redge Posted May 4, 2006 Report Share Posted May 4, 2006 I have some questions arising from my use of a SD 702T recorder, without mixer, and a Schoeps mic. I've been studying the 702T manual, and reading Geoff Martin's e-book (www.tonmeister.ca) and have followed relevant discussions here and on ramps, but am still confused. Not a little bit confused, but a lot confused. what is the point of having a Sound Devices meter display VU as well as Peak? I have discovered that the peak reading is not equivalent to loudness (a fascinating discovery), but does this matter? Is there a reason to set gain so that the meter peaks at less than 0dBFS (or, out of caution, -4dBFS) if one does not expect higher spikes in volume. Why does the SD user manual talk about clipping at -3dBFS? Why do I keep seeing references on the internet, not specific to SD recorders, to keeping peaks to -6dBFS? Do people use limiters at -6dBFS as a matter of course? My initial esperience with the recorder is that the sound clips at/just above 0dBFS, but these are rough and ready tests and maybe I need to be doing some tests that are more rigorous. The underlying assumption of some of my questions is that one should set gain so that peaks are close to, or at, 0dBFS. Is this assumption correct for a SD recorder and, more generally, for real word recording? If some of my questions demonstrate that I am missing basic concepts by a country mile, don't hesitate to say so. Thanks. Quote Link to comment Share on other sites More sharing options...
tim apter Posted May 4, 2006 Report Share Posted May 4, 2006 im in a similar position and would love some info on this topic. my understanding is that dialogue will typically be recorded around -20 to -12dBFS at the recorder with peaks going up higher than that, and if dangerously higher, they are limited (around -6dBFS? or where?) This gives a perceived loudness of -20 to -12, but with tiny transients reaching up higher, and possibly being brickwalled. a question from me would be why not record at a slightly hotter level (say -12 to -8dBFS) and limit peaks at around -4dBFS in order to get a cleaner hotter signal? im no expert, i could be talking garbage, im just curious how the "standards" came into existence... Quote Link to comment Share on other sites More sharing options...
redge Posted May 5, 2006 Author Report Share Posted May 5, 2006 Thanks Oleg. In other words: I can be as close as I want to 0dBFS, and should be pretty close to it (recognising that each -6dBFS costs one bit) as long as I don't exceed 0dBFS. The VU meter is a waste of time if one is using a digital recorder. The remaining questions are why the SD manual talks about clipping at -3dBFS (which I don't get at all) and why the input/mic limiter kicks in a -6dBFS. Thanks. Quote Link to comment Share on other sites More sharing options...
tim apter Posted May 21, 2006 Report Share Posted May 21, 2006 thanks oleg, very helpful. i enjoy hearing the more advanced theory from you. Quote Link to comment Share on other sites More sharing options...
Philip Perkins Posted May 21, 2006 Report Share Posted May 21, 2006 Some of this depends on what your recordings will be used for. If you are setting tone @ -20, and the edit system is set up for that to be the reference level, then letting the peaks go out to 0dbfs will cause a lot of extra work in post--that is too wide a dynamic range, especially at an early stage of an edit. The material will limited eventually to go no hotter than -8 or so anyway, and there is a school of though that it is better to chase the dynamic range around during recording than to do it in post by sound editing and limiters. If you have this much dynamic range, then all the edit temp dubs being viewed via VHS or QT off laptops may have clipped audio (even though your files are ok) and will require that either the editor or the viewer constantly change their volume level to both hear the quieter passages and not have the loud stuff blow up their playback system. Re: metering--I have always thought that one of Sound Devices strongest points was its metering--I use the combo peak+average mode on my SD mixers all the time. The truth about levels nowadays is that the sound people have to be very concerned about BOTH peak and average levels in their recordings. I've had a lot of back and forth with network QC depts about this issue lately, and the upshot is that they are now REJECTING mixes which might have perfectly fine peak values if the average/RMS values do not meet their spec as well. Watching only a peak meter can easily give you the idea that you are doing fine when in fact you need to be pushing those peaks up against the limiter a bit in order to get the average level into a workable place. The SD meters are GREAT for this--probably as good as any I've ever used incl Dorrough. For more drama on this peak vs average issue, look at the DUC under any topic that includes "Dolby LM100" in the subject line. Philip Perkins Quote Link to comment Share on other sites More sharing options...
Michael P Clark Posted May 26, 2006 Report Share Posted May 26, 2006 I like to use both Peak+VU on my SD, ALL the time. Using VU helps me keep the Average levels the same across scenes and setups. If I didn't have my SD mixer, and had to have a mixer with one or the other, then I want a mixer with Peak, and not VU! The other great feature that SD has is the adjustable limiter threshold, and great limiters, so I could mix with VU only if I wanted to, but only on the SD...and maybe SQN. Quote Link to comment Share on other sites More sharing options...
John Blankenship Posted June 5, 2006 Report Share Posted June 5, 2006 My thinking is pretty much in line with Philip. Unfortunately, many people who are beginning their audio education have started in the digital realm and are missing some important knowledge, since most technical literature now deals more with peaks. Let me put it this way: Peaks are about technical considerations (what's the maximum level before meltdown), average levels are about aesthetic considerations (how loud does something sound). Both are important. Watching peaks keeps you out of trouble, watching average levels keeps the sound more consistent. In the "olden days" of analogue recorders, you didn't have the same concerns about peaks because the digital "brick wall" didn't exist. As the recording tape approached saturation, distortion increased some (but it was largely second order, which means it's harmonically related, so it didn't sound so bad) and as it further approached saturation some "tape compression" took place. In those days you would watch your average levels, as well as listen to the actual playback from tape, and through experience, know what effects the peaks were having as you "pushed the tape." Nowadays (I feel like I'm sitting on a cracker barrel in front of the general store talking about "way back when"), many people new to the industry want to have everything laid out for them in nice, easy to understand, numbers: They want to hear, "If the meter hits so-and-so, then your sound will be perfect." Well, sound is complex. Every voice is different, has different timbre, different peak-to-average ratios, and different requirements. Even as a person changes inflection, the peak-to-average ratio changes. Then, when more sounds are mixed together, things get really complicated. Every bit of sound affects every other bit of sound. Phenomena like masking effect take place, and even that varies by frequency and complexity. I guess what I'm saying is that, yes there are current guidelines, but there is NO rule of thumb that takes the place of knowing what you are hearing, how sounds interact, what average and peak both mean, and how they all affect the finished product. My suggestion is practice using VU meters to learn their value. Practice using peak meters to learn their value. Compare the peak to average relationships of various sounds and different voices. Compare them under different conditions. What you will take away from all this is that there is no substitute or short cut for learning your craft. I apologize of any of this sounds preachy, but I think for anyone new to the industry, it needs to be said. Quote Link to comment Share on other sites More sharing options...
redge Posted June 5, 2006 Author Report Share Posted June 5, 2006 Nowadays (I feel like I'm sitting on a cracker barrel in front of the general store talking about "way back when")... Is that the sound of Dueling Banjos in the background? Thanks for the good advice. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.