The whole point of automatic matching is not to have it all perfect, but for those who dont have the best eQuipment and monitors, harbal is a nice hands on solution.
But still i am doing things in a automated fashoin with harbal, every time i start the same, alining the Freq Spec en bottom / mids / hi etc.
Since i automated that, i can first do that one, and afther that still do hands on things myself.... What do you do in harbal ? The same things over and over again.... ? Surely something can be automated.
And if you dont have the equipment , what would be better than a program to help you out....
suggestions and kill me now
Comment for you
Sined
When you load a track into Har-Bal each one is always going to be different. Stand back and look at the track for a second.
Only concentrate on the peaks (yellow) and valleys (green) and make the adjustments.
You will find that some songs only require a minimum adjustment. It appears that you are loading songs into Har-Bal and trying to match them to a reference file. Try to steer away from this if you can unless the track is completely screwed and needs full frequency modification
If you have a track that already sounds decent, you should not need to use a reference track. You only need to look for the peaks and valleys...
Automatic matching would completely defeat the dynamics of your track.
Earle
When you load a track into Har-Bal each one is always going to be different. Stand back and look at the track for a second.
Only concentrate on the peaks (yellow) and valleys (green) and make the adjustments.
You will find that some songs only require a minimum adjustment. It appears that you are loading songs into Har-Bal and trying to match them to a reference file. Try to steer away from this if you can unless the track is completely screwed and needs full frequency modification
If you have a track that already sounds decent, you should not need to use a reference track. You only need to look for the peaks and valleys...
Automatic matching would completely defeat the dynamics of your track.
Earle
Do you do everything exactly the same or just follow a similar process?
The adjustments require for a given track are all different even though the final spectrum shape is similar. In mathematical terms the whole process is ill defined which is why it isn't easy to implement. You could say it is a fuzzy process and perhaps fuzzy logic may one day provide a solution but that requires research time to develop which we are a bit short on at the moment.
If I had a solution that works I would have put it in but I'm not about to put in something half baked that ruins the dynamics of tracks more often than not. If you'd like your tracks to sound the way songs sound on commercial radio then perhaps you have cause to complain. I prefer my CD's to sound a bit more dynamic and not spectrally squashed. After all, when it get's played back on radio their multiband dynamics processing will give it the radio sound anyhow.
Regards,
Paavo.
The adjustments require for a given track are all different even though the final spectrum shape is similar. In mathematical terms the whole process is ill defined which is why it isn't easy to implement. You could say it is a fuzzy process and perhaps fuzzy logic may one day provide a solution but that requires research time to develop which we are a bit short on at the moment.
If I had a solution that works I would have put it in but I'm not about to put in something half baked that ruins the dynamics of tracks more often than not. If you'd like your tracks to sound the way songs sound on commercial radio then perhaps you have cause to complain. I prefer my CD's to sound a bit more dynamic and not spectrally squashed. After all, when it get's played back on radio their multiband dynamics processing will give it the radio sound anyhow.
Regards,
Paavo.
Responce to Earle
I agree, but that is not what i am afther.
I am not using a reference file of just another song to accomplic wishfull matching. At this time my program can do spectrum-analysis on multiple tracks at once, and calculate Max, Average, Min. Just a system i used (we) to do, and actually its quite similar to harbal.
As for now i make a reference file (audio track) and its no so important how its been created (maybe its our secret?) , think of it like a audio ref file or call it the calculated output of lots of music of the same genre or whatever you select to be included in the ref!
The REFfile beforehand (before i started this new programming) consisted mostly of REF tracks from other studio's , selfmade or gained from others (commercial rec etc.)
Now the REFfile is more like a database of analysis file , as opposed to real audio files, so its a analysis output REFfile now.
What i think you can always do on forehand is EQyour mix towards a good sounding sound, so what is this soundspectrum like?
Well i made a program to analyse that yourself or take an old REF and use that from the database. Anything you like...
The program suggests EQ corrections back to the user by a press of a botton (auto yeah!) and suggests it bij TEXT or a file for the PAS-EQstudio-program. As with testing this general REF suggestion is quite awkward to start seeing what is wrong with the mix, or if all is within a certain border, you can use the EQ suggestion on the material.
This balances (sorry to use that word also) the EQ in the mix, as you would do hands on.... (and everybody at home or studio will do the same on any EQ or system, make a sound generally acceptable).
Then i would start to hear and see if i can hear the spikes in EQ, i use harbal for that now, but actually its a matter that i use the PAS-32EQ cause its not a HIresulution filter but 32 Linear EQ points, that is why i cannot automaticly DE-SPIKE the EQ. Its more a EQ failure on the output that is causing that its not really automated, and things are still hands on.
I was searching for a EQ-filter and found one for VB (and C++) but did not integrate that (yet, time for programming and time for audio and fun needs also space to think and debate about the subject over here, ehe)
The deal is;
The TEXTfile output of the program has got the real deal and takes ALL EQ in consideration in steps of 2.6hz (Analyzed 7500 EQ points) when you print it out and do some hands-o EQing.
So actually you can see onscreen the analyzer, see the points in the database or print it as a textfile.
And offcourse i tried myself programming and calculating further to get better results. So its working in a kind of manner.
What actually is most intresting that the decision to cut or raise something can be done now in a automatic calculated manner, and crop as much as like like into the analyzer, make 1 Ref-filew or 1000000 Ref-files from it, calculate anything with anything, its just a matter of choice by pressing some settings and let the computer do some heavy calculations.... yes, its Visual Basic i know, not C++, but for me easy to program, maybe ill transfer it later , but first i just kinda play and check things out.
What i dont like in Harbal (much i do like , but we disagree on some points) that when i see the SPIKES , i can just use the mouse to gain or lower it myself and make mistakes with that... and if i see more than 1 spike but 5 next to eachother, i need to do 5 times the same and make more mistakes on that... its no precise.
If only harbal could reconise the spikes, you could do something better with that (is it a computer or not) , you could equally raise or lower it in a much better fashion without or with the intervention of a user.... now being an abicious Harbal sculpt artist... sorry!
I know, i dont want to ask much of harbal programming behinds, its a solution to anyone who understands the buzz. Allright.
And makes me ask myself , why did not anybody think of this before?
Its more to just explain back what i was programming here, i just love the Mastering Process... and Mastering-EQ was a long time in the darkroom of music until ?? special kind of things going on lately....
Take a look at Elevayta plugins, and see some funny EQ with Cloneboy and Spaceboy, its an eye-opener for dummys. Live and learn, we all do from time to time.
And i am not an artist in EQ or mastering, no way! Still not, thats why!
Hard lessons over many years, and still hard lessons to come.
I just love music, but do not need to spend time anymore mastering-eq, i have done it for way to long... and as to my opions the results are XXXX , well tell you that later when i finish of in VB.
The results are way better than Ozone or PAS-EQ-Studio detection thingies.... it makes them totally obsolete.
The only thing accuarte enough would be Harbal with a great user behind it, who can make better judgement cause he is the artist.
But Dummy's can do a lot with harbal if they understand it a bit and follow the rules...
What are the rules ?
That is what i was finding out and is my program about.
Yes, if i could inport a REF file in harbal and match it. Its the same.
THE REF FILE , MUST CONSIST OF RULES! and is basically doing all users functions in one go... If anybody will say , cant be done ?
Its more like Harbal, it can be done in more like a percentage towards 100% , you never get close to make the goal (can we?) but sometimes we do score one.
And even finding the way towards the goal, is as intresting as it gets towards scoring one...
Its a matter how many times you walked towards the goal, and how many times you scored and do some calculations on that, voila, a percentage hitrate will come out. If you do more of that (analysis and making rules) you can actually make the hitrate better.
That is what i was finding out and is my program about.
Hope i did some good here.
I am not using a reference file of just another song to accomplic wishfull matching. At this time my program can do spectrum-analysis on multiple tracks at once, and calculate Max, Average, Min. Just a system i used (we) to do, and actually its quite similar to harbal.
As for now i make a reference file (audio track) and its no so important how its been created (maybe its our secret?) , think of it like a audio ref file or call it the calculated output of lots of music of the same genre or whatever you select to be included in the ref!
The REFfile beforehand (before i started this new programming) consisted mostly of REF tracks from other studio's , selfmade or gained from others (commercial rec etc.)
Now the REFfile is more like a database of analysis file , as opposed to real audio files, so its a analysis output REFfile now.
What i think you can always do on forehand is EQyour mix towards a good sounding sound, so what is this soundspectrum like?
Well i made a program to analyse that yourself or take an old REF and use that from the database. Anything you like...
The program suggests EQ corrections back to the user by a press of a botton (auto yeah!) and suggests it bij TEXT or a file for the PAS-EQstudio-program. As with testing this general REF suggestion is quite awkward to start seeing what is wrong with the mix, or if all is within a certain border, you can use the EQ suggestion on the material.
This balances (sorry to use that word also) the EQ in the mix, as you would do hands on.... (and everybody at home or studio will do the same on any EQ or system, make a sound generally acceptable).
Then i would start to hear and see if i can hear the spikes in EQ, i use harbal for that now, but actually its a matter that i use the PAS-32EQ cause its not a HIresulution filter but 32 Linear EQ points, that is why i cannot automaticly DE-SPIKE the EQ. Its more a EQ failure on the output that is causing that its not really automated, and things are still hands on.
I was searching for a EQ-filter and found one for VB (and C++) but did not integrate that (yet, time for programming and time for audio and fun needs also space to think and debate about the subject over here, ehe)
The deal is;
The TEXTfile output of the program has got the real deal and takes ALL EQ in consideration in steps of 2.6hz (Analyzed 7500 EQ points) when you print it out and do some hands-o EQing.
So actually you can see onscreen the analyzer, see the points in the database or print it as a textfile.
And offcourse i tried myself programming and calculating further to get better results. So its working in a kind of manner.
What actually is most intresting that the decision to cut or raise something can be done now in a automatic calculated manner, and crop as much as like like into the analyzer, make 1 Ref-filew or 1000000 Ref-files from it, calculate anything with anything, its just a matter of choice by pressing some settings and let the computer do some heavy calculations.... yes, its Visual Basic i know, not C++, but for me easy to program, maybe ill transfer it later , but first i just kinda play and check things out.
What i dont like in Harbal (much i do like , but we disagree on some points) that when i see the SPIKES , i can just use the mouse to gain or lower it myself and make mistakes with that... and if i see more than 1 spike but 5 next to eachother, i need to do 5 times the same and make more mistakes on that... its no precise.
If only harbal could reconise the spikes, you could do something better with that (is it a computer or not) , you could equally raise or lower it in a much better fashion without or with the intervention of a user.... now being an abicious Harbal sculpt artist... sorry!
I know, i dont want to ask much of harbal programming behinds, its a solution to anyone who understands the buzz. Allright.
And makes me ask myself , why did not anybody think of this before?
Its more to just explain back what i was programming here, i just love the Mastering Process... and Mastering-EQ was a long time in the darkroom of music until ?? special kind of things going on lately....
Take a look at Elevayta plugins, and see some funny EQ with Cloneboy and Spaceboy, its an eye-opener for dummys. Live and learn, we all do from time to time.
And i am not an artist in EQ or mastering, no way! Still not, thats why!
Hard lessons over many years, and still hard lessons to come.
I just love music, but do not need to spend time anymore mastering-eq, i have done it for way to long... and as to my opions the results are XXXX , well tell you that later when i finish of in VB.
The results are way better than Ozone or PAS-EQ-Studio detection thingies.... it makes them totally obsolete.
The only thing accuarte enough would be Harbal with a great user behind it, who can make better judgement cause he is the artist.
But Dummy's can do a lot with harbal if they understand it a bit and follow the rules...
What are the rules ?
That is what i was finding out and is my program about.
Yes, if i could inport a REF file in harbal and match it. Its the same.
THE REF FILE , MUST CONSIST OF RULES! and is basically doing all users functions in one go... If anybody will say , cant be done ?
Its more like Harbal, it can be done in more like a percentage towards 100% , you never get close to make the goal (can we?) but sometimes we do score one.
And even finding the way towards the goal, is as intresting as it gets towards scoring one...
Its a matter how many times you walked towards the goal, and how many times you scored and do some calculations on that, voila, a percentage hitrate will come out. If you do more of that (analysis and making rules) you can actually make the hitrate better.
That is what i was finding out and is my program about.
Hope i did some good here.
More for answers..
I know its a deal (and i worked on and of on this program for many years) i do understand its not a thing you can program in a day or to.
I needed to recapture hours of audio-ref material into the program database before i could analyze accurately with it.
So i understand it would be a hassle on your program time (while you are busy making a plugin version of harbal (everybody would whant that and make your marker much bigger)).
I understand fully.
I would be happy to share the program if it would have any functionallity with harbal alongside or anything.
Its still not Dummy prof and pressing a wrong sequence of buttons of loading wrong files make it crash or buggy, its a matter of fashion meets garbage, but improvements are on the way.
I could make a pre-copy of the program but first need some programming time myself, think a final first beta will be around the start of 2005 or end 2004, but i have no deadline.
I am busy now doing some calculations and programming it in,
and dont know if i implement EQ, its more of finding an EQ that rocks and works well in VB/C++, i found one, but testing testing, eh eh
And while the other day , just something else came upon me!
As i know a compressor or loudness-maximiser will do a fine job, it would do better with a nice computer algorithm calculation, instead of getting cut in gain by a threshhold! bang bang, its irritating stuff.... Any math orientated wizzkidd can calculate a MATH-Domein into one other! I did learn in lower school i think ?! So why not on audio ?
Just a matter of heavy calculations again, but it would mean that the basedrum would not smack anymore into the 0db errorfield!
Simple school calculations can do better, or not?
Oeps ! And we can disagree also on this one as much as we like, oh no!
Not again.....
-----------
I am in need of a 7500 point filter or EQ, so...
As i do understand the harbal filter is 8196 points or so, if i could make use of something like that.... its fine with me and would mean a crosslink between harbal and my program.
But i think i have to convince you first.. and thats allright.
So when i have a beta version running i will let you know.
I just like fiddling the database more and do some corrections and finish up....
If you would like to share the filter C++ or any info how i can manipulate the harbal filter... you can use my email for that one.
As it is worth, my program has nothing to do with how harbal works on loudness or its EQ, my program just a big auto audio calculator with an 7500 point analyzer working from -200db to 0db and a database function and multiple calculations on source and ref (take from the databse). So i know, no harbal harmonics corrections in my program..
Its standalone and can only analyze audio files in WAV16bit and 24bit in chunks or complete...
--------------
Well its bedtime... and hope i did some answering right..
Sined
I needed to recapture hours of audio-ref material into the program database before i could analyze accurately with it.
So i understand it would be a hassle on your program time (while you are busy making a plugin version of harbal (everybody would whant that and make your marker much bigger)).
I understand fully.
I would be happy to share the program if it would have any functionallity with harbal alongside or anything.
Its still not Dummy prof and pressing a wrong sequence of buttons of loading wrong files make it crash or buggy, its a matter of fashion meets garbage, but improvements are on the way.
I could make a pre-copy of the program but first need some programming time myself, think a final first beta will be around the start of 2005 or end 2004, but i have no deadline.
I am busy now doing some calculations and programming it in,
and dont know if i implement EQ, its more of finding an EQ that rocks and works well in VB/C++, i found one, but testing testing, eh eh
And while the other day , just something else came upon me!
As i know a compressor or loudness-maximiser will do a fine job, it would do better with a nice computer algorithm calculation, instead of getting cut in gain by a threshhold! bang bang, its irritating stuff.... Any math orientated wizzkidd can calculate a MATH-Domein into one other! I did learn in lower school i think ?! So why not on audio ?
Just a matter of heavy calculations again, but it would mean that the basedrum would not smack anymore into the 0db errorfield!
Simple school calculations can do better, or not?
Oeps ! And we can disagree also on this one as much as we like, oh no!
Not again.....
-----------
I am in need of a 7500 point filter or EQ, so...
As i do understand the harbal filter is 8196 points or so, if i could make use of something like that.... its fine with me and would mean a crosslink between harbal and my program.
But i think i have to convince you first.. and thats allright.
So when i have a beta version running i will let you know.
I just like fiddling the database more and do some corrections and finish up....
If you would like to share the filter C++ or any info how i can manipulate the harbal filter... you can use my email for that one.
As it is worth, my program has nothing to do with how harbal works on loudness or its EQ, my program just a big auto audio calculator with an 7500 point analyzer working from -200db to 0db and a database function and multiple calculations on source and ref (take from the databse). So i know, no harbal harmonics corrections in my program..
Its standalone and can only analyze audio files in WAV16bit and 24bit in chunks or complete...
--------------
Well its bedtime... and hope i did some answering right..
Sined
eh more
If could,
1. import a MY-OWN-REFfile into harbal as a REF.anl file.
2. With a press of a button, match the audo and ref in harbal.
(just match the yellow peaklines or so, its easy to program)
or
1. import a My-OWN-REFfile as a filter into harbal.
(things i still need to do each time..... hands on( or use the PAS-EQ-studio instead).
Its what i asked for the first time around, and still need it anyhow.
Either i will program it myself anyhow or find somebody to help me, or you could help me do it with I would bypass my EQ output problems, and away are the hands on things i need to do each time i work with testing.
zzzzzzzzzzzzzzzzz, thx
1. import a MY-OWN-REFfile into harbal as a REF.anl file.
2. With a press of a button, match the audo and ref in harbal.
(just match the yellow peaklines or so, its easy to program)
or
1. import a My-OWN-REFfile as a filter into harbal.
(things i still need to do each time..... hands on( or use the PAS-EQ-studio instead).
Its what i asked for the first time around, and still need it anyhow.
Either i will program it myself anyhow or find somebody to help me, or you could help me do it with I would bypass my EQ output problems, and away are the hands on things i need to do each time i work with testing.
zzzzzzzzzzzzzzzzz, thx
Yes, thats solved
Doing some final testing afther programming an EQ and preset code.
Problems with filter EQ output are now solved, can now also export to Elemental Audio Firium or use an internal EQ build in the program.
FULL AUTOMATIC MATCHING! without touching any hands on EQ,,,
And it rocks.... It rocks ! Life was never this easy before!
Finishing up in the next month or so....
Sya
Problems with filter EQ output are now solved, can now also export to Elemental Audio Firium or use an internal EQ build in the program.
FULL AUTOMATIC MATCHING! without touching any hands on EQ,,,
And it rocks.... It rocks ! Life was never this easy before!
Finishing up in the next month or so....
Sya
Question
SINED
Why on earth would you want to have totally automatic matching? What about the masked frequencies that are appearing in your track.
This would not be a recommended process for an engineer to understand the many benefits of correct eq'ing.
Earle
Why on earth would you want to have totally automatic matching? What about the masked frequencies that are appearing in your track.
This would not be a recommended process for an engineer to understand the many benefits of correct eq'ing.
Earle
SINED
I am sorry, dont confuse plain basic good mixing with mastering, its not the same... and you should not tempt into going beyond the sound of the mix, .... Most clients whant Their Sound like it is, just pump it up , is what they whant mostly...
Fix it in the mix , then pump it up., that calls it well..
The nice thing about knowing HOW your mix gonna sound like when you kick it through the automatic thingy and master it, Yes there is no difference, you know beforehand where its going when you master it...
If you drag down Mastering to its basics its EQ and Volume that matters, not changing the actual sound of the recoring itself.
And i enjoy my oppenents over here....
Thx!
Fix it in the mix , then pump it up., that calls it well..
The nice thing about knowing HOW your mix gonna sound like when you kick it through the automatic thingy and master it, Yes there is no difference, you know beforehand where its going when you master it...
If you drag down Mastering to its basics its EQ and Volume that matters, not changing the actual sound of the recoring itself.
And i enjoy my oppenents over here....
Thx!
SINED
I did have an output EQ problem;
And i did find a lot of ways now to do the actual EQing, Presets for Elemental Audio Firium, Waves EQ10 are now in place cause i learned how to program VST presets (or directx, dont know).
I do have an internal EQ working but its taking alot more time to fix its internal EQ behavior i will nait that one later.... just set my mind of it now, so i learned to create presets for other plugins.
So the EQ problem is kinda fixed now.... pfeww...
And i did find a lot of ways now to do the actual EQing, Presets for Elemental Audio Firium, Waves EQ10 are now in place cause i learned how to program VST presets (or directx, dont know).
I do have an internal EQ working but its taking alot more time to fix its internal EQ behavior i will nait that one later.... just set my mind of it now, so i learned to create presets for other plugins.
So the EQ problem is kinda fixed now.... pfeww...