IPB

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
For the Music, build DSP chain effective for all kind of sound
navizan
post Dec 2 2013, 20:12
Post #1





Group: Members
Posts: 5
Joined: 29-November 13
Member No.: 112731



Hi,

I Listen at all kind of music from classical to generated sound samples and noises and I have always suffered the following :
I am playing some classical tunes fine tuned with EQ and volume settings but after a while I have taste to play some pumping electro or heavy metal for coming back again to classical and maybe passing through the acoustic jazz or Soul, why not? So whats the issue? Well, I need to entirely retouch all the EQ and volume setup again and again... style by stile...and even worste.. track by track....Yes in foobar is possible to assign a dynamic DSP for each track but come on...calibrating EQ and/or volume for each track becomes tedious and pointless. Because after that I loose the taste and the ear... coming out with unbalanced poor results and even forgetting the reason why I wanted to listen at that particular song.. and also the time that goes wasted..
I want to enjoy the music, a music that must be ready here and now without having to waste that fleeting moment ever!
I always wanted the true sound reproduction as meant to be and if it is a bad sound ok but it shouldnt neither sound better or worst.
I always wanted the sound as near as possible as the one you are getting when in front of live acoustic intruments or PA speaker when listening amplified ones, where you can feel the energy in the molecules, the electricity danger, the good vibes... And guess what? I always ended up at home with that kind of compromised consumer sound even if using several and different expensive equipments.
So I got tired and I needed to find a global solution for me since my passion for the music is always pushing behind.
I needed to build my own DSP chain setup which should work with all music genres and rms levels, giving a realistic audio impact in the entire
frequency range from DC to the upper band limit where the 50hz would not prevails or cover the 16hz or 12khz the 20khz... you know what I mean.. anyway not a kind of loudness tool/bass boost/ultra crystal ...etc available out there. Instead to be something that cleans and re-balance, something gentle when it comes to being gentle and adjuvant when is the case. Something that enchances what was there already in the beginning, something that puts it naked and raising the maximum of your sound system as it is and without having to adjust any settings...and the more you have the more you try...
This DSP chain is intended to decompress any sound and expand it without overclipping ready to shoot it at full speed & power directly to the transistors
/mosfet/valves stage of the Amp.
So in order to work properly with these samples ( or at least in order to do the same way I do and it works for me ) you should set the
final power Amp or the active speakers to the MAX output level without any gain reduction in-between and drive the desired volume only by the last volume control of your soundcard ( use the wasapi driver in order to avoid any built-in sound enhancers of the soundcard/S.O. ) The soundstream should be as simple 1=soundsource-2=Amp-3=Speakers. No middle man. And if you have some then set all EQ-Tone controls to defeat and play.
Here the music samples re-touched with this DSP chain. To me it seems working fine with all kind of sounds but please post some of your favourite tunes that sound already at top notch for you, I shall pass them through the DSP chain and let's see if they sounds even better.
When you would open the samples in wavelab you would notice a gain reduction, instead the gain missing is only gone to raise up the extremely deep and higher frequences. Its a trade off, with this DSP you will have to give up about 45% of the power of your amp because that reserve goes in part to energyze what ofter is not considered much.. and the rest to reserve plenty of dynamic for peaks. Try by yourself, at any volume level the sound should remain consistent and engaging.

* URL removed by moderation *

If you are probably asking : so why not releasing the DSP chain directly and let us try by ourselves if it works or not? Correct, but before the release I just wanted to challenge the idea behind this and put it in discussion with you expert people since I could be wrong and I could learn something new.
All 3 samples file are in .MPC Q10

Thanks to all of you.

This post has been edited by CiTay: Dec 2 2013, 22:42
Reason for edit: Removed copyrighted material. Please upload shorter samples, not entire songs.
Go to the top of the page
+Quote Post
saratoga
post Dec 2 2013, 21:26
Post #2





Group: Members
Posts: 5156
Joined: 2-September 02
Member No.: 3264



I think per track EQ and a lot of DSP effects are the opposite of accurate reproduction, and probably not what a lot of people would consider a good idea.
Go to the top of the page
+Quote Post
navizan
post Dec 2 2013, 21:35
Post #3





Group: Members
Posts: 5
Joined: 29-November 13
Member No.: 112731



QUOTE (saratoga @ Dec 2 2013, 21:26) *
I think per track EQ and a lot of DSP effects are the opposite of accurate reproduction, and probably not what a lot of people would consider a good idea.


I agree that less is better but since there are so many technical and human variants I though a kind of DSP chain to that blends the sound to my personal taste and then only to play the track in the most simple way. Did you listen at the samples? if yes what do you think?

Cheers
Go to the top of the page
+Quote Post
mudlord
post Dec 2 2013, 22:32
Post #4





Group: Developer (Donating)
Posts: 818
Joined: 1-December 07
Member No.: 49165



what.
Go to the top of the page
+Quote Post
markanini
post Dec 2 2013, 23:10
Post #5





Group: Members
Posts: 555
Joined: 22-December 03
From: Malmö, Sweden
Member No.: 10615



This is not the answer OP is looking for, but what the heck, a simple solution to the problems described:

Collect all the EQ settings you have made for each track and add the values, divide by the amount of tracks, voila you have your ideal personal EQ.

This post has been edited by markanini: Dec 2 2013, 23:50
Go to the top of the page
+Quote Post
CiTay
post Dec 2 2013, 23:19
Post #6


Administrator


Group: Admin
Posts: 2378
Joined: 22-September 01
Member No.: 3



I downloaded your samples and discovered that they contain three complete songs, at least two of which are copyrighted by the respective record companies. Therefore i had to remove the link. You may upload samples which are up to 30 seconds long each.

For the first song "Sleep", i compared your modified version to the original version, and to me it was no improvement, on the contrary. The mids were overpronounced, the bass too weak. It turned into an annoying "wall of noise"-type sound towards the end.

For the samples, it would be a good idea to include a sample of the original version and your version seperately. How do you want anyone to compare them otherwise.
Go to the top of the page
+Quote Post
db1989
post Dec 2 2013, 23:42
Post #7





Group: Super Moderator
Posts: 5275
Joined: 23-June 06
Member No.: 32180



1:
QUOTE (navizan @ Dec 2 2013, 19:12) *
I always wanted the true sound reproduction as meant to be and if it is a bad sound ok but it shouldnt neither sound better or worst.
2:
QUOTE (navizan @ Dec 2 2013, 20:35) *
QUOTE (saratoga @ Dec 2 2013, 21:26) *
I think per track EQ and a lot of DSP effects are the opposite of accurate reproduction, and probably not what a lot of people would consider a good idea.
I agree that less is better but since there are so many technical and human variants I though a kind of DSP chain to that blends the sound to my personal taste
These two sentiments can’t both be true simultaneously. So which one do you actually mean?

Presuming the latter, I’m sceptical about the potential for useful technical discussion from something as subjective as personal preference and situational setup. Especially with a shortage of real information about the process and mostly vague buzzwords instead. Which brings me to this:
QUOTE (navizan @ Dec 2 2013, 19:12) *
anyway not a kind of loudness tool/bass boost/ultra crystal ...etc available out there. Instead to be something that cleans and re-balance, something gentle when it comes to being gentle and adjuvant when is the case. Something that enchances what was there already in the beginning, something that puts it naked and raising the maximum of your sound system as it is and without having to adjust any settings...
In contrast to this, as I recall, the kind of noncommittal language that you use to describe your workflow is very, very similar to the copy employed by Creative with their Crystallizer and other companies with their equivalent offerings. Again, it doesn’t seem to offer much foundation for discussion.
Go to the top of the page
+Quote Post
DVDdoug
post Dec 3 2013, 00:23
Post #8





Group: Members
Posts: 2712
Joined: 24-August 07
From: Silicon Valley
Member No.: 46454



I think what you are trying to do is pretty-much impossible... I don't think there is a perfect solution.

What I'd suggest is this:

1. Get a speaker/room calibration system. There are home theater receivers with this feature built-in, JBL makes the MSC1, or there is software. In any case, you'll need a calibrated instrumentation mic. You might want to consider some acoustic treatment in addition to EQ (you can't EQ-out a null).


Don't go too far with this... Try to cut instead of boost. If you boost by 9 or 12dB you will probably clip your amplifier and maybe damage your speakers. If you are making extreme corrections, you may need different/better speakers or some acoustic treatment.

2. Listen to your best or favorite recordings and make any "final EQ tweaks" to your taste.

3. Leave those main EQ settings and just use your regular bass & treble controls if a song doesn't "sound right".

4. If there is a small percentage of songs that need more adjustment, use your audio editor (Wavelab?) to "permanently" tweak them a bit more.

QUOTE
When you would open the samples in wavelab you would notice a gain reduction, instead the gain missing is only gone to raise up the extremely deep and higher frequences. Its a trade off, with this DSP you will have to give up about 45% of the power of your amp because that reserve goes in part to energyze what ofter is not considered much..
Let's separate the analog & digital... Most digital recordings are normalized (maximized). So, if you boost the bass by 6dB, you have to reduce the overall level by about 6dB (re-normalize) to prevent digital clipping when you re-save.

Something similar happens on the analog side, but typically if your bass is weak you usually have some headroom and you can boost it. If you are clipping your power amp, you may need a bigger amp, more efficient speakers, or some room treatment (bass traps) to reduce acoustic bass-cancellation.

QUOTE
This DSP chain is intended to decompress any sound and expand it without overclipping ready to shoot it at full speed & power directly to the transistors
/mosfet/valves stage of the Amp.
If you are talking about dynamic expansion, I haven't had much luck with it. We don't know the parameters of the compression to un-do it, and with limiting, there is just no way to know what each peak was before limiting. And, any compression used on individual tracks before mixing can't be un-done either.

As far as your power amp & speakers, that's really just a matter of budget. I've got a pair of 15-inch subwoofers, (and about 100W per channel with my home-theater system) which is more than I need, but it may not be enough for you! biggrin.gif

QUOTE
I always wanted the sound as near as possible as the one you are getting when in front of live acoustic intruments or PA speaker when listening amplified ones, where you can feel the energy in the molecules, the electricity danger, the good vibes... And guess what? I always ended up at home with that kind of compromised consumer sound even if using several and different expensive equipments.
The acoustics have a LOT to do with it. I've used my home speakers as part of a DJ setup, and they sound a LOT better in a dance/music hall (a bigger room with nice-natural reverb).

Plus, it can get fatiguing listening at "live levels" for long periods at home. And since most modern recorded music is over-compressed (dynamically), that makes also it more fatiguing and it makes you want to turn-down the volume.

This post has been edited by DVDdoug: Dec 3 2013, 00:32
Go to the top of the page
+Quote Post
navizan
post Dec 3 2013, 07:40
Post #9





Group: Members
Posts: 5
Joined: 29-November 13
Member No.: 112731



QUOTE (CiTay @ Dec 2 2013, 23:19) *
I downloaded your samples and discovered that they contain three complete songs, at least two of which are copyrighted by the respective record companies. Therefore i had to remove the link. You may upload samples which are up to 30 seconds long each.

For the first song "Sleep", i compared your modified version to the original version, and to me it was no improvement, on the contrary. The mids were overpronounced, the bass too weak. It turned into an annoying "wall of noise"-type sound towards the end.

For the samples, it would be a good idea to include a sample of the original version and your version seperately. How do you want anyone to compare them otherwise.


ah sorry for that, it wasn't my intent to spread copyrighted material, I'd prepare the files as you said and re-upload. I am not surprised to hear that you've found the retouched version sounding with overpronounced mids and a kind of weak in the bass but thats exactly what I mean... I think that at some point we all have been addicted from the loudness war...it is difficult to explain but it is like overeating, we dispose always more food on the table since we dispose of bigger tables but at certain point I think this is degenerating because it is true that we possess a bigger table full of food but there isnt anymore space for moving so we remain seated and motionless to cram everything we can, we always crave for more and eventually become obese.
I think this metaphor can be applied to the music when it comes to loudness. We all know that trick : "loud sounds better".."looks like more and more is better" and we dont want to fall again right? I think that our ears have been and still are affected by advertising... Lets take the example with the home theathre systems but it can be applied to everyting else : 300 watt RMS - super twin drive esoteric technology - zero feedback - freq. resp 23hz-24khz +/-3db... and so on... looks attractive from a sales perspective, then you go in the shop and you see it has also a great polished appearance, looks cool, at some point they are like sports cars, they give you that sense of race even when they are parked. Virtually they sound even great when they are off..
Then you bring your new purchase in your home, setup all features and you feel satisfied for a while but then it comes again if you forget the porpose. What the purpose? For me the joy of listening the music, as simple as it is, as humans we can all agree that we can feel emotions even when listening music from a second world war old radio, or to be struck by a good movie seen in a small 15-inch TV of the 60s CRT including the antenna's interference.. What I mean is: that perfect decoder is already built in our brain and it goes far beyond everyting else because is adaptive dynamically to the events as long as you want, it does. But I think this is suffering from overload in our time. The sounds in real nature can be very loud like a thunder, and we, caught off guard could jump but hardly cover our ears because the sound will not be painful. In a stereo system instead that sound would become painful at a certain point because some frequencies tend to dominate over the others and they also occupy their dynamic space thus making a compressed sound. The intend of my DSP chain is try to restore the things by leaving headroom for the things to expand, the volume control and the capacity of the system remains the only physical factors. I do not pretend to reproduce a real thunder in my room but realistic, and realistic for me relies in proportions.
Thank you for your time.
Go to the top of the page
+Quote Post
larryfine
post Dec 3 2013, 17:29
Post #10





Group: Members
Posts: 60
Joined: 21-September 10
Member No.: 84040



Saratoga went straight to the point!

QUOTE (saratoga @ Dec 2 2013, 18:26) *
I think per track EQ and a lot of DSP effects are the opposite of accurate reproduction, and probably not what a lot of people would consider a good idea.


My personal settings for all kinds of music:

EQ = off

DSP = off

tongue.gif


--------------------
loquor mee menti: factus de materia, cinis elementi...
Go to the top of the page
+Quote Post
navizan
post Dec 3 2013, 18:53
Post #11





Group: Members
Posts: 5
Joined: 29-November 13
Member No.: 112731



QUOTE (larryfine @ Dec 3 2013, 17:29) *
Saratoga went straight to the point!

QUOTE (saratoga @ Dec 2 2013, 18:26) *
I think per track EQ and a lot of DSP effects are the opposite of accurate reproduction, and probably not what a lot of people would consider a good idea.


My personal settings for all kinds of music:

EQ = off

DSP = off

tongue.gif

but sure, at the end it all comes to personal taste smile.gif
Go to the top of the page
+Quote Post
navizan
post Dec 3 2013, 20:58
Post #12





Group: Members
Posts: 5
Joined: 29-November 13
Member No.: 112731



QUOTE (DVDdoug @ Dec 3 2013, 00:23) *
I think what you are trying to do is pretty-much impossible... I don't think there is a perfect solution.

What I'd suggest is this:

1. Get a speaker/room calibration system. There are home theater receivers with this feature built-in, JBL makes the MSC1, or there is software. In any case, you'll need a calibrated instrumentation mic. You might want to consider some acoustic treatment in addition to EQ (you can't EQ-out a null).


Don't go too far with this... Try to cut instead of boost. If you boost by 9 or 12dB you will probably clip your amplifier and maybe damage your speakers. If you are making extreme corrections, you may need different/better speakers or some acoustic treatment.

2. Listen to your best or favorite recordings and make any "final EQ tweaks" to your taste.

3. Leave those main EQ settings and just use your regular bass & treble controls if a song doesn't "sound right".

4. If there is a small percentage of songs that need more adjustment, use your audio editor (Wavelab?) to "permanently" tweak them a bit more.

QUOTE
When you would open the samples in wavelab you would notice a gain reduction, instead the gain missing is only gone to raise up the extremely deep and higher frequences. Its a trade off, with this DSP you will have to give up about 45% of the power of your amp because that reserve goes in part to energyze what ofter is not considered much..
Let's separate the analog & digital... Most digital recordings are normalized (maximized). So, if you boost the bass by 6dB, you have to reduce the overall level by about 6dB (re-normalize) to prevent digital clipping when you re-save.

Something similar happens on the analog side, but typically if your bass is weak you usually have some headroom and you can boost it. If you are clipping your power amp, you may need a bigger amp, more efficient speakers, or some room treatment (bass traps) to reduce acoustic bass-cancellation.

QUOTE
This DSP chain is intended to decompress any sound and expand it without overclipping ready to shoot it at full speed & power directly to the transistors
/mosfet/valves stage of the Amp.
If you are talking about dynamic expansion, I haven't had much luck with it. We don't know the parameters of the compression to un-do it, and with limiting, there is just no way to know what each peak was before limiting. And, any compression used on individual tracks before mixing can't be un-done either.

As far as your power amp & speakers, that's really just a matter of budget. I've got a pair of 15-inch subwoofers, (and about 100W per channel with my home-theater system) which is more than I need, but it may not be enough for you! biggrin.gif

QUOTE
I always wanted the sound as near as possible as the one you are getting when in front of live acoustic intruments or PA speaker when listening amplified ones, where you can feel the energy in the molecules, the electricity danger, the good vibes... And guess what? I always ended up at home with that kind of compromised consumer sound even if using several and different expensive equipments.
The acoustics have a LOT to do with it. I've used my home speakers as part of a DJ setup, and they sound a LOT better in a dance/music hall (a bigger room with nice-natural reverb).

Plus, it can get fatiguing listening at "live levels" for long periods at home. And since most modern recorded music is over-compressed (dynamically), that makes also it more fatiguing and it makes you want to turn-down the volume.


Thanks a lot for your detailed response
yes I also had always no luck when dealing with compressed material and trying to decompress and I am such a perfectionist, sometimes too much i think I am sick ...and maybe I really am but with this DSP chain I dont know what happened or maybe I have been just lucky this time.
I have used to manipulate with sound since i started play with music and it was long time ago with an electric organ from 70" tweaking and experimenting with the analog circuits inside to create strange sounds and I am doing that still today using the PC and sometimes creating remixes of old tunes, with the endless possibility of the multitrack and the effects.
So my ears are well trained to the differences and so many times is happened to me, with analog or digital world, where I though I have found the holy grail of my sound and then few weeks later I had to step myself back with shame, because it wasnt so good anymore. Thats why I have never posted nothing about it, just because i though this was just my pure impression, illusion, and it was only due my mood and personal circumstances and i didnt want to bother no one with that.
But this time it is lasting, after working long time and without stress and freely it seems that I have found the sound i always wanted, I mean, marriage is lasting and there is no 'divorce on the horizon blink.gif ... that sounds strange first to myself trust me.
I keep listening to find the issue but I can find , so thats why i decided to write at you expert guys, to compare the results and to hear different opinions, and thats why I have posted the samples better than many words, samples that now have been removed ( ok I'll repost them with max 30 sec lenght as said ) just to listening at them with more than two ears only...
Have a nice eve!
Go to the top of the page
+Quote Post
CiTay
post Dec 11 2013, 16:07
Post #13


Administrator


Group: Admin
Posts: 2378
Joined: 22-September 01
Member No.: 3



So, what about the samples then? I was hoping others could hear this and realize the discrepancy between what you're saying and how you've actually influenced the sound. Because to me it appeared that the end result of your work was exactly what you were trying to avoid; a more compressed, noisier sound that is more stressful to listen to. I have a setup with only slight adjustments on my amp, no EQ or driver enhancements, and (as far as consumer parts go) neutral speakers. So if your samples sound more annoying on my setup, one of us is doing something wrong.
Go to the top of the page
+Quote Post
2Bdecided
post Dec 11 2013, 17:06
Post #14


ReplayGain developer


Group: Developer
Posts: 5362
Joined: 5-November 01
From: Yorkshire, UK
Member No.: 409



EQ is useful for fixing bad systems and bad recordings.

It can be used to fine-tune the balance of a decent recording, but few people have the skill to do this.

On a really decent audio system, for the vast majority of recordings, using EQ will make them sound worse.

Cheers,
David.
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



RSS Lo-Fi Version Time is now: 22nd December 2014 - 00:19