Olive / Harmon stuff

Discussion in 'Measurement Techniques Discussion' started by purr1n, Mar 14, 2019.

  1. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    42,674
    Dislikes Received:
    68
    Trophy Points:
    113
    Location:
    Antarctica
    How was this arrived at? It would seem that the Olive target compared to the SBAF compensation would be upward sloping at 1db per octave with the addition a few db extra from 2kHz onwards based three factors:
    1. The Olive curve below.
    2. The SBAF curve is supposed to be perceptive neutral. This meaning flat on-axis at 1m away which typically translates in-room response to a 1db downward slope (either from 20Hz or 200-300Hz, the latter being the B&K curve)
    3. Note that loudspeakers equalized flat in-room (per below) will sound fricking bright.
    4. Therefore the Olive curve below if applied to speakers in a living room would be upward sloping (fricking bright) with a fuckton of bass below 100-150Hz.
    160503_Blog_AcousticBasisHarmanTargetCurve_Photo_Img5.jpg
     
    Last edited: Mar 14, 2019
  2. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    42,674
    Dislikes Received:
    68
    Trophy Points:
    113
    Location:
    Antarctica
    Here, this is what I mean. My drawing is not accurate, but you should get the idea. Hopefully think gives you guys a better idea why I think Olive is a turd and should shove his consumer preference curve up his ass.

    160503_Blog_AcousticBasisHarmanTargetCurve_Photo_Img5.jpg

    BTW ^ since those are raw measurements, that stuff may not work unless you have Olive's exact measurement rig. That is a GRAS or EARS won't work with those targets. They have different responses.
     
    9suns likes this.
  3. Vtory

    Vtory Illogical Spock

    Friend MZR
    Joined:
    Sep 12, 2016
    Likes Received:
    2,224
    Dislikes Received:
    18
    Trophy Points:
    93
    Location:
    Fort worth, TX
    The following filter set was used to harmanize hd800.

    upload_2019-3-14_19-32-17.png

    According to the guy who proposed this eq, the resulting fr would be (harman2018-compensated) the below:

    upload_2019-3-14_19-32-53.png

    My graph is the calculated fr based on my hd800 (measured with EARS and sbaf-compensation v1.0) and the mentioned eqs applied via REW.

    Btw afaik harman target has several versions. Original Olive target and Harnman 2018 were different -- indeed every version was meaningfully different in bass and treble gains.

    Ref link:
    https://www.dropbox.com/s/v999jbft39506ma/Sennheiser HD800.pdf?dl=0
     
    Last edited: Mar 14, 2019
  4. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    42,674
    Dislikes Received:
    68
    Trophy Points:
    113
    Location:
    Antarctica
    Ahh gotcha. Did Ratory use the same head / torso / ears simulator as Olive? I don't know how the GRAS 43AC compares to the 45AC used by Olive.

    In any event, I don't think the HD800 is a good sample to arrive at an SBAF to Olive mapping. The HD800 is just too screwy with the cup design. FR is way too easily affected by the artificial ear. Might be a good idea to get more data points using more normal headphones like the LCD2C, HD600, etc, and maybe arrive at an average.
     
    Last edited: Mar 14, 2019
    Vtory likes this.
  5. purr1n

    purr1n Finding his inner redneck

    Staff Member Friend BWC
    Joined:
    Sep 24, 2015
    Likes Received:
    42,674
    Dislikes Received:
    68
    Trophy Points:
    113
    Location:
    Antarctica
    Dammit. So his targets are like PS Audio's firmware upgrades? The OE 2018 actually looks decent, except for the bass. Ugh. We might need to wait for the 2022 curve.

    To some extent, the extra bass suggested by OE2018 is explainable: I like the B&K curve which flattens out below 180Hz. The 1db / octave slope target keeps increasing and by 20Hz is about 3db higher than B&K. But even if we use 1db / octave slope instead of B&K, OE2018 still calls for 6db moar bass at 20Hz.

    I just wish Olive would quit or tell people that this is work in progress that he won't get right until 2029. Too many idiots have been using too many of his ever-changing curve targets, often with dissimilar simulators, and getting super oddball results suggesting Audezes need a massive +5db bass increase starting from 200Hz down.
     
    Last edited: Mar 14, 2019
    9suns, FlySweep and monacelli like this.

Share This Page