Page 6 of 7
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 7:40 am
by gearboy
magicd wrote:
Third party plugs sound different with different levels going in? Sounds plausible. Do you have any specific examples?
Dave
I have the Waves Ren Max Bundle and the IKM T Racks Plug-in bundle. Both of these bundle's plug-ins will clip from too hot a signal at input. I simply insert a Trim plug-in and drop by -12dB and this usually does he trick.
Jeff
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 7:59 am
by tommymandel
Wow, thanks for that post, Gearboy - I sometimes run into distortion in my TRacks, but I never thought of inserting the Trim, and at -12db.
Really gonna help, I appreciate the tip.
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 8:12 am
by Phil O
Incidentally, if any of you remember the old Elemental Inspector XL (which still works with DP6 by the way), its level meters keep a running total of full scale (or over) samples. If one of these phantom spikes comes through, it logs it. My mix bus always has one of these meters on it and any overs (even one lonely sample) is noticed. I haven't seen the Roger Nichols equivalent of this, but I assume it is the same. I should have mentioned this earlier as it pertains to the original poster's question.
BTW, I'm not just some "home studio guy working on a project." I do this for a living. I hold two engineering degrees, have taken several graduate courses, and have been in the field for 31 years. Thank you.
Phil
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 9:01 am
by Phil O
magicd wrote:...Put a trim plug in before and after the plug. Add 30 db of gain on one side and take it away on the other side. You should get the same result as if there were no trims in the path...
FYI the numeric display on the trim plug's meter goes to over 360 dB. I didn't test it beyond that. I assume it goes much higher. Who would ever need more?
Umm, and no clipping of the signal.
Phil
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 9:18 am
by magicd
Phil O wrote:FYI the numeric display on the trim plug's meter goes to over 360 dB. I didn't test it beyond that. I assume it goes much higher. Who would ever need more?
Well, if you're mixing cowbell...
Dave
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 9:26 am
by Splinter
Just to throw some sand in the ointment. No ones mentioned the fact that even some of the most revered pre amps have a 70dB s/n ratio at best. Once we throw in mic noise, cable noise, and air conditioners, we have maybe a 55 dB s/n ratio. So, the noise floor of our converters and DAW are really irrelevant. That said, why were we recording hot, again? Huh? I couldn't hear that. Recording as hot as possible affords us NOTHING and only creates gain staging and distortion issues later. I've been recording with peaks at -12dBFS with good success. That still gives me 132 s/n ratio far beyond the output capability of any analog system on the planet! I'm only annoyed by the lack of DP's graphic resolution at lower levels. I don't know if it will make a big difference since I'm already recording conservatively, but I will experiment with Tom's -20dBFS = 0VU.
Regarding MD's comment about recording to tape at 0VU: You must have not recorded rock and roll. Rock and roll was ALL about tape compression. That was the beauty of Ampex 499. You could soak that stuff and it would just get creamier and creamier. 0VU was a given... if not hotter! Obviously, hi-hats were not at 0VU.
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 9:49 am
by James Steele
Splinter wrote:Regarding MD's comment about recording to tape at 0VU: You must have not recorded rock and roll. Rock and roll was ALL about tape compression. That was the beauty of Ampex 499. You could soak that stuff and it would just get creamier and creamier. 0VU was a given... if not hotter! Obviously, hi-hats were not at 0VU.
I can't find this comment. I don't think he was suggesting this. You're taking it somewhat out of context. Everybody knows that you could slam tape hard and DID for rock and roll recordings. Just remember to store your tapes "tails out."

Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 9:56 am
by tripit@earthlink.net
magicd wrote:tripit@earthlink.net wrote:
My only testing was done while mixing IB for a live album for an indi label. So, in my test, I was using PT with a ton of plugs
A ProTools HD system works in an entirely different way than a native system. Up until now, this discussion has been specifically about Digital Performer running under the MAS engine. If you run Digital Performer using ProTools HD hardware, or if you use ProTools running HD hardware, you are using hardware based DSP processing on the Digidesign farm cards.
DP native MAS processing is 32 bit floating point resolution all the way through the data path. Depending on the design of the plug-in, resolution can go up to 64 bit float within the plug-in. DPs Masterworks series of plug-ins are an example of 64 bit float processing. Therefore, the test described earlier is a valid, real world test. Essentially what we are talking about is the difference between 3+2 vs 2+3.
TDM plug-ins and the TDM data path do not use floating point resolution processing. The hardware based system uses fixed integer processing. At this point I am going to shut up and step back because I don't want to get details wrong and misrepresent the Digi HD system. However, go check it for yourself. TDM/HD systems do not use 32 bit float processing. Therefore the way TDM works internally is fundamentally different than the native MAS system.
So what does that means for the way you mix? It could mean a lot on the hardware-based system. I don't know. That's not my area of expertise. As I said, everything discussed up to this point has been based on the native processing of the MAS engine.
Dave
Yes you are correct. The TDM system doesn't use 32 bit floating, it uses 48 bit. The PT LE system does use 32bit floating, like DP.
This isn't my area of expertise either, so I'm not going to dive in head deep. I have a fundemental understanding but leave others to be experts. PT TDM HD uses a 48 bit mixer (288 db dynamic range) and most all the TDM plugs are essentially 48 (24 double precision - 24bit in, 48bit processs with 288db range, brought back to 144db for mixer), of course the RTAS plugs are native. I do know that the Digi mixer at 48 bit allow enough headroom for the digi mixer to have a +12 fader (DP is +6) while keeping low level resolution, still contributing to the bus even as low as -90db. I also know that you can mix a large number of tracks without clipping the input side of the mixer, but you can clip the master output.
But I'll leave this at that.
I still think that in a real world mix, you'll get better results with a lower level mix on any system, but hey, to each his own.
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 10:54 am
by wurliuchi
Good or bad (but I think it's good) this is my process. Many should find it familiar.
Musicians play.
Set levels. (Depending on the source, peaks anywhere from -12 to -4. I try not to go over -4 dBFS.) I find ADCs tend to ring (that's my unofficial term for it) when pushed. Some are better than others, of course, but I've never found a need to push any of them).
Record at 24-bit.
Mix. (Peaks at -4 dBFS. RMS around -22 to -18 overall average -- depending on genre and song. If the mix buss goes over -4, I go looking for the culprit(s) and manually adjust.)
I never move the master fader off of unity. I know I can do that and it will sound fine (I've tested it). But I like going in and adjusting individual tracks to get the wayward levels under control and to get as much average level out of the tracks as pleasingly or satisfyingly as possible without going over the -4 ceiling on the mix buss. It's a challenge.)
Master (done by others).
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 10:55 am
by Splinter
James Steele wrote:Splinter wrote:Regarding MD's comment about recording to tape at 0VU: You must have not recorded rock and roll. Rock and roll was ALL about tape compression. That was the beauty of Ampex 499. You could soak that stuff and it would just get creamier and creamier. 0VU was a given... if not hotter! Obviously, hi-hats were not at 0VU.
I can't find this comment. I don't think he was suggesting this. You're taking it somewhat out of context. Everybody knows that you could slam tape hard and DID for rock and roll recordings. Just remember to store your tapes "tails out."

Here you go James:
magicd wrote:Also, when was the last time you printed to analog tape with your analog meters hovering at 0VU? Anybody I know that has worked with analog considers 0VU as a reference point for gear calibration, not an ideal recording level.
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 11:00 am
by magicd
Splinter wrote:
Regarding MD's comment about recording to tape at 0VU: You must have not recorded rock and roll. Rock and roll was ALL about tape compression. That was the beauty of Ampex 499. You could soak that stuff and it would just get creamier and creamier. 0VU was a given... if not hotter! Obviously, hi-hats were not at 0VU.
Dang. I thought I was done with the cowbell comment.
Yes I have recorded rock n roll to analog tape. A properly calibrated deck and board will mean that a signal that is 0 on the board is 0 on the tape deck VU. That's the starting point. Then it is a question of how far above that signal can you go before you hear saturation to the point you don't like it. How far above 0VU you can go is headroom, and how much of that you get in the analog world depends on the rest of the circuitry and what the tape can hold. You could argue that there is no absolute ceiling with analog tape. You may not get any more volume, but you may still hit the tape harder to get that good compression and distortion.
Digital converters are not like that. The A/D converter has an absolute ceiling. That ceiling is "digital zero". Digital zero and analog zero are not the same thing. That's why you do need to know the full scale range of the A/D converter. If the A/D converter has a full scale gain range of 18dbu, that means that the signal that registers 0 on the analog VU will register -14db on the digital meter. The 18dbu FS range of the converter means you have 14db of available headroom above the analog 0 +4dbu VU reference.
All this exists to make the cowbell louder...
Dave
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 11:02 am
by James Steele
Splinter wrote:James Steele wrote:Splinter wrote:Regarding MD's comment about recording to tape at 0VU: You must have not recorded rock and roll. Rock and roll was ALL about tape compression. That was the beauty of Ampex 499. You could soak that stuff and it would just get creamier and creamier. 0VU was a given... if not hotter! Obviously, hi-hats were not at 0VU.
I can't find this comment. I don't think he was suggesting this. You're taking it somewhat out of context. Everybody knows that you could slam tape hard and DID for rock and roll recordings. Just remember to store your tapes "tails out."

Here you go James:
magicd wrote:Also, when was the last time you printed to analog tape with your analog meters hovering at 0VU? Anybody I know that has worked with analog considers 0VU as a reference point for gear calibration, not an ideal recording level.
Okay... I still don't get from A to B from his statement. It really depends on the source as well. When recording analog you'd hit it hard enough so peaks went past 0. Last time I did analog recording I think we were using 456. I only think I got to use 499 or whatever the very last analog project I did over ten years ago. I vaguely remember 499 took a lot more level than 456.
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 11:37 am
by Splinter
magicd wrote:It means the 0 on the meter is 4dbu of actual signal. Now notice all that space to the right of the zero on the analog meter? That's where you go to get your best S/N performance, tape saturation, etc. That's where you do the analog slam. Not "hovering around the 0"
Yes, I understand all this technically, but I don't understand what
you are saying. If the signal "hovers" at 0VU that means yours peaks are rising considerably above that giving plenty of saturation. But I guess it all depends on how you calibrate your machines. It's been 15 years since I've recording to tape - yes, I was an early embracer of digital recording and, yes, I understand digital metering and the conversation at hand - so, I don't remember the calibration settings we used, but I know we changed them depending on the tape formulation.
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 1:01 pm
by gearboy
tommymandel wrote:Wow, thanks for that post, Gearboy - I sometimes run into distortion in my TRacks, but I never thought of inserting the Trim, and at -12db.
Really gonna help, I appreciate the tip.
My pleasure!
Lower levels!
Re: Question for Magic Dave
Posted: Thu Aug 28, 2008 2:28 pm
by tripit@earthlink.net
James Steele wrote:
Okay... I still don't get from A to B from his statement. It really depends on the source as well. When recording analog you'd hit it hard enough so peaks went past 0. Last time I did analog recording I think we were using 456. I only think I got to use 499 or whatever the very last analog project I did over ten years ago. I vaguely remember 499 took a lot more level than 456.
Typically, at least for me, we used +6 for 456 and +9 for 499. 499 could really take a beating - I know guys who went higher. Mind you this was on studer 800's at 30ips. Ahh....the smell of fresh tape in the morning.....don't get me started on editing with razor blades.