TAK 1.0 - Final release of the new lossless codec, offers superior performance Options

[pepoluan]

At first, you end up with scepticism from people like me, who'd like to just take a peek at the code, just to be sure that the benefits claim are true and to be able to see the perks...

Now I'm not a programmer, so I might be way off base here. But why do you need to look at the source code to alleviate your skepticism when TAK has been heavily tested by many people and, more importantly, has been publically available for anyone to try for the last 3 iterations (2 betas and the final)? I can certainly understand preferring open source software, and this post is in no way a comment on whether TAK should be open or closed, but the notion that you need to look at the source code "to be sure that the benefits claim are true" strikes me as nonsensical.

I agree. Even looking at the source code is no way to determine if a complex algorithm (which I believe TAK is) works. The only way to know if it works or not is to test.

Thomas has provided the binaries. All you need to "... be sure that the benefits claimed are true ..." is to compress a WAV file using the binaries, and time it with whatever means you have. Then decompress it, again timing it. Then compare the result of the decompression with the original.

I am very sorry to say that some open source advocates in some way resemble audiophooles.

[boombaard]

At first, you end up with scepticism from people like me, who'd like to just take a peek at the code, just to be sure that the benefits claim are true and to be able to see the perks...

Now I'm not a programmer, so I might be way off base here. But why do you need to look at the source code to alleviate your skepticism when TAK has been heavily tested by many people and, more importantly, has been publically available for anyone to try for the last 3 iterations (2 betas and the final)? I can certainly understand preferring open source software, and this post is in no way a comment on whether TAK should be open or closed, but the notion that you need to look at the source code "to be sure that the benefits claim are true" strikes me as nonsensical.

Some people perhaps don't understand the notion 'bit identical before and after compression & at the stated compression level when compressed'?

I suppose it is logically conceivable that the TAK encoder was actually a rootkit that made any files with the extension .tak have an arbitrary cool-looking %age compared to encodes with other lossless codecs of the same input data, but it seems more likely the individual in question here has been effectively conditioned by the (imo sometimes somewhat idiotically zealous F/OSS crowd that is commanded by the FSF).

It's really odd how think about how his question would have sounded lots more reasonable if he had said he was "just curious to see how and what algorithms were implemented", in order to understand why FLAC(wavpack too¿?) was so much slower and compressed so much worse (no offense josh ), rather than this rather idiotically skeptical sounding question "just because the source isn't available YET" (mind you, it's not as if it hasn't already been pretty much guaranteed it will at some time in the reasonably near future be released).

yes, the FSF has its uses.. but the crusading it inspires seems almost medievally unreasonable

This post has been edited by boombaard: Jan 29 2007, 13:44

[PabUK]

Congratulations Thomas on the first final release of your incredible new lossless codec. I'm yet another person who has been silently watching the development of TAK since your first post on April Fool's day (and at that time was sure you were just joking). It's been really enjoyable to watch it grow from there to this point, and I look forward to seeing what comes next.

This post has been edited by PabUK: Jan 29 2007, 14:24

[MaB_fr]

That's a lot to answer and many misunderstanding to correct.
[beside, calling me idiotic and stupid will in no way affect me...so you can go on ]

First, i'm in no way related to any work and/or people involved in ANY activity in the FSF or ANY OpenSource movement (apart from the fact that i've a computer with Ubuntu installed...one on so many, and that i've released WMPTSE under a "public domain" license).

Anyway, my point wasn't about "i dont believe your codec isn't doing what your are claiming".
My point was, if you choose to not release the source SAY it right now.
Everybody can endure a closed source codec, and as the developper of WMPTSE, i'm totally open to this.

We just want to know without having to browse all Hydrogen Audio Thread about your codec.

For the "rootkit part", it is true that i believe a piece of closed source software is a security danger in ANY operating system and especially under Windows. There's ways to ensure a program is not doing anything wrong (anti virus and such) to a certain extend.

I DO NOT say that's what your program is doing, i'm just saying that you will endure SUSPICION (mostly from paranoid people like me) as soon as some user have any kind of strange configuration problem with it.

Now, if you are already bored of answering stupid question, i'm sorry to predict nightmarish days when you widely release it (outside HA).

Last, i don't need the source code to compare a file bit-to-bit, thank you for the tips...
I need the source code to have confidence. I can trust a program to behave as deeply as i am permitted to understand its inner working...
You don't give source, i don't give full thrust...That was my point.

Again, i can live with it. But i NEED TO KNOW your politics... in the related thread, you give no answers TAK & License. First you say no to "publish code on release", then you say yes to "publishing code someday", and you conclude by "i've no time to spend on deciding"...

In a way, i was trying to put pressure on you to decide.
To me, it seems Release is the time to make these choices.

Maybe you don't agree...

At least, let's discuss it.


MaB_fr

This post has been edited by MaB_fr: Jan 29 2007, 14:42

[TBeck]

Again, i can live with it. But i NEED TO KNOW your politics... in the related thread, you give no answers TAK & License

Well, i should be more service oriented... It wasn't sufficient to guide you to the previous page of this thread. I better had copied the post from page 1 for you:

My current plan:
...
3) Source code conversion will be performed in small (bearable) portions, when i have time. Absolutely no promises about a release date!

In a way, i was trying to put pressure on you to decide.

Your "Pressure" has been the reason for me to react less reserved than usual. Probably it on it's own would not have been sufficient, but trying to put pressure on me while putting absolutely no effort into own (easy) evaluation of this question simply is too much!

Now, if you are already bored of answering stupid question, i'm sorry to predict nightmarish days when you widely release it (outside HA).

Thanks for beeing so careful! But 'already' seems a bit misplaced here, if you take into account that i am answering thoses questions since about 9 months. Usually no problem. There must be something special with your post. But i told you above.

Thomas

This post has been edited by TBeck: Jan 29 2007, 15:13

[MaB_fr]

I'm about to make a volcano of yourself

But i'm sorry, you didn't answer the question just yet...

What will it be (i simplify to make all things crystal clear) :
- Totally close (ok, you say no to this one, so we just pass it) ?
- Closed source (other will not be able to modify it [even if you publish the code]) ?
- Open source (other will be able to modify it under certain condition) ?
- Public domain (you let your code in the wild without condition) ?

Are you exploding yet ?

MaB_fr

[Synthetic Soul]

For someone who wants to look at the source code of every application they use you seem incredibly bad at reading posts. You said that you'd looked at the TAK - Source code release and conversion thread, in which is written (on whatever it is they write on up there):

When i release the source code i want it to be used by others. I will choose a license which makes this easy. Probably GNU. But i have to admit, that i don't know too much about the differences of open source licenses. I will deal with this when the source code is ready.

Now, this may not be a definitive or conclusive answer, but if you bothered to read the single page in that thread, and the first in this, you would have all the answers you are going to get at this time. If you don't have faith in Thomas' word then please, move on.

[Martin H]

TBeck has said numerous times that the sources will be opened when he has ported them from Delphi to C/C++, but that this project will take some time, so no release date is promised(for the sources, i mean).

Sorry, Synthetic Soul was faster than me

This post has been edited by Martin H: Jan 29 2007, 15:40

[TBeck]

I'm about to make a volcano of yourself

Any indication for this happening?

But i'm sorry, you didn't answer the question just yet...

What will it be (i simplify to make all things crystal clear) :
- Totally close (ok, you say no to this one, so we just pass it) ?

And this already answers your earlier question:

I need the source code to have confidence. I can trust a program to behave as deeply as i am permitted to understand its inner working...
You don't give source, i don't give full thrust...That was my point.

- Closed source (other will not be able to modify it [even if you publish the code]) ?
- Open source (other will be able to modify it under certain condition) ?
- Public domain (you let your code in the wild without condition) ?

Why raise new questions?

I will not answer them now, simply because i am not sure about the details. I am prefering to write (hopefully useful) code instead of dealing with open source bureaucracy.

Are you exploding yet ?

Continue dreaming...

This post has been edited by TBeck: Jan 29 2007, 15:46

[MaB_fr]

Then, sorry for your users and the willing developer...

I will not bother you anymore...just enjoy the release

Good luck anyway, keep containing the fumes


MaB_fr

[Shade[ST]]

I'm making a FAQ for you, Thomas. I'll link it here when I'm done.

[TBeck]

' date='Jan 29 2007, 17:16' post='467940']
I'm making a FAQ for you, Thomas. I'll link it here when I'm done.

Thank you!

Put please let me read it first, before publishig it.

Nice to see(read) you again

Thomas

edit1: You have been faster! It's perfectly ok!

edit2: Now i have been too fast.

Two corrections:

1) My name is Thomas Becker, not Thomas Beck.

2) "it is most likely to be very easy to decode on hardware, even in its most compressing modes"

"even in its most compressing modes" is questionable.

This post has been edited by TBeck: Jan 29 2007, 17:40

[Shade[ST]]

Yeah, sorry. I edited it again. I do need a description for forward prediction.

Here's the link. Keep it in your sig?

http://www.hydrogenaudio.org/forums/index....showtopic=52276

I updated to your 11:40 edit.
I was wondering... what's the take on 32-bit floating point files? Can you compress 'em? Should I read the readme? Probably, eh?

This post has been edited by Shade[ST]: Jan 29 2007, 17:45

[Eli]

Sounds pretty good. Well will it be added to the lossless comparison chart? It will need alot more support before I could consider migrating. To bad its not OS, or the support may be able to be added faster.

[rjamorim]

http://www.rarewares.org/

http://www.rarewares.org/lossless.html

[jido]

Glad that version 1.0 is out. Congratulations!



Looking forward to see someone experiment with a freepascal port...

[vhl]

Is it possible to make version with multicore support? Almost all new processors is multicore, why developers do not use this advantages?

[Fandango]

I'm pretty sure that this has to wait.

[jcoalson]

Is it possible to make version with multicore support? Almost all new processors is multicore, why developers do not use this advantages?

it's complicated, difficult to get right, and non-portable. better to spawn multiple processes for individual files and let the OS handle it.

[Shade[ST]]

TAK's actual implementations are generally only limited by IO speed, in case of "turbo" mode. I'm not sure that a multicore implementation would be that useful.

[vhl]

Is it possible to make version with multicore support? Almost all new processors is multicore, why developers do not use this advantages?

it's complicated, difficult to get right, and non-portable. better to spawn multiple processes for individual files and let the OS handle it.

What about WinRar? What about video codecs? It's true and it's present! Why i must use only one core in my multicore system? Why i must run multiple copies instead of put all files in one queu and just run it?

' date='Jan 30 2007, 13:09' post='468209']
TAK's actual implementations are generally only limited by IO speed, in case of "turbo" mode. I'm not sure that a multicore implementation would be that useful.

Then it's possible improve speed in other modes!

This post has been edited by vhl: Jan 30 2007, 20:27

[sPeziFisH]

Thanks Thomas, the coding/decoding-abilities of the new codec seems to be amazing *thumbs up*
I am really astonished that you get it to achive better compression and speed than the other lossless-codecs which have been optimized over a long period - due to a better basic structure/concept or due to smartish coding?

Feinste Grüsse

[Firon]

Is it possible to make version with multicore support? Almost all new processors is multicore, why developers do not use this advantages?

Properly multi-threaded code is not trivial to do.

[JunkieXL]

Just wanted to say Thank You Tbeck!

I'm really happy to see new lossless codecs being developed and improved upon. I'll probably have to wait until the SDK is out and some playback support is available to fully test it out, but I'm definately interested/excited about this project.

Keep up the good work!
JXL

[Martin H]

Why i must use only one core in my multicore system? Why i must run multiple copies instead of put all files in one queu and just run it?

Would you please STFU and stop whining! Nobody is forcing you to use non-SMP enabled software in the first place and it doesn't make it better that you're whining at a freeware developer, where you haven't even paid a nickle for using his work!