KVR Audio

tony tony chopper · Post by **tony tony chopper** » Mon Oct 01, 2007 8:18 am

quoting michael j. fox in 'inside the actors studio':
'2+2 always equals 4.. where's the fun in it?'

(or maybe it was 'two plus two equals four every time… so what good is it?' to be exact)

hibidy · Post by **hibidy** » Mon Oct 01, 2007 10:23 am

C00kie wrote:Because you set something to fire, acting like you don't care, and then go watching in the crowds, giving indications to the firemen

DAW summing threads are definitely dead. Noone needs to proof you that 0 + 0 = 0; 0 + 1 = 1 and 1 + 1 = 2.

It just isn't necessary to try and beat that into me

hibidy · Post by **hibidy** » Mon Oct 01, 2007 10:26 am

the bitch of it is there was a very specific reason I posted this, there is not need to attack anyone (me or otherwise) by posting

there was indeed a certain tongue in cheek-ness to it.

hibidy · Post by **hibidy** » Mon Oct 01, 2007 10:33 am

I can't think of one thing that I've said that is offensive. Oh, maybe I'm wrong, but not offensive

Sorry to have pissed everyone off

C00kie · Post by **C00kie** » Mon Oct 01, 2007 11:19 am

Calm down, no offense intended, Hibidy. Maybe I should have inserted some emoticon.

Anyway, back on topic:

hibidy wrote:provide a link with step by step procedures and the results.

Already discussed was http://www.kvraudio.com/forum/viewtopic ... 48&start=0

But it's performed multiple times:
http://www.kvraudio.com/forum/viewtopic.php?t=151960
http://www.kvraudio.com/forum/viewtopic.php?t=147203
http://www.kvraudio.com/forum/viewtopic.php?t=129341
http://www.kvraudio.com/forum/viewtopic.php?t=129685
http://www.kvraudio.com/forum/viewtopic.php?t=75112
spawned into The Host Comparison Test: http://www.kvraudio.com/forum/viewtopic.php?t=75793
(February/March 2005, so time goes faster than one imagines...)

spaceman · Post by **spaceman** » Mon Oct 01, 2007 11:56 am

all hosts are the same but some are better

Penis Arcade · Post by **Penis Arcade** » Mon Oct 01, 2007 5:15 pm

There are a multitude of reasons while a mix performed in one host will sound different from results produced by another, so this argument is not completely moot and performing tests is a good idea.

However it doesn't fall into the straight "summing" category - this is a simple mathematical operation and everyone here went to comical length to illustrate that.

However, here are a few reasons why a host can sound different, while apparently doing the same thing - let's say triggering a particular beat composed of multiple samples.

- Resampling - to maximize the likelyhood of producing the difference between host, deliberately pick files of different sampling rates, as well as insure that the host clock itself is set to something odd, like 48000. The host will likely be forced to resample the files to an internal sampling rate. I don't think bit depth matter much here, unless you files are in 64bit float, and host is in 32bit float. Then bit conversion will take place as well. Otherwise integer to float conversion is not lossy (afaik).

Timing - this is where the "sample accurate" part comes in - not all hosts trigger samples identically. I think nowadays this will likely be identical among leading popular hosts.

Panning levels - pan your samples, and the host will adjust their levels depending on the panning law that it uses. Different hosts use different panning laws.

Trigger velocities - when a note is placed in different hosts - it may sound at different volume, depending the chosen default velocity. Also depending on the user interaction - there might be velocity curves at play, and other similar stuff.

I am sure this list can be expanded. I think it is important for a particular user to understand the existance and interaction of all those things, instead of choosing a mentality of ignorance and subscribing to some dogma of "summing law is different" or some other "why music made in this sounds worse that in that, except for tarnce".

lowkey · Post by **lowkey** » Mon Oct 01, 2007 11:24 pm

Most of the best music sounds (technically) crap. I thought my stereo was broken when I first heard "Crazy" by Knarls Barkley and "Since I left you" by The Avalanches. Use the tools YOU are comfy with to make YOUR sound.

backpage · Post by **backpage** » Mon Oct 01, 2007 11:33 pm

All the Daw's playback engines sound differently. Are you gonna say with a straight face you don't hear a difference between floating point and integer? Play a single mono track in PTLE and PTHD and tell me you don't hear a difference.

Cordelia · Post by **Cordelia** » Mon Oct 01, 2007 11:53 pm

hibidy wrote:It just isn't necessary to try and beat that into me

That seems waaay over the top. All C00kie said was "you asked for it" and he was literally correct.

You started this thread with the statement "I don't care what the answer is", but you obviously did care, you had an agenda to begin with. When the thread didn't run your way, you responded to Art Evan's very calm, thoughtful post by saying "I have a headache", which was another way of saying "I physically can't tolerate anyone disagreeing with me anymore". That was the first rude post in this thread.

You can run your own tests if you don't like the way these tests have been done before.

Post the results when you're done. If you find something that refutes what's already been proven, we'll be interested. Seriously.

Edit: I see now that backpage has chimed in with another logical, well reasoned post about this subject.

I should have been done with it a dozen idiotic threads ago, so sorry for lengthening this conversation.

Meffy · Post by **Meffy** » Tue Oct 02, 2007 12:12 am

backpage wrote:All the Daw's playback engines sound differently.

Did you read the quotations earlier in this thread from the people who actually wrote several of the major hosts? If not, go back and take a look at what the software's authors say on this.

backpage wrote:Are you gonna say with a straight face you don't hear a difference between floating point and integer?

Did you take the integer-versus-float blind test in a previous thread on this subject? :-}

joshhunsaker · Post by **joshhunsaker** » Tue Oct 02, 2007 12:21 am

It's a good point. Summing two mono files together using a floating point engine (as long as the sample floats are not longer than the actual supported summing calculation...) will always yield the same results. If you were daft enough you could perform all the math yourself and input it bit by bit (literally) into a new file and it would also match exactly what the computer has done.

Now (in addition to types of bit reduction) there also also different methods of downsampling (including taking and average of the samples if you are using say...88.2 to go to 44.1) which may or may not sound comparatively better.

abi · Post by **abi** » Tue Oct 02, 2007 12:29 am

freaks

hseiken · Post by **hseiken** » Tue Oct 02, 2007 4:14 am

You just sum a couple of audio files (we used 16 IIRC) the exact same way in several hosts. Take two resulting files,flip the polarity on one of them and all that is left is the difference between them.

I'm one of those 'idiots' that thinks of you're so caught up in this kind of crap that it keeps you up at night, you're using audio programs for a reason that I can't see would make any sense...

'Audiophiles' are so silly...having sex with soundwaves...as if!

joshhunsaker · Post by **joshhunsaker** » Tue Oct 02, 2007 6:50 am

ultradia frequencies are really good in bed i hear

dead horse (the resurrection)