Dick I think you are referring to variation in group delay, a constant group delay is simply a constant delay. If you are simply listening to a recorded performance no problem, if you are trying to play along with what's recorded a problem. Aligning the delay of digital filters is not rocket science but it does require that you understand the basic math.
High order (3 or 4) Sallen-Key is your best bet. Unfortunately, nothing that you can conveniently do with passives or even discretes. With op amp(s) is rather trivial, but I am sure it will sound, in the ears of the audio review chimps, like crapola.
Quad did it with some LC filters in series in one of their speakers, the ESL-63.
The Cooper Time Cube did it with a garden hose in a box. Duane Cooper was one of my college profs.
Not sure if it would sound better than the Sallen-Key op amp filters, though.
http://lcweb2.loc.gov/master/mbrs/r...sal Audio Cooper Time Cube (Model 920-16).pdf
Last edited:
There is a not-too-bad discussion of delay equalization of filters in Williams and Taylor, Electronic Filter Design Handbook, 3rd ed., ISBN 0070704414. There is also plenty about Bessel filters, which are decent without requiring separate equalization. The big red Zverev has lots of material for things made out of Ls as well as Cs and Rs.
Of course passive filters will need some large L at these frequencies.
Yes, I am envisioning a high end audio filter, hosted in a 19" 5RU box (of course milled from an aerospace grade aluminum billet) full of passive stuff (including carefully shielded, 1ft. long, air core inductors) only to avoid a couple of 50p opamps.
Or, for the budget-minded, just use carefully selected machined panels of different thicknesses "to increase the mechanical damping factor":
Burson Conductor Virtuoso | AudioStream
Oddly, despite the "already-improved IC-free power supply" and the alleged black background, the reviewer hears hiss. But he gets rid of a perceived hardness in the audio by swapping USB cables.
This is a very important subject for audio besides the <20Hz LP junk removal:
View attachment GD 1.PDF
and - View attachment GD 7.pdf
THx-RNMarsh
Last edited:
yes, and as the amp used is just a buffer, a discrete circuit can be easily made with compl jFET's. One or two of these S-Key discrete switchable filters-- in or out -- would not be hard to do and could be very transparent.
Some other options might be to implement it in the Line level stage as a (switchable) filter with gain. Can you put up a circuit with values for GD of less than 2.5mS? Cut-off for HP might be 20Hz. Lets see what that would look like.
THx-RNMarsh
I took a different approach and built a phono preamp with flat response from the 50 Hz pole to DC. What took me by surprise is how much better it seemed to sound in the bass and (the elephant in the room) less prone to acoustic feedback.
All this talk about LF resonance in the tonearm and etc. doesn't address a 12" diaphragm, not always damped, connected to a sensitive vibration pickup and in turn connected to large speaker with lots of gain. Just because it dosn't oscillate does not mean there is no acoustic feedback happening.
The lack of feedback (and I have woofers with flat acoustic response to below 20 Hz) suggests that maybe the phase shift from the high pass filter could be exacerbating acoustic feedback problems.
There is a new school of phono preamp design that insists on using a flat preamp and EQ executed in digital. My instint is that the headroom, dynamic range and SNR would suffer with this approach. Have I missed something? its important since I'm about to execute a phono to digital design and don't want my prejudices and resistance to new stuff to get in the way of a valid solution. Adding a DSP (The AD Sigma Studio Family SigmaStudio | product info The SigmaStudio | Analog Devices family are very good, simple to execute and cheap in quantity) would be not much more expensive than the 1% caps necessary for EQ. And it makes alternate EQ's for collectors easy.
All this talk about LF resonance in the tonearm and etc. doesn't address a 12" diaphragm, not always damped, connected to a sensitive vibration pickup and in turn connected to large speaker with lots of gain. Just because it dosn't oscillate does not mean there is no acoustic feedback happening.
The lack of feedback (and I have woofers with flat acoustic response to below 20 Hz) suggests that maybe the phase shift from the high pass filter could be exacerbating acoustic feedback problems.
There is a new school of phono preamp design that insists on using a flat preamp and EQ executed in digital. My instint is that the headroom, dynamic range and SNR would suffer with this approach. Have I missed something? its important since I'm about to execute a phono to digital design and don't want my prejudices and resistance to new stuff to get in the way of a valid solution. Adding a DSP (The AD Sigma Studio Family SigmaStudio | product info The SigmaStudio | Analog Devices family are very good, simple to execute and cheap in quantity) would be not much more expensive than the 1% caps necessary for EQ. And it makes alternate EQ's for collectors easy.
Dsp is a nice way to do it. I dont think the dynamic range will suffer unless it is worse than an LP's 60-70dB.
THx-RNMarsh
Manley massive passive anyone? Manley Massive Passive Stereo Tube EQ Eq is too small a word, it is tone shaping!!!
Demian,
Does that mean we simply decouple the phono player by moving it into another
room or out of the sonic field of speaker?
OR alternatively
Can we follow the guildlines in the Audio Cyclopedia and build a mini soundstage
for the phono device, thus isolating it from 12 " or other diaphragms that might
co exist in said elephant room?
Richard, I'd be interested in the GD 2, 3, 4 ,5 6, if you'd also care to share.
Thanks for the VK5BR link. Some of this is starting to sink in especially
after I recently found Tektronix Cookbook of Standard Audio Tests.
The tests were required in 1966 for stereos and receiver's sold in the United States.
The 1974 Federal Trade Commission regulation to measure amp power output,
(1/3 power, 1 hour, 1KHz). There are some others.
While y'all are doing the advanced discussion, the Tektronix Cookbook would help the DIYer understand with demonstration examples, screen shots etc, how to do the basic test procedures and what they look like on a scope and/or FFT.
Once those are understood it will give meaning to Richard's VK5BR link.
Rayma, I guess Prof Cooper didn't get much respectability with his two
hoses in box, so his buddy had to do it electrically? That was really funny
reading that. Prof Cooper must have been a good hands on instructor.
ONE BIG QUESTION
Why hasn't anyone just gone out and created an analog disc that was 50KHz or 75KHz
bandwidth per channel read it with a laser and be done with all the digital stuff that
claims 24 bit...Why not just put the real thing down optically and read it with
a laser. Heck we used to put audio tracks on the NTSCs Raster Scan lines for longer
playback etc. So while you were looking at a video still, with animated overlay, computer graphiscs, Text etc....You would also be listening to how to perform a task or something. The audio and be good. You put it in a buffer and get it all ready for play back and off you go. Non of the other bit issues, no mechanical problems with tone arms etc, we just
pick up the analog signal via laser.
GOOODDDDDD NNNNIIIIIIIGGGGGGGHHHHHHTTTTTTTTT....DIY.
Does that mean we simply decouple the phono player by moving it into another
room or out of the sonic field of speaker?
OR alternatively
Can we follow the guildlines in the Audio Cyclopedia and build a mini soundstage
for the phono device, thus isolating it from 12 " or other diaphragms that might
co exist in said elephant room?
Richard, I'd be interested in the GD 2, 3, 4 ,5 6, if you'd also care to share.
Thanks for the VK5BR link. Some of this is starting to sink in especially
after I recently found Tektronix Cookbook of Standard Audio Tests.
The tests were required in 1966 for stereos and receiver's sold in the United States.
The 1974 Federal Trade Commission regulation to measure amp power output,
(1/3 power, 1 hour, 1KHz). There are some others.
While y'all are doing the advanced discussion, the Tektronix Cookbook would help the DIYer understand with demonstration examples, screen shots etc, how to do the basic test procedures and what they look like on a scope and/or FFT.
Once those are understood it will give meaning to Richard's VK5BR link.
Rayma, I guess Prof Cooper didn't get much respectability with his two
hoses in box, so his buddy had to do it electrically? That was really funny
reading that. Prof Cooper must have been a good hands on instructor.
ONE BIG QUESTION
Why hasn't anyone just gone out and created an analog disc that was 50KHz or 75KHz
bandwidth per channel read it with a laser and be done with all the digital stuff that
claims 24 bit...Why not just put the real thing down optically and read it with
a laser. Heck we used to put audio tracks on the NTSCs Raster Scan lines for longer
playback etc. So while you were looking at a video still, with animated overlay, computer graphiscs, Text etc....You would also be listening to how to perform a task or something. The audio and be good. You put it in a buffer and get it all ready for play back and off you go. Non of the other bit issues, no mechanical problems with tone arms etc, we just
pick up the analog signal via laser.
GOOODDDDDD NNNNIIIIIIIGGGGGGGHHHHHHTTTTTTTTT....DIY.
Based on my own experience with a 6 dB/Oct. 20 Hz subsonic filter and my bass reflex speakers with its port tuned to 35 Hz, what you wrote is exactly what happens.
Switch the filter on and the bass lnes become cleaner, clearer and better controlled. No doubt some phase shift is added, and that seems to be the key trade-off, perhaps a very slightly volume reduced sound at the very lowest of frequencies (that I have), some added phase shift for better control of the bass driver and better overall clarity of the bass lines.
Speaking strictly with the preamps I own and know (Luxman C-03, H/K Citation 21, Marantz 3256b and Philips AH286), all of which have very similar high pass filters centered at 20 Hz, the effect is virtually the same, their other differences notwithstanding. It's a trade-off, but in my view, a very acceptable one because the pros are greater than the cons.
Better yet, this is further confirmed when using speakers with less bass extention, like AR94 (nominally -3 dB @ 44 Hz) and JBL Ti600 (nominally -3dB @ 45 Hz).
- Status
- Not open for further replies.