Field recording equipment - Is MD the way to go?

[Cornucopia]

Well, if you've got the $$$ and you want it rugged, long lasting (in all senses of the phrase), and full-featured, the king is probably still a NAGRA. See http://www.nagraaudio.com/pro/index.php. Connects to the best mikes, many choices for battery, sample rate, resolution, connections, but they aren't cheap.

Scott

[Raga]

The "low-end" piece was 3250CF, or $2600 USD.

[joeg04]

Check out the iRiver mp3 player/recorders. I use the 895 model (512 MB)for wedding gigs using a lapel mic and it can record over 8 hours at 44KHz 128kbps. Runs on 1 AA battery and the battery lasts a very long time. You can find them at Best Buy, I think I paid $149 for mine a couple months ago.

I read that it is impossible to copy high bitrate MP3 recordings (>96 kbit/s) digitally from recorder to PC:

The UMS firmware doesn't allow you to record with more than 96kb/s.
The Manager firmware doesn't allow you to download music from your player back to your PC.

Source:
http://www.hydrogenaudio.org/forums/lofiversion/index.php/t33515.html

Does anyone know if that is true?

[Raga]

There ar several firmware versions. Some allow you to record 320kbps, some allow less. At any rate this has nothing to do with copying them from the recorder to the PC. The person writing may have some weird firmware version. I don't have the device myself, but others seem to have had no problems with this, for what I've been reading.

[joeg04]

It seems you are right. I read now that you can upload recorded audio files, but with the manager firmware no MP3 files that were before downloaded from the PC.

However, I found a report about another issue. All recordings of the iRiver recorder seem to have a 16 kHz lowpass limitation. Source:
http://www.hydrogenaudio.org/forums/index.php?showtopic=35259

[Raga]

16kHz cut? Hello, it's mp3 we're talking about. Mp3 means lossy. If you're worried about such issues, don't record mp3, don't record atrac. Record lossless.

And, a good many mics in the > $100 price range only go up to 15 or 16 kHz, some even as low as 10 kHz!

[joeg04]

Mp3 means lossy.

MP3 with 44.1 kHz sample rate and >=128 kbit/s bit rate should lose only information that humans can't here anyway, if the encoder is good. The frequency range should go almost up to half of the sample rate, so about 20 kHz in this case.

You can get small $100 stereo mics including preamp with 20 - 20,000 Hz, for example:
http://www.soundprofessionals.com/cgi-bin/gold/item/SP-BOOSTER-1

[Raga]

One rarely sees an mp3 file without a cut-off above 16kHz.

We're talking about two entirely different things here, both measured in hertz, or cycles per second. You cannot compare sample rate hertz to frequency response hertz any more than you can compare it to your processor's megahertz reading.

Sample rate means the quantity of samples taken per second in the recording. For example 44.1kHz sample rate means 44100 samples are taken per second.

Frequency response, on the other hand, has to do with the height of the sound. The more frequent the vibration, the higher a sound is produced. Low sound vibrates less times in a second, hence the Hz (vibrations per second) reading is lower.

I hope that clarifies the concepts for you.

[joeg04]

We're talking about two entirely different things here, both measured in hertz, or cycles per second.

Both is connected. See:

The theorem states that: when sampling a signal (e.g., converting from an analog signal to digital), the sampling frequency must be greater than twice the bandwidth of the input signal in order to be able to reconstruct the original perfectly from the sampled version.

Source:
http://en.wikipedia.org/wiki/Nyquist-Shannon_sampling_theorem

[Raga]

Thanks for the interesting link - all right, they are connected in that sense. Makes sense.

However, regardless of the sampling rate of the mp3, the fact is most mp3 rarely go beyond 16kHz in frequencies. I just opened some random mp3s and looked at the spectral view. 192kbps was invariably cut above 16kHz, in one of the 256kbps tracks cutting was occasional, not constant, and in another one (where the sound world was very simple) there wasn't much cutting at all until 20kHz. Likewise, one 320kbps piece I found had most of it left up to 20kHz, another showed some cutting and a third showed substantial cutting.

Whether there will be noticeable cutting at the high end or not depends largely on what you're recording.

At any rate, with all that's been said, iRiver isn't really the recorder of choice if you're going to record things for serious use. However, it seems to be a pretty good replacement of a convenient size you can keep with you and use in case your actual recorder isn't on you.

[joeg04]

The MP3 algorithm seems to be inefficient above 16 kHz.

Scalefactor band 21 problem:
The last scalefactor band (sfb21 for long blocks or sfb12 for short blocks) has no own scalefactor. This scalefactor band covers the range from 16kHz up to the higher frequency limit, when using 44.1 or 48kHz sampling frequency.
...
When encoding sfb21 content, it is common to encounter some scalefactor bands that are encoded with a too high resolution just to accomodate the coding needs of sfb21.

Source:
http://www.mp3-tech.org/content/?Mp3%20Limitations

The following link shows similar observations as you have reported (better frequency response at higher bitrates):
http://www.mp3-tech.org/tests/pm/index.html

If the iRiver cuts the spectrum generally at 16 kHz, then it does not take full advantage of the higher bitrates or of easy to compress content. But I agree, it's still an interesting small device.

[Raga]

The thing is, I don't think iRiver was really meant to be a super high-end recording device. If the 16kHz cut is an actual fact, they probably did it to improve the quality of the compression of what's getting recorded, to save all the bits for the more audible <16kHz frequencies

[joeg04]

I found another MP3 recorder that is similar to the iRiver IFP, the iAudio G3.
- Plays 50 hours MP3 on 1 AA battery,
- Recognized by PC as USB2.0 UMS device with 1 GB,
- MP3 128 kbit/s recording from int mic, FM radio, or line-in.
http://eng.iaudio.com/product/product_G3_feature.php
What do you think about it? Does anyone have experience with it?

[DereX888]

I'd rather find some old analog tape recorder and have better quality (albeit on analog media) than any of those MP3 recorders... But thats me

Seriously, I saw one of the great Revox portable professional recorders for like $50 on ebay or some pawn shop. Or find some used old Nagra, they were always popular and shouldnt be hard to find. Anyway any old-time analog professional recorder would do. Those things used to cost thousands in their time, and there was a reason for it - their recording quality is still no match for any of those walkman-alike mp3 toys youre asking and asking about here

I suggested this already before:
make up your mind, man
Either you want quality or you want MP3

If its quality you want - stop whining about short battery times or insufficient recording time capabilities of some devices. Its obvious you'll need 2 of them if you want to have continuus live recording of any longer venue, and you will need an external source of power for them as well (hint: motorcycle batteries have large enough capacity, yet they are often way smaller than car batteries; it doesn't take rocket science to use it as source of power even if the device doesn't have anu external power source plugs... just use your brain ).

If its MP3 you want - whats there to discuss? Does the filter or cut above 16kHz really matters if youre recoirding at 128kbps in layer III compression? ROTFL

[joeg04]

Either you want quality or you want MP3

I want quality MP3 @20-20,000 Hz.

Therefore I don't want the iRiver IFP, if it cuts at 16 kHz and if it's limited to 96 kbit/s with UMS (universal mass storage) device firmware.
I don't know the frequency spectrum of the iAudio G3 yet. Therefore I asked if someone has experience with it. At 128 kbit/s, it depends very much on the quality of the encoder.

old analog tape recorder and have better quality...than any of those MP3 recorders

I think the following MP3 example from Edirol-R1 (44.1 kHz, 160 kbit/s, internal mics) has a better sound quality than old Revox tape recorders, especially regarding signal to noise ratio:
http://www.edirol.com/products/info/demo/r1/Violin24.mp3

[DereX888]

Either you want quality or you want MP3

I want quality MP3 @20-20,000 Hz.

Therefore I don't want the iRiver IFP, if it cuts at 16 kHz and if it's limited to 96 kbit/s with UMS (universal mass storage) device firmware.
I don't know the frequency spectrum of the iAudio G3 yet. Therefore I asked if someone has experience with it. At 128 kbit/s, it depends very much on the quality of the encoder.

Read the MPEG-1 Layer III specs (i.e. www.mpeg.org have them).
It is designed to cut frequency past 16kHz, you wanna reinvent the wheel again or something?
There is Layer II if you want broader frequency. Then there is MPEG-2 Layer I and MPEG-2 Layer II. All have different purposes. Just educate yourself first before you ask again for "quality" MP3 (MPEG-1 Layer III), at 128kbps bitrate, and with a full frequency range. Its just an oxymoron!
On a lighter note example:
If you want Mercedes Benz but bought a Hyundai because they have same technical specs on a paper (lets just assume) you still can't and shouldn't expect it to ride like a Benz because it is still a Hyndai

old analog tape recorder and have better quality...than any of those MP3 recorders

I think the following MP3 example from Edirol-R1 (44.1 kHz, 160 kbit/s, internal mics) has a better sound quality than old Revox tape recorders, especially regarding signal to noise ratio:
http://www.edirol.com/products/info/demo/r1/Violin24.mp3

I have no comments ROTFL

[Dragonsf]

Now I got my M-Audio MicroTrack.
This thing is cheap/small/light/powerfull. Just made some recording test:no noise at 24/96 or mp3. Size is about an iPod but much lighter. Price is 39000 Yen.
Easy menu and handling.

[That Bastid]

Raga,

After listening to your recordings, I think that your focus on the ability of your recorder to resolve very high fidelity is misplaced. A far more serious problem is your distant mike placement, exacerbating ambient noise and echoes that destroy intelligibility. You just don't have a decent signal to record.

I would suggest putting your energy into figuring out a miking/mixing strategy, perhaps with rf mikes for unobtrusiveness if you can afford them and believe you can keep track of them (and take care of them) in your environment. (Admittedly unlikely in rural India.) Alternatively, a kludgy and ungainly method would be several iRiver recorders started and left running wild in front of the key musicians. They can run for hours and should hold sync, and later can be integrated into a multitrack sound editing program for a mix. The benefit of the iRivers for your application is that they're considerably cheaper. If some of them walk off, you will be able to replace them in the major cities (unlikely if the Edirol or DAT goes missing.)

But even a single iRiver in the middle of the musicians would be better than what you're getting. It really doesn't matter what your bit depth or s/n ratio is if your feed is a single mike at great distance from the subject in a large room with all hard surfaces. The audio is very hard to listen to. You may be so familiar with it that you're no longer hearing this.

[Cornucopia]

Good point. I concur.

Scott

[Raga]

Following up on this - the Edirol has now arrived. Quick specs from the battery life test.

Two 2200 mAh batteries. 4 hours 20 minutes of recording time in 16-bit WAV with plugin-powered mic, recorded from headphones placed next to the mic with medium input levels. No effects used.

Batteries were six months old, recharged perhaps 20 times in the past, so pretty fresh still. On top of that, 2 GB worth files copied from the CF card to the PC and a thorough browse of every switch and option in the device. Not bad.

* * * * *

Bastid - I agree, the main problem is the environment itself. On top of that, I have to be somewhat discreet with the whole operation, I gain more than my desired share of attention even without an elaborate setup.

I'll see what comes out with Edirol and a Delta mic with the MiniPod to direct it, and will try to get it placed as well as possible. If only they would turn off those damn loudspeakers!



(Yes, that's a LOUDspeaker!)