So at 26 Hz it will be 38.5 ms delayed. Don't look at the wiggles, Smooth it a bit to see you're below 5ms at 55 Hz. The wiggles aren't static, they move with time.
I'm sure Greg will eek every ounce of Group Delay out of it. 🙂
Couldn't you seal off the mains? Let the 3 subs do their job there?
The main subject of this thread has been GD, which has led us to DRC. All of that is good. Ive learned alot. But it is not the only criteria for improving my sound. Ive had to restrain myself a bit not to get into other unrelated topics, for I wanted this discussion to stay focused. And it has. But when it comes to what Greg is doing, I dont want the the DRC shaping to merely see how far down the GD can be driven down to. Its an important part of things, but far from being the only thing.
As it happens, what ive seen from DRC so far DOES address issues beyond GD. I wouldn't be much interested if that was the only thing it could fix.
Rega DAC D/A processor Page 2 | Stereophile.com
In the case of direct USB connection, the data go to that Burr-Brown PCM2707 receiver mentioned above. With S/PDIF inputs, the data go to a Wolfson Microelectronics WM8805 receiver chip.
Thanks for the link.
Reading the article, nobody says what is done with the audio data once decoded from USB.
But going on in the article to the measurements section, John Atkinson says this:
Turning to the Rega's USB input, Apple's USB Prober utility identified the device as a "USB Audio DAC" by "Burr-Brown from TI," and confirmed that it operated in "Isochronous adaptive" mode. Bit resolution was listed as "16," and the sample rates supported were 32, 44.1, and 48kHz. The Rega's performance via its USB input was identical to that from the S/PDIF inputs, other than the restricted bit depth and sample rate. This was true also of the DAC's jitter spectrum with USB data (fig.14), which was marred by the same supply-related sidebands. However, sidebands at the adaptive-polling interval–related frequency of ±1kHz are absent in this graph. In addition, the central spike that represents the 11.025kHz tone with USB data in this graph is as well defined and narrow as it is with S/PDIF data, suggesting that the large amount of the random low-frequency jitter that is often evident with adaptive USB products is also absent. The Rega DAC's USB input is well sorted, as its English designer would say.
Can you be more specific about how your USB connection is inferior to S/PDIF connection?
Thanks for the link.
Reading the article, nobody says what is done with the audio data once decoded from USB.
But going on in the article to the measurements section, John Atkinson says this:
Can you be more specific about how your USB connection is inferior to S/PDIF connection?
I cant offer any further technical details. But as I already said, the sound of my system is best via Coax, then Toslink, with USB coming in a distant third.
Perhaps not every Rega DAC owner would say the same. Certainly, no other system is going to be exactly like mine. I can only relate what I hear.
The main subject of this thread has been GD, which has led us to DRC. All of that is good. Ive learned alot. But it is not the only criteria for improving my sound. Ive had to restrain myself a bit not to get into other unrelated topics, for I wanted this discussion to stay focused. And it has. But when it comes to what Greg is doing, I dont want the the DRC shaping to merely see how far down the GD can be driven down to. Its an important part of things, but far from being the only thing.
As it happens, what ive seen from DRC so far DOES address issues beyond GD. I wouldn't be much interested if that was the only thing it could fix.
Glad to hear that, DRC indeed does offer more than that alone. Basically it tries to correct the signal at the listening position with you being the judge.
It also tries to correct the speaker.
While we're on a wider scope here, what I wanted to ask you, that Haas Kicker, does it work well? The reason I ask is that I've read a lot about faking such a signal and the bandwidth mentioned for such a thing was usually quoted about 200 to ~3 KHz. Yours seems to reflect much higher in frequency than that. I know you mentioned it's tail was a bit to short (low extension), but do the higher frequencies help with it's effectiveness? Just curious.
Our high frequency perception is mostly based on level if I remember that right.
Glad to hear that, DRC indeed does offer more than that alone. Basically it tries to correct the signal at the listening position with you being the judge.
It also tries to correct the speaker.
While we're on a wider scope here, what I wanted to ask you, that Haas Kicker, does it work well? The reason I ask is that I've read a lot about faking such a signal and the bandwidth mentioned for such a thing was usually quoted about 200 to ~3 KHz. Yours seems to reflect much higher in frequency than that. I know you mentioned it's tail was a bit to short (low extension), but do the higher frequencies help with it's effectiveness? Just curious.
Our high frequency perception is mostly based on level if I remember that right.
I think things sound better with it than without it. In my case, where I have tons of absorption, the sound gets a little dead. The kicker brings needed life back into the room without most of the negative effects of early reflections. For a room not so dead, I am not sure it (the kicker) would be as welcome.
As far as the bandwidth of the kicker, ive read 500hz to about 6k is optimum (not sure everyone agrees what this should be). In any case, if your doing this acoustically, that is using reflective panels to bounce the energy around, your going to get the higher stuff whether you want it or not. But I have experimented with having less or more >5K and I like the higher frequency energy being there.

Measuring where the ear actually is, a spectrograph shows pretty well what I am getting.
When it comes to kickers, not only is bandwidth important, but when it occurs and from what angle of incidence as well as magnitude. Timing wise, 20-25ms seems optimum. Incidense wise, 110 +/- 10 degrees would be (about where you would put rear speakers in a 5.1 setup). Magnitude wise, I like it around -8db (-12db is required for the Haas effect).
Here is a link to a good read on Terminators: https://www.gearslutz.com/board/7014534-post11.html
But this quote summarizes its benefits the most clearly:
The termination of the ISD does several things as well psycho-acoustically. It provides the sense of liveliness to the listening space. Second, it aids in localization by 'removing' the localization and tonality shifting cues from later arriving signals - causing the focus and localization to lock onto the direct signal.
Last edited:
Thanks for the answer. The processed version proposed to work very well in a car usually lift out the voice by using L-R and R-L and use limited bandwidth.
Your example is the first I have seen as used in a room, very cool!
Your example is the first I have seen as used in a room, very cool!
New tracks and mdats
New processed tracks as well as mdat showing predicted response for 3 and 4 cycles of correction with and without a flat response above 5khz:
https://www.dropbox.com/s/qax0n16bc4g6c34/jim1961.mdat?dl=0
https://www.dropbox.com/s/vpr3ze5v49s8ppf/No%20photo%20%283%20cycles%20knee%29.flac?dl=0
https://www.dropbox.com/s/xc02k9644nox5yq/No photo (3 cycles).flac?dl=0
https://www.dropbox.com/s/4e4ufb9bkocml76/No photo (4 cycles knee).flac?dl=0
https://www.dropbox.com/s/2kh324awmoswuk4/No photo (4 cycles).flac?dl=0
New processed tracks as well as mdat showing predicted response for 3 and 4 cycles of correction with and without a flat response above 5khz:
https://www.dropbox.com/s/qax0n16bc4g6c34/jim1961.mdat?dl=0
https://www.dropbox.com/s/vpr3ze5v49s8ppf/No%20photo%20%283%20cycles%20knee%29.flac?dl=0
https://www.dropbox.com/s/xc02k9644nox5yq/No photo (3 cycles).flac?dl=0
https://www.dropbox.com/s/4e4ufb9bkocml76/No photo (4 cycles knee).flac?dl=0
https://www.dropbox.com/s/2kh324awmoswuk4/No photo (4 cycles).flac?dl=0
Last edited:
New processed tracks as well as mdat showing predicted response for 3 and 4 cycles of correction with and without a flat response above 5khz:
https://www.dropbox.com/s/qax0n16bc4g6c34/jim1961.mdat?dl=0
https://www.dropbox.com/s/vpr3ze5v49s8ppf/No photo (3 cycles knee).flac?dl=0
https://www.dropbox.com/s/xc02k9644nox5yq/No photo (3 cycles).flac?dl=0
https://www.dropbox.com/s/4e4ufb9bkocml76/No photo (4 cycles knee).flac?dl=0
https://www.dropbox.com/s/2kh324awmoswuk4/No photo (4 cycles).flac?dl=0
Thanks! I may need a few days to evaluate all this.
An off the wall question here, but does anyone know the "Q" value for the individual bands on a 32 band GEQ ?
5 maybe?
5 maybe?
An off the wall question here, but does anyone know the "Q" value for the individual bands on a 32 band GEQ ?
5 maybe?
I think Q=5 is more like a 2/3 octave eq (15 bands), and Q=10 would be more like a 1/3 octave eq....
edit: there is a calc for this here: http://www.sengpielaudio.com/calculator-bandwidth.htm
According to this, it looks like I was mistaken and it's actually 4.32...........
edit 2: unless you're using a "constant q" equalizer, the amount of boost/cut will affect the q....
Last edited:
I think Q=5 is more like a 2/3 octave eq (15 bands), and Q=10 would be more like a 1/3 octave eq....
edit: there is a calc for this here: Q factor vs bandwidth in octaves band filter -3 dB pass calculator calculation formula quality factor Q to bandwidth BW width octave convert filter BW octave vibration mastering slope dB/oct steepness EQ filter equalizer cutoff freqiency - sengpielau
According to this, it looks like I was mistaken and it's actually 4.32...........
edit 2: unless you're using a "constant q" equalizer, the amount of boost/cut will affect the q....
Thanks. Q= 10 wasnt jiving with what I was seeing. 4.32 sounds good.
A good example that group delay is no problem is the CD itself. It can sit for years on a shelf and all of a sudden, sound very good 😉
//
//
This is probably obvious by now, but when I started this thread, I wasn't really sure how much GD correlated with tight sounding bass. I am now convinced it is to a pretty substantial degree.
A good example that group delay is no problem is the CD itself. It can sit for years on a shelf and all of a sudden, sound very good 😉
//
But then the big moment finally arrives for the waveform when it's time to re-enter the world as the multidimensional wavefront it longs to be. Except, the low frequencies, for whom it seems were forced to take a longer route than the others, start yelling to the rest of the group "hey guys, wait up!!".
I can agree with that! Boosting the low end can't make up for timing.
No. But you can shift the delay to the lowest region where its matters less.
I just realized that when I made the "knee" target versions, I had the FR target section of DRC set to interpolate a certain way between my control points causing the knee came out a bit softer than I anticipated. This is causing the response to differ below 5khz compared to the straight target. I wonder if this will be too "thick" sounding?
Corrected test tracks and mdat with hard knee @5khz
Here are new 3 and 4 cycle processed tracks using a target FR with a hard knee at 5khz along with the mdat showing predicted responses:
https://www.dropbox.com/s/5upfewufnac40bq/No photo (3 cycles hard 5khz knee).flac?dl=0
https://www.dropbox.com/s/fpah3ruto4bmp2v/No photo (4 cycles hard 5khz knee).flac?dl=0
https://www.dropbox.com/s/7t49o481pdxf5tz/3 and 4 cycles correction with hard knee target.mdat?dl=0
Here are new 3 and 4 cycle processed tracks using a target FR with a hard knee at 5khz along with the mdat showing predicted responses:
https://www.dropbox.com/s/5upfewufnac40bq/No photo (3 cycles hard 5khz knee).flac?dl=0
https://www.dropbox.com/s/fpah3ruto4bmp2v/No photo (4 cycles hard 5khz knee).flac?dl=0
https://www.dropbox.com/s/7t49o481pdxf5tz/3 and 4 cycles correction with hard knee target.mdat?dl=0
Graphs are looking nice, you can see some differences between 3 cycle and 4 cycle.
Jim, don't look! Listen first! 😉
You don't want to manipulate yourself into hearing things.
Jim, don't look! Listen first! 😉
You don't want to manipulate yourself into hearing things.
Yes, it seems like that bump in the very low bass range needs at least a 4 cycle window to be flattened out...
That being said, it's amazing how little difference there is between the two window lengths in this case. It goes to show how well treated the room is.
That being said, it's amazing how little difference there is between the two window lengths in this case. It goes to show how well treated the room is.
- Status
- Not open for further replies.
- Home
- Loudspeakers
- Full Range
- Group Delay Questions and Analysis