Tascam DR-60D problem

[Glen Trew]

Hi All,

 

A customer brought in a Tascam DR-60D and Canon 5D Mark II camera to figure out a noise problem worth mentioning here. It seems that the DR-60D is emitting more digital noise from its converters than the Canon 5D can shield.

 

Details are here: http://www.trewaudio.com/audioflow/?p=4978

 

Glen

 

[Jay Rose]

Well yeah, that's bad design... but if you're using a DR60 then the 5D's track is just for reference. This will force the editor to ingest the DR60's tracks right from the start.

 

Anybody test whether the digital hash on the internal track is enough to mess up Plural Eyes?

[Glen Trew]

Well yeah, that's bad design... but if you're using a DR60 then the 5D's track is just for reference. This will force the editor to ingest the DR60's tracks right from the start.

 

Anybody test whether the digital hash on the internal track is enough to mess up Plural Eyes?

The DR60 is not just for double system, but is also intended to be a mixer/interface for DSLR camcorders. Did not check Plural Eyes function, but I imagine it would work as normal.

 

gt

[Jay Rose]

I did sweeps / thd / dynamic range on a 5D Mk III yesterday, recording at -12 dBFS, using a Beachtek HDX DXA for interface. 

 

Please do not use the 5D's internal recording for anything other than reference. 

 

THD (thru the Beachbox's transformer, mind you) is a respectable < .05%, IM+Noise < .1%, dynamic range a just-tolerable 81 dB...

 

but low end starts rolling off at 200 Hz. -4 dB at 80 Hz, -10 dB at 40 Hz. 

 

(Yes, that's with no wind filter. The manual says it's not even available with external input.)

 

Could I fix the bass in post? Sure, if we can tolerate a little additional distortion and noise. But somebody who's shooting single-system on a 5D probably will never go into an audio post suite... and trying to fix that signal on a NLE in most edit suites will be a disaster.

[Glen Trew]

I did sweeps / thd / dynamic range on a 5D Mk III yesterday, recording at -12 dBFS, using a Beachtek HDX DXA for interface. 

 

Please do not use the 5D's internal recording for anything other than reference. 

 

THD (thru the Beachbox's transformer, mind you) is a respectable < .05%, IM+Noise < .1%, dynamic range a just-tolerable 81 dB...

 

but low end starts rolling off at 200 Hz. -4 dB at 80 Hz, -10 dB at 40 Hz. 

 

(Yes, that's with no wind filter. The manual says it's not even available with external input.)

 

Could I fix the bass in post? Sure, if we can tolerate a little additional distortion and noise. But somebody who's shooting single-system on a 5D probably will never go into an audio post suite... and trying to fix that signal on a NLE in most edit suites will be a disaster.

 

Hi. Jay,

 

That info may be interesting, but it is not related to the topic of original post. Also, without testing and stating the Beachtek box's performance (hardly considered a laboratory test instrument), the implications about the Canon 5D camera are meaningless. Further, a low end rolling off by -4dB at 80Hz, then -10dB at 40Hz (which, by the way, could have been entirely due to the Beachtek box) does not necessarily make a device unacceptable for dialog recording. For the sake of comparison, a respected old standard for sync dialog recording (still current in the opinion of many professionals) is the Schoeps CMC 41 mic element with CUT-1 filter that intentionally reduces low freqs much more steeply that the slope you described. Likewise, while the dynamic range of 81dB stated in your test is low by current standards, but it is still much better than the Nagra IV-S recorder was. So, the specs from your test aren't something that necessarily needs to be fixed. Lastly, the cameras used in the original article were the Canon 5D Mark-II, which may or may not be different that the Mark-III in inducing noise.

 

The statements in the original post were about audible noise significant enough to make otherwise good tracks unacceptable for professional or commercial use. The noise was generated by the Tascam DR60, induced into the Canon 5D Mark-II camera.

 

To put it all in perspective: The dialog signal through a Beachtek interface into a Canon 5D Mark-II Dialog was acceptable, but going through the DR60 into the 5D Mark-II, it was not acceptable.

 

gt

[Jay Rose]

Glen, we agree that single system on a 5D with a DR60 under isn't usable. Is it a bad design or lack of testing on Tascam's part? Probably yes, though I'd want to test a whole bunch of other DSLRs for comparison. If they're all noisy, it's definitely Tascam. If the 5D is noisy when you get it near an iPad or any other CPU, I'd wonder how carefully they shielded it...

 

And a Beachbox is hardly a piece of lab equipment. But I have gotten much better results using that box on other cameras... even in this past week's tests. So the rolloff is Canon's, not Beach's.

 

The comparison with a CUT-1, however, is specious. That's something you intentionally put on to cut the lows. I said in my post that I'd turned all the filters OFF for my test. (Would you use your Schoeps as much if the CUT-1 was glued on and could never be defeated?)

 

N-IV dynamic range less than 81 dB? At 15 IPS with good tape, and considering the headroom? Remember, the 5D's dynamic range includes its 12 dB headroom, so the noise floor is 67 dB below nominal... within a few dB of what you'd get from a Nagra in good tune.

[Glen Trew]

Glen, we agree that single system on a 5D with a DR60 under isn't usable. Is it a bad design or lack of testing on Tascam's part? Probably yes, though I'd want to test a whole bunch of other DSLRs for comparison. If they're all noisy, it's definitely Tascam. If the 5D is noisy when you get it near an iPad or any other CPU, I'd wonder how carefully they shielded it...

 

And a Beachbox is hardly a piece of lab equipment. But I have gotten much better results using that box on other cameras... even in this past week's tests. So the rolloff is Canon's, not Beach's.

 

The comparison with a CUT-1, however, is specious. That's something you intentionally put on to cut the lows. I said in my post that I'd turned all the filters OFF for my test. (Would you use your Schoeps as much if the CUT-1 was glued on and could never be defeated?)

 

N-IV dynamic range less than 81 dB? At 15 IPS with good tape, and considering the headroom? Remember, the 5D's dynamic range includes its 12 dB headroom, so the noise floor is 67 dB below nominal... within a few dB of what you'd get from a Nagra in good tune.

 

This is getting away from the original post (large amounts of data noise induced from the DR60 to the 5D Mark-II), but I'm a sucker for dynamic range discussions...

 

I've serviced hundreds of Nagra 1/4" recorders, and know them to be amazing machines, but the dynamic range of Nagra 4.2 was around 70dB at 7-1/2ips, and the distortion was typically around 1% for a new machine at 320nW/m. Keep in mind that those specs were on the best day for a new machine with low noise, low print tape, and would often vary from machine to machine and from one tape batch to the another, and usually declined with age to never be scene or heard again. Even when new they had some odd peaks and dips, particularly in the low freqs. Dynamic range specs include the full range, right up to maximum (0dBfs for digital), and M.P.L (Maximum Peak Level for analog). So, a recommended 0VU reference of -12dB will have no affect on a recorder's dynamic range spec. Analog Nagra recorders had only 12dB of headroom above their 0VU reference (8dB below the modulometer's 0, then another 4dB to reach it's M.P.L. = 12dB). This was largely because of the need to counter the evils analog tape hiss.

 

But even then, as we all know, Nagra recorders were used to successfully record thousands of great music and dialog tracks that are still enjoyed today.

 

Now, back on topic:

 

The noise created by the DR60 that was induced into the 5D Mark-II (tested on two cameras with the same results) was blatantly obvious and didn't require any test instruments other than untrained human ears. It was so obvious, in fact, that a camera operator brought the system in to find out what was wrong. (Hilarious, sad, and true)

 

I want to emphasize a point in the original post: The noise was only a problem when the camera was bolted onto the DR60 with the treaded adapter, as intended by the manufacturer. Locating the camera just a couple of inches away made the problem go away completely. But this removes the convenience of mounting the DR60 between the camera and tripod, and the owner (a camera operator) decided that was not an solution he could live with.

 

Glen Trew

[dfisk]

Did you try taking that same output into another input device that can take a stereo mini input? Maybe it's a specific issue with just these two devices? I have a DR-60D and a 7D, but I don't have the stereo mini to stereo mini cable to put between them, so I can't test this on my end. 

[rafaeldelauz]

Glen, I own a 5d mark iii and the tascam and I've been using both, screwed to each other without issues, but after reading your article I realize I always use the line ouput instead of the tape out. Did you try different outputs on the tascam?

BTW, and I know this is no the issue you are talking about, I'm pretty happy with the quality of the internal recording in the d60

[Tom Duffy]

We haven't been able to reproduce the reported problem here at TASCAM with a 5DmkII (or any other camera), trying to reach out to Trew to share a sample of the noise.

 

No-one from Trew Audio has contacted us directly about this issue.

 

Tom Duffy (TASCAM)

[Glen Trew]

Glen, I own a 5d mark iii and the tascam and I've been using both, screwed to each other without issues, but after reading your article I realize I always use the line ouput instead of the tape out. Did you try different outputs on the tascam?

BTW, and I know this is no the issue you are talking about, I'm pretty happy with the quality of the internal recording in the d60

Yes, the line out was tested. Keep in mind that the customer's problem was when using the DR-60 the 5D-MkII (not the MkIII). The problem with using the line out is that the 5D-MkII input cannot handle the DR-60 line out with enough headroom, even when barely cracked open. But, even though the higher voltage of the line level signal reduced the significance of the noise, the noise was still there.

 

gt

[studiomprd]

"using hte DR-60 the 5D-MkII, not the MkIII."

so maybe it is a Canon 5D-MkII, problem..?

[Glen Trew]

"using hte DR-60 the 5D-MkII, not the MkIII."

so maybe it is a Canon 5D-MkII, problem..?

It is possible that the 5D MkII isn't shielded as well as it should be. However, the noise was produced by the DR60, not the 5D MkII.

 

GT 

[Glen Trew]

We haven't been able to reproduce the reported problem here at TASCAM with a 5DmkII (or any other camera), trying to reach out to Trew to share a sample of the noise.

 

No-one from Trew Audio has contacted us directly about this issue.

 

Tom Duffy (TASCAM)

Hi Tom,

 

I'm still waiting on a return call from Gary Jones, which I placed on September 3rd. I'm traveling in Europe now, but I left my cell phone number for him, without response. Email is also available. The problem was reported because it was, indeed, a problem that caused serious problems for a user of the Tascam DR-60.

 

Glen Trew

[dfisk]

were multiple DR-60D units tried? Could be a "ghost in the machine". I've seen that happen a lot with all kinds of products. 

[studiomprd]

" However, the noise was produced by the DR60, not the 5D MkII. "

acknowledged,  but now-a-days everything has some EMI radiation...

unless this was a particularly bad DR-60 interfering with several different camcorders, and different camcorder models(thus needs service), I'd still say is is potentially as much a camcorder issue as it is a sound recorder issue.