Korg Forums

[menara]

As AntonySharmman cleverly found out, the Pa4x sample compression uses a lossy compression algorithm. This method is supposed to eliminate only the sounds which are not noticeable.

But AntonySharmman also stated that he could hear the difference between a compressed and an uncompressed sample. So i was curious to see if i could also hear this difference.

After compressing some multi-samples, at first, it seemed to me that their sounds became flat, muddy and their bass were less punchy. But the next day, i was used to the compressed sounds and couldn't remember the difference with my uncompressed sounds.

I thought it was just a subjective impression but after reloading the uncompressed sounds, i realized that they are indeed much better.

To dig deeper and let you make your own opinion, i extracted with an audio editor the differences between 3 compressed samples and their uncompressed versions (16bits-41.1 Khz : low, mid and high notes). The comparison waveforms are copied in the files named DeltaLow, DeltaMid and deltaHigh.

The compression is made under the V3.1 OS and the decompression is also done by the Pa4x's undocumented functionality when saving/exporting a single compressed samples.

The 9 files can be downloaded here : https://www.4shared.com/rar/eItFliWEiq/Test_Pa4x_v31_Compression.html

As you can see and hear, the difference is noticeable and all the frequencies are altered by the compression. No wonder why the sample compression wasn't implemented when the Pa4x was released and we definitely deserve a better compression algorithm.

Its up to you to decide if, the approximately 50% ratio compression, justifies this loss in quality.

In conclusion, and as confirmed by the manual, always save a copy of your uncompressed multi-samples.

kind regards
_________________
Pa4x 76 - Sy77

[siebenhirter]

Hello,
thank you very much for your documents
_________________
kind regards
- siebenhirter, austria -

Interesting facts about styles and stylePlayer functions can be found at http: www.elmarherz.de

[Biggles]

Thankyou for posting your research.

A self degrading keyboard is a good enough reason not to spend £2500 on a new PA4X.

There is no way I am spending £2500 an old model.

In the last couple of years Yamaha have released the Genos and three SX models, what has Korg done in the arranger market?

Zilch.

The PA4X v3 Next OS is coming up for two years old, the OS of the 700 and 1000 is even older.

Yet Korg released three uodated or new Synths and two DP’s at NAMM 2021, so that they now have 21 synths or variations listed in their website.

Korg Arranger users are being very badly neglected.

Rant over.

PS
Don’t bother responding, the only thing that will change my negative views of Korg is Brand New Korg Arrangers being released.

Until that happens Korg sucks.
_________________
Biggles
Lancashire, UK

[siebenhirter]

I'm much more humble there - my negative views will change a little bit, if finally for my favorite keyboard (Pa1000) Korg would fix the missing or faulty features with an update, so that it would offer the same comfortable styleplay functionality as its predecessor (ie Pa800/500)

---> lack of support - forgotten again for a NEXT YEAR
---> http://www.korgforums.com/forum/phpBB2/viewtopic.php?t=123875
_________________
kind regards
- siebenhirter, austria -

Interesting facts about styles and stylePlayer functions can be found at http: www.elmarherz.de

[AntonySharmman]

[Asena]

Great review

Thank you.

I have allways told that it,s not a god thing to make compression on sounds .
_________________
www.globalsound.se

KORG PA 5-X/YAMAHA GENOS 2/YAMAHA A 5000
LIONSTRACK X 76 & GROOVE XR
MEDELI AKX-10




MacbookproM2-Ssd/Logic/Neuman/Kali Audio8/Komplette14SDD/ Apollo Twin/PIONEER XDJ RX 2
LOTS OF SAMPLE SOUNDS!
KorgPaManager V 5

[Aripearlmusic]

Since the people in the audience are not audiophiles and have never heard the original sample before compression i dont see any reason to leave anything uncompressed
_________________
Ari Pearl Music
Sound Developer for the 5x and 4x at https://performersheaven.com/en/shop-en/by-author-en/ari-en

[siebenhirter]

[Korghelper]

[Aripearlmusic]

lol my hearing range is from 7hz to 38.6khz due to a surgical error. Let me dumb it down for you what I meant was if the audience hasn't heard the "before" vs "after" they will have no idea that there even is one. Also if your samples don't sound as good after file compression you can EQ it or sample it better with more saturation and then use the settings to tweak it after compressing
_________________
Ari Pearl Music
Sound Developer for the 5x and 4x at https://performersheaven.com/en/shop-en/by-author-en/ari-en

[fResH_]

I never use the compression of pa4x due to I noticed a difference in sound quality (worse).

And forgett about the audience understandable, Its for my own satisfaction to know I get the best sound quality out of the system.
_________________
// Matthew

"Its not how good you are, its the impression you leave behind."

[BR]

[Aripearlmusic]

as long as the samples are recorded with the right gain staging and audio compression to maximize the normalizing but not be boosted as much there will be no noticeable difference after file compression
_________________
Ari Pearl Music
Sound Developer for the 5x and 4x at https://performersheaven.com/en/shop-en/by-author-en/ari-en

[AntonySharmman]

Skilled Sound engineering is strictly attached to advanced knowledge of Physics and ears experience in sound is a very secondary thing to mention
in rules of physics that can be calculated with Maths or even can be illustrated by scientific lab gear and will always reject people's subjective speculations
of what is the correct result of re-produced sampled sound since the solution of equation is crystal clear ... reproduced sound is as good as the output
waveform is geometrically identical to original recorded one that spectrum analyzer will result.
Our ears that are mostly effected by our Psychoacoustic mood Phenomenon can never be objective , but lab instruments will always be !

Therefore I will note a couple of things :

- Audience can mostly understand the difference of real & fake , 'cause they have already listened to a real Band , Drums , Piano etc , in another performance
scene or even by their home Hi-Fi audio system from decent Music production stuff , but who really cares if you already know that as a performer and you can
be compromised with less than best !
- You can never recover lost frequencies by EQs and FX tricks or you can never "just" enhance those frequencies before compression due to preserve them ...
you will totally fail , so in that case of rich/detuned and high frequencies , you never compress samples !
- People do not hear ultrasonic frequencies no matter of what they claim , so I give you an example of 2 frequencies 20KHz & 28KHz that a real time cymbal
crash will produce among others.
Fourier analysis declare (and can be measured) that those 2 frequencies will produce another 2 derivatives , one outside spectrum 48KHz and a second audible
one at 8KHz that might be the essential reason for a drummer's sonic taste to buy this Zildian cymbal , that's why all frequencies of 15Hz-35KHz spectrum
must be reproduced after Pro recordings in higher sampling rate.

Summary , things in electroacoustic as many people might imagine , are not so simple to be subjective without deep knowledge in Physics and help of scientific lab gear !
Skilled ears are always an asset , but never enough !
_________________
Music Conductor - Sound Engineer & Developer - Automotive SMPS/RF R&D - Electronics Engineer
PaSeries Demos - WavesArt.eu - KorgPa.gr <> Facebook

Keyboards : Steinway-D, Kronos X, Pa5X 76, Pa4X 76, Montage M7 , Roland-XV88, Emu3,Emax II,Synclavier II,Yamaha DX Series, ΟΒ-8V

[midimusa]

I am sorry, but I can not download from this site: "The 9 files can be downloaded here : https://www.4shared.com/rar/eItFliWEiq/Test_Pa4x_v31_Compression.html. " without registrations.
_________________
Korg PA1000.