waves TDM vs. Native wankery!

[stefancrs]

What is this thread about? Establishing that different implementations of an algorithm might sound different?

Under certain circumstances, it's quite expected that you'll get audible results from having a bit depth difference. Which you do have here. 56bit fixed point vs 64bit floating point implementations of the same algorithm. It'd be interesting to have a reference implementation at a ridicously high bit depth, so that one could see whether the 56 bit fixed point or the 64 bit floating point is the most correct one.

It seems however like Waves messed up when transporting the L2 plugin, since the difference is way too huge. Especially when you compare to for instance rencomp.

[championrabbit]

ProRec and GearSlutz are the kinds of forums to read, and not post in unless you really, really know your stuff.

I'd be inclined to take opinions aired at those places fairly seriously.

[bmanic]

What is this thread about? Establishing that different implementations of an algorithm might sound different?

Exactly. Too many people refuse to believe that there is a difference at all. Now if TDM is better than RTAS everybody can decide for themselves but one can't deny there isn't a difference.

It seems however like Waves messed up when transporting the L2 plugin, since the difference is way too huge. Especially when you compare to for instance rencomp.

Indeed.. which makes waves statement of Native being identical to TDM very suspicious, no?

Cheers!
bManic

[Spaceman Sounds]

best way to hear a difference with music, is to have a 2 bar loop of each .wav and place them one after the other for loads of bars. Then you get to hear if there really is a difference. Will try these when I'm in the studio later.

[defjamm]

What is this thread about? Establishing that different implementations of an algorithm might sound different?

Exactly. Too many people refuse to believe that there is a difference at all.

i think that's not the problem, it is something like: waves is better in tdm than in rtas and stating it as a fact for their whole line of plug-ins. you should go from plug-in to plug-in and compare, there could be big differences, there could be small differences, or no differences at all. let's forget better and worse for this and point the differences out, this way we can learn way more.

It seems however like Waves messed up when transporting the L2 plugin, since the difference is way too huge. Especially when you compare to for instance rencomp.

Indeed.. which makes waves statement of Native being identical to TDM very suspicious, no?

that's why i want a new test with a native daw(and not rtas in tdm), then we can go to the waves-board and ask some questions.

btw i asked this several times now, which version of waves was used?

[stefancrs]

What is this thread about? Establishing that different implementations of an algorithm might sound different?

Exactly. Too many people refuse to believe that there is a difference at all. Now if TDM is better than RTAS everybody can decide for themselves but one can't deny there isn't a difference.

It seems however like Waves messed up when transporting the L2 plugin, since the difference is way too huge. Especially when you compare to for instance rencomp.

Indeed.. which makes waves statement of Native being identical to TDM very suspicious, no?

Cheers!
bManic

Indeed. The only way they could do that is by having exactly the same implementation on all platforms. Which basically would mean 56bit fixed point, if I've understood everything correctly.
On all platforms that is. Which would seem a bit dumb, because 64bit floats are probably better.

[defjamm]

I suspect one of the reasons these subtle differences can not be discussed on KvR but can and are discussed with great interest and civil (usually) behaviour on Gearslutz and Prosoundweb forums is because the majority there have a good monitoring chain (high quality soundcard and good monitors).

that's not true at all, i know both webboards, and even a lot of 'pros' have a different taste and don't agree with your opinion:

http://recforums.prosoundweb.com/index. ... /9341/96/0

http://recforums.prosoundweb.com/index. ... msg_127996


i think a rme-soundcard with adam 2.5s and beyerdynamic-headphones should be good enough for a hearing-test.
but a hearing-test can fool one pretty fast, even if it isn't a series of blind-tests, where the results are clear.
people are fast with words like better and worse. imho better and worse don't say much about a sound.

Umm.. read my post once more. I was talking about discussing the results in general, not about which is better or worse, which is rather rare here on KvR. It usually ends up into an argument of "so what? does it make my music better?" instead of discussing the finer details of the differences and what may be causing them.

Even the links you provided prove this quite well. They talk about stuff that are not discussed here on KvR.. like, umm, ever.

Cheers!
bManic

my answer was pointed to your remark about using a good monitoring-chain:
even with good monitoring, people will still have different views.
i agree that they talk and act more civilised than some of us at kvr, but their average age is also way higher i believe.

until now this topic has more substance than the summing-stuff, let's find the truth.

[bmanic]

There's nothing wrong with having different views. The point is to HAVE some views. Doesn't make a difference even if they are completely opposite to mine. What I don't like in these kinds of threads are people who chime in just to say "but it doesn't make crap music better" or "no, there is no difference" (as it can be proved beyond doubt that there is). Maybe this explains me better?

Anyways, I'll get some new examples up ASAP. Anybody else got access to a PT TDM rig with the RTAS plugins and want to run some tests?

Cheers!
bManic

[Kriminal]

What I don't like in these kinds of threads are people who chime in and say "no, there is no difference" (as it can be proved beyond doubt that there is). M

but is the diff good or bad? Its all opinion, and im guessing most ppl here are not technically minded enough to notice or even care, they just make music as best they can.

Diff or not, good or not, if it sounds ok, it is.

Thems the facts.

[bmanic]

[Jason Brian Merrill]

I hear a difference that I cannot define as good or bad on either side. However, if the phase test reveals transients, well then yet, there obviously is a difference. What proves that the transients are being destroyed though?

[el davo]

What proves that the transients are being destroyed though?

Certainly the transients are being changed from the original in both cases. This is heavy compression/limiting after all.

As to which of the two "destroyed them less":
Bmanic subjectively compared them to the original waveform. The orig wav is in the "all about compression" thread, haven't checked it myself

[fizbin]

They aren't exactly the same. I can't hear a difference though (tin ears).

1. Load them both into Wavelab
2. Cut the first 64 samples off the L2 RTAS wav. (they are offset differently)
3. Run global analysis on both - there are slight differences on the "loudness" tab under RMS, even thought the maximum loudness for both is -0.8 db.
4. Invert the phase on one, select all, copy, switch to the other wav then go to the Edit menu and pick "Paste Special" and pick "Mix" (two identical wavs would result in a blank wav file after this)

There isn't much left after that, but there is a slight difference.

[Aleksey Vaneev]

There's no base for comparison if Waves have not transferred their algorithm to TDM in full accordance with RTAS version. It's like comparing two different limiters that share similar processing stages.

[Alan]

What I don't like in these kinds of threads are people who chime in just to say "no, there is no difference"

Like I pointed out in the other thread concerning the new demos,I didn't hear enough of a difference working with both to promt me to do a comparison,besides the fact that I don't have a TDM system at home to even do a back to back comparison besides the fact that I got rid of Waves almost 2 years ago. Anytime I heard or used them in the studio I never heard much difference than what I used at home,if any,and that's more important to me than trying to analyze the difference in a test. Did I hear the difference in your test? A little,but not much. I also suspect other factions here like hardware,host summing and host math. This type of test can get very anal if you wan't exact references. In the end I never heard enough of a difference to care anyway. Good luck.