|
|
Post by leonski on Aug 28, 2021 2:09:45 GMT -5
LC?
Any chance you are just getting BETTER at making such measurements?
Share your proceedure, please.
|
|
|
|
Post by marcl on Aug 28, 2021 4:48:07 GMT -5
I probably have said this before, but the procedure I was using with 3.0.14 no longer works the same way with 3.1.1. It ends up okay with 3.1.1, but Dirac definitely changed some things wiht how they take measurements based on the volume settings and how they calculate the filters. I think Dirac has been changing the measurement process since a year ago when many people complained about test sweeps being too loud and there were reports of blown speakers.
|
|
cawgijoe
Emo VIPs 
"When you come to a fork in the road, take it." - Yogi Berra
Posts: 5,032
|
Post by cawgijoe on Aug 28, 2021 9:21:31 GMT -5
I probably have said this before, but the procedure I was using with 3.0.14 no longer works the same way with 3.1.1. It ends up okay with 3.1.1, but Dirac definitely changed some things wiht how they take measurements based on the volume settings and how they calculate the filters. I think Dirac has been changing the measurement process since a year ago when many people complained about test sweeps being too loud and there were reports of blown speakers. I need to rerun Dirac when I get some quiet time this weekend. As I said in an earlier post, I downloaded the latest Dirac version and imported my last run into the XMC-2. The sound has been muffled especially from the center channel and Iβve had to increase the volume considerably to get good sound. This on Preset 2. I switched to preset 1 which was already in the XMC-2 and the sound quality and volume improved immensely. I also noticed that on Preset 2 the Loudness function didnβt work. No change. I donβt use that function but wanted to mention it. Works fine on 1. Iβll report back once I run Dirac again for Preset 2. Weird. |
|
|
|
Post by marcl on Aug 28, 2021 10:23:49 GMT -5
I probably have said this before, but the procedure I was using with 3.0.14 no longer works the same way with 3.1.1. It ends up okay with 3.1.1, but Dirac definitely changed some things wiht how they take measurements based on the volume settings and how they calculate the filters. I think Dirac has been changing the measurement process since a year ago when many people complained about test sweeps being too loud and there were reports of blown speakers. I need to rerun Dirac when I get some quiet time this weekend. As I said in an earlier post, I downloaded the latest Dirac version and imported my last run into the XMC-2. The sound has been muffled especially from the center channel and Iβve had to increase the volume considerably to get good sound. This on Preset 2. I switched to preset 1 which was already in the XMC-2 and the sound quality and volume improved immensely. I also noticed that on Preset 2 the Loudness function didnβt work. No change. I donβt use that function but wanted to mention it. Works fine on 1. Iβll report back once I run Dirac again for Preset 2. Weird. I noticed a weird change in levels yesterday after having done a full calibration on Wednesday, and set all the levels using REW. This morning I confirmed that everything was out of whack ... rear channels loud and front channels quieter ... like by as much as 6db! I did a cold reboot and confirmed. So I set all levels back to 0.0 and I'm reloading my Dirac filters and will set them again. Weird ... I was very careful to get them within .5db Wednesday and confirmed with REW frequency sweeps. All of a sudden Friday they changed ... though I probably had done a cold reboot somewhere in the middle. |
|
|
|
Post by geebo on Aug 28, 2021 10:49:47 GMT -5
I need to rerun Dirac when I get some quiet time this weekend. As I said in an earlier post, I downloaded the latest Dirac version and imported my last run into the XMC-2. The sound has been muffled especially from the center channel and Iβve had to increase the volume considerably to get good sound. This on Preset 2. I switched to preset 1 which was already in the XMC-2 and the sound quality and volume improved immensely. I also noticed that on Preset 2 the Loudness function didnβt work. No change. I donβt use that function but wanted to mention it. Works fine on 1. Iβll report back once I run Dirac again for Preset 2. Weird. I noticed a weird change in levels yesterday after having done a full calibration on Wednesday, and set all the levels using REW. This morning I confirmed that everything was out of whack ... rear channels loud and front channels quieter ... like by as much as 6db! I did a cold reboot and confirmed. So I set all levels back to 0.0 and I'm reloading my Dirac filters and will set them again. Weird ... I was very careful to get them within .5db Wednesday and confirmed with REW frequency sweeps. All of a sudden Friday they changed ... though I probably had done a cold reboot somewhere in the middle. I usually do a cold reboot after running Dirac because I've seen similar results that you saw. I think that maybe changing to to other preset and then back may accomplish the same thing. Sometime ago I had noticed the center image after running Dirac was pretty bad and required some rather large level changes but after a reboot I had to change levels again and that time they were all within 1db of each other and the center image was good. So I always reboot before adjusting post Dirac levels. |
|
|
|
Post by marcl on Aug 28, 2021 11:32:32 GMT -5
I noticed a weird change in levels yesterday after having done a full calibration on Wednesday, and set all the levels using REW. This morning I confirmed that everything was out of whack ... rear channels loud and front channels quieter ... like by as much as 6db! I did a cold reboot and confirmed. So I set all levels back to 0.0 and I'm reloading my Dirac filters and will set them again. Weird ... I was very careful to get them within .5db Wednesday and confirmed with REW frequency sweeps. All of a sudden Friday they changed ... though I probably had done a cold reboot somewhere in the middle. I usually do a cold reboot after running Dirac because I've seen similar results that you saw. I think that maybe changing to to other preset and then back may accomplish the same thing. Sometime ago I had noticed the center image after running Dirac was pretty bad and required some rather large level changes but after a reboot I had to change levels again and that time they were all within 1db of each other and the center image was good. So I always reboot before adjusting post Dirac levels. Yes, exactly ... depending on the phase of the moon, of course. Sometimes I turn the lights out when I reboot. Can you tell? I. M. SOFA. KING. TIRED of the hoops we have to jump through just to get this machine to play the same volume twice in a row! And the Levels being useless to actually get the speaker levels aligned. And. And. And. That volume change was spontaneous yesterday. I had listened a lot on Thursday after recalibrating on Wednesday. Everything was fine. Between Thursday and Friday I did not recalibrate or change any levels. I put some multichannel music on and all of a sudden the surrounds are loud! Same with 5.1 video. So I just reloaded Dirac filters, then did a cold reboot. Came up Init:EDA2 error so I pulled all the HDMI and line cord, let it sit and brought it up again. Set all the Levels to 0.0. Measured levels with REW full range pink noise and the SPL meter. Adjusted Levels based on those measurements to get all channels equal. And in between - whenever I went into Settings - I cycled the audio mode through and back to Surround. BTW, since REW doesn't measure tops I use the Left channel as a reference and set the tops using Levels ... best I can do. So now it's all leveled out again ... until .....  |
|
|
|
Post by geebo on Aug 28, 2021 11:46:49 GMT -5
I usually do a cold reboot after running Dirac because I've seen similar results that you saw. I think that maybe changing to to other preset and then back may accomplish the same thing. Sometime ago I had noticed the center image after running Dirac was pretty bad and required some rather large level changes but after a reboot I had to change levels again and that time they were all within 1db of each other and the center image was good. So I always reboot before adjusting post Dirac levels. Yes, exactly ... depending on the phase of the moon, of course. Sometimes I turn the lights out when I reboot. Can you tell? I. M. SOFA. KING. TIRED of the hoops we have to jump through just to get this machine to play the same volume twice in a row! And the Levels being useless to actually get the speaker levels aligned. And. And. And. That volume change was spontaneous yesterday. I had listened a lot on Thursday after recalibrating on Wednesday. Everything was fine. Between Thursday and Friday I did not recalibrate or change any levels. I put some multichannel music on and all of a sudden the surrounds are loud! Same with 5.1 video. So I just reloaded Dirac filters, then did a cold reboot. Came up Init:EDA2 error so I pulled all the HDMI and line cord, let it sit and brought it up again. Set all the Levels to 0.0. Measured levels with REW full range pink noise and the SPL meter. Adjusted Levels based on those measurements to get all channels equal. And in between - whenever I went into Settings - I cycled the audio mode through and back to Surround. BTW, since REW doesn't measure tops I use the Left channel as a reference and set the tops using Levels ... best I can do. So now it's all leveled out again ... until ..... View AttachmentNever got the EDA2 error (knock on wood) but I usually set levels to 0 before or after running Dirac and before adjusting after a reboot. I use the internal RMC levels and except for the bass they seem to do a good job. Unfortunately a further adjustment of the subs of 6 ~ 7 dB is necessary but the end result has been satisfying for both music and Movies. (For movies I use my second sub as LFE only as Preset 2 but for music I use the main SVS SB-16 only as Preset 1). |
|
okjazz
Minor Hero
Posts: 55
Member is Online
|
Post by okjazz on Aug 28, 2021 11:56:18 GMT -5
I usually do a cold reboot after running Dirac because I've seen similar results that you saw. I think that maybe changing to to other preset and then back may accomplish the same thing. Sometime ago I had noticed the center image after running Dirac was pretty bad and required some rather large level changes but after a reboot I had to change levels again and that time they were all within 1db of each other and the center image was good. So I always reboot before adjusting post Dirac levels. Yes, exactly ... depending on the phase of the moon, of course. Sometimes I turn the lights out when I reboot. Can you tell? I. M. SOFA. KING. TIRED of the hoops we have to jump through just to get this machine to play the same volume twice in a row! And the Levels being useless to actually get the speaker levels aligned. And. And. And. That volume change was spontaneous yesterday. I had listened a lot on Thursday after recalibrating on Wednesday. Everything was fine. Between Thursday and Friday I did not recalibrate or change any levels. I put some multichannel music on and all of a sudden the surrounds are loud! Same with 5.1 video. So I just reloaded Dirac filters, then did a cold reboot. Came up Init:EDA2 error so I pulled all the HDMI and line cord, let it sit and brought it up again. Set all the Levels to 0.0. Measured levels with REW full range pink noise and the SPL meter. Adjusted Levels based on those measurements to get all channels equal. And in between - whenever I went into Settings - I cycled the audio mode through and back to Surround. BTW, since REW doesn't measure tops I use the Left channel as a reference and set the tops using Levels ... best I can do. So now it's all leveled out again ... until ..... View AttachmentMarcl, the SPL graph for your center speaker shows quite some sounds level variation (not flat) between 400 Hz and 5kHz, especially with a nasty drop at 2kHz. Are you sure you are not missing some dialogs on that center channel because of that 2kHz level drop? |
|
|
|
Post by marcl on Aug 28, 2021 12:21:51 GMT -5
Yes, exactly ... depending on the phase of the moon, of course. Sometimes I turn the lights out when I reboot. Can you tell? I. M. SOFA. KING. TIRED of the hoops we have to jump through just to get this machine to play the same volume twice in a row! And the Levels being useless to actually get the speaker levels aligned. And. And. And. That volume change was spontaneous yesterday. I had listened a lot on Thursday after recalibrating on Wednesday. Everything was fine. Between Thursday and Friday I did not recalibrate or change any levels. I put some multichannel music on and all of a sudden the surrounds are loud! Same with 5.1 video. So I just reloaded Dirac filters, then did a cold reboot. Came up Init:EDA2 error so I pulled all the HDMI and line cord, let it sit and brought it up again. Set all the Levels to 0.0. Measured levels with REW full range pink noise and the SPL meter. Adjusted Levels based on those measurements to get all channels equal. And in between - whenever I went into Settings - I cycled the audio mode through and back to Surround. BTW, since REW doesn't measure tops I use the Left channel as a reference and set the tops using Levels ... best I can do. So now it's all leveled out again ... until ..... View AttachmentMarcl, the SPL graph for your center speaker shows quite some sounds level variation (not flat) between 400 Hz and 5kHz, especially with a nasty drop at 2kHz. Are you sure you are not missing some dialogs on that center channel because of that 2kHz level drop? That was with Psychoacoustic Smoothing which may hide or exaggerate certain things. Here's a case where - in the mid-upper frequencies - I'll agree with what many people say ... the narrow peaks and dips will change if you move the mic a foot. That said, I don't exactly know why the center behaves like that, but it sounds really good! Here's the reality of measurements above the Schroeder Frequency ... 1/48 octave smoothing this time ...  |
|
okjazz
Minor Hero
Posts: 55
Member is Online
|
Post by okjazz on Aug 28, 2021 12:34:17 GMT -5
Marcl, the SPL graph for your center speaker shows quite some sounds level variation (not flat) between 400 Hz and 5kHz, especially with a nasty drop at 2kHz. Are you sure you are not missing some dialogs on that center channel because of that 2kHz level drop? That was with Psychoacoustic Smoothing which may hide or exaggerate certain things. Here's a case where - in the mid-upper frequencies - I'll agree with what many people say ... the narrow peaks and dips will change if you move the mic a foot. That said, I don't exactly know why the center behaves like that, but it sounds really good! Here's the reality of measurements above the Schroeder Frequency ... 1/48 octave smoothing this time ... View AttachmentNice to hear that it sounds really good for you. That is the ultimate goal we all seek. When it come to displaying SPL graphs, I normally use the 1/12 octave smoothing. I read somewhere that the 1/12 smoothing is what is the closest to human hearing representation. It do not know what it truly means, but that has been my preferred smoothing since. |
|
|
|
Post by AudioHTIT on Aug 28, 2021 12:36:51 GMT -5
Iβve just noticed something of interest to this thread. Dirac has now been broken out on all of the processors, in other words you can now buy them with or without it, and the value is $200.
So the RMC-1 & 1L can be had for $200 less without Dirac (or the same price with), the XMC-2 comes without Dirac at the original price, and is $200 more with Dirac.
|
|
|
|
Post by marcl on Aug 28, 2021 12:42:38 GMT -5
That was with Psychoacoustic Smoothing which may hide or exaggerate certain things. Here's a case where - in the mid-upper frequencies - I'll agree with what many people say ... the narrow peaks and dips will change if you move the mic a foot. That said, I don't exactly know why the center behaves like that, but it sounds really good! Here's the reality of measurements above the Schroeder Frequency ... 1/48 octave smoothing this time ... View AttachmentNice to hear that it sounds really good for you. That is the ultimate goal we all seek. When it come to displaying SPL graphs, I normally use the 1/12 octave smoothing. I read somewhere that the 1/12 smoothing is what is the closest to human hearing representation. It do not know what it truly means, but that has been my preferred smoothing since. I generally use Variable and Psychoacoustic.. Variable declutters the upper frequencies, while retaining bass detail where you need to know what the room is doing. Psychoacoustic uses different techniques to simulate our perception. Per the REW site .... "Variable smoothing applies 1/48 octave below 100 Hz, 1/3 octave above 10 kHz and varies between 1/48 and 1/3 octave from 100 Hz to 10 kHz, reaching 1/6 octave at 1 kHz. Variable smoothing is recommended for responses that are to be equalised." "Psychoacoustic smoothing uses 1/3 octave below 100Hz, 1/6 octave above 1 kHz and varies from 1/3 octave to 1/6 octave between 100 Hz and 1 kHz. It also applies more weighting to peaks by using a cubic mean (cube root of the average of the cubed values) to produce a plot that more closely corresponds to the perceived frequency response." |
|
|
|
Post by geebo on Aug 28, 2021 12:45:18 GMT -5
Iβve just noticed something of interest to this thread. Dirac has now been broken out on all of the processors, in other words you can buy them with or without it, and the value is $200. So the RMC-1 & 1L can be had for $200 less without Dirac (or the same price with), the XMC-2 comes without Dirac at the original price, and is $200 more with Dirac. Interesting. And it raises the question will one be able to purchase Dirac if a processor was originally purchased without? |
|
okjazz
Minor Hero
Posts: 55
Member is Online
|
Post by okjazz on Aug 28, 2021 13:14:38 GMT -5
Nice to hear that it sounds really good for you. That is the ultimate goal we all seek. When it come to displaying SPL graphs, I normally use the 1/12 octave smoothing. I read somewhere that the 1/12 smoothing is what is the closest to human hearing representation. It do not know what it truly means, but that has been my preferred smoothing since. I generally use Variable and Psychoacoustic.. Variable declutters the upper frequencies, while retaining bass detail where you need to know what the room is doing. Psychoacoustic uses different techniques to simulate our perception. Per the REW site .... "Variable smoothing applies 1/48 octave below 100 Hz, 1/3 octave above 10 kHz and varies between 1/48 and 1/3 octave from 100 Hz to 10 kHz, reaching 1/6 octave at 1 kHz. Variable smoothing is recommended for responses that are to be equalised." "Psychoacoustic smoothing uses 1/3 octave below 100Hz, 1/6 octave above 1 kHz and varies from 1/3 octave to 1/6 octave between 100 Hz and 1 kHz. It also applies more weighting to peaks by using a cubic mean (cube root of the average of the cubed values) to produce a plot that more closely corresponds to the perceived frequency response." Thank you for the smoothing info. From now on, I will be Variable smoothing if I have to generate EQ in REW, and Psychoacoustic to assess deep and peak as they can be perceived by human hear. So if one could know the type of smoothing done by Dirac, it would be easier to compare it to the REW measurements smoothed out equally/similarly. Right? Regardless of course of the phases realignment. |
|
|
|
Post by AudioHTIT on Aug 28, 2021 13:15:07 GMT -5
Iβve just noticed something of interest to this thread. Dirac has now been broken out on all of the processors, in other words you can buy them with or without it, and the value is $200. So the RMC-1 & 1L can be had for $200 less without Dirac (or the same price with), the XMC-2 comes without Dirac at the original price, and is $200 more with Dirac. Interesting. And it raises the question will one be able to purchase Dirac if a processor was originally purchased without? Right, Iβd expect the answer is yes, though maybe at a premium? It seems to be following the model of some other manufacturers, who support it on the platform, but require you to buy it separately. It does allow the RMCs to be advertised at a lower price which may be attractive to some. |
|
|
|
Post by marcl on Aug 28, 2021 13:23:36 GMT -5
I generally use Variable and Psychoacoustic.. Variable declutters the upper frequencies, while retaining bass detail where you need to know what the room is doing. Psychoacoustic uses different techniques to simulate our perception. Per the REW site .... "Variable smoothing applies 1/48 octave below 100 Hz, 1/3 octave above 10 kHz and varies between 1/48 and 1/3 octave from 100 Hz to 10 kHz, reaching 1/6 octave at 1 kHz. Variable smoothing is recommended for responses that are to be equalised." "Psychoacoustic smoothing uses 1/3 octave below 100Hz, 1/6 octave above 1 kHz and varies from 1/3 octave to 1/6 octave between 100 Hz and 1 kHz. It also applies more weighting to peaks by using a cubic mean (cube root of the average of the cubed values) to produce a plot that more closely corresponds to the perceived frequency response." Thank you for the smoothing info. From now on, I will be Variable smoothing if I have to generate EQ in REW, and Psychoacoustic to assess deep and peak as they can be perceived by human hear. So if one could know the type of smoothing done by Dirac, it would be easier to compare it to the REW measurements smoothed out equally/similarly. Right? Regardless of course of the phases realignment. Here's a comparison I did a month ago.   |
|
|
|
Post by marcl on Aug 28, 2021 13:30:01 GMT -5
Interesting. And it raises the question will one be able to purchase Dirac if a processor was originally purchased without? Right, Iβd expect the answer is yes, though maybe at a premium? It seems to be following the model of some other manufacturers, who support it on the platform, but require you to buy it separately. It does allow the RMCs to be advertised at a lower price which may be attractive to some. The interface kit is listed separately for $199 and it says it's required to use Dirac with the processors. One would assume you get the kit if you opt in for Dirac when you buy the processor. It's a simple way to sell you the upgrade later by just selling you the interface kit and registering your Mac. |
|
okjazz
Minor Hero
Posts: 55
Member is Online
|
Post by okjazz on Aug 28, 2021 19:22:15 GMT -5
Thank you for the smoothing info. From now on, I will be Variable smoothing if I have to generate EQ in REW, and Psychoacoustic to assess deep and peak as they can be perceived by human hear. So if one could know the type of smoothing done by Dirac, it would be easier to compare it to the REW measurements smoothed out equally/similarly. Right? Regardless of course of the phases realignment. Here's a comparison I did a month ago. View AttachmentView AttachmentTo my amateur eyes, the Dirac smoothing on your graphs seems more accentuated than the 1/48 smoothing of REW: the peaks and low points, including the slopes (rises and falls) are more exaggerated in Dirac smoothing. This is telling that Dirac smoothing may be higher than the 1/48. |
|
|
|
Post by omuracada on Aug 28, 2021 21:54:36 GMT -5
Quick question. Should I set the levels of all speakers to zero before running Dirac?
|
|
LCSeminole
Global Moderator 
Res firma mitescere nescit.
Posts: 20,830
|
Post by LCSeminole on Aug 28, 2021 21:58:05 GMT -5
Yes, exactly ... depending on the phase of the moon, of course. Sometimes I turn the lights out when I reboot. Can you tell? I. M. SOFA. KING. TIRED of the hoops we have to jump through just to get this machine to play the same volume twice in a row! And the Levels being useless to actually get the speaker levels aligned. And. And. And. That volume change was spontaneous yesterday. I had listened a lot on Thursday after recalibrating on Wednesday. Everything was fine. Between Thursday and Friday I did not recalibrate or change any levels. I put some multichannel music on and all of a sudden the surrounds are loud! Same with 5.1 video. So I just reloaded Dirac filters, then did a cold reboot. Came up Init:EDA2 error so I pulled all the HDMI and line cord, let it sit and brought it up again. Set all the Levels to 0.0. Measured levels with REW full range pink noise and the SPL meter. Adjusted Levels based on those measurements to get all channels equal. And in between - whenever I went into Settings - I cycled the audio mode through and back to Surround. BTW, since REW doesn't measure tops I use the Left channel as a reference and set the tops using Levels ... best I can do. So now it's all leveled out again ... until ..... View AttachmentNever got the EDA2 error (knock on wood) but I usually set levels to 0 before or after running Dirac and before adjusting after a reboot. I use the internal RMC levels and except for the bass they seem to do a good job. Unfortunately a further adjustment of the subs of 6 ~ 7 dB is necessary but the end result has been satisfying for both music and Movies. (For movies I use my second sub as LFE only as Preset 2 but for music I use the main SVS SB-16 only as Preset 1). This is also my procedure before/after running Dirac. I ran Dirac again today, this time the 17pt variety, and I still think this latest v3.1.1 is processing the filters faster, and just all around a smoother run...........or maybe this is all just my mind playing tricks, but the outcome is highly pleasing to my ears!  |
|
