Behringer ADA8000 pre bypass questions

Advertise with us
shop deals on amazon
Shop deals on ebay
GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Here are a couple more listening samples if anyone is interested.

The first one is the source file, and the second one has been passed through one of the stock channels set on its internal clock.

http://www.musicians-samples.com/Original - Palo Alto.wav
http://www.musicians-samples.com/Stock x 1 internal clock.wav

So after MANY different listening test, I've decided that these converters sound just fine without any mods...

The ONE things, however, that really bugs me about the stock channels is the apparent high-end boost I hear after running something through them... it's subtle... but its there and I find it a bit abrasive. (Flip the phase on the source file and line it up the stock x 1 file to hear the extra high frequencies).

Question:

1. How could I smooth the high end on these converters? Does the small 3n3 cap (actually a 1n8 cap on mine) right before the a/d inputs have anything to do with this?

Any comments would be greatly appreciated!
 
Is it really a high end boost or is it really just distortion and the harmonics created by it? Maybe some nicer coupling caps would help out? I'm just guessing...

Dude just buy the Aurora unit.

Brad
 
Here is a quote from the datasheet of the A/D converter:

"The digital section of the AL1101 compensates for the passband amplitude deviation of an external single-pole 80kHz anti-alias filter (@ Fs=48k, scaling with Fs). To remove highfrequency noise at the differential inputs, the capacitor between the differential inputs should be located as close as possible to the input pins."

I interpret this as meaning that the 4700pf capacitor shown in the schematic in the datasheet is part of an analog anti-aliasing filter whose frequency is set at 48KHz. Your capacitor being set to 1800pf would put the pole of the filter almost 3 times too high. High frequency content that is aliased will appear as noise with the frequency distribution being inversely distributed starting at the sample frequecy and going down from there.
 
Is it really a high end boost or is it really just distortion and the harmonics created by it? Maybe some nicer coupling caps would help out? I'm just guessing...

Dude just buy the Aurora unit.

Brad
Click to expand...

Haha.. Thanks Brad...but I'm on a mission to make these things *good* :green:

The high end boost is actually a part of the frequency balance... up about .5db at 20khz... I found this in the other thread

http://rockthemountain.com/~watkins/ada8k-da7-lavry-Comparison.htm

Here is a quote from the datasheet of the A/D converter:

"The digital section of the AL1101 compensates for the passband amplitude deviation of an external single-pole 80kHz anti-alias filter (@ Fs=48k, scaling with Fs). To remove highfrequency noise at the differential inputs, the capacitor between the differential inputs should be located as close as possible to the input pins."

I interpret this as meaning that the 4700pf capacitor shown in the schematic in the datasheet is part of an analog anti-aliasing filter whose frequency is set at 48KHz. Your capacitor being set to 1800pf would put the pole of the filter almost 3 times too high. High frequency content that is aliased will appear as noise with the frequency distribution being inversely distributed starting at the sample frequecy and going down from there.
Click to expand...

Burdij,

Thank you very much for confirming what I suspected. :guinness: Over the weekend I replaced *all* 8 of these caps w/ 4700pf wimas. So far I'm really liking this for the top end... I'll try to post some samples. I have one 8 chan unit that is now modded and one completely unmodded unit... so I'll try to do a side by side comparison.

- Damon
 
[quote author="burdij"]"The digital section of the AL1101 compensates for the passband amplitude deviation of an external single-pole 80kHz anti-alias filter (@ Fs=48k, scaling with Fs). ... I interpret this as meaning that the 4700pf capacitor shown in the schematic in the datasheet is part of an analog anti-aliasing filter whose frequency is set at 48KHz. [/quote]

:? :roll:
Sure? I think the text (clearly) says (analogue) 80kHz filter (6dB/oct) when sampling at 48k. This would be typical for an oversampling converter (most current products are) and gives a gentle (analouge) filter slope, with good phase carateristics. After conversion, a digital filter kicks in, with a steep cutoff 'round 21k - however this can be modeled as a FIR filter, and optimal phase response can be designed in.
I don't know these converters though, this is just general stuff.
 
?
:?
I'm sure that my ADA80000 delivered healthy level when at 0dbFS
compared to the other interfaces ... better than other interfaces

and I don't remember there being any issues into 600 ohms
 
Odb full scale shows the max. level before (hopefully!) clipping, this has nothing to do with what I am talking about. What is at issue is the analog output level when a reference level digital input at 1K (-18 digital scale, as used in our studios) is sent to the unit. That should be a +4 signal, 0Db on a VU meter, or 1.23 volts AC. Many interfaces have adjustable in / out level controls for calibrating levels, as there are several accepted standards (-16, -18-, & -20 are the most common) for what digital level = 0Db @+4 analog.
I am surprised that no one else has encountered this, has anyone else tried to use one of these to add more outputs to a Pro Tools HD system (or the like), or are most using it as their only interface?
I will be back at the studio on Monday & I will measure the output then, & post it.
 
I think +16dB means that the maximum output voltage that will be generated when the peak of the sine wave is at the maximum converter full scale would be 4.90 Volts RMS, assuming a reference of 0dBm at .776 V RMS (1 milliwatt, 600 Ohms). A +4 dBu signal is a reference of 1.23 V RMS, unterminated.

Your studio, correctly, assumes a reference for 0dB at -18 dBu which is the international (EBU) standard.
 
+16 Db analog out at 0Db full scale digital is quite low, as well. +20 to +28 is more what one expects to see on Pro Audio gear, this is telling of the problem. I am thinking the output gain is set by the feedback resistors in the drive circuit, but as they are most likely SMT, not easy to change.....
 
+20 to +28 is a lot of level for your analogue gear, if you follow the common trend of pushing everything close to digital zero. That standard was set keeping in mind 18-20dB headroom at all times. But real-life recording does not work like that.

We have ended up setting our studio levels more conservatively around here - at 0dBfs=+12dBu - and that gives significantly better subjective results with most our analogue outboard gear.

+28dBu is around 30V pp - and even though most modern gear will accept and output such levels, this is not the same as it does so in an elegant way.

Jakob E.
 
[edit] sorry ... looks like Jakob got in first

there is a lot of stuff out there that is not at 28dBu and 30+ any more

this is why I've been banging on about the DIY passive adapter box to bring the OLD PRO gear down from 28 and 30 +
down to the piss poor NEW PRO levels of 22 and sub 20dBu

The New HD interfaces don't have the grunt the old 888/24 had

I do use Ai3 and ADA8000 with the new interfaces
and I used them with the old interfaces and the 001

The Ai3 is even worse ... hence I use it for a Keyboard sound module input AND ... cheap FX returns

AND
many audio based systems use -12 and -14dBFS as a zero VU line up and not the Sony / Broadcast styled -18 and -20dBFS

IF you do a search here you will find many threads where many of us have complained about the subject
...
but to be honest we have moved on and just developed solutions to suit our individual needs

surface mount and RoHS and piss poor levels and low current ability is here to stay
...so develop your work arounds and stock up of old school gear and parts as fast as you can
 
Talk with most any mastering engineer, and they will plead with the engineer to leave some headroom......one should only be getting close to 0db digital full scale on transient peaks. A 24 bit 96k system no longer requires the effort to capture maximum bits, and any digital mix buss is going into dither so quickly that the common overly hot levels just make things worse. There is no way to accomodate bad engineering, so why deviate from the accepted professional norms that allow for interchange between studios? Another case of the lowest common denominator ruling over all?
When ever I have had to accomodate analog gear that could not tolerate hot digital gear outputs (SSL consoles, anyone?), I went for higher levels, +20 digital or more...going to lower levels just makes it worse.
 
12dBs of headroom is not enought... I think 16dBs would be minimun, and depending of the program material 18-22dBs of headroom will be better.
I have my studio adjusted that 0dBVU = -20dBFS = +24dBm
When tracking, I keep peaks around -10 to -12dBFS and during mixing at -8 to -6dBFS, so no problem if I use modern gear, I don`t need to push gear harder and record hotter signals, with 24bit the digital noise will be way behind the gear noise or room noise.
When mastering I uswe gear that can handle +24dBms without problems, but also take in account that I don´t need to stay with those levels from the begining, usually I use near +24dBm only as the last step just before the AD, and in the worst case, if I´m forced to use a limiter that can´t handle such hi levels, I can readjust the AD input levels in 5 minutes...

BUT, here we are speaking of the B unit, not about a high-end converter, nor PRO gear... more a kind a prosumer gear that sound decent... What do you want for 200 bucks? +22dB output and trimmers in each channel?

Also, is not a good idea using different brands of converters at the same time for the same aplication: for example using the ada8000 for monitoring and the PT converter for sending channels to a summing box is ok, but using a 8 channel digi converter (or whaterver converter you want) and a ada8k (or whatever converter you want) for sending 16 outs to a summing mixer will be problematic, not because in this specified case the levels dont mach (you can easily adjust the PT converter levels to match the Beh outs), but because each brand and model of converter will have a different latency depending of the converter IC used, etc... and you can´t get accurate up to sample sync between them.

So better get another PT , or another Beh... hmmm better get 16 channels from mytek! :thumb:

Best regards,

Synthi
 

Latest posts

Back
Top