DIRAC VS. Audyssey | What Does DIRAC Do on the NAD T778?

NAD T778: bit.ly/NADT778AA
If you want to check out the REW mdat files, they're here: / rew-mdat-files-54248452
I've always wondered what the differences were between what DIRAC Live does on my NAD T778 vs. what Audyssey MultEQ XT32 does on my Denon AVR's. I hear a difference, and after doing these tests, I have a bit more insight as to why. Now, these tests don't take into account the differences in how they do level matching, time-alignement, and phase alignment. Maybe I can compare those in a future video.
Watch my video taking a look at how Audyssey works: • What Does Audyssey REA...
My Speaker Leaderboard: bit.ly/SpeakerL...
My Subwoofer Leaderboard: bit.ly/subwoofe...
My Headphone Leaderboard: bit.ly/headphon...
Become a Patron: / joentell
My Recommended Products: kit.co/joentell
Gear I Use: www.amazon.com...
Channel Sponsors:
OSD Black: bit.ly/OSDBlackOSD
Monoprice/Monolith: bit.ly/Monopri...
Follow me on:
📸Instagram: / joentell
🐦Twitter: / realjoentell
👥Facebook: / realjoentell
Disclosure: The Affiliate links above earn me a commission if you purchase using those links. This is at no additional cost to you. AFFILIATE LINKS:
Crutchfield: howl.me/ckjnBv...
Best Buy: howl.me/ckjoam...
Target: howl.me/ckjoaH...
Disclosure: The Affiliate links above earn me a commission if you purchase using those links. This is at no additional cost to you. This product was sent free for the purpose of this review.

[@danielwander605]

Really good idea for a video. I am curious why you chose to do a near field calibration? Besides correcting for different targets I thought the main purpose of room correction is to correct for the room. So i would think a nearfield calibration would give less for the software to do and would yield similar results for both calibrations. I’d like to see the same thing with 2 speakers, in room, with a 3 point calibration. I know that makes it harder to keep the tests identical but I’d still be interested in the results. Either way, is a great idea. Also, can you link to that shock mount for the Umik? I’d like to grab one.

[@Untilitpases]

This

[@FURognar]

You confirmed what I suspected: Audyssey calibrates to flat which is more accurate (to whats in the source) where Dirac (and other calibration profiles) targets a slightly warmer curve which most people find to be a more pleasant sound.

[@BriBCG]

Yes, very interesting. I created a similar filter in Equalizer APO to the Dirac curve and even though it doesn't look like much it actually makes a substantial difference to the tone of the sound. I would guess this is the biggest difference people are hearing between the two.

[@rolandrohde]

This confirms that my custom curve with the Audyssey app is pretty close to what Harman suggests and Dirac does.

[@joentell]

According to the research at Harman, a speaker that measures flat anechoically (reproduces sound accurately) will have that downward tilt when in-room. That's also saying that a speaker that has a flat response in the listening position will have an upward tilt when measured anechoically.

[@rolandrohde]

@@joentell Yes, and Audyssey messes with that. It pulls the Bass down.too far without DynEQ and raises Bass and Treble (and the Surrounds) with DynEQ far too much in my experience.

[@tomaszoslizo9226]

@@rolandrohde use movie curve or modify default with the app.

[@rorymells3405]

How do you know what level in db to set the speakers at before you run Dirac? Is there a measurement you take or just go by ear?

[@vailvon]

Joe, can you add the ARC system from Anthem in the mix. I got Anthem receiver and would like to know how it compares against other systems. Thanks

[@joentell]

If I had access to one, I would

[@beaten_tech]

That weird correction at 5:30 is most likely due to that being the speaker crossover point.

[@arthurprs]

I think Audyssey is shooting itself in the foot with these defaults. Flat in room response may be "accurate" but will sound weak/unrealistic to most listeners. Instruments are mostly captured nearfield BUT a live listener (and music producers to some extent) are not and also hear the room interaction (gain, etc..). Forcing it to be that flat in room is BAD.

[@tomaszoslizo9226]

There is dynamic eq with audyssey which then adapt the curve to the volume level. This solutin is missing form dirac.

[@peetiegonzalez1845]

I always do the Audyssey first, then tweak it to get a sound I like.

[@peetiegonzalez1845]

Which invariably means cranking up the low-end a little. Whilst taking care not to get resonant boom.

[@Nickceid]

Hey Joe great comparison! I have the Nad t778 paired to the Svs prime bookshelves. I have tried both Nad target curve and Dirac target curve. I noticed that in movies I prefered the nad target curve but in 2 channel stereo the dirac target curve had a little more live sound. What is your opinion? Do you think full range correction worths the extra money?

[@joentell]

I'm still figuring it out myself. For movies I typically like the NAD curve. For music DIRAC target curve.

[@Nickceid]

@@joentell Exactly! We have the same ears 😂 I wonder how better movies and music can sound. This receiver combined with the dirac is the best thing I have heard. I was wondering what will happen if I upgraded my speakers to something else.I still don't know what this something else could be. Maybe Focal or Arendal..

[@joentell]

@@Nickceid I keep telling people that DIRAC is no joke. The sound will improve with better speakers. I look for speakers with smooth directivity as they perform better when DSP EQ is applied.

[@Nickceid]

@@joentell Could you please give an example of speakers with smooth directivity?

[@joentell]

@@Nickceid typically KEF and Revel. Genelecs. Kali Audio. When looking at a spinorama graph, what you're looking for is a DI directivity index that doesn't change slope drastically. After looking at enough speakers, you can see what is good, great, and bad.

[@adamjj85]

I would assume Dirac would pull more ahead of Audyssey if the measurements were taken further away and the room played more of a part. Then its impulse correction would shine. I was impressed just running the trial on my PC in my office. Cleared up things that REW and Equalizer APO could not fix.

[@joentell]

I would need to show the difference in impulse responses using impulse response graphs. I think the obvious difference would be more in how they level match and how the phase is aligned between speakers and the sub(s).

[@tomaszoslizo9226]

@@joentell do not expect phase correction from DL below 100Hz. It would require long FIR filters. NAD most likely uses filters below 8k samples. It's one of the reason why DL uses hybrid filters - IIR for low freq and FIR for mid and high. Phase changes can be visible due to amplitude correction with IIR however it's not a phase correction.

[@joentell]

@@tomaszoslizo9226 I knew about the mixed-phade filters but your explanation helps to clarify. Thank you!🙏

[@JerryRutten]

@@joentell You can definitely do phase correction with IIR, but not independently from amplitude correction. If you flatten out an amplitude response, you also flatten out the phase response (when the signal is minimum phase), but you cannot correct reflections (which FIR can). Dirac also resolves time alignment issues. I think that the time and phase correction is the strong point of Dirac. What Dirac does for the sound image is amazing. If your room is not perfectly symmetrical or your furniture arrangement is not perfectly symmetrical, Dirac will help. Also what Dirac does to room modes, not only in the listening position, but also elsewhere in the room is stunning, especially with a multipoint measurement. Another nice “feature” is that the target curve is just a file. I made a spreadsheet to adjust the tilt of the curve, to adjust the balance of the midrange compared to the lows and highs and to adjust the level of certain areas, for example to create more or less depth in the soundstage. The default target curve is only a starting point and most often not even that!

[@NirreFirre]

@@JerryRutten this is imho the comment to be lifted by Joe and perhaps even Dirac themselves, Dirac's USP a bit too scientific o enginerdy for most audio and music shoppers and being Swedes they don't get the points across very well IMO. I do live in Sweden, studied at Uppsala University and understand the basics of signal processing in the time domain but you put some more positive words on the implications that capability adds. For example. Is it correct to say Audyssey is just operating in that limited FR (amplitude) domain subset and only have intra speakers delays when it comes to time? I would like Joe to read up and really get gritty on these details since I think the dsp side of audio will take over most markets (small BT speakers, every new car audio, headphones, in stores and at sporting events audio, and so on)

[@juandoe7802]

I don't understand any of this. I just purchased a Denon X3700 because I couldn't locate a Onkyo TX RZ50 which uses Dirac. I have a normal living room...rather on the small side. Does this room correction really make a difference for small rooms/living rooms? I will be connecting a pair of SVS ultra bookshelves, and Ultra Center, and dual PB3000.... I'm working on rears and eventually heights. I just wanna know if I made a good purchase?

[@joentell]

I prefer Dirac, but you can use the editor app to change the sound of Audyssey. You may want to watch this video I made kzbin.info/www/bejne/f4CoqZ2Flp6qbK8

[@janbenschop7597]

About that correction at 3K... Is the speaker crossed over there?

[@joentell]

Perhaps. That can be found online. It's been a while.

[@Methamill]

How did you get the ranges for the spectrum (sub bass, bass, low bass...) at the top of the spectrum frame?

[@hdmoviesource]

Do you think that you can use the pre-outs on a receiver, go into MiniDSP, then measure with REW and room correct using REW instead of these other room corrections?

[@joentell]

Yes, but it takes a lot of time, knowledge, and experience to do a better job. Lots of room for error.

[@hdmoviesource]

@@joentell Yeah. Do you know if you can use multiple EQ measurements on one channel using MiniDSP? So, bass up to 250hz, then 2 midrange measurements, and then one tweeter. So 4 measurements. I wonder if we can use 4 at once?

[@joentell]

@@hdmoviesource I know what you're saying. Short answer, no. That's done often when someone uses these to design active speakers with multiple drivers, but a 3-way would take up 3-channels even though it's a single speaker. BUT, if you want, you can do as many measurements as you want in REW for various frequency ranges, then create a combined correction filter that you would put into one of the channels on the MiniDSP

[@hdmoviesource]

@@joentell that's pretty interesting. I could start slow and just see if I can calibrate stereo speakers that way. Like you say, it's going to take longer, but I bet the final result would be more accurate.

[@keepingupwiththejones2933]

I wonder why Audyssey continues to use a flat target. People's reactions seem to be more positive to room correction software that makes the sound livelier and warmer.

[@joentell]

I think a slight downard tilt is more correct

[@SealedOrPorted]

I used Audyssey when I began this hobby in the 90’s and never liked the sound - always fiddling. Even with XT32, I could never find a setting I liked. Once the Anthem MRX-1120 came out, I migrated to ARC and found a sound I liked better. I have yet to try Dirac, but have learned ARC enough to get a pleasing response similar to the tilt and gains afforded by Dirac. The miniDSP really made the best impact in sound so far in my room. Hoping to be able to audition Dirac some day.

[@peaceandrelaxationwithgodscrea]

How about a receiver leader board?

[@joentell]

:-) I don't review enough receivers. And there's way too many factors aside from just sound quality.

[@NakkiranSunassee]

Do you plug the mic in the AVR or in your computer?

[@joentell]

Computer.

[@manumichael4424]

Which AVR is good for bose accoustimas 10 series v Marantz or denon Mostly I'm using for streaming music 70%music 30%movies

[@peetiegonzalez1845]

Either of those would work. They're the same company, after all. More importantly I would definitely recommend using Audyssey (or Dirac) with those speakers. Bose are notorious for using DSP to correct the shortcomings in their speakers.

[@MasterApex1]

This room correction is worthwhile if you are have a lot if "room mode" effecting the response. For HiFi listening to SACD/High res recording with good recording from 2 channel stereo on highly resolving speakers/amps (B&W 802D3 / Levinson)...in my room, it messed up with the good coherency and soundstage of the recording. For multi-channel sound track where the stereo soundstage is not there anyway, I found that correcting up to 300Hz is all I need to fix the bass "room" mode...beyond that I prefer to hear the direct sound from the speakers...the correction seems to mess with group delay so it really not good for stereo soundstage.

[@NakeanWickliff]

I’m curious what would’ve happened if you turn on dynamic EQ. I know that odyssey reference is set up for reference level listening and I doubt you measured at reference level, I wonder if the dynamic EQ would have brought back some of that bass. Just an idea.

[@joentell]

Dynamic EQ is really helpful for compensating for human hearing. Microphones don't react the same way as how we perceive sound.

[@NakeanWickliff]

@@joentell I guess my point is audyssey is calibrated for reference level where as I think I remember hearing dirac is calibrated for 86db. The mics do not hear differently but like you said our ears do, therefore you would have different slopes for different volume reference points. If you were designing a slope for 86db it would be a different slope that we would design for reference level.

[@drsuresh9]

Very good video... I did this sometime back but didn't thought of doing KZbin video... Good work. Much appreciated.

[@roccobruno8027]

Looks like the customization of Dirac is comparable to Anthem ARC while Audyssey is lagging behind.

[@TylerStout]

Thanks for the video. Don't know if you saw my other comment but I do have some comparison charts between my CSA/miniDSP cal for the UMIK, just wasn't sure how to share with you currently in a google sheet. Wasn't sure if I could share the link. I think another interesting video idea could be taking an anechoic measurement from Erin of a speaker you have and comparing how each room correction suite fairs in removing issues in the room. Obviously you'd want to be using the same target curve and also same microphone placement such as maybe both you only go 1.5 feet from MLP in the suggested pattern the correction recommends. Would be quite a bit of work trying to figure out what you're seeing in terms of what's correctable what isn't and what each are doing.

[@joentell]

That's a tough one because there's different philosophies about what should and shouldn't be corrected for. I do often overlay my in-room measurements over Erin's predicted in-room responses of the same speaker and it's awesome how closely they match consider we're in different states completely, using different measurement mics and different methods, yet you can see the correlation between the two. It's really cool. You can share you findings with me at [e-mail]

[@pauledwards2817]

I don't think people will appreciate this but as a two channel listener, apart from SACD then I don't think any of these DSP solutions really make anything more involving. On the whole for music the subtleties just seem to get lost and replaced with a dead acoustic and so what performance compared to having DSP options on. for music, just turn them off. As for cinema, well the choice less obvious. To me, movies mixes are really quite contrived sound stages constructed wholly in a mixing studio and get away with it because the visuals add so much. Just listen to movie on it's own without the screen on and the illusion gets lost, never have a been even remotely convinced I am in the action right there because the sound mix is hyper over produced. As far as shaping I have a nasty feeling that is is loved, in a similar manner to the old loudness control as a way of suppressing midrange to make it easier to ignore the failings of reproduction where the ear is most sensitive, rather like elevator music does, cut the midrange to make it less annoying and intrusive.

[@joentell]

I'm currently using DIRAC for my 2 channel system and I think it helps a lot. It's actually a 2.1 system with a sub and DIRAC does a great job blending the sub so it sounds like one cohesive system.

[@pauledwards2817]

@@joentell I think it is nice it has worked well for you. Perhaps I am a bit set in my ways. I have used REW quite a lot just really to experiment with levels and response. When I did the measurements this kind of supported what the receivers were saying, not much needed to be done and of the combinations I can put together the most enjoyable is very close to as flat as is reasonable with or without a DSP, and a slight tilt upwards perhaps but that helps my old and abused ears. The sub is just used for movies and the few SACDs that support one. Perhaps I should open my ears, and must confess I am not using the best iterations of the technologies as I have gone for older used equipment to be able to get near top of the range models at much lower cost.

[@deepcuttabeats356]

Awesome analysis

[@mattgiunt]

Excellent comparison.Would love to see what Dirac does on a Yamaha receiver.

[@danielmiller469]

Not as much of a difference as I thought there would be

[@TylerStout]

I mean there's only so much you can extrapolate with just looking at target curve. I would wager to bet the impulse response would be the big difference you were looking for.

[@ubacow7109]

Can you also do a comparison video with Dirac and ARC or Audessey & the latest YPAO?

[@thelonewolf666]

graphs dont mean anything--- it what it sounds like that is important

[@joentell]

If you understand graphs, you'll understand how what you hear and what the graph shows are correlated. If you don't understand the graphs, then it's a good idea to learn. If you're unwilling, you can resort to trusting your ears I guess. It doesn't matter to me.

[@joelopez7459]

Dang..........wonder if it's all hype.

[@joentell]

If what is?

[@mattknoff]

This is just a bad test, I gained nothing from this except you need 2 speakers to calibrate.

[@joentell]

And I gained nothing from this comment. Explain why it's a bad test.

[@mattknoff]

@@joentell Your setup seems like you are not taking this seriously. I'm not saying I know more than you or anything, but I do believe Dirac is way more suited to a surround setup than it is just placing it right in front of a speaker. If the system need two speakers to properly adjust them then I would place the other speaker literally anywhere else besides on the ground next to you. Does this make sense?

[@seanlacroix]

There's no way that Bookshelf speaker reaches down to mid 30 he.

[@joentell]

It definitely does in room. Dirac detects it as a speaker that can go down to 36hz -3dB in-room. I've measured it numerous times in different locations.