aydyssey because of the DEQ, the best way to make it better is also use Multiq-X or doesnt worth it?

[@ocaudiophile]

No, you don't need MultEQ-X. You only need MultEQ Editor app. Follow this method: kzbin.info/www/bejne/nWOZmJV8maZ8rqc

[@HarryM2022]

@@ocaudiophile I do appreciate your response. Many thanks indeed!!!

[@HarryM2022]

Keep up the good work!! This hobby needs people like you! You are valuable to everyone that cares about good sound even if we don't have all the expensive bells and whistles

[@AdryDoic]

thank you so much i learned so much ❤

[@ocaudiophile]

Glad I could help!

[@Goose_Lord]

Do you know if it is possible to complete proper manual calibration with Emotiva processors? They have parametric graphical EQ.

[@ocaudiophile]

kzbin.info/www/bejne/pouQh6Knp5mVZrs kzbin.info/www/bejne/gGjQhaKFjbmnY9E

[@mikeortiz2139]

The manual calibration in the next video, is that going to be Audyssey based? As in, we run a Audyssey Calibration and then Umik measurements, and then hack Audyssey for better curves? Also will this apply to subwoofers too?

[@ocaudiophile]

No, it's independent of Audyssey, full manual calibration. It will include subwoofer time alignment but not sub response correction because Marantz/Denon don't offer sub in the graphic equaliser. All other brands do.

[@ocaudiophile]

I'd keep Audyssey at preset 1 and set up a manual graphic EQ calibration in preset 2

[@urbanlarsson8252]

Audyssey didn't add 3dB boost out of the blue. As it was a peak of 3dB at the measured 1000Hz it would logically have though it was too loud and had lowered the whole scale by 3dB and all other frequencies would had been 3dB too low. So somehow it was smart enough to know that it was just a peak at 1000Hz and not a good reference to use and let it be a 3dB peak to keep all other frequencies at a more correct level. Audio measurements are done using a dB A weighted scale to somewhat mimic how we perceive the sound level at different frequencies. It's a curve that looks like an upside down U. But in reality our hearing looks a little bit different, and are probably different from person to person. But the more correct curve are Fletcher-Munson or the more modern ISO equivalent. It shows that we have more sensitive hearing around 2KHz and the sound level from our sound system should have a dip in that area. We also have better hearing around 12KHz and needs our sound system to drop off there as well. It looks as if Audyssey have manage to cater for those differences and adapt the EQ more to our ears. We also have bad sensitivity to low frequencies at lower sound levels so that low frequencies need to be boosted more the lower sound level we use. And even the highest frequencies at 12KHz needs a little 6dB boost at lower volume levels. If we listen at a 110dB level we need a 10dB boost at 30Hz but at a 60dB level we will need a 40dB boost to make it sound as a flat EQ. upload.wikimedia.org/wikipedia/commons/thumb/4/49/Lindos4.svg/400px-Lindos4.svg.png upload.wikimedia.org/wikipedia/en/thumb/c/c2/Lindos3.svg/400px-Lindos3.svg.png hyperphysics.phy-astr.gsu.edu/hbase/Sound/imgsou/eqlou2.gif

[@ocaudiophile]

We're on the same page on varied human perception of volume levels of different frequency bands but I'm certain that Audyssey only checks 1000Hz SPL level and because there's a dip there (not a peak), boosts the full range by 3dB. Dynamic EQ also applies a boost way above what's suggested by equal loudness contour all the way until 0dbFS which is a level which can cause permenant hearing loss quite quickly. I'll not even start to tell what it does to surrounds' levels ;)

[@urbanlarsson8252]

@@ocaudiophile But if you look at the revised ISO curve of the Fletcher-Munson you'll see that it's a huge boost needed of the bass frequencies that I believe even Audyssey doesn't do enough. The boosted general levels of surrounds when using DEQ are going to be removed in newer Denon models according to Audyssey, and I believe that it's also removed or reduced to proper levels in the latest firmware for older models like my X3700.

[@andreemilsen369]

This makes so much sense, thank you😊 I now understand why I can not (not for the lack of trying🤪), to get Audyssey to sound the way I want, and also why Audyssey does not measure my subs flat (as I have made them in MiniDSP/REW). I am considering getting another MiniDSP to correct my 4 towers (LCR and rear C), that have external amps. That way I can correct them better I think. Would this be a good idéa? 🤔 What do you think sir?

[@ocaudiophile]

You can use all types of EQ filters that REW offers with any Minidsp or even use FIR filters and do phase correction with some models.

[@andreemilsen369]

@@ocaudiophile Thanks man😊 Was thinking to use the dsp for the towers, measure all the other speakers in rew, but using the peq of the reciever for theese (until I can afford external amps and more dsp's 🙊) serious illness this aufiophilia🤪

[@andreemilsen369]

@@ocaudiophile You man... I was thinking...the audyssey measures wrong, is that only because of software, or could I make the meassurements better by getting a mic that corrects it? 😊

[@ocaudiophile]

@@andreemilsen369 You can do full manual calibration for free with Audyssey's own mic and REW software. In my experience, the results far exceed Audyssey but you will lose low volume listening tools of Audyssey (Dyn. vol & Dyn. EQ). You can purchase Audyssey's calibrated microphone but you need to also invest in MultEQ-X software to use it. With the default mic and Audyssey, your results will improve if you keep the measurement points very close to each other.

[@andreemilsen369]

@@ocaudiophile Okay😊 I will use my U-mik1 then. And for low listening, I will try using tonecontroll, make a V i.e bass 3, treble 1 😊 (not that low level is often used😂). Thanks for great content man 😊

[@hdmoviesource]

How do you work out the frequency dependent windowing?

[@ocaudiophile]

Depends on distance from speaker but needs different cycles for LF and HF. I use 5-6 cycles plus variable smoothing and trim if there're still peaks left.

[@hdmoviesource]

@@ocaudiophile Thank you. Looking forward to your new video.

[@ocaudiophile]

@hdmoviesource I've been hearing a lot about your new target curve recently ;)

[@hdmoviesource]

@@ocaudiophile It's the best. It's amazing!!!!!

[@ManCaveAudio]

How I can calculate FDW for my room? Any video or guideline on the same?

[@ocaudiophile]

Hard to say. 15 cycles is a good starting point.

[@ManCaveAudio]

@@ocaudiophile Then how can I find correct cycle? Can you please guide?🙏

[@taheralikhan5771]

Hi this is amazing video Lots of learning from you thank you sir Please do pre pro 2 channel eq and analog setting

[@ocaudiophile]

Thank you

[@Unknown-1709]

Excellent!

[@ocaudiophile]

Glad you liked it!

[@msoles30]

Do you know the q value of the Yamaha avr geq

[@ocaudiophile]

It's user definable

[@rts100x5]

to be fair , DIRAC LIVE is light years ahead of Audessy and YPAO in correction capacity

[@ocaudiophile]

True but you can hack and fully customise Audyssey. Automated systems can never beat a proper manual calibration due to variations in rooms, speakers, etc.

[@AdryDoic]

are you some kind if sound engineer or its all diy knowledge? just asking

[@ocaudiophile]

I have an aeronautical engineering degree.

[@msoles30]

And how to do a real level matching the right way

[@ocaudiophile]

After equalisation remeasure and align SPL at 1000Hz 2 octaves. You're still limited to 0.5dB min step size with my Marantz

[@olgagreb7554]

Thanks for uploading!!! This channel needs Promo*SM!!

[@ocaudiophile]

Thank you too!