Thoughts on Dirac Correction for a Less-than Loudspeaker ?

[Eric Natural]

I guess my question is : can something like Dirac make a purse out of a sow's ear ? Can a slightly technically-flawed or inexpensive loud be made more ideal w/correction ?
Second question - izzit pronounced DIE-rac or DEE-rac : = )

 

[Power Pop 23]

I think it depends on what flaws need to be corrected but is possible. It is my understanding dips can be boosted only so much while peaks can be reduced more readily.

Wikipedia say 'dih-Rak' is the pronunciation for the last name of Paul Dirac - after whom the software is named.

Paul Dirac - Wikipedia

en.wikipedia.org

 

[ZolaIII]

If it's truly broken it can't. By that I mean even moderate deeps and physical limitations. Like broken or bad crossovers, fluffy let too far woofers and so on. It can make a half deacent speakers that aren't broken sound much, much better (actually most in such particular case) and with a lot of help from sub's of course.
Edit: that's how ratings work on Spinorama with ideal sub and EQ what's again ideally achievable but that's vogue example.
To really understand the essence watch hosts videos about speaker measurements and in generally and take your time.

Understanding Speaker Measurements (Video)

Per popular demand, I recorded a long (1 hour) video on how to read speaker measurements but also, the underlying research (and physics) that mandate some of those concepts. Also covered is equalization and listening tests (briefly). As usual, my videos are "one take" directly recorded and...

www.audiosciencereview.com

 

[kyuu]

The issue is that what Dirac and other room-correction systems are measuring is the in-room response. This works fine for correcting response below Schroeder (generally 200-300Hz in most rooms). However, above that you can't really correct the speaker's response by way of an in-room measurement, as the measurement includes not just the speaker's response but all the reflections as well as sources of noise in the room. Any corrections of a speaker's frequency response should be based on its anechoic response (as you get in speaker reviews on this site among other sources).

Having said that, some people seem to have good experiences using full-range correction. Others do not. So in practice, the answer is likely that it depends on one's specific equipment, environment, and wetware.

Edit: And I think it's pronounced more like DEER-ACK.

 

[Laika2]

I pronounce it the first way, the answer is normally no, massive corrections normally don't turn hohum into a high end speaker, that said will it sound better, likely yeas, quite a bit, especially if you have DIRAC LBC and a sub. It's always made real improvements to my ears but not end game improvements.

 

[ZolaIII]

I pronounce it the first way, the answer is normally no, massive corrections normally don't turn hohum into a high end speaker, that said will it sound better, likely yeas, quite a bit, especially if you have DIRAC LBC and a sub. It's always made real improvements to my ears but not end game improvements.

End game is just a myth same as unicorn. However objective good is close enough. If you think you will be able to clearly distinguish very good system scoring rather high (above 8) to one scoring little more in double blind test, well think again.

 

[Laika2]

End game is just a myth same as unicorn. However objective good is close enough. If you think you will be able to clearly distinguish very good system scoring rather high (above 8) to one scoring little more in double blind test, well think again.

My response to turning a sows ear into a end game speaker, either subjective by one's owns ears or objectively through measurements appears to be the question at hand. I made no comment about endgame speakers in general, only in the context one can not change a poor speaker into a endgame speaker with processing but can make a serious improvement.

" something like Dirac make a purse out of a sow's ear"

I really think we are saying the same thing. End game speakers to me is a term like purse and sows ear. My end game speaker is where I stop, not the best speaker ever made. Its the end of the game in my speaker hunt. It's where I am totally happy. (Though after 15 to 20 setups in 30 years finding my end game is and will be a Unicorn, there's always some frequency area I would like to improve)

 

[ZolaIII]

@Laika2 you can turn something that many hire consider broken to almost ideal but they are wrong that it's broken in the first place. I don't talk about Dirac but human brain use to do it with everything comprehensive DSP have to offer. As long as it's a good capable speaker with out obvious flows. But you really don't want to do it or become speaker designer for that manner so you follow the path of less resistance as possible to have it there with least fiddling by your self to get it there as possible. Trends as it goes are; close/open buffle design sub's, cardioid alike response, DSP all crossover, speaker and room correction and so on but no end.

 

[Laika2]

@Laika2 As long as it's a good capable speaker with out obvious flows.

That's where we differ, though vague a sows ear does not sound like a good capable speaker without obvious flaws.

 

[Keith_W]

The issue is that what Dirac and other room-correction systems are measuring is the in-room response. This works fine for correcting response below Schroeder (generally 200-300Hz in most rooms). However, above that you can't really correct the speaker's response by way of an in-room measurement, as the measurement includes not just the speaker's response but all the reflections as well as sources of noise in the room. Any corrections of a speaker's frequency response should be based on its anechoic response (as you get in speaker reviews on this site among other sources).

This is the correct answer. Loudspeakers can be corrected by DSP, provided the correct measurements are taken (i.e. free from room influence!) and you know what you are doing. Rooms can only be partially corrected by DSP - i.e. the frequencies below Schroder.

Dirac works by taking measurements from multiple positions and averaging them. Whilst this is better than a single measurement at the listening position, it is not loudspeaker correction. It may improve matters, or it may not.

 

[kemmler3D]

As others have said, using Dirac in particular is technically meant to correct the influence of the room, not issues with the speaker, but it does a little of both.

EQ / DSP in general can (IMO) take a speaker one notch up in quality. You can go from bad to okay, or good to great, but not usually more than that.

Speakers don't usually have big flaws that are correctable with EQ without having other uncorrectable flaws, they tend to have common causes.

Think of it like color correction for a TV. It can't make the TV bigger, brighter, or sharper, it can't fix dead pixels, and it can't get rid of glare on the screen. But it can get the TV working better within its basic capabilities.

 

[Eric Natural]

Very helpful, thank you all. I'm taking away that this type of processing is very powerful/impactful and seems critical to achieve serious fidelity in our flawed listening environments. My anecdote : it's taken what were pretty good-sounding louds and made them better !
Someone above mentioned endgame and drc sounds like it is one.

 

[bluefuzz]

can something like Dirac make a purse out of a sow's ear

In a word, yes. Obviously, miracles don't happen, but in my experience with Dirac on 4 very different pairs of loudspeakers the improvement has been from small but worthwhile to positively spectacular. I'm currently using a pair of ~$100 2-way plastic bookshelf speakers that are close to unlistenable without EQ together with a DIY sub. With Dirac (fullrange) I can't say I've ever hear anything better. Louder perhaps, but not better. I'm constatly amazed how good this setup sounds ...