Seems like comparing apples to oranges................

I use multiple iriver recorders to backup audio from events we've videoed.
Normally we get the backup audio from the soundboard. We also use shotgun mics to get audio. I am just wondring about the quality of the small Iriver recorders. Sound is great to us, and when I looked up the figures on the net, it seems to be on par with FM radio broadcasts but somewhat below the quality of an audio CD.

Trying to compare specs, I found the following which is a little confusing since the same unit of measure is not used.

i river mp3 digital recorder(line in or ext. mic in) = 44kHz 160kBPS

Canon camcorders = PCM 16 bit (48kHz/2ch)

Just wondering if one of you can interpret these numbers for me and maybe make it a bit easier to understand ?

...I use multiple iriver recorders to backup audio from events we've videoed.
Normally we get the backup audio from the soundboard. We also use shotgun mics to get audio. I am just wondring about the quality of the small Iriver recorders. Sound is great to us,...
If it sounds good, it is good. This is the most important measure.
...Trying to compare specs, I found the following which is a little confusing since the same unit of measure is not used.

i river mp3 digital recorder(line in or ext. mic in) = 44kHz 160kBPS

Canon camcorders = PCM 16 bit (48kHz/2ch)

Just wondering if one of you can interpret these numbers for me and maybe make it a bit easier to understand ?
Here's an outline of the basics of these measurements. Most importantly, the PCM indicates an uncompressed signal, we'll start there.

16 bit - this indicates a sample of 16 bits, that is, a collection of 16 ones and zeros, that's the length of the digital "word".
48KHz - this indicates that a new sample is being created 48,000 times per second.
So, for your canon camcorder, a 16-bit sample 48,000 times per second, with no compression. This is the audio standard for DV-MiniDV-DVCAM recordings.

On your iRiver - MP3 is a compressed format. This means information is thrown away. The MP3 method is perceptual, that is, it attempts to throw away info that we wouldn't perceive anyways, more or less.
44KHz - indicates that a sample is taken 44,000 times per second.
160Kbps - is a measure of bitrate, in this case 160,000 bits are created every second.

We can do some conversions that will tell us "this particular MP3 stream is 1/x the size of uncompressed", but, nobody I've met in production work really understands the deeper math behind MP3 compression. I certainly don't. Never bothered to do those conversions, I'd rather listen, since a statement like "this MP3 is 1/12th the size of the uncompressed signal" is only good for impressing people, it's not really that useful.

MP3 started out as a delivery codec. Yes, I too have used it in recording, when I needed some small and light recorder. The big question for all of us to answer individually is whether it's good enough for the purpose at hand. IMO, a 44KHz/160Kbps recording is not good enough for critical music recording, but, can be good enough for incident music and efx recording, and is usually fine for dialog.

Your mileage will vary. If it sounds good, it is good.

PS. Depending you your editing environment, it may or may not be important to either record all sources at 48KHz, or, convert them to 48. 44KHz source on a 48KHz timeline is a frequent source of sync problems.

I posted this recently on another forum when there was some confusion over formats, file sizes, editing & quality issues. Some of which may be relevant here.
Use a Wave format for the initial save and subsequent editing saves and only then, encode the MP3 for your final delivery format if needed.
To clarify, and for instructional purposes, *I am referring to spoken word material, and where there is no spacial information, such as with stereo music.
In the WAVE format, Mono/Stereo affects only the file size.. A one minute spoken word 44.1KHz, 16 bit 'stereo' .wav file, would be approx.10MB, whereas the same wave file in 'mono', would be approx. 5 MB. There would be no gain or loss in quality.*
However the MP3 format is quite different. A 64kbps 44.1 CBR MP3 encoded in 'mono' would have the file size of approximately 450KB, but would have the equivalent quality of a 128kbps MP3 encoded in 'stereo', which has the file size of about 975KB. To put it another way, a 128kbp/s CBR MP3 encoded in EITHER stereo OR mono, would be the same file size. BUT, the mono version would have superior quality.
Addendum,
In some informal tests, a mono 44.1/16 bit 328kbps MP3 and a 44.1/16 PCM file completely cancel each other out when realigned and one is flipped 180 degrees out of phase. Artifacts were audible in any encoding less than 328kbps.
Filesizes:
60 second 44.1/16 PCM mono = 5.1KB (stereo=10.2KB)
60 second 44.1/16 MP3 328kbp/s CBR = 2.3KB
Encoder: winLAME rc3: filters 18K/30Hz

Another issue frequently brought up, "Why can't I loop MP3 files"
The MP3 encoding process adds approx. 40 milliseconds at the head and tail of the file.

If it sounds good, it is good. This is the most important measure.

NOTE: That methodology is only valid in the FINAL MIX. You would do yourself a serious disservice by trying to evaluate raw tracks by that measure.

Note that professionals typically avoid using ANY kind of lossy compression (like MP3) for production recording. Lossy compression BY DEFINITION irretrievably discards audio information. Furthermore decompressing lossy recordings and then re-compressing the audio (for distribution over the internet or on DVD, for example) frequently creates ugly artifacts because of the repeated lossy compression processes. I would NEVER use lossy compression for recording production audio tracks and I don't know anyone who would do that.

If it sounds good, it is good. This is the most important measure.


Define good.

Then realize half the people on this forum think a Zoom sounds good.

Now redefine good.

Then realize half the people on this forum grew up listening to MP3s and have never heard a .wav or .aiff file.

Now redefine good.

I don't wish to seem anti-pro-sound, because I'm not. Certainly I agree that lossless field recording at 24-bit/48KHz is part of a best-practices approach to professional acquisition.

However, the original poster is using iRiver recorders for backup in live-event recording, and wanted to make some sense of how to interpret those specs vs. the specs of his DV-camcorders.

"If it sound good, it is good" is really so important to evaluating gear or workflows, and its importance frequently gets diminished in these online discussions, in which well-intentioned people solemnly pronounce their judgement that what the O.P. wants to do really can't be done without a Sound Devices recorder and Schoeps microphones.

Here's my point - there are all sorts of markets for audio and video recording, program creation, and distribution. People who are trying to make their living in weddings or event recording are working in a legitimate market, with real opportunities. The iRiver MP3 recorder is a favorite "spot" or "plant" recorder in those markets, and people who are looking for success there don't need pros from a different market telling them how wrong they are to be using a lossy codec for acquisition.

If they are to make a go of things, they need to carefully consider "best practices" for their market, because these practices have been shown to satisfy customers at a price-point that allows the production service to make money.

So, I'll modify Duke Ellington's immortal quote: If it sounds good enough, it is good enough. "Good enough" as delivered to client, "good enough" as perceived by the client.