Deeper level questions on current state of speaker calibration Answered

Comments

7 comments

  • Avatar
    Kārlis Stenders (Edited )

    Hi Dave, 

    Yes, the results will have slight differences depending on the level you're measuring at. We try our best to make sure the calibration is done at an optimum level, but personal preference is difficult to address. In theory - yes, you will have better results if calibrating at your mixing volume. But it's your own judgement call as to what that level is. The dB meter measurable feature you suggested is worth considering, it could be useful.

    It seems you've done a lot of experimenting already, which is what I'd suggest in the first place. There are endless amounts of room dimension, treatment, speaker, positioning, etc. possibilities, not to mention personal preferences regarding mixing response, level, and so forth. This makes it extremely difficult for us to advise on fine-tuning on such a tiny level without actually being in your room and evaluating the results ourselves.

    The reality is that you'll be better off trusting your own ears and experimenting with different settings. Luckily you seem very confident in doing this; I'm sure you'll find the best calibration setting yourself and at this point and it seems we'd only be confusing you since you already have assessed so many variations and have your own conclusions. There are fundamental limits to what digital calibration can do too. 

    As for your speakers - I'm pretty sure the drivers in your ADAM's aren't the same. You might want to do some further investigation on this, but I think the actual mid-range drivers in your setting (horizontal) are the inner ones. So you should measure pointing at those and specify the distance between the centers of those too. This could potentially improve the results noticeably. 

    Thanks for your feedback, much appreciated!

    0
    Comment actions Permalink
  • Avatar
    David Snow

    Thank you Karlis,

    I am conscienciously working very hard to use SW and get great results. Things are getting better and better. But I am still noticing a lower mid range mud, just a touch.

    The way I am determining this lower mud and other small things I’d like improved, is by listening to a ton of commercial releases in all genres. If various pro tracks sound a little muddy, well, there is something in my system doing that.

    Now, you and I could go back and forth with many correspondences about how I can pull this slight mud out by changing small things when calibrating. But my goal is clear, and it seems like a lot of unnecessary work to guess at calibration manuevers, compared to simply having a capability IN the program to ALTErRthe EQ.

    If you guys could just allow the end user to manipulate the end result by having user EQ change available, I really think it would solve more problems than its worth to refuse to allow for this feature.

    I totally understand that SW wants the software to be accurate, and doesn’t want end users screwing up the results by goofing with the EQs. But professionals aren’t always dumb. Sometimes a professional sees something the software is missing and wants to adjust for it. The software has done a fabulous job But not a perfect job. As you say, it is a complicated subject.

    But if a true professional notices a 2 db bump at 400 k, that professional should be able to pull down 400 by a couple db.It would be way easer to have an EQing capability than me doing 20 more calibrations, guessing at what I need to do to lower some of the mud.

    I would add a “User EQ” button. And when it is “off”, the system goes back to the EQ curve made from pure analysis. And when it is “on”, you can get a slightly altered version. 

    Right now, I am using the “tilt” function. And I have thoroughly decided it is the best setting available. But that tilt function is way more of a change to the default calibration compared to me just pulling 2 db out of 400 k.So actually, with the way the system is currently programmed, I am altering the calibration EQ way more than I WOULD if I could just make small EQ adjustments. 

    Think about it...:)

    1
    Comment actions Permalink
  • Avatar
    Kārlis Stenders (Edited )

    Thanks, David, 

    All good points! Implementing such a feature however is a different story. I wouldn't want you to get the impression that we're holding back on this because we don't think our average user would be capable of meaningfully using it. In fact quite the opposite - the Reference crowd is usually very competent and capable. 

    But still, the room for error here is huge, so it's extremely difficult to make it work. Again, we ourselves would like to have a feature for more control over EQ, but it has to be bulletproof and so far we haven't finalized it. We will keep trying though, there are always things in development!

    Thank you for your input!

    0
    Comment actions Permalink
  • Avatar
    David Snow (Edited )

    Hi Karlis,

    I kind of feel you guys are just putting off the inevitable. I cannot imagine a future in which Sonar Works does not have user EQ options available. I am almost positive you guys will have to put this in, so why not sooner than later? The ‘bulletproof” aspect of this feature you speak of is a user EQ  “on/off” switch. The analysis file can never be altered or screwed up in this instance. If a user shuts OFF the user EQ, the user get back to the original analysis. If a user wants to be so retarded as to add 50 db of 1k, it will do nothing to screw up anything but the user’s own ears. Which any EQ would do at such stupid settings. SW is not responsible for such idiocy, and should not worry about such. Same thing with ANY adjustment made by a user. Any argument that “the software sucks because my speakers sound bad”, would be obfuscated by just turning off the user’s EQ adjustments and going back to the analysis default file.

     

    So i see too much worry about unrealistic user stupidity, and not enough worry about reality. The reality is, i have to make huge TILS to the EQ instead of having control over individual frequencies. To me, that kind of inaccurate EQ option is something that could more harm the reputation of SW the having individual EQ’s or 8 band parametric EQs. The only no brainer decision that need to be made by SW is “what kind of EQ”? I would recommend 32 band with adjustable frequencies. I have attached  link to a very interesting EQ I would love to see SW implement. Sys EQ doesn’t have to have “matching” capabilities like the EQ. I just like how adjustable the bands are in this EQ. Thanks Dave-----https://www.youtube.com/watch?v=YeXPbIRg4Qk

    1
    Comment actions Permalink
  • Avatar
    Jonas Olofsson

    Hi, I would like to have the option to turn off EQ above a certain frequency.

    There's no idea to try to compensate for comb filtering in the upper register above say 5k cause the filtering will change if you move your head or change your position slightly.

    And I don´t like the idea to have Sonarworks try to compensate for that with eq.

     

    1
    Comment actions Permalink
  • Avatar
    Andreas Dahle Aagård

    Hi. I would like to see an EQ option to use after correctiom, as mentioned in this this thread. It would help a lot. I’ve callibrated my studio approximately 20 times now (after moving around/isolating), but I would like to attenuate more from certain frequenzies than the software says.

    Come on, add an EQ option;)

    Anyway, I’m so glad that I found this Software - will never stop using it.

    -Andreas

    1
    Comment actions Permalink
  • Avatar
    Michael Weinstein

    I mean while waiting around for SW to do "the inevitable" and build in an eq that behaves well with their correction schema, why not insert your favorite eq plugin after SW and adjust to taste?  

    I'm sure a lot of proprietary "Latvian Special Sauce" went into the development of this piece of software.  When I manually linearize my speakers using REW or implement other auto room/speaker correction products I notice many interesting things. First and foremost is in each case the end results all sound different to various degrees.

    SW's main strength, IMHO, are linearizing headphones because, well, they're the only game in town.  But even in Headphone Land there are lingering mysteries: take two headphones with abysmal FRs - like, oh I dunno...Sony 7506's and Audio-T's M40's.  Apparently, what SW is doing is analyzing the rollercoaster ride of those headphones' FR and applying an inverse filter so that the result is a flat (and boring rollercoaster) line. In theory and in plain sight on the SW's window - yay!  

    BUT! The 7506's respond MUCH better to whatever SW's is actually doing. They become reference grade mix check tools after correction. It's remarkable. The AT's? Not so much. This... is weird. And for obvious reasons Rudi or Karlis are not going to publicly explain in any exactitude why this is.

    (But, Latvian friends, feel free to PM/email me if you're feeling subversive ;-)

    Thanks guys!

    0
    Comment actions Permalink

Please sign in to leave a comment.