A couple things I noticed in Dirac that might be helpful.

[gabe]

Hi All. I was running Dirac again, again, again..... Anyway. I am one of those poor souls who loses quite a bit of level with Dirac engaged. I was looking over a very helpful screenshot that tech support sent me regarding setting levels and I stumbled across something that was new, at least to me. I started thinking I might not be choosing the correct speaker as my least efficient to set the master level @ -20db with the mic gain @-37db per the instructions I was given. What I did was leave all the channel sliders all the way up and the master volume quite low. I clicked on my right front height speaker that I always thought would be least efficient because the front heights are reflective and started the test tone. I increased the master output until the right front was @-20db and turned the tone off. NOW, if I click on any channels level control with no test tone playing, I can bring the level down to -20 with no tone playing. After doing so and playing the tone for that channel confirms the accuracy of the setting, at least with my junk. Not curing cancer here, but kinda cool. Also, this almost happened to me so I thought I would share here just as food for thought especially for those who are uncertain if they love their Dirac results. When looking at the person/people in the imaging selection section, remember the seated individual (s) are facing YOU in the picture. In other words, your left is their right and vice versa. If you get it backwards, the microphone placement will be wrong for every location. Maybe this seems stupid, and if so I apologize, but sometimes it is the little things that are hard to catch, at least in my experience.

[ttocs]

May 18, 2022 14:13:15 GMT -5 gabe said:

Hi All. I was running Dirac again, again, again..... Anyway. I am one of those poor souls who loses quite a bit of level with Dirac engaged. I was looking over a very helpful screenshot that tech support sent me regarding setting levels and I stumbled across something that was new, at least to me. I started thinking I might not be choosing the correct speaker as my least efficient to set the master level @ -20db with the mic gain @-37db per the instructions I was given. What I did was leave all the channel sliders all the way up and the master volume quite low. I clicked on my right front height speaker that I always thought would be least efficient because the front heights are reflective and started the test tone. I increased the master output until the right front was @-20db and turned the tone off. NOW, if I click on any channels level control with no test tone playing, I can bring the level down to -20 with no tone playing. After doing so and playing the tone for that channel confirms the accuracy of the setting, at least with my junk.

You "stumbled" across something interesting, not playing tones while making adjustments, which is very helpful. However, it can be done with less effort. At least, I think so.

It's easy to find how each channel compares. This is what I've been doing for several years.

1. Keep the mic at 100%.
2. Set the Master low enough (after doing this a once or twice it's easy to discover how low is low enough) so that no channel is loud at first.
3. Then I play the pink noise in every channel for a couple seconds each, Dirac now has an actual sample from every channel to set levels with.
4. Look at each channel to find the one with the lowest output and leave that one alone.
5. Set the Master for the desired amount of output from that lowest output channel.
6. Set every other channel's slider to match the one with lowest output.

All of this is accomplished with only playing pink noise in each channel just for the couple seconds each in step 3. This makes it quick. Takes about a minute for 13 channels.

[marcl]

May 18, 2022 15:05:33 GMT -5 ttocs said:

May 18, 2022 14:13:15 GMT -5 gabe said:

Hi All. I was running Dirac again, again, again..... Anyway. I am one of those poor souls who loses quite a bit of level with Dirac engaged. I was looking over a very helpful screenshot that tech support sent me regarding setting levels and I stumbled across something that was new, at least to me. I started thinking I might not be choosing the correct speaker as my least efficient to set the master level @ -20db with the mic gain @-37db per the instructions I was given. What I did was leave all the channel sliders all the way up and the master volume quite low. I clicked on my right front height speaker that I always thought would be least efficient because the front heights are reflective and started the test tone. I increased the master output until the right front was @-20db and turned the tone off. NOW, if I click on any channels level control with no test tone playing, I can bring the level down to -20 with no tone playing. After doing so and playing the tone for that channel confirms the accuracy of the setting, at least with my junk.

You "stumbled" across something interesting, not playing tones while making adjustments, which is very helpful. However, it can be done with less effort. At least, I think so.

It's easy to find how each channel compares. This is what I've been doing for several years.

1. Keep the mic at 100%.
2. Set the Master low enough (after doing this a once or twice it's easy to discover how low is low enough) so that no channel is loud at first.
3. Then I play the pink noise in every channel for a couple seconds each, Dirac now has an actual sample from every channel to set levels with.
4. Look at each channel to find the one with the lowest output and leave that one alone.
5. Set the Master for the desired amount of output from that lowest output channel.
6. Set every other channel's slider to match the one with lowest output.

All of this is accomplished with only playing pink noise in each channel just for the couple seconds each in step 3. This makes it quick. Takes about a minute for 13 channels.

+ gabe  Yes and ... reiterating NOT to touch the mic gain ... do not increase it to try to get its level to -40db as is recommended in some tutorials.  

Also, I notice you mentioned reflective front heights ... note that Dirac will not calculate the distance/delay properly and our processors do not allow us to tweak the distance after Dirac.  So it will play, but delays will not be correct.