1. Register NOW and become part of this fantastic knowledge base forum! This message will go away once you have registered.

Feeding a test signal into multiple input channels

Discussion in 'Recording' started by djmukilteo, Nov 22, 2009.

  1. djmukilteo

    djmukilteo Well-Known Member

    Wondering if someone has a fast easy way to get a test tone into 26 channels at the same time to test for recording dropouts and performance.

    I have a ZED16R with 16 channels of Adat running into a FF800 with the remaining 10 inputs. I have a sine wave generator for the test tone.

    I'm sure I could make a 26 to one breakout cable, but that seems like alot of work for just a test...
    Is there some way using TotalMix in the FF800 that you could route one of the ZED Adat channels to all the other 25?
     
  2. Boswell

    Boswell Moderator Distinguished Member

    Yes, TotalMix can do that. Just check (green) all the output intersects along the row corresponding to the input that has the tone generator connected.

    However, that will only check the section of your system from the FF800 mix matrix through the FireWire interface and on to the PC disk, but maybe that's all that you need. While recording all 26 channels, you could complement the first test with a single wandering TRS plug from the generator into each R16 input and FF800 input in turn. When it comes to loading the system, there is no distinction between signal and no signal, and it will enable you to check that you get the same signal amplitudes on all channels.
     
  3. djmukilteo

    djmukilteo Well-Known Member

    Thanks Boswell,
    So if I inject my test signal into say Analog 1 (left side row???) of the FF800 then go into the TotalMix matrix view and click each intersecting output box (bottom row??) then they will all have the test tone available in Cubase so I can record a 15 minute test.
    Does that end up looking like a diagonal when your done?
    What I'm trying to test for is dropouts at 44.1khz and 48khz while recording 26 tracks in Cubase simultaneously and I would like to see if my computer can handle a lengthy (10-15min) session without crapin out...I was going to have the CPU Performance Meter up and running while in record.
    Moving the signal around to the different inputs makes sense too I will try that as well....
     
  4. djmukilteo

    djmukilteo Well-Known Member

    OK well the first test worked fine...I did 28 channels and recorded around 5 minutes worth and saw no dropouts. Scaled out the tracks all the way out to the sine wave and had a nice digitized staircase sine wave on all tracks....no clip, nice and repetitive.....everything looked good.....but

    I don't think this was exactly the test I wanted to do.
    What I think I want to test is both interfaces together using all 28 of their A/D converters at the same time and record that for awhile and see if anything goes sideways.

    I think what I did in this first test was just take the input from the one channel that had the analog test tone and used it as the input to each track in Cubase and basically recorded a bused or duplicated signal to all 28 tracks.
    So what I tested was Cubase (and to some extent my computer) and found out that my system is capable of recording that successfully but not the actual interfaces???
    Well except for the one input that actually had an analog signal injected.... Does that sound correct?
    So now I'm still trying to figure out what the easiest way would be to hook up the test tone into all the analog inputs on both interfaces and then setup the 28 tracks again in Cubase and record each input on a separate track.
    That should test all A/D converters at the same time....right?
    Maybe the big Y cable is the only way to do this.....I just thought there was a clever way to accomplish that I hadn't thought of..
    FWIW
    I'm not completely sure why I'm doing this....it's just something another forum person said was not possible and that his RME FF800 was flawed and that the FF800 can't handle that sort of input channel count without having clicks or dropouts... maybe there are some persons on here that have an opinion on that claim or have done this test?
    I've never had any dropouts or clicks with my RME FF800 by itself and now with the ZED16R routed in with it, I still haven't had one problem on any of the channels....they all work great....
    But....I've never tracked that many tracks simultaneously, so I thought I would try this experiment....any suggestions or comments?
     
  5. djmukilteo

    djmukilteo Well-Known Member

    Well at least I have a 28 track 5 minute Cubase test file
    now.
    Looks pretty LOL
     
  6. TheJackAttack

    TheJackAttack Distinguished Member

    The Fireface won't have any problem handling the load. The computer on the other end might be another story. You can actually chain two FF800's together via 1394b and get double the track count-again providing your computer is up to the task.
     
  7. Boswell

    Boswell Moderator Distinguished Member

    The first test is exactly what I said earlier - a test that your FireWire interface-PC-DAW-disk combination is up to the job. That was the crucial one.

    Now do the second test of inputting the tone one channel at a time into all 16+10 inputs. It does not matter that you do not have the same signal in all channels - all the ADCs are converting all the time and as long as each channel comes through cleanly without dropouts or breaking through into another channel, you are good to go.

    The FF800 will cope with all 26 channels on a FireWire 400 connection. As John said, a pair of FF800s daisy-chained can double that track count without problems if you use a FireWire 800 connection.

    By the way, you can't run audio from the Zed-R16 via FireWire at the same time as the FireFace 800, if that's the way you were thinking of getting 28 channels. You had it correct in taking the ADAT outputs of the R16 and connecting them to the ADAT inputs of the FF800, so the FF800 is the only FireWire device transferring audio to the computer. You can continue to use the R16's FireWire interface for patching and control. If you need the R16's mix output as a pair of your recorded tracks, these are not available on ADAT, so you need to take them via analog cables to two of the line inputs of the FF800.
     
  8. djmukilteo

    djmukilteo Well-Known Member

    Ya that was my opinion as well. I have been watching this thread on another forum (GS)....sorry I'm weak...."All RME Fireface 800 are defective" and this guy is getting random clicks on a FF800/OctamicD using ADAT from the Octamic D into the FF800 and claims all FF800 are flawed in this regard and that they lose sync and have dropouts during a 10 minute recording or longer. I'm basically calling bulls#*&hit on that claim and think he's either doing this test wrong, has a PC problem or something is completely wrong with his approach and what he's trying to do..I wanted to test my 16 channels of ADAT from the ZED into my FF800 (which is my current setup and works perfectly for me) and see if I could duplicate this and see a similar anomaly.
    He has a Youtube video where he runs his tests and shows the dropouts..
    http://www.youtube.com/watch?v=iLiAW3gS67M
    It pretty entertaining in itself...
    So...in order for me to run a test like that I figured I would have to inject my function generator at 1khz into each ZED input and each FF800 input simultaneously and then setup Cubase with 26 tracks and run the recording.....should be easy to spot a dropout in Cubase considering you can zoom all the way into the sine wave easily....
    Don't know why I feel the need to prove this, my setup works fine....
    But I think there is something fishy with the whole claim?!!
    He's had it repaired and returned to RME so I kinda wonder what it is he's seeing?
     
  9. djmukilteo

    djmukilteo Well-Known Member

    I agree the first test was a crucial test and interesting at the same time.
    I can move the test signal around like you mentioned....I'm still not convinced I'm exercising all channels of the A/D converters simultaneously that way?.....if there is no analog signal inserted at each input how would there be any conversion?

    I am running the ZED ADAT1 and ADAT2 into the FF800 @44.1 FW400 and that seems to work great....There are things like the MIDI control on the ZED that haven't been fully tested and flushed out yet...but I need to re-evaluate my MIDI setup with the keyboards and conflicts with my Kore2 controller as well.
    I'm really using the ZED as a summing mixer and have the main stereo outputs on the ZED feeding back to two inputs on the FF800....
    What I like about my setup so far is I can record on any input on the FF800 or ZED into Cubase and then after recording route any of those tracks back into the ZED (up to 16 tracks) and mix either pre or post EQ which is very nice....it feels like a a real console....and I don't have to touch the levels at all in Cubase once levels are set.
    Anyway so far so good and I am extremely happy with the preamps on the ZED for all my equipment....really is a nice mixer with a smooth rich sound....hard to tell the difference between it and the FF800 they are so close....I might try putting my O-scope on some samples and see if I see any differences.....
     
  10. TheJackAttack

    TheJackAttack Distinguished Member

    I have not experienced the issues your buddy at GS and the two others at the end describe. I'm like the 99% of respondents that never had an issue and I have two FF800 units.

    I will say that the idea a single Octamic is going to sync two pipes worth of digital data to the same interface (FF or otherwise) is a little skewed. It would be a much better test for him if he used two Octamics or other 8 channel AD like the Onyx 800R. Then the FF can provide master clock and actually lock both light pipes rather than locking only one ADAT channel and the other one going squirrely.

    Just my opinions of course.
     
  11. djmukilteo

    djmukilteo Well-Known Member

    Exactly....My FF800 has worked flawlessly and I am also one of the 99% that has had zero issues with RME.
    What was interesting was this person "buddy" (I deserved that JA...good one!) has been trying to resolve this "quirk" he has found in this testing for over a year now and still hasn't found the root cause of the problem!
    WOW! Talk about a dog with a bone!
    When I came across this thread I thought....this should a couple easy posts, eliminate a few variables..figure out what his problem was!!? Rule out a few things....you know help out a fellow RME user.
    And this was why I found his claim hard to believe...and why I wanted to do a more accurate and thorough test on my own...I thought it might also provide me with some idea of how far I can push my own system with the 26 channels I now have.
    Now I could be wrong and he may have a serious problem...I don't know.
    I do know an OctamicD only has one ADAT output 1-8 and no ADAT inputs so the premise and hookup in his video is flawed from the get go.
    Too many other weird connections going on there too!
    And don't even get me started on using Digicheck as your reference multitrack recorder....your just asking for glitches from that! and I'm not clear what that is testing anyway!

    Anyhoo
    I know I can make up a jumper cable assembly and plug it into all 26 channels, but before I go and dig up 26 TS jacks from my junk box and commence to soldering...I thought someone would have a easier suggestion...
     
  12. Boswell

    Boswell Moderator Distinguished Member

    There is conversion all the time the device is switched on. Whether you plug a signal in or not merely changes it to converting your test tone instead of converting the noise on the input. If a single channel converts correctly when plugged in turn round all the inputs, then they would all convert correctly if the same tone were plugged into all channels at once. The sync is not data dependent, at least, not at the level you are investigating here.

    The GS thread was more interesting for it being civil than for the technical content.
     
  13. djmukilteo

    djmukilteo Well-Known Member

    Yes that thread was pretty civil..
    I really had no dog in that fight, I was just trying to help the poor guy out....I feel sorry for someone who continues to beat a dead horse and still find no concrete solution!!!

    One question I have with MY test....
    In TotalMix matrix view, I created a "horizontal" line across the matrix at the input I had the test signal at which was In13 (ADAT1-1), "channel1 on the ZED". When I did that the signal appeared nicely at all the other channel meters on the bottom row of the mixer view.
    When I went into Cubase and created 26 audio tracks it would only see signal on the In13 channel...none of the other channels had signal.
    I was expecting to see the test tone on each of the input channels??....
    So maybe my test was screwed up because I was merely recording the same input signal on 26 Cubase audio tracks...
    Maybe I had the TotalMix setup wrong??
    Seems like that still only tests Cubase and one channel of analog input? and not the conversion of 26 analog signals coming from the FF800...

    Does that make sense..or am I still missing something??
     
  14. TheJackAttack

    TheJackAttack Distinguished Member

    And his bone is his obsession with using only one ADAT device. If he added a second Octamic his problems would disappear.

    As to your test tone not working like you thought I'll give it a look this weekend if I get a minute. My daughter is only 11 weeks old so I'm too tired to envision the scenario-I'll have to physically run the test on Tracktion and Audition. Cucco could probably answer without even thinking about it as he has many more years of use than I do with the FF.
     
  15. djmukilteo

    djmukilteo Well-Known Member

    Well congrats on your new daughter!
    I'm not in any hurry or really have anything to prove by this...after trollin that thread I thought I'd give it a try as an experiment but I'm perfectly happy with the way my setup runs. Just curious how it might perform....
    To me it's real simple put a signal into all your inputs and run a record. it will either crap out or it won't. I just don't have a simple way to input 26 channels...
    FWIW...The GS guy was using the "Global Record" tool in RME's Digicheck software which interleaves channels as one file rather than streaming separate audio channels like most DAW software does....
    Now RME does state that Global Record can record 200+ tracks simultaneously for hours on end and I think this was what he was looking into...
    It was never that clear to me what he thought he was really testing in practical terms in his video's...He was doing stuff that didn't make any sense to begin with...It might be he's trying to prove RME's software can't do what they claim it can do without dropouts....I will continue to monitor his antics and breakout the solder station....LOL
    Have a great Thanksgiving!
     

Share This Page