EX-1r recording quality vs. Fostex

ullanta said:
Yes, and I understand that the quantization error is scaled up as we add gain, and that the much-smaller quantization error of a 24 bit recording makes a difference when you have to crank up the gain in post. A 24-bit recording with peaks at about -48dBFS has the same quantization error (theoretically, and after being raised by 48dB) as a 16-bit recording peaking at a perfect 0 dBFS... Or (another fun way to think about this for the non-mathematically inclined) is that a 16-bit recording with peaks at -48dBFS is essentially an 8-bit recording in terms of quantization; a 24-bit recording with peaks at -48dBFS is essentially a 16-bit recording. Sure, this is an extreme example, but it illustrates the difference. Whatever the bit-banging capabilities of the DAW/NLE, acquisition at an effective 16bit sample size will be better than acquisition at an effective 8 bits. The final product tends to be 16 bits, so 24-bit recording gives you tremendous leeway at least insofar as quantization error is concerned, while 16-bit recording gives you relatively little.
Click to expand...

There is a big THEORETICAL advantage in using 24 bit recorder over a 16 bit recorder, but in real life the difference is not so huge, or it does not exist at all. No recorder can achieve nowhere near 24 bit real life performance.

With best 24 bit field recorders (Sound Devices, Nagra, Zaxcom in the $3000-9000 range) the actual noiseless bit depth is around 20 bits (120 dB dynamic range). Certainly better than 16 bits, but does not allow -48 dB undergains, only 24 dB (which is GREAT!).

When talking about cheap "24-bit recorders" the sad fact is that too often they do not provide even 16 bit resolution even if set at 24 bits. The analog electronics (mic-pres) and converters are just too noisy. So using a cheap recorder in a double system because it records 24 bit files and camera only 16 bit is pointless. If the camera has over 90 dB dynamic range with 16 bits and the recorder has the same 90 dB range with 24 bit files, why bother?

With all this worry about the files being 24 or only 16 bit, I wonder how the soundmen of the past were able to make any movies even in the sixties or seventies or eighties, when the best open-reel Nagras struggled to get 65 dB dynamic range, just over "10 bits" in digital scale. Still the soundtracks sound fine! Maybe it is the user not the machine, after all?
 
The bottom line here is that the majority of the popular small plastic prosumer portable digital audio recorders have average but unremarkable performance that is equivalent to the better prosumer camcorders (by which I mean cameras with native XLR inputs and which record with uncompressed or lossless audio codecs). To be sure, there are some recorders that have less than average performance, so caveat emptor. There are many tests, reviews and discussions of this genre of equipment, even including a "sticky" in this very forum.

The exercise of simply substituting one average but unremarkable mic preamp for another will not significantly improve the final product. This is axiomatic common sense. You don't need an elaborate experimental test to determine that there isn't a dime's worth of difference between Tweedledum and Tweedledee.

And if that experiment takes us to double-system sound recording, it increases the complexity (and risk) and makes it come up negative in the cost/benefit equation (at least IMHO).

OTOH, if it also improves the monitoring and/or control of the audio signal, one might make an argument that it has some positive benefit. THere is also the whole discussion of external mic preamps feeding into the camcorder (and/or separate audio recorder). Here again, better monitoring and control are undeniable benefits. And using cleaner external gain to feed a hotter signal into the camcorder (or recorder) MAY give you better signal-to-noise ratios (SNR), but is nothing like a sure-fire panacea, which seems to be a popular conception in "indie" circles.

Clearly, there are some audio recorders (such as those identified by Petrus above) which exist at a higher celestial plane. They have demonstrably better mic preamp performance than the average recorders commonly discussed in these forums. They also have many other performance/usability/stamina features that make them good investments for production sound professionals. IME, there are NO camcorders that have equivalent mic preamp performance. So using one of the premium products will definitely give you better (i.e. quieter) mic preamp performance.

Jay Rose and his son published some exhaustive audio tests of various camcorders a few years back. I believe the reports are available online. He discovered that even the most expensive video camcorders had very pedestrian audio performance and (IIRC) there wasn't a significant difference among any of the professional and prosumer models. If you think about it, there is an economic explanation for this phenomenon. This has been discussed here (or in similar forums) before more than once. I believe they discontinued this testing exercise simply because there was nothing remarkable to report and it wasn't worth the effort. No camcorder manufacturer has shown any interest in baking in premium audio performance in a camcorder product, and the economics predict that we shouldn't hold our breath waiting for such a sea change, either.

Remember, however, that having a nice quiet, low-distortion mic preamp is only one part of the total puzzle.

And clearly, there are some applications (such as serious music production) where the camcorder audio track is never the first choice, no matter how good(~) the camcorder.

The flip-side of that is the growing segment of people who are shooting with cameras (such as the better consumer handy-cams and DSLRs) which have such bad audio performance, that it makes sense to ALWAYS consider using a separate audio recorder because even "average but unremarkable" is demonstrably better than the audio performance of the camera.

I am trying to advocate common sense here. I don't have an ax to grind for or against any product. But I don't believe in supernatural properties of equipment, either. I only believe what I can see and hear. And experiments that can be measured quantitatively and are repeatable by others (i.e. the "scientific method"). I dont buy into the "golden ears" theories. If you can hear something others can't, then good for you. But I want to be able to see/hear some benefit before I will spend any of my $$$.
 
IMHO, because of the better controls, limiters, and monitoring a mixer provides, one would get more bang for the buck sticking with single system and adding a high-quality mixer to the kit than going double system and spending the same amount on a recorder. A mixer allows one to better exploit the mic and camera recording chain to its best advantage. Only once one has the in-camera recording as good as it can possibly get and find they're still falling short does it make sense to consider adding the complexity, etc, of double system to an external recorder. Of course I'm assuming a decent quality pro or prosumer camcorder here - if shooting on a DSLR or cheap consumer handicam all bets are off.
 
You must log in or register to reply here.
Back
Top