Adding mic or upgrade recorder?

Gohanto

Gohanto

Veteran
Currently I own an ME66 (Although for most shoots I can use my school's MKH 416). And I use a Fostex MR8 mkII (designed for music, recorders 2 tracks at 16bit/44.1kHz) for recording audio and sync in post.

And I'm currently looking at buying a AKG CK93 for interior dialog. I'm curious though what would be the benefit of upgrading the recorder to a SD model instead (or lower priced option, not really savvy with sound recover companies)? I'm guessing my mics are way below the price point where the higher quality recording makes much difference, but I was hoping to get some other opinions.

Also, what are the advantages of pro recorders beyond 192kHz/24bit and timecode additions? Or are those the bulk of the expense in high end units?
 
It is highly controversial if anybody can hear any difference between 16/48 and higher "resolution" formats. For audio recording 24 bits is useful for not having to set the levels as exactly, but the final format is 16 bits anyway. For some audio effects higher sample rates are used for slowing down later, but if the mics are not special hi-frequency mics there is no point in even this.

So, for standard video production having 24 bits is usefull, higher sampling rates than 48 khz is not.

It is worth noting that only better 24 bit recorders have even 20-21 bit true resolutions, as cheap units are so noisy that getting even 16 bits worth of true resolution from them is difficult, even if the output file is nominally 24 bits. To get real benefit from 24 bit sample depth you need something like Sound Devices 7xx series recorder (around 120 dB S/N = 20 bit true resolution).

Timecode is nice for fast bulk editing, but it is possible to do fairly complex edits also without, just use slate, slap your hands...
 
and sprnd te money on location first , as nothing sounds good with bad sounding stage .
foforget the 24 /192 . you can live with 16 bit and 48 .
the 441 is bad ( not sound wise) as you will need to resample it to your final project which is 48khz
 
Hxd Ped said:
What do you mean by "special hi-frequency mics"?
Click to expand...

If you think you have to have 2 times oversampling to get good 50% slowdown, you must have microphones which actually record something between 20 and 40 kHz. Normal mics do not. If you use normal mics, there is no need to oversample for slow-down, as there is no audio content above 20 kHz anyway.
 
Last edited:
Um, I don’t mean to sound like a jerk, but I think you’re misunderstanding what the sample rate of recorders refers to.

Having a sample rate of 48 kHz (48,000 Hz) means the Analog to Digital converter takes a “sample” of the sound wave’s amplitude 48,000 times a second. A 96k sample rate takes samples 96,000 times a second. It has nothing to do with the frequency of the sound wave being recorded. (The bit depth it the resolution of the sample. Higher is always better.)

Most A-D converters have a low-pass filter set to a little over 20 kHz (or so). The Nyquist theorem states that the sampling rate must be at least twice that of the highest frequency recorded. Normal humans can’t hear tones over 20 kHz, so the lowest “good” sampling rate is 40,000 times a second. (They put the low pass filter before the sampler to avoid aliasing.) Wikipedia has a good, if not technical, explanation of the Nyquist theorem.

So, you don’t need fancy, expensive mics to record at high sample rates. (I’ve recorded a Sure SM-57 at 176.4 kHz. Defiantly NOT a fancy expensive mic!) Like I said before, A-D converters have low pas filters on them to filter out sound waves over 20k or so anyway. I don’t mean to insult anyone if they didn’t understand what sample rates represent. I didn’t understand until I learned it in school.

I hope I could clear things up


BTW: In case anyone wanted to know.

44.1k, 88.2k, and 176.4k sample rates are for music
48k, 96k, and 192k sample rates are for film/video

Why? I have no idea.
 
Last edited:
Hxd Ped said:
Having a sample rate of 48 kHz (48,000 Hz) means the Analog to Digital converter takes a “sample” of the sound wave’s amplitude 48,000 times a second. A 96k sample rate takes samples 96,000 times a second. It has nothing to do with the frequency of the sound wave being recorded.
Click to expand...

Why record at 96 kHz or similar, if the mics pick up nothing past 20 kHz anyway?

Nice to learn new things every day.
 
Petrus said:
Why record at 96 kHz or similar, if the mics pick up nothing past 20 kHz anyway?
Click to expand...
The technical answer is it will sound closer to analog by giving you higher resolution audio, like taking pictures on higher resolution film. Weather or not people can hear the difference or not is a matter of much long-winded debate.

I take the position that you should edit your audio at higher resolutions. Hopefully, your edits will be more precise and less noticeable in the final product. I have no idea if that is true or not, but it makes sense to me. I believe that if you change the audio with special effects, like time stretching that kind of thing, it’s better to have higher resolutions to work with. Again, I have no idea if that’s true, but it makes sense to me.

FYI: We did a test at school with sample rate conversion with ProTools. We found that audio recorded at 44.1k sounds better then the same audio recorded at 48k and converted down to 44.1k. So, if you record at higher sample rates, record at dubble (or quadruple) the final sample rate. The math works out cleaner that way. In other words, if the final product is 48k, record at 96k or 48k instead of 88.2k.

Hope that helps.
 
Hxd Ped said:
The technical answer is it will sound closer to analog by giving you higher resolution audio, like taking pictures on higher resolution film. Weather or not people can hear the difference or not is a matter of much long-winded debate.
Click to expand...


I thought that if there is no high frequency content in the signal, there is nothing there that could even be "higher resolution". Sampling rate only affects the frequency range, nothing else. Please enlighten me.
 
I think he misunderstood your question. For all practical purposes, mics pick and record 20-20kHz max. That's frequency range from rumbling bass to nails on a chalkboard.


Sampling rate, what people debate, is a problem when recording in digital. Here's a good explanation of it...
http://www.cepece.info/amstrad/docs/sound.html

But basically, sound constantly changes an infinite number of times every second. Analog records by having a needle push onto a material so it captures EVERY moment. Digital, on the other hand, puts everything in to 1's and 0's so it has to listen to the sound and record what level it's at many many times every second to simulate a continuous recording.

That's why, technically, digital can never be as good as analog recording. They question is, how many times do you have the sample the sound every second to mimic analog the point where you can't tell the difference. Some think 48kHz (48000 samples every second) is enough. Some think 192kHz is needed instead. 192 being the "higher resolution" of the two.
 
Gohanto said:
But basically, sound constantly changes an infinite number of times every second. Analog records by having a needle push onto a material so it captures EVERY moment. Digital, on the other hand, puts everything in to 1's and 0's so it has to listen to the sound and record what level it's at many many times every second to simulate a continuous recording.

That's why, technically, digital can never be as good as analog recording.
Click to expand...

Interesting. Which analog recording system has better S/N ratio, dynamic range, wow and flutter, distortion figures than even a decent 16/44.1 digital recorder?

If analog recording is better than any digital, why nobody uses them anymore?

This is a great forum for learning.
 
Higher resolution recording pick ups on high frequency harmonics that help enrich below 20kHz tonality. Harmonics are the overtones that distinguish a flute from a piano from a bullfrog all creating the same fundamental pitch/frequency. Some say these upper frequency harmonics are subtle and perceived as near liminal audio enhancments, and therefore needed for better quality recordings. Others don't hear it and think it's a marketing gimick.

Bit depth does make a difference, like it does with your computer; 16 colors vs. millions of colors. Normal 16 bit recordings are good. Above 16 bit, recordings are even more dynamic simply because there are more levels of amplitude in the given dynamic range. In analog there are no steps, it's just one sweeping ramp from silence into amplitude overload.

Both "well made" digital & analog recordings are good, if it's pleasing to the ear, it's good! Some old mono and 3 channel recordings from the 50s blow people away today when heard as they were designed to be heard.
 
If you're recording for video, then you want to record at 48KHz or some multiple - the issues of the conversion between 44.1 (or 88.2, or 176.4) and 48k (or 96, or 192) outweigh the benefits of higher sampling rates.

There'a a lot of argument in the field about the benefits of higher sampling rates (88.1, 96, 176.4, 192). There's very little argument about the benefits of 24-bit recording. As an acquisition format, 24-bit recording allows for more maniupulation in post, especially in regards to boosting levels. Given widespread availability of decent 24-bit recorders, it's pretty much a no-brainer given any choice.

By upgrading to a 48Khz, 24-bit recorder I think you'll hear significant improvement in your final product. While it's true that in theory most mic/preamp combinations don't take advantage of the full 24-bit range, when you factor in headroom and real-world "safe" recording levels in many situations, the added resolution of 24-bit recording makes a noticeable difference.

Finally, I don't think the Fostex unit you mention runs on batteries, has limiters, or is as convenient for film production as a "bag" device. Once you try working with equipment designed for production sound, you'll realize the great benefits of it. Having the extra time, better access to controls, and fewer worries about powering, etc., that a well-designed production recorder provides allows you to focus more on other factors that will get you better sound.

The expense in higher-end units is everywhere - the design, powering, preamps, analog circuitry and converter quality, features (limiters, ms decoders, timecode), monitoring options, redundancy (in the SD units, at least. you can record to 2 or 3 different media at a time), and more limited market.

So, in short, I think at this point you would see most overall improvement from a better and more convenient recorder. Its mobility and convenience will actually help you concentrate on getting the mic positioned well, which is a bigger factor than mic choice. It'll give you better ability to boost/manipulate sound in post. It'll give you more time to deal with acoustic issues of the location, and will let you capture sound in a wider variety of situations. It'll also give you more reliability and security of your recordings.

The HD-P2 has a great reputation as a price/performance champ... that and saving up for later purchase of better mics might be the best idea. The Sound Devices units are sweet, and have many benefits, but are pricier... so it sounds like in your situation you can wait a while and dream of those!
 
Oleg is right about the Tascam HD-P2 and the AKG mic.

I own the hd-p2 and bought it with the Oade super upgrade from Oade Brothers in Florida. The upgrade gives you 6db more gain with much better new potts than what Tascam puts in. One of my engineering friends owns a sound devices 7 series and we put the two side by side using Sennhiser shotgun mics with and without a pre amp. They are very close and we both like each a little better depending on the mic and on the pre amp. For the money difference I chose the hd-p2 and am happy with it and it has timecode in, but there are reasons to go sd as well.

AKG's top of the line mics are very good as well, it all depends what your recording and where.

You have to call Doug Oade and see if he still does the super upgrade which includes both upgrades. Sometimes the good pots are hard to get.
http://www.oade.com/
 
Petrus said:
Interesting. Which analog recording system has better S/N ratio, dynamic range, wow and flutter, distortion figures than even a decent 16/44.1 digital recorder?

If analog recording is better than any digital, why nobody uses them anymore?

This is a great forum for learning.
Click to expand...


So here we all are using digital cameras w/ 35mm adapters to try and get a 16mm film look! Why don't we just all get 16mm cameras?

It's all the same audio/video...analog/digital...the debate will go on forever. The best way I've been able to describe it is like this:

We have 2 paintings of the ocean.
Monet did one (analog - soft edges) & the other is a paint by number (digital - sharp edges) ...when you step back and look, they both look like the ocean.

We keep raising sample rates for higher resolution and we're lucky if we can even hear 20Hz - 18KHz (men...women can hear higher than us...), then we buy tube processing to 'warm' it up (this means add distortion), because distortion sounds good to the ear.

...and after all that attention to detail, we crunch it into an MP3's that we can sell on iTunes.

...and the beat goes on....
 
ullanta said:
If you're recording for video, then you want to record at 48KHz or some multiple - the issues of the conversion between 44.1 (or 88.2, or 176.4) and 48k (or 96, or 192) outweigh the benefits of higher sampling rates.

There'a a lot of argument in the field about the benefits of higher sampling rates (88.1, 96, 176.4, 192). There's very little argument about the benefits of 24-bit recording. As an acquisition format, 24-bit recording allows for more maniupulation in post, especially in regards to boosting levels. Given widespread availability of decent 24-bit recorders, it's pretty much a no-brainer given any choice.

By upgrading to a 48Khz, 24-bit recorder I think you'll hear significant improvement in your final product. While it's true that in theory most mic/preamp combinations don't take advantage of the full 24-bit range, when you factor in headroom and real-world "safe" recording levels in many situations, the added resolution of 24-bit recording makes a noticeable difference.

Finally, I don't think the Fostex unit you mention runs on batteries, has limiters, or is as convenient for film production as a "bag" device. Once you try working with equipment designed for production sound, you'll realize the great benefits of it. Having the extra time, better access to controls, and fewer worries about powering, etc., that a well-designed production recorder provides allows you to focus more on other factors that will get you better sound.

The expense in higher-end units is everywhere - the design, powering, preamps, analog circuitry and converter quality, features (limiters, ms decoders, timecode), monitoring options, redundancy (in the SD units, at least. you can record to 2 or 3 different media at a time), and more limited market.

So, in short, I think at this point you would see most overall improvement from a better and more convenient recorder. Its mobility and convenience will actually help you concentrate on getting the mic positioned well, which is a bigger factor than mic choice. It'll give you better ability to boost/manipulate sound in post. It'll give you more time to deal with acoustic issues of the location, and will let you capture sound in a wider variety of situations. It'll also give you more reliability and security of your recordings.

The HD-P2 has a great reputation as a price/performance champ... that and saving up for later purchase of better mics might be the best idea. The Sound Devices units are sweet, and have many benefits, but are pricier... so it sounds like in your situation you can wait a while and dream of those!
Click to expand...

Awesome, thanks for all the information. 24-bit I'm sure would help a lot especially as we're often short a dedicated mixer to watch the levels closely. Never really considered the what the advantages of having a limiter built into the recorder would be (or in a mixer for that matter). If dialog was ever loud enough to hit the limiter though, wouldn't that distort it to the point of needing to do it again anyway? Granted not as much as clipping, but how often is limited dialog usable? The Fostex actually does run off batteries, although a lot of people guess it wouldn't as it's more of a studio model. Granted it kills batteries rather fast...
 
Last edited:
10s said:
Above 16 bit, recordings are even more dynamic simply because there are more levels of amplitude in the given dynamic range. In analog there are no steps, it's just one sweeping ramp from silence into amplitude overload.
Click to expand...

There are no steps in digital audio output, the wave is smoothed by the low-pass filters used after the DAC.

Having more bit depth does not add any more levels of amplitude in the given dynamic range; digital sampling does not work that way. Higher bit dept only adds more dynamic range*. It is a scientific fact not changed by popular misconceptions.

-----------

*) There are only two parameters in digital recording which could be called "resolution": amplitude and frequency. Having more bit depth means smaller variations in amplitude can be recorded. This can be either the difference between the loudest and the quietest part of the performance, or the amplitude difference between the strongest fundamental tone and the harmonics "piggybacking" it (or the smallest harmonic details of quietest instruments). Practical limit is around 22 bits, as the analog components are never better than that. Pracitical limits for frequency are human hearing, microphones and reproduction systems. None of those are really capable of utilizing frequences above 20 kHz, thus 48 kHz is plenty enough.

Small detail: if the mics can not pick up anything above 20 kHz, there will be no harmonics above 20 kHz that can be recorded. So why oversampling?

--------

Against ingnorance the Gods themselves fight against in vain.
 
Of course the digital is smoothed out once it becomes analog after Digital-Analog-Conversion and outputs to the amplifiers as a smooth analog signal, but before it goes through the conversion the audio signal is made of digital samples glued together with dither. Digital is all steps, look at a sample up close, it's small tiny snapshots of audio frequencies & intensities. Analog is continuous, digital is incremental. Of course when digital is converted back into analog it becomes, continuous.

The other comment: I think we are saying the same thing only using different words. I use the word "levels" to describe these multitude of dynamic dB intensities available for each sample within the range. Your description of amplitude below is what I was attempting to communicate. Oh well... :)
 
Last edited:
You must log in or register to reply here.
Back
Top