Gearslutz

Gearslutz (https://www.gearslutz.com/board/)
-   Product Alerts older than 2 months (https://www.gearslutz.com/board/product-alerts-older-than-2-months/)
-   -   TBProAudio releases DSEQ, a dynamic spectral equalizer for Windows and Mac OS X (https://www.gearslutz.com/board/product-alerts-older-than-2-months/1306847-tbproaudio-releases-dseq-dynamic-spectral-equalizer-windows-mac-os-x.html)

DSK 18th June 2020 09:10 AM

Quote:

Originally Posted by Endor (Post 14803405)
Will DSEQ go on sale in a near futur ?

I tell myself DSEQ is on sale everytime there's a new feature added :cop:

Endor 18th June 2020 02:10 PM

Quote:

Originally Posted by DSK (Post 14806855)
I tell myself DSEQ is on sale everytime there's a new feature added :cop:

Good point ! :lol:

PitchSlap 18th June 2020 03:00 PM

Quote:

Originally Posted by TBProAudio (Post 14806628)
We do this with the following setting:
  • quality high/ultra
  • selectivity sharp
  • threshold moderate
  • and finally delta mode (hear removed content)
Now you can play with the threshold parameter and GR strength.

Thanks.

I'll try that, it seems to be similar to what I've been doing, I assume ultra is important since the higher resolution would affect the maximum sharpness? I haven't used GR strength much, but I assume that's just a multiple of the gain reduction, which would still affect all frequencies caught by the threshold.

It might be due to how the reduction spectrum is displayed vs. EQuilibrium.

In the image below I've manually found just the 7 most offending frequencies (loudest when sweeping with a very tight Q).

I think it would be hard to get this level of precision with DSEQ, but if I understand you I'd want to try catch those frequencies, but rather than lower the threshold (which would also lower adjacent frequencies) I'd want them minimally reduced, but then use the strength to to reduce them further which should leave the others alone since they didn't reach the threshold.

*EDIT* I think I got the hang of it. :)

https://i.imgur.com/4M9cJIy.jpg

JfromRVA 18th June 2020 04:39 PM

Quote:

Originally Posted by PitchSlap (Post 14807268)
Thanks.

I'll try that, it seems to be similar to what I've been doing, I assume ultra is important since the higher resolution would affect the maximum sharpness? I haven't used GR strength much, but I assume that's just a multiple of the gain reduction, which would still affect all frequencies caught by the threshold.

It might be due to how the reduction spectrum is displayed vs. EQuilibrium.

In the image below I've manually found just the 7 most offending frequencies (loudest when sweeping with a very tight Q).

I think it would be hard to get this level of precision with DSEQ, but if I understand you I'd want to try catch those frequencies, but rather than lower the threshold (which would also lower adjacent frequencies) I'd want them minimally reduced, but then use the strength to to reduce them further which should leave the others alone since they didn't reach the threshold.

*EDIT* I think I got the hang of it. :)

https://i.imgur.com/4M9cJIy.jpg

If you don't mind sharing, what instrument/sound are you de-resonating in the photo provided? I ask because that range is our 'sensitive' area. Boost and sweep through there and a lot of things will sound 'bad'. The notches seen could cause some comb-filtering like artifacts that could 'smear' the range, which is diserable from time to time.

TBProAudio 18th June 2020 05:50 PM

Quote:

Originally Posted by PitchSlap (Post 14807268)
I think it would be hard to get this level of precision with DSEQ, but if I understand you I'd want to try catch those frequencies, but rather than lower the threshold (which would also lower adjacent frequencies) I'd want them minimally reduced, but then use the strength to to reduce them further which should leave the others alone since they didn't reach the threshold.

As already mentioned static,steep filter could cause all kinds of artefacts (e.g. comp filtering), so you should avoid them.

Have you tried DSEQ with quality "high" or even "ultra" and a reasonable threshold to fight them? Its the job of DSEQ to "translate" annoying resonances to "normal" resonances.

And don't let your ears fool you: set-up DSEQ, make a rest and check again ;)

Edit: default GR strength (x1.0) should be OK.

Spoff 19th June 2020 01:54 AM

Something I have wondered since I bought this is does it just pick out the frequencies that are louder than others, or does it actually intelligently look for certain frequencies that are known to be more harsh or unpleasant to the human ear? Just referring to the default as I know one can get manually intelligent with the filters and other controls. Love it either way but just curious.

MogwaiBoy 19th June 2020 02:12 AM

Unpleasant resonance can happen at just about any frequency, depending on the instrument(s) involved - for example low-mid harmonics that occasionally pop out way too far on cellos or synth basses. So resonance suppression shouldn't just be about upper midrange sibilance etc.

Although, I don't actually know so TPProAudio will have to answer, maybe there is some kind of Fletcher-Munson type weighting happening in the detection underneath. I wouldn't think so though (unless it was a special setting that you could turn off).

TBProAudio 19th June 2020 06:03 AM

Quote:

Originally Posted by Spoff (Post 14808425)
Something I have wondered since I bought this is does it just pick out the frequencies that are louder than others, or does it actually intelligently look for certain frequencies that are known to be more harsh or unpleasant to the human ear? Just referring to the default as I know one can get manually intelligent with the filters and other controls. Love it either way but just curious.

DSEQ does not use AI to find resonances :-)
Its plain threshold (or custom threshold) in combination of slope and pre-filter.
Pretty much predictable because it is transparent for the user. And it works.

The DSEQ slope parameter is the connection to human listening.

Wikipedia: "Pink noise is one of the most common signals in biological systems."

Pink Is Hearing: "So, it’s not that pink noise is calibrated to the human ear’s frequency response. It’s just calibrated to how we hear, which is very well grounded in math. Each time the frequency doubles, we hear that as an octave. From one octave to the next, we expect to hear an appropriate amount of sound energy (depending upon the program material), which is why we calibrate our audio systems to pink noise. Octave bands are easier for our hearing mechanism to understand."

So following pink noise does not need a big AI, I think :)

TBProAudio 19th June 2020 06:12 AM

Quote:

Originally Posted by MogwaiBoy (Post 14808449)
Unpleasant resonance can happen at just about any frequency, depending on the instrument(s) involved - for example low-mid harmonics that occasionally pop out way too far on cellos or synth basses. So resonance suppression shouldn't just be about upper midrange sibilance etc.

Although, I don't actually know so TPProAudio will have to answer, maybe there is some kind of Fletcher-Munson type weighting happening in the detection underneath. I wouldn't think so though (unless it was a special setting that you could turn off).

Well, as stated above DSEQ can be set to follow pink noise (slope parameter). And pink noise is connected to human hearing.
So yes, there is a relation to Fletcher-Munson as well. Maybe even closer than any static EQ, as DSEQ handles the dynamics. kfhkh

JfromRVA 19th June 2020 06:21 AM

Quote:

Originally Posted by TBProAudio (Post 14808689)
Well, as stated above DSEQ can be set to follow pink noise (slope parameter). And pink noise is connected to human hearing.
So yes, there is a relation to Fletcher-Munson as well. Maybe even closer than any static EQ, as DSEQ handles the dynamics. kfhkh

To expand. The greatest advantage to having 'AI' is the ability to 'tell' it what to 'think'. If the user does not know what a sound should sound like...what is 'AI' to think? It has only math. It's how a user directs 'AI'. There is the power. Power is saving time. AI will live forever. Sweeping and notching ticks our time away.

Philosophy rant over. Use 'AI' to save time and use it to create more music because if you don't know where you want to go....the the computer can't tell you the way either!

TBProAudio 19th June 2020 08:55 AM

Quote:

Originally Posted by JfromRVA (Post 14808703)
Philosophy rant over. Use 'AI' to save time and use it to create more music because if you don't know where you want to go....the the computer can't tell you the way either!

kfhkh Thats why we added the slope value display. It helps to find the right slope parameter value (aka pink noise, aka Fletcher-Munson)

Circle closed, I guess ;)

rocketdirk 19th June 2020 10:11 AM

Is the underlying concept of DSEQ unique? If so, maybe it's worth patenting?

TBProAudio 19th June 2020 10:55 AM

Quote:

Originally Posted by rocketdirk (Post 14808898)
Is the underlying concept of DSEQ unique? If so, maybe it's worth patenting?

Interesting question :)

Well, we have not seen any tools like DSEQ have a slope parameter and therefore considering concepts like pink noise or Fletcher-Munson.

But I don't think that this can be patented.

What can be patented is the calculation of the gain reduction curve, which enables artifact-free processing with a relatively low threshold.

But it makes no sense to publish the algorithm and spend money to defend it.
So we keep it to ourselves, enjoy the advantage and spend the money on new products cooge

Software patents are tricky thing and make only fun if you are a big big company with money for lawyers.

rocketdirk 19th June 2020 11:37 AM

Quote:

Originally Posted by TBProAudio (Post 14808948)
Interesting question :)
But I don't think that this can be patented.

What can be patented is the calculation of the gain reduction curve, which enables artifact-free processing with a relatively low threshold.

But it makes no sense to publish the algorithm and spend money to defend it.
So we keep it to ourselves, enjoy the advantage and spend the money on new products cooge

Thanks for explaining this. DSEQ is doing wonders to my mixes, btw. kfhkh

no genre 19th June 2020 12:02 PM

Quote:

Originally Posted by TBProAudio (Post 14808948)
Interesting question :)

But it makes no sense to publish the algorithm and spend money to defend it.
So we keep it to ourselves, enjoy the advantage and spend the money on new products cooge

And since you bring up new products, anything new and exciting about to drop from the new product pipeline?

Herzton 19th June 2020 04:33 PM

So I bought DSEQ a couple of days ago, and as someone suggested here to me, used it mainly to remove the muddy bits on acoustic guitars so far, which works just great. Always a struggle before with static or dynamic EQ, and I think overall I use less gain reduction overall with DSEQ and it sounds better.
Used it on vocals, too, effect was not as impressive as with guitars, but definitely useful.
Next thing to test is uneven acoustic bass...

TBProAudio 19th June 2020 05:11 PM

Quote:

Originally Posted by no genre (Post 14809039)
And since you bring up new products, anything new and exciting about to drop from the new product pipeline?

Not yet ;)

I need to take a rest and rebuild our studio. Just to get HD wiped kfhkh

But yes, there are some ideas, lets see...

TBProAudio 19th June 2020 05:12 PM

Quote:

Originally Posted by Herzton (Post 14809490)
Used it on vocals, too, effect was not as impressive as with guitars, but definitely useful.

Please watch the slope value display, and adjust the slope parameter accordingly kfhkh

stripealipe 19th June 2020 11:47 PM

Quote:

Originally Posted by TBProAudio (Post 14809561)
Please watch the slope value display, and adjust the slope parameter accordingly kfhkh

Or indeed, capture the vocal threshold if you generally like how it sounds, and then bring the threshold down on it. It will be very transparent and smooth...

mcbpete 20th June 2020 04:21 PM

Been using and loving DSEQ since my purchase a month or two ago. Now that it's added so much there's no reason I need to get Zynaptiq Unflilter on sale or Gulfoss right .... right ?!

cooker 20th June 2020 09:50 PM

Quote:

Originally Posted by mcbpete (Post 14811281)
Been using and loving DSEQ since my purchase a month or two ago. Now that it's added so much there's no reason I need to get Zynaptiq Unflilter on sale or Gulfoss right .... right ?!

I have experience with unfilter in a mastering perspective.

Though it can be considered quite similar to DSEQ in terms of use, my thought is its better as a miracle worker for terrible sounds. This also means it can go quite extreme pretty fast, for me it was too easy to overdue what I wanted resulted in a not so similar sound compared to a mix...also the fact that likely its too intense to be used in multiple tracks in a mix.

So compared DSEQ has a finer touch, fixing but not changing anything and can easily be used in a track.

I don't have experience with gullfoss but I believe its not really similar to DSEQ in use.

JfromRVA 20th June 2020 10:11 PM

Quote:

Originally Posted by cooker (Post 14811801)
I have experience with unfilter in a mastering perspective.

Though it can be considered quite similar to DSEQ in terms of use, my thought is its better as a miracle worker for terrible sounds. This also means it can go quite extreme pretty fast, for me it was too easy to overdue what I wanted resulted in a not so similar sound compared to a mix...also the fact that likely its too intense to be used in multiple tracks in a mix.

So compared DSEQ has a finer touch, fixing but not changing anything and can easily be used in a track.

Agreed. I use Unfilter to quickly handle drum room mics or to super clean a reverb return when using room reverbs like IK's Sunset (then re-dirty it up with saturation like BlackBox or AA's Celestial to taste). Unfilter can zap standing nodes like no other but has a heavy handed sonic imprint (the price of linearization 'perfection'). But if you have a terrible sounding recording it's the only plugin 'suite' that can be a one-stop shop and rescue whatever music is in the 'haze'!

Herzton 21st June 2020 03:02 AM

Quote:

Originally Posted by mcbpete (Post 14811281)
Been using and loving DSEQ since my purchase a month or two ago. Now that it's added so much there's no reason I need to get Zynaptiq Unflilter on sale or Gulfoss right .... right ?!

I think while Gullfoss covers some similar ground as DSEQ (you can e.g. also de-ess with it), it is different enough that it warrants having both. Gullfoss for example doesn’t only do eq cuts but also boosts at the same time. It is also less “technical” than DSEQ, which gives you the traditional controls of a dynamic equalizer with lots of options for tweaking. Gullfoss offers a simple UI with much less knobs (basically “tame” and “recover”).

So a bit simplified you could say: DSEQ is for fixing problems , Gullfoss for giving the final touch. You can absolutely use both on the same track.

b0se 21st June 2020 09:59 AM

DSEQ in tracks, Gullfoss on busses.

abduction

alibling 21st June 2020 11:59 AM

Quote:

Originally Posted by b0se (Post 14812499)
DSEQ in tracks, Gullfoss on busses.

abduction

Really need to test that.

mcbpete 21st June 2020 02:56 PM

Quote:

Originally Posted by b0se (Post 14812499)
DSEQ in tracks, Gullfoss on busses.

abduction

Is it that Gullfoss is better at processing busses than DSEQ or that's the workflow one should be employing? As there's a mixture of individual instrument and 'Master' presets I was under the impression DSEQ handles both tasks

Quote:

Originally Posted by Herzton (Post 14812173)
Gullfoss for example doesn’t only do eq cuts but also boosts at the same time.

Not sure if this is the same thing but I know when 'expansion' was raised a few weeks ago this was suggested by TBProAudio:

Quote:

Originally Posted by TBProAudio (Post 14711657)
What about dimming by -6dB (threshold) and increasing master gain by +3dB?
This dims the spectrum hills by -3dB and lifts the valleys by +3dB?
Same effect?


TBProAudio 22nd June 2020 06:05 AM

Quote:

Originally Posted by mcbpete (Post 14812812)
As there's a mixture of individual instrument and 'Master' presets I was under the impression DSEQ handles both tasks

DSEQ can handle both cases: surgical intervention on tracks, broad on buses/master. Just a matter of DSEQ set-up ;)

Quote:

Originally Posted by mcbpete (Post 14812812)
Not sure if this is the same thing but I know when 'expansion' was raised a few weeks ago this was suggested by TBProAudio:

Threshold -6 and master gain +3 creates some kind of expansion, right.
Another approach is to enable AB-LM Lite (slow) instead of adjusting the main gain.

In any case: Removal is preferable to adding while forming the spectrum (EQing)...

b0se 22nd June 2020 10:44 AM

Quote:

Originally Posted by mcbpete (Post 14812812)
Is it that Gullfoss is better at processing busses than DSEQ or that's the workflow one should be employing? As there's a mixture of individual instrument and 'Master' presets I was under the impression DSEQ handles both tasks

Not sure if this is the same thing but I know when 'expansion' was raised a few weeks ago this was suggested by TBProAudio:

In my testing I've found that Gullfoss is best used as a 'spatial enhancer', a final touch. When fed a bus it's capable of carving out more separation of the elements, via clarity around the edges, making the source sound more 3D.

I never use it to 'fix' things, as I would DSEQ.

alibling 22nd June 2020 11:18 AM

Quote:

Originally Posted by b0se (Post 14814214)
In my testing I've found that Gullfoss is best used as a 'spatial enhancer', a final touch. When fed a bus it's capable of carving out more separation of the elements, via clarity around the edges, making the source sound more 3D.

I never use it to 'fix' things, as I would DSEQ.

Thats how i use it normally, but only on the Mixbus. Will try to use it on more Busses and see how it works out for me.

dan98 22nd June 2020 11:29 AM

Yikes, this is a great plugin.

Would it be possible to have a keyboard modifier for adjusting Q without resorting to the numeric display? This would greatly speed things up.

Also, a version with the ability to make static 'regular' EQ adjustments would be really nice, rather than using a separate plugin for that.