Kramer Tape TDM 10db noisier than RTAS

[zakco]

I have discovered an oddity on my system and was wondering if someone could confirm this phenomenon. As the thread title suggests, I am getting a strange result when I compare the TDM version of Waves MPX to the RTAS version.

-24/44.1 session with just one Stereo Audio Track (no input -bussed directly to i/o) - no audio regions in session at all.

-Insert 1 - Signal Generator RTAS - pink noise @ -96db peak (just to get the meters showing something).

-Insert 2 - Waves MPX TDM - Input all the way down, linked, Flux, Noise, Wow etc all off.

-Insert 3 - TL master meter shows -79.59 db

Now, when I switch the waves tape to RTAS the noise floor drops to -89.62 db at output. More than 10db less noise with RTAS...odd.

But not as odd as this:

If I raise the signal generator to -60db peak, the meter shows -52.59db for BOTH tdm and rtas. No difference.

It's NOT a metering anomaly. When I crank my monitors, I can clearly hear the 10 db difference.

Any thoughts?

HD2 8.0. cs3
OSX 10.5.8
5 gb ram

[zakco]

I just recieved confirmation from Waves that their test machine exhibits the same behavior.

Could somehere that's running v9 and/or v10 test this and see if it's system/version specific?

Thanks!

z

[Bill Denton]

Given that TDM and RTAS use totally different code bases and languages I would not be at all surprised to see some anomalies.

Note that I am a native guy so I really don't know exactly what happens, but I have read that putting RTAS plugins before TDM plugins is not recommended. I wonder if the result would change if you were using a hardware tone generator.

As I write this, I'm eating some of my "World Famous Sleazy-Greasy Popcorn"* so there are plenty of grains of salt in this post, so caveat emptor, even if it's free...

*I invented my "World Famous Sleazy-Greasy Popcorn" back in the 60's to solve a major problem people have had for years. "World Famous Sleazy-Greasy Popcorn" alleviates the problems associated with eating stuff when you're so stoned that you forget to swallow...the popcorn has so much butter on it that it will just slide on down all by itself...

[Top Jimmy]

Thanks Bill.

[JFreak]

Originally Posted by Bill Denton

I have read that putting RTAS plugins before TDM plugins is not recommended.

That is because first RTAS plugs go to bypass when you hit record. Only if you put a TDM plug first, then followed by RTAS plugin(s) (and wasting voices), the RTAS plugs are active while you track. But then you need to take that Native HW Buffer into account, not that low-latency anymore...

Nothing to do with possible difference in sound.

[phatbeatstudio]

Originally Posted by JFreak

That is because first RTAS plugs go to bypass when you hit record. Only if you put a TDM plug first, then followed by RTAS plugin(s) (and wasting voices), the RTAS plugs are active while you track. But then you need to take that Native HW Buffer into account, not that low-latency anymore...

Nothing to do with possible difference in sound.

So are you saying RTAS will not go into bypass if there is a TDM plug before it.

[Drew Mazurek]

Originally Posted by phatbeatstudio

So are you saying RTAS will not go into bypass if there is a TDM plug before it.

yes. But you will be subject to twice the buffer in latency.

[Firechild]

Originally Posted by Bill Denton

but I have read that putting RTAS plugins before TDM plugins is not recommended.

You should always put RTAS before TDM to keep voices and latency down.

[JFreak]

Originally Posted by Firechild

You should always put RTAS before TDM to keep voices and latency down.

Always = whenever you're not tracking!

First RTAS plugs go to bypass once you hit record. The reason behind this (and the voice issue) is that as long as the signal is on the Native domain (hard drive, RTAS) it is processed with another thread than the TDM mixer. Once the signal goes to TDM domain (the first TDM plug) it obeys the normal rules, and yes, if you insert a RTAS plug after TDM plug it will not be bypassed but as the signal needs to get out of the TDM domain (1 voice) and be processed on a CPU and get back to the TDM domain (1 voice) you would have used multiple voices for a single signal.

The issue gets even worse if you have TDM/RTAS/TDM/RTAS/TDM as you are using a voice every single time the signal gets out or back in to the TDM domain.

Yes, it is doable. Just place a 1-sample-latency TDM plugin such as Trim into the first slot and place your RTAS plugs after that. But try to avoid mixing formats multiple times.

And check your playback engine buffer settings! As mentioned, inserting a first RTAS plug introduces the Native Buffer to the whole session. If you only have TDM plugs, you can calculate the total latency very much easier...

[Firechild]

Yes, I know that, but still in my world you should ALWAYS put RTAS before TDM.
Why? I am using a HD TDM system ONLY for near zero ms latency when recording and if I for some reason have a RTAS on my tracking channel I thank Pro Tools big time for bypassing that plugin for me, to keep my latency at minimum.
If there is a RTAS only plugin you just need when tracking no matter what, then I suggest you to go native 100 percent. Because then you don´t appreciate what TDM gives you