Topic: Mysterious pops on a recording ?!?

[temis]

I taped a show a while back and when editing it (downsampling, track-marks etc) I noticed A LOT of little "pops" throughout the recording. The source of the recording is: dpa4061 > mps6010 > pcm-m1 (no tape, used just as an A/D) > 7pin to toslink mini-plug > njb3 (optical in).

The pops are all over the place, they seem to be quite random and maybe at about every 30 seconds (not on regular intervals, though). The usual pattern is two "zero"-samples on the left and two "zero"-samples on the right with usually a preceeding sample that is "higher" or "lower" than "should" be (see picture 2). There are also some worse examples (see picture 3) but not nearly as many as the simple two-zeroed-sample-ones.

These are of course quite easy to fix on the "sample adjustment"-tool on Audition, for example but naturally I would like to avoid this in the future. It IS a pain in the ass to fix these pops as they go into hundreds!

Any ideas what could cause this? I doubt that the A/D on pcm-m1 is faulty, also the njb3 is not known for dropping samples. That leaves only the 7pin cable but can a loose connection or something cause "missing" samples? It seems weird to me that there's (usually!) only TWO (or three) samples missing...

I have quite a few other recordings made with the same setup - I will check them next to see if the same problem appears but any ideas and/or questions are very welcome!

[spreadheadtom]

are they evident on the master recording? 

[temis]

are they evident on the master recording? 

Yes, well, it was only recorded on the njb3 - so the file I listen to and you see in the pictures is the master...

[spreadheadtom]

if you listen to the JB3 with headphones........are the pops there.

i ask this because I had a similar situation occur.  only evident on my laptop after transfer.  ended up having to defragment the laptop to fix the problem.

[temis]

if you listen to the JB3 with headphones........are the pops there.

i ask this because I had a similar situation occur.  only evident on my laptop after transfer.  ended up having to defragment the laptop to fix the problem.

Hmm, interesting... I don't have the recording on the njb anymore, so I can't check.

The faulty samples were zeros, so the pops are not very audible and I only listened to the recording with the njb3 on the move and not in a quiet situation... I didn't notice anything then but this might be something to consider - thanks!

[cleantone]

Well the problem is misplaced samples. The cause it harder to pinpoint. The solution is to find ALL of them and fix them. I have never used audition. Some software has a feature to fix that. If it does, you would select withing a few samples of the flaw, and process. The hard way is to redraw the sample with an editing tool, very tedious. Again I have never used that software but the images I see do NOT look like what I've seen in other software. Maybe it is stray voltage getting into your chain. Is it fizzy or clicky sounding?

Good luck...

[keepongoin]

tommy's right - i had two recordings in June have issues like this, that were generated in the transfer from JB3 > desktop... a simple re-transfer, and they were fine.  I did both transfers with firewire... don't know what caused the problem.

i always save my JB3 recordings on there until (1) i am happy with how they look as waves on my desktop, and (2) someone else has copies of the tracked, finished show in FLAC format.

[temis]

tommy's right - i had two recordings in June have issues like this, that were generated in the transfer from JB3 > desktop... a simple re-transfer, and they were fine.  I did both transfers with firewire... don't know what caused the problem.

i always save my JB3 recordings on there until (1) i am happy with how they look as waves on my desktop, and (2) someone else has copies of the tracked, finished show in FLAC format.

Thanks! That's very good to know! Were yours just like this or similar? (Just trying to verify that this indeed is the real reason... BTW, I used a firewire as well...)

I've gotten the habit of removing recordings from the njb right after the transfer to make the space available for other recordings. Maybe I'll have to check the transfers more carefully in the future... On the other hand, I would expect the transfer to have some "error correction" embedded to avoid this...

If the real reason is a fragmented HD would that imply that the firewire/usb is feeding the information too fast for the HD to save it properly - shouldn't there be a mechanism to hold the transfer if it cannot be saved to the HD? Actually, now that I think about it - one way to verify would be to transfer the files twice (no big deal with the firewire) and create md5s for both and check that they match...

Thanks again!

[Swampy]

Hmm I got little clicks on my MULE show (which sounds freakin awsome) after I dropped my JB3... Go figure I wonder if its the same thing???

[temis]

Well the problem is misplaced samples. The cause it harder to pinpoint. The solution is to find ALL of them and fix them. I have never used audition. Some software has a feature to fix that. If it does, you would select withing a few samples of the flaw, and process. The hard way is to redraw the sample with an editing tool, very tedious. Again I have never used that software but the images I see do NOT look like what I've seen in other software.

Yes, the Adobe Audition (formerly CoolEdit Pro) has a Click/Pop Eliminator option "Fill Single Click Now" for the task but I found that a bit "harsh" at times. It is of course easier than adjusting the samples one by one by hand but think I get more "accurate" results by doing it that way.

Maybe it is stray voltage getting into your chain. Is it fizzy or clicky sounding?

Good luck...

I am quite sure it's in the digital domain but I'm not really familiar with stray voltage and don't know if it can affect digital signals as well. The "pops" sound like clicks but it's since they are zeros it's quite hard to spot them especially on the quiteter parts. They are very easily spotted with the "spectral" view, however...

[hexyjones]

I wonder if the problem isn't having the DAT in Record mode....?

Do you have to have it in Pause/Record? or can you just put it on "Rec" and have it run the A/D?

Not sure if I would call it faulty...but it may not work as you might hope...(ie a standalone AD) You may need to be running tape for it to work properly...

Wasn't there some way to generate clicks when digi-chaining Sony D8s together...? Been awhile so I forget...?

[F.O.Bean]

bet its the sony optical digital cable going from the m1>jb3

[temis]

I wonder if the problem isn't having the DAT in Record mode....?

I don't think that's a problem - the M1 is in the "record" mode.

Do you have to have it in Pause/Record? or can you just put it on "Rec" and have it run the A/D?

No, you actually cannot leave it on "pause-record"-state (with a tape inside) because the deck stops and unloads after a few minutes. Without tape you can just push REC and it stays on the "rec" mode which can be used as an A/D...

Not sure if I would call it faulty...but it may not work as you might hope...(ie a standalone AD) You may need to be running tape for it to work properly...

I do believe it can be used as a standalone A/D, I think it was on the manual even. I'll have to check... Anyway, I've used it for quite a while without problems.

Wasn't there some way to generate clicks when digi-chaining Sony D8s together...? Been awhile so I forget...?

Maybe, in this case I really wouldn't call them "clicks" though. Usually "clicks" are MUCH more than a couple of missing samples...

[temis]

bet its the sony optical digital cable going from the m1>jb3

Yep, that seems to be it.

BTW, I did loads of testing trying to get somehow corrupted transfers from NJB3 to my HD but always ended up with matching md5's. The transfer definitely stalls and takes longer when it can't write to the HD so I doubt the problems came from there.

Wiggling the 7pin cable causes increase in the dropped samples.. I decided to ditch the 7pin altogether and modify my dat decks to have toslink outputs on the deck instead of the 7pin. The D7 is done already - I will post pictures soon. M1 is a bit trickier because of the 3V operating voltage. (...that why I posted http://www.taperssection.com/yabbse/index.php?topic=30773.0)

Thanks again - I'll do some more testing (also on the field) and let you know how it goes...