Correct SPL correction values?

Discussion in 'Archived Threads 2001-2004' started by Ranga, Feb 28, 2002.

  1. Ranga

    Ranga Stunt Coordinator

    Joined:
    Nov 30, 2000
    Messages:
    107
    Likes Received:
    0
    There seems to be some confusion with the SPL correction values. brucek posted these values for 1/6 octave readings. Some posters use the value below these.
    Which is the right one?
    Brucek's values
    10Hz +20.5
    11Hz +18.0
    12.5Hz +16.5
    14Hz +13.5
    16Hz +11.5
    19Hz +8.0
    20Hz +7.5
    22Hz +6.5
    25Hz +5.0
    28Hz +4.0
    31.5Hz +3.0
    36Hz +2.5
    40Hz +2.5
    45Hz +2.0
    50Hz +1.5
    56Hz +1.5
    63Hz +1.5
    71Hz +1.5
    80Hz +1.5
    89Hz +1.5
    100Hz +2.0
    111Hz +1.0
    125Hz +0.5
    142.5Hz +0.5
    160Hz -0.5
    200Hz -0.5
    250Hz +0.5
    315Hz -0.5
    400Hz 0.0
    500Hz -0.5
    630Hz 0.0
    800Hz 0.0
    1.0KHz 0.0
    OR
    Alternate correction values
    20 6.2
    254.4
    31.53
    402
    501.3
    630.8
    800.5
    1000.3
    1250.2
    1600.1
    2000
    2500
    3150
    4000
    5000
    6300
    8000
    1k0
    Also, I find the readings for 1/3 octave using the Stereophile Test CD 2 different from the Stryke Bass One CD. Is this normal?
    Ranga
     
  2. brucek

    brucek Second Unit

    Joined:
    Dec 29, 1998
    Messages:
    335
    Likes Received:
    0
    Ranga,
    Those values are the ones I use, but I certainly have no verification. They seem to be the most used values on fairly reputable sites, but I've seen your other set around also. Actually, I think the actual frequency value between 20Hz and 16Hz should be rounded to 18, not 19. [​IMG]
    brucek
     
  3. ColinM

    ColinM Cinematographer

    Joined:
    Dec 9, 2001
    Messages:
    2,050
    Likes Received:
    0
    Set # 1.

    Don't know what to tell ya abt the 2 CD's.

    - C
     
  4. Glen_L

    Glen_L Stunt Coordinator

    Joined:
    Sep 10, 2000
    Messages:
    86
    Likes Received:
    0
    The chart is in the Software Archive (not sure why not hardware) at the bottom, here's a link. As far as the two CDs, the frequencies may be the same but they may be different in output level. If one measures at, say, 5dB higher than the other it should do so at all frequencies though.
     
  5. Cliff

    Cliff Stunt Coordinator

    Joined:
    Aug 4, 1998
    Messages:
    55
    Likes Received:
    0
    Ranga,

    Your first list of correction values is the one to use. The second list of "alternate correction values" is actually the C weighting curve built into the SPL meter which is user-selected (A or C) and intended, I believe, to allow the SPL meter to operate with a responsivity that approximates the response curve for human hearing.

    Have the SPL meter set for C weighting when you take the measurements, and then apply the first list of correction factors.

    Cliff
     
  6. Ranga

    Ranga Stunt Coordinator

    Joined:
    Nov 30, 2000
    Messages:
    107
    Likes Received:
    0
    Thanks..

    I took a look at the CD covers, and the Stryke Bass CDs are recorded at 0db while the Stereophile CD tones are at -20db.

    Ranga
     
  7. Kerry Hackney

    Kerry Hackney Stunt Coordinator

    Joined:
    Oct 18, 2000
    Messages:
    206
    Likes Received:
    0
    Ranga, Here is a chart with more values filled in that may help. I used MS Excel to do a non-linear progression for the in-between values...

    Frequency-Correction Factor

    10Hz.20.50

    11Hz.18.80

    12Hz.17.23

    12.5Hz.16.50

    13Hz.15.67

    14Hz.14.13

    15Hz.12.75

    16Hz.11.50

    17Hz.10.33

    18Hz.9.29

    19Hz.8.35

    20Hz.7.50

    21Hz.6.92

    22Hz.6.38

    23Hz.5.88

    24Hz.5.42

    25Hz.5.00

    26Hz.4.62

    27Hz.4.27

    28Hz.3.95

    29Hz.3.65

    30Hz.3.38

    31Hz.3.12

    31.5Hz.3.00

    32Hz.2.97

    33Hz.2.91

    34Hz.2.84

    35Hz.2.78

    36Hz.2.72

    37Hz.2.67

    38Hz.2.61

    39Hz.2.55

    40Hz.2.50

    41Hz.2.26

    42Hz.2.18

    43Hz.2.09

    44Hz.1.99

    45Hz.1.90

    46Hz.1.81

    47Hz.1.73

    48Hz.1.66

    49Hz.1.58

    50Hz.1.50

    51Hz.1.50

    52Hz.1.50

    53Hz.1.50

    54Hz.1.50

    55Hz.1.50

    56Hz.1.50

    57Hz.1.50

    58Hz.1.50

    59Hz.1.50

    60Hz.1.50

    61Hz.1.50

    62Hz.1.50

    63Hz.1.50

    63Hz.1.50

    64Hz.1.50

    65Hz.1.50

    66Hz.1.50

    67Hz.1.50

    68Hz.1.50

    69Hz.1.50

    70Hz.1.50

    71Hz.1.50

    72Hz.1.50

    73Hz.1.50

    74Hz.1.50

    75Hz.1.50

    76Hz.1.50

    77Hz.1.50

    78Hz.1.50

    79Hz.1.50

    80Hz.1.50

    81Hz.1.52

    82Hz.1.54

    83Hz.1.57

    84Hz.1.59

    85Hz.1.61

    86Hz.1.64

    87Hz.1.66

    88Hz.1.68

    89Hz.1.71

    90Hz.1.73

    91Hz.1.76

    92Hz.1.78

    93Hz.1.81

    94Hz.1.83

    95Hz.1.86

    96Hz.1.89

    97Hz.1.92

    98Hz.1.94

    99Hz.1.97

    100Hz.2.00

    125Hz.0.50

    160Hz.-0.50

    200Hz.-0.50

    250Hz.0.50

    315Hz.-0.50

    400Hz.0.00

    500Hz.-0.50

    630Hz. 0.00

    800Hz. 0.00

    1000Hz. 0.00

    1250Hz. 0.00

    1600Hz.-0.50

    2000Hz.-1.50

    2500Hz.-1.50

    3150Hz.-1.50

    4000Hz.-2.00

    5000Hz.-2.00

    6300Hz.-2.00

    8000Hz.-2.00

    10000Hz.-1.00

    125000Hz.0.50

    16000Hz.0.00

    20000Hz.1.00
     
  8. JoelO

    JoelO Stunt Coordinator

    Joined:
    Dec 5, 2001
    Messages:
    92
    Likes Received:
    0
    Hi guys,

    Can any of you explain what do you mean by 1/3 or 1/6 octave? Are the sine wave tone generated from that program NCHTONE considered Pink noise? Also, do you really have to set the volume so that your SPL measure 80dB at 1KHz, before taking any measurement (it sounds kinda loud to me)? Or any volume should be ok?

    Thanks!!

    Joel
     
  9. Kerry Hackney

    Kerry Hackney Stunt Coordinator

    Joined:
    Oct 18, 2000
    Messages:
    206
    Likes Received:
    0
    Joel,

    1/3 and 1/6 octave are deviding the frequencies in a given octave in 3 or 6 parts respectively. If you look at the first chart you will see 6 points from 10 to 20 hz which is one octave then you'll see 6 more from 20 to 40 which is the second octave. NCHtone outputs true sine waves at various frequencies. Pink noise is sound at all frequencies at the same time. The main reason to pick 80db is that is "reference" by some standards. As long as the level is high enough that ambient noise doesn't interfere it should not matter. The level could be lower.
     
  10. JoelO

    JoelO Stunt Coordinator

    Joined:
    Dec 5, 2001
    Messages:
    92
    Likes Received:
    0
    Thanks Kerry for the explanation. More question... Is it necessary to do it in that pattern 1/4, 1/5, 1/6 or whatever octave to measure the response? Or can I choose random frequency to measure?

    Joel
     
  11. Kerry Hackney

    Kerry Hackney Stunt Coordinator

    Joined:
    Oct 18, 2000
    Messages:
    206
    Likes Received:
    0
    Joel, There is no real magic in choosing those increments other than it spreads your data points evenly. You could also do it in 5hz steps or even randomly. As long as you have enough readings to determine the response. In the bass octaves you can have pretty wide swings in a small range of tones so, IMHO the more readings the better. YMMV.
     
  12. JoelO

    JoelO Stunt Coordinator

    Joined:
    Dec 5, 2001
    Messages:
    92
    Likes Received:
    0
    Thanks again for the explanation, Kerry!

    I measured my room response using my own created tone sweeps (using NCHTone) in 1/10 octave. The list below is the room response of the sub & my 2 mains speakers (Sorry, I can't post the pix since I can't host it anywhere). I know that some of you measured the sub only, can someone care to explain why do it that way? (Unless you want to tame the peak using BFD or something, but can't you do that having the sub & main running?)

    Hz Response(dB)

    1083.5

    1181.8

    1281.23

    1383.67

    1482.13

    1587.75

    1686.5

    1788.33

    1894.29

    1999.35

    2099

    2299.38

    2498.42

    2696.62

    2894.95

    3093.38

    3293.47

    3493.34

    3694.72

    3893.61

    4089.5

    4485.99

    4882.66

    5292.5

    5697.5

    6093.5

    6489.5

    6889

    7287.5

    7681.5

    8091.5

    8894.68

    9690.89

    10093.5

    12584.5

    16083

    Any comment on how I measured my room response is greatly appreciated.. especially if I'm doing it wrong. Thank you guys.

    Joel
     
  13. brucek

    brucek Second Unit

    Joined:
    Dec 29, 1998
    Messages:
    335
    Likes Received:
    0
    Joel,
    Testing your frequency response is generally carried out using 1/6 octave tones. Any more is uneccessary. You may sometimes use finer tones if you're trying to establish some specific EQ filter data.
    With regard to the sub/mains question, see the following thread for some more information. It may help.
    http://www.hometheaterforum.com/htfo...=BFD+sub+mains
    brucek
     
  14. BruceD

    BruceD Screenwriter

    Joined:
    Apr 12, 1999
    Messages:
    1,220
    Likes Received:
    0
    Joel,

    Depending on the frquency response of your mains, you don't want to measue both sub and mains together.

    Why? because there is likely both destructive cancellation and constructive re-inforcement of certain frequencies between the sub and mains. This makes it very difficult to identify where to actually apply EQ.

    If you wanted to use a parametric EQ to correct for the frequency peaks and/or valleys it is best to start with the sub only. This way you can clearly identify the problem areas and correct for them.

    Then if you want to measure with the sub+mains, you can identify where the main's influence changes things.
     
  15. Kerry Hackney

    Kerry Hackney Stunt Coordinator

    Joined:
    Oct 18, 2000
    Messages:
    206
    Likes Received:
    0
    Depending on how you are crossing over the mains to the sub you will have some interaction, particularly in the area of the crossover. I think it is valuable to do the measurements both with the mains and without. You should also do some experiments with room placement if possible as well as phase adjustment. Phase will effect the response through the crossover point the most. If you decide to add a Behringer eq, remember that you can cut the peaks but should not try to fill in the nulls. You can run your amp out of steam in a hurry without much success in a lot of cases. Trying to fill nulls is usually more problems than benefits. Pull the peaks down and then raise the entire level back to proper calibration usually works well. If you can get the response to within 4-5 db that would be pretty good, down to 3 db of flat would be excellent.
     
  16. JoelO

    JoelO Stunt Coordinator

    Joined:
    Dec 5, 2001
    Messages:
    92
    Likes Received:
    0
    Thanks guys for all your info, I'm starting to understand this stuff now... slowly.. [​IMG]
    I'll take some more measurement this weekend and will keep you guys updated.
    I'm not really looking to buy a BFD for now (no budget).. but will certainly look into it in the future, especially if there're nasty peaks in my rooms.. [​IMG]
    Joel
     

Share This Page