Topic: Tascam DR-70D 4-channel audio recorder (Part 2)

[voltronic]

I have not come across any liniter noise

I was just using that as an example.  Maybe a better one could be the three different gain level settings - if one is shown to be significantly better than the other, in the way that LOW gain is the preferred setting on the M10.

High Gain would likely be too loud for most rock shows in my experience.  I have always used low gain even when not running a pre in front.

Good to know for when I take this out during marching band season.  Most of my year is choir season though, so I'll be getting a lot of use out of high gain.

[Rally_AK]

Well, I'm sorry to report that the limiters do raise the noise floor about 12dB (according to Sound Forge Pro) It's most noticeable by ear at the High and Hi+Plus gain settings but the noise is still there on the Low setting. Whats weird is that the noise is present even when the gain pots are turned all the way down. This isn't a huge issue for me because I'll rarely use the limiters but it's still bothersome. Can someone else check their deck to see if they hear the same thing?

On the "Write Timeout" front, I'm still getting the errors recording at 24/96 BWF but when I drop down to 24/48 it's rock solid. I suspect that my SD card doesn't have a high enough write speed to record (4) 24/96 files at once. Next step is to see if recording standard WAV at 24/96 has the same issue. If any of you record at 24/96 I'd stay away from the SanDisk Ultra cards for now.

[sperho]

Well, I'm sorry to report that the limiters do raise the noise floor about 12dB (according to Sound Forge Pro) It's most noticeable by ear at the High and Hi+Plus gain settings but the noise is still there on the Low setting. Whats weird is that the noise is present even when the gain pots are turned all the way down. This isn't a huge issue for me because I'll rarely use the limiters but it's still bothersome. Can someone else check their deck to see if they hear the same thing?

Can you describe exactly how you performed your noise floor test?

[F.O.Bean]

Well, I'm sorry to report that the limiters do raise the noise floor about 12dB (according to Sound Forge Pro) It's most noticeable by ear at the High and Hi+Plus gain settings but the noise is still there on the Low setting. Whats weird is that the noise is present even when the gain pots are turned all the way down. This isn't a huge issue for me because I'll rarely use the limiters but it's still bothersome. Can someone else check their deck to see if they hear the same thing?

On the "Write Timeout" front, I'm still getting the errors recording at 24/96 BWF but when I drop down to 24/48 it's rock solid. I suspect that my SD card doesn't have a high enough write speed to record (4) 24/96 files at once. Next step is to see if recording standard WAV at 24/96 has the same issue. If any of you record at 24/96 I'd stay away from the SanDisk Ultra cards for now.

Thanks for checking that out! And that's a bummer about it not being able to handle all four channels at 24/96. I hope its just card specific or the regular wavs work at 24/96, because I planned on running at 24/96 at non-festie shows.I was going to use Sandisk ultra microsdhc cards as well, but in a micro>SD card adapter

[F.O.Bean]

Well, I'm sorry to report that the limiters do raise the noise floor about 12dB (according to Sound Forge Pro) It's most noticeable by ear at the High and Hi+Plus gain settings but the noise is still there on the Low setting. Whats weird is that the noise is present even when the gain pots are turned all the way down. This isn't a huge issue for me because I'll rarely use the limiters but it's still bothersome. Can someone else check their deck to see if they hear the same thing?

On the "Write Timeout" front, I'm still getting the errors recording at 24/96 BWF but when I drop down to 24/48 it's rock solid. I suspect that my SD card doesn't have a high enough write speed to record (4) 24/96 files at once. Next step is to see if recording standard WAV at 24/96 has the same issue. If any of you record at 24/96 I'd stay away from the SanDisk Ultra cards for now.

Forgot to ask what class card you're using? A class 10 card should be able to handle four channels at 24/96

[Rally_AK]

Well, I'm sorry to report that the limiters do raise the noise floor about 12dB (according to Sound Forge Pro) It's most noticeable by ear at the High and Hi+Plus gain settings but the noise is still there on the Low setting. Whats weird is that the noise is present even when the gain pots are turned all the way down. This isn't a huge issue for me because I'll rarely use the limiters but it's still bothersome. Can someone else check their deck to see if they hear the same thing?

Can you describe exactly how you performed your noise floor test?

I loaded two clips in to Sound Forge, one with limiters off and one with limiters on, both recorded with all 4 mic gain pots at 0. Then I did and dB level scan of each file. The clip with limiters off had an Average value of -90dB, the clip with limiters on had an Average value of -78dB.

[Rally_AK]

Well, I'm sorry to report that the limiters do raise the noise floor about 12dB (according to Sound Forge Pro) It's most noticeable by ear at the High and Hi+Plus gain settings but the noise is still there on the Low setting. Whats weird is that the noise is present even when the gain pots are turned all the way down. This isn't a huge issue for me because I'll rarely use the limiters but it's still bothersome. Can someone else check their deck to see if they hear the same thing?

On the "Write Timeout" front, I'm still getting the errors recording at 24/96 BWF but when I drop down to 24/48 it's rock solid. I suspect that my SD card doesn't have a high enough write speed to record (4) 24/96 files at once. Next step is to see if recording standard WAV at 24/96 has the same issue. If any of you record at 24/96 I'd stay away from the SanDisk Ultra cards for now.

Forgot to ask what class card you're using? A class 10 card should be able to handle four channels at 24/96

It's a class 10, SanDisk's newest version of the Ultra. http://www.amazon.com/gp/product/B00M562LOK/ref=oh_aui_detailpage_o01_s00?ie=UTF8&psc=1

I went with SanDisk because they seem to be universally approved by Tascam, at least according to the tested media list for the DR-680.

This is the first time I've gone multi channel to SDXC so maybe I'm doing it wrong. I formatted it directly on the 70D like I've done with all my other cards on the DR-100mkII. Is it necessary or recommended to reformat from exFAT to a different file system like FAT32?

[voltronic]

Speaking of the 680, I found this thread on GS about the same error message:
https://www.gearslutz.com/board/remote-possibilities-acoustic-music-location-recording/558184-dr680-write-timeout-error-massage.html

[voltronic]

Well, I'm sorry to report that the limiters do raise the noise floor about 12dB (according to Sound Forge Pro) It's most noticeable by ear at the High and Hi+Plus gain settings but the noise is still there on the Low setting. Whats weird is that the noise is present even when the gain pots are turned all the way down. This isn't a huge issue for me because I'll rarely use the limiters but it's still bothersome. Can someone else check their deck to see if they hear the same thing?

On the "Write Timeout" front, I'm still getting the errors recording at 24/96 BWF but when I drop down to 24/48 it's rock solid. I suspect that my SD card doesn't have a high enough write speed to record (4) 24/96 files at once. Next step is to see if recording standard WAV at 24/96 has the same issue. If any of you record at 24/96 I'd stay away from the SanDisk Ultra cards for now.

Forgot to ask what class card you're using? A class 10 card should be able to handle four channels at 24/96

It's a class 10, SanDisk's newest version of the Ultra. http://www.amazon.com/gp/product/B00M562LOK/ref=oh_aui_detailpage_o01_s00?ie=UTF8&psc=1

I went with SanDisk because they seem to be universally approved by Tascam, at least according to the tested media list for the DR-680.

This is the first time I've gone multi channel to SDXC so maybe I'm doing it wrong. I formatted it directly on the 70D like I've done with all my other cards on the DR-100mkII. Is it necessary or recommended to reformat from exFAT to a different file system like FAT32?

Many devices like these recorders want cards to be FAT32, but card manufacturers might be using ExFAT out of the box.  Theoretically if you're formatting the card in the recorder than it should be formatting properly to the correct filesystem, but I've sometimes found this to not work if the card is already in the "wrong" filesystem or there were errors on the card - I ran into this issue specifically with high-capacity microSD cards used in my phone.  Sometimes you need to format it to FAT32 first on your computer for it to work properly.  It's not intuitive that step is necessary, but that's been my experience.

The problem is you can't format FAT32 with higher capacities in Windows natively - you need a third-party program.  I've had success with both of these in formatting high-capacity cards to FAT32:
http://www.ridgecrop.demon.co.uk/index.htm?guiformat.htm
https://www.sdcard.org/downloads/formatter_4/

[F.O.Bean]

Yea I agree! Maybe you just need to format the card in your computer so that its a FAT32 format I'm HOPING that's all it is. So, don't use limiters, check. Thankfully I will NEVER do that anyway lol. And always run the mic/line inputs on LOW gain! And format from the computer to get the file form at to FAT32 instead ExFAT!

Id love it if someone would keep testing this card error thing going on, because I was planning on using a 32gb Sandisk Ultra micrdosdhc card in an sd adapter until I can save up and get a nice 32gb Sony Class 10 SD card

[sperho]

I loaded two clips in to Sound Forge, one with limiters off and one with limiters on, both recorded with all 4 mic gain pots at 0. Then I did and dB level scan of each file. The clip with limiters off had an Average value of -90dB, the clip with limiters on had an Average value of -78dB.

An average of a series of low amplitude, zero-crossing data points may simply be reflective of how much DC offset the trace has. The value that is of interest here is RMS level, not average level... This way, DC offset doesn't "contaminate" the calculation.  To get this in SoundForge, select several seconds of waveform and then view the statistics using a menu command (or the statistics button on one of the optional menu bars).

[Ozpeter]

About extra noise when the limiter is used -

Mention of 12dB rang bells here, and the following is a quote from Edirol in relation to the R-44.

" The R-44's limiter is analog and digital mixed like the R-4Pro. When you turn on the limiter, analog gain goes automatically -12dB down for prevention of digital distortion and after that passes through the digital limiter. After that the signal level goes +12dB back up again for matching total signal level."

 It seems highly likely to me that the DR-70D is using a similar scheme and without having thought it fully through, I could imagine that the manipulation of the signal as described could impact signal to noise, by the amount of the manipulation (12dB).

I have commented in the past that this method of limiting is a bit pointless - you might as well record with the level set 12dB below what it might have been, and then apply digital limiting to taste in your DAW during post production.  The end result should be the same, except you have full control of the limiter settings rather than accepting what the device happens to offer.

Some guesswork here, as I am assuming that the Tascam and Edirol devices work similarly, but I bet I'm right.

[flipp]

It's a class 10, SanDisk's newest version of the Ultra. http://www.amazon.com/gp/product/B00M562LOK/ref=oh_aui_detailpage_o01_s00?ie=UTF8&psc=1

I went with SanDisk because they seem to be universally approved by Tascam, at least according to the tested media list for the DR-680.

Be careful of cards bought from non-authorized dealers. Some can be outright fakes, others can be the real card but with a lower capacity than what the label shows. I believe tonedeaf had a recent exprience with the latter. If Amazon sold the card you should be okay (if not Amazon will probably give you a refund), if it was a third-party seller you may have experienced the same thing tonedeaf did.

[Gutbucket]

About extra noise when the limiter is used -

Mention of 12dB rang bells here, and the following is a quote from Edirol in relation to the R-44.

" The R-44's limiter is analog and digital mixed like the R-4Pro. When you turn on the limiter, analog gain goes automatically -12dB down for prevention of digital distortion and after that passes through the digital limiter. After that the signal level goes +12dB back up again for matching total signal level."

 It seems highly likely to me that the DR-70D is using a similar scheme and without having thought it fully through, I could imagine that the manipulation of the signal as described could impact signal to noise, by the amount of the manipulation (12dB).

I have commented in the past that this method of limiting is a bit pointless - you might as well record with the level set 12dB below what it might have been, and then apply digital limiting to taste in your DAW during post production.  The end result should be the same, except you have full control of the limiter settings rather than accepting what the device happens to offer.

Some guesswork here, as I am assuming that the Tascam and Edirol devices work similarly, but I bet I'm right.

I was just about to post that this is most likely what is going on with the limiter function.

Nice recorder.  Saw one first hand a couple weeks ago at Suwannee Springfest.  Very compact.  Controls seem straight forward.

I really like that it has individual channel delay and EQ functions, selectable for playback only if desired I presume.  Those functions were absent on the DR-680 and I don't think they've been added to the new DR-680II, unfortunately.  Many here will never use those funcitons, but they are very important for me.  I'd also really like it to have and a polarity invert switch per channel on both these machines as well. 

Hey Tascam, put those things on your firmware update lists.

[Rally_AK]

Speaking of the 680, I found this thread on GS about the same error message:
https://www.gearslutz.com/board/remote-possibilities-acoustic-music-location-recording/558184-dr680-write-timeout-error-massage.html

Great find voltronic! I have another identical card and a pile of different ones that I'll try out. After reading a couple of the comments, I highly doubt the write speed of my card has anything to do with the "Write Timeout" error. In fact, last night I was able to replicate the error a couple times by clapping in front of my mics and getting all 4 channels to peak at the same time while recording at 24/96.

Many devices like these recorders want cards to be FAT32, but card manufacturers might be using ExFAT out of the box.  Theoretically if you're formatting the card in the recorder than it should be formatting properly to the correct filesystem, but I've sometimes found this to not work if the card is already in the "wrong" filesystem or there were errors on the card - I ran into this issue specifically with high-capacity microSD cards used in my phone.  Sometimes you need to format it to FAT32 first on your computer for it to work properly.  It's not intuitive that step is necessary, but that's been my experience.

The problem is you can't format FAT32 with higher capacities in Windows natively - you need a third-party program.  I've had success with both of these in formatting high-capacity cards to FAT32:
http://www.ridgecrop.demon.co.uk/index.htm?guiformat.htm
https://www.sdcard.org/downloads/formatter_4/

I reformatted the card to FAT32. The 70D couldn't recognize the card after that so I had to roll back to exFAT.

@F.O.Bean - I'll keep playing with this until I figure out what's going on. This is a huge problem because when it happens, the files it's recording disappear... all 4 channels!!!

An average of a series of low amplitude, zero-crossing data points may simply be reflective of how much DC offset the trace has. The value that is of interest here is RMS level, not average level... This way, DC offset doesn't "contaminate" the calculation.  To get this in SoundForge, select several seconds of waveform and then view the statistics using a menu command (or the statistics button on one of the optional menu bars).

Here are a couple screen shots of the statistics with limiters on and off. These were both 15 second clips. I have screenshots of the 2 high gain settings, as well as sound clips if you're interested.

Low Mic Gain - Limiters Off


Low Mic Gain - Limiters On

Be careful of cards bought from non-authorized dealers. Some can be outright fakes, others can be the real card but with a lower capacity than what the label shows. I believe tonedeaf had a recent exprience with the latter. If Amazon sold the card you should be okay (if not Amazon will probably give you a refund), if it was a third-party seller you may have experienced the same thing tonedeaf did.

Both cards I bought are genuine SanDisk, thanks for the warning though. I ran one through a chkdsk like program for mac, called F3, it had full capacity and zero corrupted sectors.

Nice recorder.  Saw one first hand a couple weeks ago at Suwannee Springfest.  Very compact.  Controls seem straight forward.

I really like that it has individual channel delay and EQ functions, selectable for playback only if desired I presume.  Those functions were absent on the DR-680 and I don't think they've been added to the new DR-680II, unfortunately.  Many here will never use those funcitons, but they are very important for me.  I'd also really like it to have and a polarity invert switch per channel on both these machines as well. 

I totally agree, the amount of features Tascam crammed in to this tiny little thing is astounding, especially for under $300. I know it'll be great in the field, just need to figure out what's going on with this "Write Timeout" issue.