Topic: Tascam DR-70D 4-channel audio recorder (Part 6)

[voltronic]

See reply 189.  I already said I got it wrong, and Jim was talking about input transistors.

[soundsinteresting]

Most of Tascam high end audio products claim they use Tascam HDDA input amplifier tecnique and 5532 amplifiers (except Dr701)).
I have not seen any diagrams but to me it is interpreted as Tascam input amplifiers has a discrete transistor input stage combined with 5532 opamps which together set input noise levels obtained.
That may explain why it did not change a lot to switch 5532 amplifiers with even less noisy amplifiers as discrete components are still a major part of noise levels.
Anyway I would personally not really care of the present noise as that is hardly the limit for most users compared to microphones used generally.

However i have seen in other forums that people struggle with camera/line out especially camera out which seems to have severe noise problems due to some design flaw with output amplifiers.
Most users referred with this problem has solved this S/N problem by neglecting camera out (which should have been an attenuated line signal) but using the line out signal and purchase a cable with builtin attenuator (25dB in this case).
Camera out was probably intended to be used with DLSR cameraes which have a microphone input with high sensitivity so lineout needs to be attenuated to fit this input signal level.
By using the line out with proper gain settingsfor this it seems to be working but using cameraout it seems to be a degraded lineout signal with lower S/N level that is not of decent quality. That is a pity :-(.

Does anyone in here have any experience on this problem with Tascam line out versus camera out settings and how it could be related to lineout gain settings eventually and why it is better to use an attenuator cable instead of internal gain setting? Again this could be a design flaw similar to camera out.
 

[soundsinteresting]

Most of Tascam high end audio products claim they use Tascam HDDA input amplifier tecnique and 5532 amplifiers (except Dr701)).
I have not seen any diagrams but to me it is interpreted as Tascam input amplifiers has a discrete transistor input stage combined with 5532 opamps which together set input noise levels obtained.
That may explain why it did not change a lot to switch 5532 amplifiers with even less noisy amplifiers as discrete components are still a major part of noise levels.
Anyway I would personally not really care of the present noise as that is hardly the limit for most users compared to microphones used generally.

However i have seen in other forums that people struggle with camera/line out especially camera out which seems to have severe noise problems due to some design flaw with output amplifiers.
Most users referred with this problem has solved this S/N problem by neglecting camera out (which should have been an attenuated line signal) but using the line out signal and purchase a cable with builtin attenuator (25dB in this case).
Camera out was probably intended to be used with DLSR cameraes which have a microphone input with high sensitivity so lineout needs to be attenuated to fit this input signal level.
By using the line out with proper gain settingsfor this it seems to be working but using cameraout it seems to be a degraded lineout signal with lower S/N level that is not of decent quality. That is a pity :-(.

Does anyone in here have any experience on this problem with Tascam line out versus camera out settings and how it could be related to lineout gain settings eventually and why it is better to use an attenuator cable instead of internal gain setting? Again this could be a design flaw similar to camera out.

Well...
I made a few measurements with the DR70 line out, and  guess what: I did not recognize the problem that a lot of other people have met when using the lineout signals and found too much hiss with camera out.
Are there any information that the latest version 1.13 should have removed this hiss problem or are we up against some basic operator problems here incl me?

Basic is that hiss should be measureable like any other signal.
Difference of line out and camera out is a level difference of 30db (camera out is line out -30db)
so i made a test of measuring noise at line and camera out and it shows a difference of 30dB as stated and no more or less  noise at camera out.
a simple setting is that at monitor you may set lineout to -18dB and camera out to +12 dB and voila you get the same volume output at line/camera out. (measuring AC at output)

So how can this be explained.
Does this problem not exist other than people abusing level settings when connected to DSLR's and similar or am i missing something.
Why would people go out paying for expensive cabling (attenuators) to solve a problem that may not exist?
I just cannot see it, sorry.

This is one of the forums that claim this to be a problem and some reports here say that problem is present at least with ver 1.12:
http://www.dvxuser.com/V6/showthread.php?329630-Tascam-DR-70D-what-do-you-think/page19

[voltronic]

FWIW, the mini jack line in and outs use different opamps: NJM 4580.  The NE5532A are only used for the 4 main XLR combo inputs. 

I never tried using the line or camera outs until after I got my unit modded, which replaced all the opamps with LME49720.  Both outputs are noiseless for me, but maybe others can share their experiences with a stock unit.

[mfrench]

I found out, this weekend, that the Aux-in in Sarges Pontiac Vibe head unit doesn't like the line-level output from the DR70D. 
I plugged it in to have a listen to the show on the way home, and started experiencing severe distortion.  I thought; How the heck did I hammer this recording so badly? [eyes bugging out smilie]  I then unplugged, and ran headphone out, and it all cleared up.
No problems with any issues in any recording so far.....

[voltronic]

I found out, this weekend, that the Aux-in in Sarges Pontiac Vibe head unit doesn't like the line-level output from the DR70D. 
I plugged it in to have a listen to the show on the way home, and started experiencing severe distortion.  I thought; How the heck did I hammer this recording so badly? [eyes bugging out smilie]  I then unplugged, and ran headphone out, and it all cleared up.
No problems with any issues in any recording so far.....

The line out has an adjustable level in the menu - did you go into the settings and try lowering it to see if you were just brickwalling your aux input on your head unit?

[mfrench]

I was on the road home from the gig. Menu fiddling was not the thing to be doing at that point.  So at a traffic signal stop, I re-jacked into the headphone jack, and, enjoyed the rest of the trip home.  Next time, I'll check the output.
Yeah, definitely was brickwalling the head unit.

[soundsinteresting]

Most of Tascam high end audio products claim they use Tascam HDDA input amplifier tecnique and 5532 amplifiers (except Dr701)).
I have not seen any diagrams but to me it is interpreted as Tascam input amplifiers has a discrete transistor input stage combined with 5532 opamps which together set input noise levels obtained.
That may explain why it did not change a lot to switch 5532 amplifiers with even less noisy amplifiers as discrete components are still a major part of noise levels.
Anyway I would personally not really care of the present noise as that is hardly the limit for most users compared to microphones used generally.

However i have seen in other forums that people struggle with camera/line out especially camera out which seems to have severe noise problems due to some design flaw with output amplifiers.
Most users referred with this problem has solved this S/N problem by neglecting camera out (which should have been an attenuated line signal) but using the line out signal and purchase a cable with builtin attenuator (25dB in this case).
Camera out was probably intended to be used with DLSR cameraes which have a microphone input with high sensitivity so lineout needs to be attenuated to fit this input signal level.
By using the line out with proper gain settingsfor this it seems to be working but using cameraout it seems to be a degraded lineout signal with lower S/N level that is not of decent quality. That is a pity :-(.

Does anyone in here have any experience on this problem with Tascam line out versus camera out settings and how it could be related to lineout gain settings eventually and why it is better to use an attenuator cable instead of internal gain setting? Again this could be a design flaw similar to camera out.

Well...
I made a few measurements with the DR70 line out, and  guess what: I did not recognize the problem that a lot of other people have met when using the lineout signals and found too much hiss with camera out.
Are there any information that the latest version 1.13 should have removed this hiss problem or are we up against some basic operator problems here incl me?

Basic is that hiss should be measureable like any other signal.
Difference of line out and camera out is a level difference of 30db (camera out is line out -30db)
so i made a test of measuring noise at line and camera out and it shows a difference of 30dB as stated and no more or less  noise at camera out.
a simple setting is that at monitor you may set lineout to -18dB and camera out to +12 dB and voila you get the same volume output at line/camera out. (measuring AC at output)

So how can this be explained.
Does this problem not exist other than people abusing level settings when connected to DSLR's and similar or am i missing something.
Why would people go out paying for expensive cabling (attenuators) to solve a problem that may not exist?
I just cannot see it, sorry.

This is one of the forums that claim this to be a problem and some reports here say that problem is present at least with ver 1.12:
http://www.dvxuser.com/V6/showthread.php?329630-Tascam-DR-70D-what-do-you-think/page19

I am still puzzled of this problem that some people claim Tascam makes noise out of LINE/CAM output, welll I cannot make it happen, sorry guys.
May be it is because i do not have a GH4 that is part of the problem, but i try to be objective and get some value facts.
Actually no one claims it to be noisy but claim that S/N is worse when CAM Output is used, but my claim is that it is NOT, problem must be elsewhere in the understanding of connecting a second MIC amp (GH4)

As S/N is the clue i do not care of signal levels but try to measure noise levels of Tascam.
my setup is:
Input CH1 is shunted by 150 ohm impedance simulating a low noise microphone at XLR.
GAIN is set at max HIGH PLUS (+63dB)
LINE Output is taken into some recorder, I have used Sony M10 but it could be any lownoise amplifier/recorder but it could also have been a quality AC Voltmeter but we are talking of very levels (uV).
Output at LINEOUT is Input amplifiers selfnoise (amplified 63dB) which is claimed to be better than -120dBu (<1uV)

The picture below is showing noise output at LINE OUTPUT at LINE GAIN settings of +10 dB/0 dB/-10dB/-20dB/-30dB
Same picture can be made for CAM output which is LineOutput -30dB.

These steps on recorded noise show exactly 10 dB attenuation for all steps so no additional noise is added at different attenuator settings  nor for CAM Output.
Claim was that noise is increasing on LINE/CAM Output when attenuation is set lower and therefore attenuation cable would make a difference.
Well it is not TASCAM that is producing this noise as you may see at picture which is total noise coming out of LINE OUT with input shorted and displayed with Audition from noise file recorded with M10.

I am not able to conclude what makes GH4 a problem with Tascam, which seems to be a fact as well, but it is not explained by a noisy output of TASCAM unit
Conclusion is that you will have the same noise and S/N whether you use Tascam attenuation on LINE OUT or CAM (-30dB) and it would not normally need any attenuation cable

Maybe others could dig into this problem with GH4 and have a conclusion that is eatable, but I am convinced that it is not caused by TASCAM DR70.
TASCAM DR70 is a fine recorder but have to be used with care as any recorder.
PERIOD.

I hope you forgive me being stubbern on this :-)

a little "googl'ing" gave this, which might explain a lot:
http://suggestionofmotion.com/blog/panasonic-gh4-audio-buzz/

[Life In Rewind]

GAIN is set at max HIGH PLUS (+63dB)

Have you tried this one of the lower gain settings?

Others have said the HI-PLUS setting is noisy - regardless of how you are routing the signal...

[soundsinteresting]

GAIN is set at max HIGH PLUS (+63dB)

Have you tried this one of the lower gain settings?

Others have said the HI-PLUS setting is noisy - regardless of how you are routing the signal...

No
The noise measured is the absolute value of noise generated in input amplifiers, amplified at 63 dB (1400x) and routed to LINEOUT, where it could be further amplified (12dB) or attenuated.
This is ofcourse the same condition for any signal added to input.
Are you suggesting that noise at LINEOUT could be different dependant on input amplifier gain range selected? It is somehow simulated by setting LINEOUT GAIN to -10-60dB

HIGH GAIN+ is not very noisy. Most people are interpreting this wrong because microphones generally are very noisy (20dB(NOT dBA)) so what you se on display with most mics is mic noise amplified 63dB and that is not looking very usefull. With mic replaced with dummy 150ohm, noise is not even visble at display at max gain 63dB  (-57-60dBu or app 1mV) .
 if you take a low noise mic (5-10dB) like MKH50/NT1A you will not see any noise at display if used at very low sound levels or at low level sound chamber:

[soundsinteresting]

i may be the last person on this channel that has any curiousity left with DR 70.
It would be nice if somebody could try to do the same test or maybe correct me if i am doing anything wrong at the following test.

i have a problem with LINE OUT signal that I am trying to understand but not succeeded sofar.
I have verified that LINEOUT has a permanent GAIN of 8.2 dB (2.5X) from XLR input to the display when LINEOUT GAIN set to 0dB.
It is verified with specs as LINEOUT has a max output of 6dBV (2V RMS) and as display has a max of 0dBu (0.775V RMS) this would be a difference of2/0.775=2.5X (8.2dB).
This has been verified at different gain settings so if you put a signal to the XLR input you will see a difference between signal displayed and LINEOUT of 8.2 dB (2.5X) . (LINEOUT GAIN set to 0dB)

Sofar so good. this could influence some interface understandings but ok.

My problem is a bit further.

If i try to use SLATE I get a different result and that seems strange or at least I do not understand it.
If I set SLATE to -12dB it would show -12 dB on display which would be the same as 0dBu-12 = 0.775V/4 which is the same as 0.19375V.
As remembered from above LINEOUT has a gain of 8.2dB so the expected output would be 0.19375V X 2.5 = 0.48V right?
Well TASCAM has another result as output is double up 0.96V (+6dB) which puzzles me because why should gain from input to LINEOUT be different for SLATE to LINEOUT.
It is fairly easy verified if you set LINEOUT gain to compensate for this. Set LINEOUT GAIN to -(6+8)dB (-14dB) and you will se that LINEOUT now show the expected 0.194xxV at output (app).
This could mean that you are not getting the  calibration you intended as calibration signal is +6dB too high compared to the actual signal when you connect somthing to LINEOUT.
 
To me it looks like a design bug but if anyone could clarify this to me i would sleep better :-).

[voltronic]

You're certainly not the only person here still interested in this recorder, but you may be one of the few here who both own one of these and have the measurement capabilities.

Perhaps one could use a test tone generator with a known output level into the line input and compare that to the internal slate tones?  I could do that with my FP24.

[soundsinteresting]

You're certainly not the only person here still interested in this recorder, but you may be one of the few here who both own one of these and have the measurement capabilities.

Perhaps one could use a test tone generator with a known output level into the line input and compare that to the internal slate tones?  I could do that with my FP24.

Ok then :-)
I have not yet been playing with LINE inputs so that could be interesting as well, I am not sure I even have a cable that fits.
That FP24 looks like a fine piece of HW  .

OK, found a cable and made the same test with input at LINE IN.
I set LINEOUT GAIN to 0 dB. Input signal at LINE IN 1kHz at -12 dBu.
Set SLATE at -12dBu as well.
Start recording with LINE IN at -12dBu and measure at LINEOUT which measures 480mV RMS (LOW GAIN). (that is 0 dBu/775mv -12dB + 8.2dB LINEOUT GAIN)
When SLATE is pressed,display also shows -12dBu, same as signal but LINEOUT raises +6dB to 960mV with SLATE active and falls back to 480mV when released.

So conclusion is that if you use SLATE for calibration of camera or similar connected equipment, you should be aware that SLATE causes a +6dB higher signal at LINEOUT than expected.

Unless I am blind to something this is definitely a severe error which should not have  passed, but maybe TASCAM made this for marking only and did not intend it to be used for calibration.
I will wait for someone to confirm this before I report this to TASCAM   

[Life In Rewind]

Unless I am blind to something this is definitely a severe error which should not have  passed, but maybe TASCAM made this for marking only and did not intend it to be used for calibration.
I will wait for someone to confirm this before I report this to TASCAM   

I dont think it was meant for calibration - just marking start points.

[2manyrocks]

The 70d product page says the slate tone is an aid to sync audio and video. I don't recall if a TS member figured out you could also use it to match levels or if that's in the manual.   But it wouldn't hurt to mention to Tascam that the tone is very useful to set matching levels between the 70d and camera, and to ask if they could address this in future firmware releases.