Tascam DR05 - New Rebate Offer!

[Kyle Root]

I was looking around on BH this morning and saw that the Tascam DR05 has a $20 rebate, making it $59!

I already have one, and the bigger DR-40, and love them both for wedding work. At this price, I'm going to pick up another!

If you're in the market for a small, pocketable (with an attached lav), or easily hidden recorder, the DR-05 does the trick.

http://www.bhphotovideo.com/c/product/757065-REG/Tascam_DR_05_DR_05_Portable_Handheld_Digital.html

[Greg Miller]

That looks like an irresistable price... can't have too many toys! Thanks for posting that.

The literature says the mics are omni, but I wonder whether anyone has confirmed that. If they're really omni, they are too close together to capture a useful stereo image. But if they're cardioid, they might be useful in a pinch. Has anyone here tried the DR-05's integral mics? If so, please report your observations.

[Jay Massengill]

I rarely use the internal mics on my DR-05, but in the simple testing when I first got it you can clearly tell moving sound sources are passing through the stereo image.

Unlike the Zoom H4n with X/Y coincident mics, I guess the 2 inch spacing of the Tascam's A/B mics is enough even if they are omni.

I didn't test for sound pickup from behind the mics.

[Greg Miller]

Jay, thanks for the comments.

I have had an original DR-07 for a long time, and also got a DR-03 for "in-pocket" use. The DR-03 has mics which are definitely omni, and with about 1.5" spacing. If I listen to playback on headphones, I can hear a bit of stereo image, but it's not at all useful for normal playback on loudspeakers. The DR-07 mics are a bit of a mystery: they produce a more obvious stereo image, but visually appear to be aimed 180º apart which is obviously wrong.

Perhaps the DR-05 mics produce a better sound field. I hope your impressions are correct, Jay, thanks again! And perhaps someone else will chime in...

[Kyle Root]

Unfortunately, I've never used my DR-05 built in mics. I use mine just for backup audio with a lapel mic and drop the DR-05 in a jacket pocket of the minister.

Now I'm curious though. Maybe I'll test them out.

[Greg Miller]

Just thought I'd bump this topic after a week of "dead air" here.

Has anybody tested the directionality of the integral mics?

Thanks!

[Seth Bloombaum]

It's not just directionality that contributes to the perception of stereo. It's also timing, which in these very small differences is also known as phase.

Our perception of directionality relies on several perceptual cues that are processed to create "direction". These include:

* Clean vs. reverb-y sound. If I hear a sound that is cleaner in my right ear, and has more reverb in my left ear, I think it originates from the right, and have a sense of space to my left.

* Early vs. late sound. If I hear a sound earlier in my right ear, and later in my left ear, I think it orginates from the right.

* Clean vs. reflected sound. If I hear a sound that is cleaner in my right ear, and is later in my left ear, and also has different frequency characteristics in the left, consistent with reflection off hard surfaces, or absorbtion off soft surfaces, I think it originates from my right, and have a sense of space and objects to my left.

That there is mass and distance between the two mics helps bring out the small timing differences, emphasizing the sense of space and direction.

These timing differences are exploited even more in ORTF near-coincident and Binaural recording techniques. They really work! It's amazing what the brain does in processing these cues for directionality and space. We're talking extremely small timing differences, on the order of half a millisecond.

[Greg Miller]

Either phase, or level, or both, can create a stereo image. Whether the location of the sound, relative to the microphone pair, causes phase (timing) difference, or level difference, or both, depends on the microphones used.

An X/Y (coincident) pair of directional mics produces mainly level difference (and also some difference in frequency response between mics, since the response is different at different angles relative to the mic axis).

If the mics are both omni, then presumably they will both pick up the same amount of reverberation, or the same ratio of direct to reflected sound (whichever way you want to describe it).

A properly spaced pair of omnis produces only phase/timing difference, since the level will be very nearly the same. Quite adequate for binaural listening on headphones, but not always so obvious with playback on speakers. Omnis work better with significant baffling between them, e.g. a dummy head, or a jecklin disc. But the DR-05 has very small spacing, and insignificant baffling.

As I stated above, I have listened to recordings made with two omnis, spaced roughly two inches apart (similar to the DR-05 spacing). Listening on headphones, there is just barely any stereo image, because the phase of the two mics is very nearly the same. Listening on speakers, it's so close to a mono image as to be indistinguishable. So I believe that a two inch spacing (as with the DR-05 mics) is not adequate to produce a useful stereo image if the mics are omni.

And while the literature says "omni," I note that there are some side openings, not just a grille in the front of the capsule. Usually I associate side openings with some sort of directionality. Any omnis that I can think of do not have side openings. This is not very scientific, but it gives me reason to wonder about these specific mics.

Thus my question: has anyone paid enough attention to the DR-05's integral mics, to say whether they seem to be directional, or whether they really are entirely omni.

[Jay Massengill]

I've carried my DR-05 and a tripod around with me a couple of times this week. Made some quick and dirty recordings of things I thought would demonstrate the built-in mics' stereo response.

Certainly the noise floor is higher than I would like when recording very quiet subjects. Really this device was made for recording stereo in much louder environments, like a music ensemble or band practice.

I need to do more experimentation, but for serious work clearly a pair of really quiet condenser mics and a good preamp would be the ticket. This is how I normally use the DR-05 anyway when working indoors in a fixed location.

Attached are two audio files using the built-in mics, one minute each.

The first is at my radio control club field near a large lake, with woods behind me. You can hear tree frogs, a distant jet coming into the airport 15 miles away, a few guys 100 yards away chatting, and most importantly some passes and maneuvering of a small electric RC model that I was flying.
The overall recording level was too low, but this was really just to judge the model's movement in the stereo field. You can tell it's moving and which direction it's going, but the effect on typical computer speakers isn't dramatic.
I didn't apply any noise reduction.

I did pick up major interference from my 2.4Ghz RC transmitter when standing right at the recorder. The antenna was only a foot away from the recorder at that point, with the transmitter hanging around my neck while I made adjustments on the DR-05. At 20 feet away while I was flying I didn't hear any interference in the recording.

The second file is just 1 minute of tree frog and small bird ambiance. Interestingly this file really sounds noticeably "in stereo" to me even though there is no movement.

This file could certainly be used in a production where evening swamp sounds are needed as a quiet bed. The levels and noise floor are ok for use under the main dialogue, and the track loops well if more time is needed.
Feel free to use it if you need it.

As mentioned earlier in the thread, the mics are very close to omnidirectional.

[Greg Miller]

Jay, thanks for posting those.

I don't hear as much stereo image as I'd hope to hear on the RC fly-by. Of course I wasn't there and don't know how wide an angle the plane filled (e.g. maybe it was quite far from the recorder, and the entire fly-by only encompassed 15º of angle as seen by the recorder). However, I'm inclined to agree with your conclusion that the mics really are not very directional.

I'm also puzzled about the tree frogs. They seem to have a bit of a stereo image, but when I look at the playback on a vectorscope, they are very significantly out of phase... almost as if either the phase was reversed on one of the internal mics (very unlikely) or else this was matrixed and the (L+R) was lowered relative to the (L-R). Assuming you didn't process the file at all, I wonder whether Tascam has incorporated some artificial widening in their electronics, to compensate for the poor directionality of the mics?

Some things to ponder, that's for certain.

[Jay Massengill]

You'll be really disappointed then to learn the model plane was passing no more than 30 feet from the recorder, on as close to 180 degrees across as I could fly.

However, this airplane has a 6" (diameter) x 4" (pitch) propeller mounted at the rear and spinning at around 20,000+ rpm.

The pusher configuration, especially with twin tail fins close to the prop, has a very different sound signature than a "normal" airplane and the small prop size and high rpm are setting up lots of odd sound wave interactions I'm sure.

The tree frog results are interesting. I didn't do any processing. Perhaps rendering to wma at 256kbps has introduced some oddities.

I can say the frogs were not a uniformly distributed sound source. The loudest one on the left was only 8 feet away and distinctly noticeable in person. Some of the others at slightly greater distances and different angles could also be isolated when standing next to the recorder and listening with bare ears.
Perhaps my experience of listening to the real thing is influencing my listening of the recording. It really sounds like it did in person when I listen to the recording.

[Rick Reineke]

Tree Frogs: A slight stereo image.
Sound Forge indicates 'in phase'. The same rendered to mono, no audible abnormalities. Other than that I have no comments.

[Greg Miller]

Very interesting, Rick. Thanks for providing some more food for thought. Were you sampling any particular place in the file?

I selected just one brief section of the file where the frogs were much more audible than the motor, somewhere around 22.5 seconds into the file. Other places, where the motor sound predominates, the picture is much different, with the motor sound quite visible and changing in phase and channel balance as the plane moves past. With the exception of the short times when the motor is much louder, the vectorscope always seems to be tilted into the second quadrant, i.e. more out-of-phase than in-phase material.

To verify that the vectorscope isn't fooling me, I recorded identical pink noise onto both tracks of a stereo file. (Not stereo pink noise. I recorded mono noise onto one track, then duplicated that onto the other track.) I then converted this to a .wma file, closed it, re-opened it, and viewed the vector display. A thin solid line in the middle of the first quadrant, as expected.

It occurs to me that if the frogs are at a very nearly constant frequency, and if the time difference between mics (based on the off-axis angle to the frogs) is "just right," it might create tracks that are significantly out of phase. Offhand I can't think of a good way to test that, without resorting to a much tightly controlled situation. (e.g. a large non-reverberant space, a point source of sound, ability to rotate the recorder on a tripod, etc.)

At any rate, given the lack of separation (especially if the fly-by crossed nearly 180º of the sound stage), I don't think it's worth pursuing further tests with this recorder. So maybe $59.00 really is a fair price, and not a great deal, after all.

EDIT: I just looked at the "frogs and birds" file and that one is very strongly in phase! I wonder what accounts for the difference between the two files. It might be interesting to pursue the vector display of this file, if anyone else has any additional thoughts.

[Jay Massengill]

The tree frogs that are audible in the first recording of the model airplane are the same frogs as in the actual frog recording. However they are 170 degrees to the left of center and some 100 feet behind the recorder when I was recording the airplane.

During the frog recording, I'm very close to the edge of the wetlands and forest, and essentially on-axis with the general sound coming from this area. I'm guessing at least 5 frogs were within 20 feet of the recorder and no more than 35 degrees combined Left and Right from the center axis of the recorder.

In other words, while it is the same frogs making the sounds in both recordings, the recorder has been rotated 170 degrees to the left, moved 90 feet away from the flying field and toward the edge of the wetlands, and tilted down instead of up. Two totally different recorder placements for two totally different recordings.

[Rick Reineke]

"Were you sampling any particular place in the file?"
Yes, but it's pretty consistent all the way through with no phase issues at all.

Same with these scopes: