THAT4301 compressor

[hazel]

I'm here again,

for those of you who are already building this compressor i've recorded some samples playing around with some Constant Time. Dramatic changes with extreme values...

https://soundcloud.com/hazel/sets/that-4301-compressor-constant-times-mod

 

[shot]

Excellent comparison!
THANK YOU!

 

[hazel]

Glad you like it!

I'm still toying around the idea of adding manual Attack & Release of just play with RMS / PEAK / AUTO. I'm having some problems with THAT DN115 AR circuit...

By the way here you have some screenshots of Sidechain filter options and capacitor values. 50Hz peak was my finger hanging capacitor leads 

http://s1070.photobucket.com/user/hazel/library/#/user/hazel/library/?sort=3&page=1&_suid=140353150128301394994717231668

PS 1: i know there are a couple us "useless" HPF but i was just playing with different values. I'll experiment with HPF, LPF, BPF and NOTCH FILTER so i could:

- not compress lows, mids, highs
- only compress lows, mids, highs

I just have to switch LPF and HPF between series and parallel configuration

PS 2: To ungifted and moderators: don't want to hijack ungifted's thread, just add info to potential builders. If you consider please move or delete my posts

 

[ungifted]

Hi hazel

The thread is open and anyone can post his experience! I will be glad if somebody will find some time for testings and new features to basic design.
Many thanks to all!

 

[L´Andratté]

Cool ungifted! hazel, please report!

Sc filtering is most influential onthe compressor response and yields really different results (note that I am an experimental musician, not recording or mastering person).

Just to make sure: you are actually compressing the whole audio band, just only when low, band, or high are triggering the sc.

I´m afraid to bore everybody with promoting the 3dBslope filter, which nobody seems interested in  ;D  but gives an even response over the audio band with regard to uneven power in bass/high frequencies...

Edit: hazel, are you using a state variable filter or what?

 

[shot]

Just to be sure, I want to check with you...

1. CT cap is C7, right?
2. Sidechain filter is before VCA pin 15, right?

correct me if I'm wrong please!

 

[hazel]

I´m afraid to bore everybody with promoting the 3dBslope filter, which nobody seems interested in  ;D
Edit: hazel, are you using a state variable filter or what?

L'Andratté: i never tried 3dB slope filter, in fact, i never heard about it. I really find 6dB/oct to be very gentle, subtle and phase-friendly  i'm not afraid of HPF voice at 150Hz to just remove rumble and not killing voice's body. Keep in mind that 6dB/oct at 100Hz just indicates where bass rolloff reaches 3dB. Is such a smooth curve.

It's just a rotary switch with caps attached following this schematic:

 

[hazel]

Just to be sure, I want to check with you...

1. CT cap is C7, right?
2. Sidechain filter is before VCA pin 15, right?

correct me if I'm wrong please!

Shot:
1. it depends on every schematic, CT cap is connected to pin 5. In THAT Demo Board is C7 10uF.
2. SC Filter is connected at the input...sorry don't rememeber pin now   

 

[L´Andratté]

This is what I meant (LP version, HP attached)
http://groupdiy.com/index.php?topic=55857.msg712753#msg712753
(reference here:
http://dt.prohosting.com/hacks/what.html
http://mnats.net/what_compressor.html)

 

[hazel]

This is what I meant (LP version, HP attached)
http://groupdiy.com/index.php?topic=55857.msg712753#msg712753
(reference here:
http://dt.prohosting.com/hacks/what.html
http://mnats.net/what_compressor.html)

Oh i see L'Andratté,

I don't use opamps at SC filter, just capacitors.

Must say i didn't add SC filter option yet, just sending pink noise in playback loop through caps to determine which values i'll end up using in finished building. This afternoon i'll try HPF and LPF values and experiment with NOTCH and BPF.

Do i connect SC filter options at 4301 pin 1?? it's RMS input so it should be there?

 

[hazel]

After reading and testing LPF and HPF circuits i ilke the most comes from http://sound.westhost.com/dwopa2.htm

As simple as it gets:

PASSIVE LPF (rotary switch selects caps from 12kHz down to whatever you want) --> TL072 BUFFER --> PASSIVE HPF (rotary switch selects resistors from 50Hz up to whatever you wish) --> TL072 BUFFER --> THAT4301 DETECTOR

 

[L´Andratté]

I´m afraid to bore everybody with promoting the 3dBslope filter, which nobody seems interested in  ;D  but gives an even response over the audio band with regard to uneven power in bass/high frequencies...

The sc filter I proposed is known (and patented) as the ´THRIST´, no,  ´THROST´, no, but something like that filter.
It has been discussed at this forum, but the discussion rather died in a way smelling of lawyers. Thanks to the best of all possible worlds... :-[

 

[druu]

Can someone advise on how I would alter the ratio range on this design? Aiming for a maximum of 4:1 if possible.

 

[hazel]

Can someone advise on how I would alter the ratio range on this design? Aiming for a maximum of 4:1 if possible.

Hi Druu,

i'm not an expert but if i would like to limit ratio from 1:1 to 4:1 i would measure resistance on Ratio pot at 4:1 and this would be pot value...or you can make it stepped rotary switch by measuring resistance on 1:1, 1:1'2, 1:1'5, 1:1'75, etc...

 

[druu]

Thanks hazel. How would I know ratio was at 4:1? I've never performed compression tests before in software so any insight appreciated.

 

[hazel]

Thanks hazel. How would I know ratio was at 4:1? I've never performed compression tests before in software so any insight appreciated.

You're welcome Druu

Years ago i was told about doing measurements with at least one leg disconnected off circuit so:

STEPPED RATIO SWITCH

- Connect Ratio potentiometer into circuit, adjust for 4:1 ratio
- Disconnect Ratio pot off circuit without rotating
- once isolated measure resistance between wiper (middle pin) and one lug. Sorry i cannot remember if you have to measure between wiper and pin 1 or wiper and pin 3
- Substitute pot with a resistor according to measurement


RATIO POT LIMITED TO 4:1

Ratio pot is 10k lin corresponding center position (12 o'clock) to 4:1 ratio in THAT 4301P compressor circuit.

So we have 10k lin pot at half rotation is 5k. Try a 5k lin and you're good to go!

 

[druu]

Thanks again, I will experiment and see as I'll be using stepped controls.

BTW, any idea what the optional R27 1K4 is for?

 

[ungifted]

It's for 4:1 when ratio pot is centered. Please check DN115.
"The network comprising R25, R31, and either R16 or R28 combine with the variable
source impedance of VR7 to yield a 4:1 compression ratio when VR7 is centered."

 

[ForthMonkey]

How can i add indicator like this?

http://medias.audiofanzine.com/images/normal/dbx-that-4301-568623.jpg

 

[druu]

Try this:

http://www.thatcorp.com/datashts/dn132.pdf