Author Topic: PCM-M10 Glitch--and of course it was during Radiohead.... (Read 9893 times)

earmonger · « **on:** June 17, 2012, 05:10:03 PM »

I'm sure the sound of the glitch in the attached snippet is familiar to someone here: Clicks that are not just noise on top of the recording, but actually skips in the recording, so there are parts of the music that just aren't there.

It was PCM-M10, 24/96, onto the external memory, a SanDisk Class 2 8GB without previous problems. And it started pretty close to the end of the set so only the last 4 or 5 songs were glitched out.

My guess is that the Class 2 card wasn't fast enough to handle all the information being written in what was turning into a big file. I wonder if it could have been avoided by starting a new file midway through.

It was Bonnaroo though. Somehow I think someone else got a good pull.

Any thoughts, folks?

vanark · « **Reply #1 on:** June 17, 2012, 05:31:24 PM »

24/96 on a Class 2? Not sure I would try that.

earmonger · « **Reply #2 on:** June 17, 2012, 05:40:18 PM »

Actually, it was 24/44.1 . I only record 24/96 if dogs will be listening.

newplanet7 · « **Reply #3 on:** June 18, 2012, 06:35:28 PM »

Has the light gone out for you? Cause the light's gone out for me.

Write speed error IMO.

Fatah Ruark (aka MIKE B) · « **Reply #4 on:** June 18, 2012, 08:02:47 PM »

I ALWAYS record to the internal memory first. I have found that it is much more reliable.

Rarely I go to a show long enough that it spills over to the external card...and even then I made sure I got a card that will record at 24/96 (and I usually only record at 24/48 to be on the safe side).

F.O.Bean · « **Reply #5 on:** June 19, 2012, 01:26:48 AM »

I would bet the class 2 card was/is the culprit

adrianf74 · « **Reply #6 on:** June 19, 2012, 10:11:03 AM »

Class 2 is a BIG NO-NO. Is it a quality brand name or some off-brand crap?

I would stick with a class 6, if not 10, card by Transcend, Kingston, Patriot, SanDisk, etc., otherwise you're running the risk of what happened here.

Todd R · « **Reply #7 on:** June 19, 2012, 12:05:43 PM »

It does sound to me that it was probably the card that was the culprit. But I wouldn't say Class 2 is a big no-no. A true Class 2 card (2 MB/s write rate) should be fine for 24/96 sampling, which is less than 1MB/s for 2ch.

I use a Class 2 Sandisk microSD card in my M10 and it works fine. I also have found that Sandisk cards (if you get real ones and not bunk ones on ebay) are what they say they are. I also have found Kingston to be good. But I've had Transcend and A-Data cards that were supposedly Class 6 or Class 10, but operated far below Class 2 standards.

I'd spend the money on a good Class 2 card before I got a shoddy A-data or Transcend Class 6 or Class 10. ymmv of course, but that has been my experience.

adrianf74 · « **Reply #8 on:** June 19, 2012, 12:46:25 PM »

Quote from: Todd R on June 19, 2012, 12:05:43 PM

It does sound to me that it was probably the card that was the culprit. But I wouldn't say Class 2 is a big no-no. A true Class 2 card (2 MB/s write rate) should be fine for 24/96 sampling, which is less than 1MB/s for 2ch.

I use a Class 2 Sandisk microSD card in my M10 and it works fine. I also have found that Sandisk cards (if you get real ones and not bunk ones on ebay) are what they say they are. I also have found Kingston to be good. But I've had Transcend and A-Data cards that were supposedly Class 6 or Class 10, but operated far below Class 2 standards.

I'd spend the money on a good Class 2 card before I got a shoddy A-data or Transcend Class 6 or Class 10. ymmv of course, but that has been my experience.

One brand I would endorse is Lexar. I've got a Lexar Class 10 32GB SDHC card in my M10 and it has *NEVER* let me down. And it's super fast, too.

hi and lo · « **Reply #9 on:** June 19, 2012, 12:47:44 PM »

Quote from: Todd R on June 19, 2012, 12:05:43 PM

...I wouldn't say Class 2 is a big no-no. A true Class 2 card (2 MB/s write rate) should be fine for 24/96 sampling, which is less than 1MB/s for 2ch.

I'd spend the money on a good Class 2 card before I got a shoddy A-data or Transcend Class 6 or Class 10.

Facts... I sure do hate it when they get in the way.

Brand is far more important than the class rating and I don't recommend any brand other than SanDisk. Your problem may or may not have been the memory card, but as suggested its the first place to troubleshoot.

kleiner Rainer · « **Reply #10 on:** June 19, 2012, 05:52:45 PM »

earmonger,

get this program:

http://www.heise.de/download/h2testw.html

It will tell you the sustained write rate (and read rate) of your Micro SD card and if there are any bit errors.

I use it to test all new cards. Highly recommended! But remove all data you need from the card, it will overwrite everything on it.

I do not have any commercial connection with the publisher except for buying their mag for a quarter century

Greetings,

Rainer

earmonger · « **Reply #11 on:** June 19, 2012, 10:54:11 PM »

Thanks, Rainer, but I always use h2testw and this card--a SanDisk Class 2 (see original post) from Amazon--tested out fine. As noted, I was only going 24/44.1. Still, it was a long set and I think the file just got too big.

If I try this experiment again I'll break the set in half. But maybe I'll just spring for a class 4.

Fatah Ruark (aka MIKE B) · « **Reply #12 on:** June 20, 2012, 01:14:17 AM »

The size of the file shouldn't be an issue. Just recorded a 24/96 file that went a hair over the 2GB limit...no issues at all. Do it all the time with no problems.

ScoobieKW · « **Reply #13 on:** June 20, 2012, 03:36:33 AM »

Quote from: earmonger on June 19, 2012, 10:54:11 PM

Thanks, Reiner, but I always use h2testw and this card--a SanDisk Class 2 (see original post) from Amazon--tested out fine. As noted, I was only going 24/44.1. Still, it was a long set and I think the file just got too big.

If I try this experiment again I'll break the set in half. But maybe I'll just spring for a class 4.

By design the M10 splits files at 2GB due to older computer file systems having a 2GB file size limit.
The file split is seamless. No samples are lost.

Cross memory recording, when turned on in the M10 menu, is also seamless.

I use Sandisk, but I need new glasses and can't read the speed.

Sebastian · « **Reply #14 on:** June 20, 2012, 04:17:59 AM »

Quote from: ScoobieKW on June 20, 2012, 03:36:33 AM

By design the M10 splits files at 2GB due to older computer file systems having a 2GB file size limit.

This is actually a limitation of the WAVE file format, not the file system.

robeti · « **Reply #15 on:** June 20, 2012, 12:02:01 PM »

I always use class 10 or class 16. I think the problem is the speed of the card.

aaronji · « **Reply #16 on:** June 20, 2012, 12:08:10 PM »

A stereo recording at 24/96 produces ~ 0.55 MB/s. As Todd noted, a class 2 card should be good for 2 MB/s...

ScoobieKW · « **Reply #17 on:** June 20, 2012, 02:13:56 PM »

Quote from: Sebastian on June 20, 2012, 04:17:59 AM

Quote from: ScoobieKW on June 20, 2012, 03:36:33 AM
By design the M10 splits files at 2GB due to older computer file systems having a 2GB file size limit.

This is actually a limitation of the WAVE file format, not the file system.

FAT32 had 4GB limit. Wikipedia quotes 4GB as a limit to the WAV format as well due to using 32bit unsigned intergers. (http://en.wikipedia.org/wiki/WAV)
It goes on to say that some software had a 2GB limit due to implementing the file length with a signed integer, IIRC Soundforge was one of them. (more info at http://www.hydrogenaudio.org/forums/index.php?showtopic=32422)

Real world? 2GB is the least common denominator. It works well everywhere. With seamless splits between files, it's not a big deal.

Ekib · « **Reply #18 on:** June 20, 2012, 10:15:02 PM »

What mic's are you using ?

Someone on the PCM thread mentioned the card needs to be formatted everytime you use it . I don't know if you did that , but I did find it very useful to know.

earmonger · « **Reply #19 on:** June 21, 2012, 03:21:31 PM »

Those were Church Audio CA-14 omnis.

I don't think the card needs to be reformatted for each use. There are times when I'm too busy to upload, and the card can store a lot of shows.

The card does need to be formatted by the PCM-M10 before you start recording on it--but just once, not every time.

Ekib · « **Reply #20 on:** June 21, 2012, 11:00:50 PM »

Quote from: earmonger on June 21, 2012, 03:21:31 PM

Those were Church Audio CA-14 omnis.

I don't think the card needs to be reformatted for each use. There are times when I'm too busy to upload, and the card can store a lot of shows.

The card does need to be formatted by the PCM-M10 before you start recording on it--but just once, not every time.

Why is it you choose Omnis if I my ask ?

I format my MicroSd everytime I make a new recording . I don't think it can hurt and uploading only takes 5- 10 minutes . I use a lexar 16 GB class 10 card. So far every recording came out flawless.

vanark · « **Reply #21 on:** June 21, 2012, 11:13:18 PM »

I reformat my card in the M-10 after removing the recordings. I find it simply easier and simpler than deleting them.

I've used class 4 to class 10. I'd never consider using a class 2, regardless of brand.

earmonger · « **Reply #22 on:** June 22, 2012, 05:01:36 PM »

Well, 40+ hours of Bonnaroo recordings all need to be named and uploaded...and I haven't gotten around to that. Once everything is uploaded, obviously formatting is the way to go. Agreed that it doesn't hurt.

Ekib, I use omnis because I like the sound better. The few recordings I have made with cardioids gave me the impression through headphones that there was nothingness behind me--a weird flat void, which didn't feel natural. Also, since I usually clip the microphones to me, semi-stealthy, omnis don't sound like they are swinging around if I have to move.

Omnis also give you more bass response for the buck, and they are less susceptible to wind noise. Cardioids affixed to a stand, up over people's heads and aimed at the PA, are probably an ideal choice for some people, but I prefer omnis for my more low-profile method.

Ekib · « **Reply #23 on:** July 02, 2012, 04:27:56 PM »

Quote from: earmonger on June 22, 2012, 05:01:36 PM

Well, 40+ hours of Bonnaroo recordings all need to be named and uploaded...and I haven't gotten around to that. Once everything is uploaded, obviously formatting is the way to go. Agreed that it doesn't hurt.

Ekib, I use omnis because I like the sound better. The few recordings I have made with cardioids gave me the impression through headphones that there was nothingness behind me--a weird flat void, which didn't feel natural. Also, since I usually clip the microphones to me, semi-stealthy, omnis don't sound like they are swinging around if I have to move.

Omnis also give you more bass response for the buck, and they are less susceptible to wind noise. Cardioids affixed to a stand, up over people's heads and aimed at the PA, are probably an ideal choice for some people, but I prefer omnis for my more low-profile method.

I find the bass response with omni's the thing I don't like. I find that very unnatural. And for me mostly recording metal , the bass drums are always the big problems with mic's who have more bass response. Now I know you have bass roll off's , filter's etc. But in the end I ended up mixing my recordings the same way my current card's are doing . The way omni's give that bass response is something you never hear in a concert at all .
I am very happy with my card's ( Core Sound Stealthy Cardioid ) and I won't change them for anything else . Not even for mic's that are supposed to be better . This is the sound I like ( and I use them since 1998 ).
Still I find it interesting to know someone else's opinion ( like yours ) and it's good everyone's taste can be different !

earmonger · « **Reply #24 on:** July 08, 2012, 02:56:45 AM »

The sound you like is the sound you are accustomed to. Many concerts give you huge bass in terms of real experience. The omnis don't increase the bass but they do record it. The cards roll it off. I think the omnis are more accurate--even at a bass-heavy show--and after they capture the full spectrum, you can roll off the bass if it's too much.

Ekib · « **Reply #25 on:** July 08, 2012, 01:59:34 PM »

Quote from: earmonger on July 08, 2012, 02:56:45 AM

The sound you like is the sound you are accustomed to. Many concerts give you huge bass in terms of real experience. The omnis don't increase the bass but they do record it. The cards roll it off. I think the omnis are more accurate--even at a bass-heavy show--and after they capture the full spectrum, you can roll off the bass if it's too much.

Don't you have a problem on the bass drums etc. starting to distort when there's heavy drums ? That was my big problem the time I used binaural's.
I also like card's because they are more "direct".

I totally dig my mic's and not because I am used to them . It's because they record the way I like it . Metal is extremely hard to record , especailly thrash metal . And the Coresound's have never failed on me.

It's all taste though and I wouldn't argue at all . I am just interested why people choose this or that mic .

F.O.Bean · « **Reply #26 on:** July 09, 2012, 10:09:13 PM »

I reformat my cards after they are full and I need to record a new festival. I have been very lucky and have had ZERO problems with my 2 x M10's

I JUST formatted my cards from electric forest after i xferred them to both my INHDD AND EXHDD. so now its safe to record on them again

MrMuse · « **Reply #27 on:** May 27, 2013, 05:28:58 PM »

Torn between digging up an old thread or starting a new one. Figured it might be better to keep things all in the same place...

Anyway, I had exactly the same problem reported by the OP.

A two hour+ concert recorded onto a Kingston class 4 16gb micro SD, recorded at 96/24.

This has been automatically split by the M10 into three files, the first two of which are 2,097,148 kb and each of these files is 1 hour, 2 minutes and 8 seconds in duration. The last one is a smaller file of only a few minutes duration.

Here's the interesting part, I have two sections, one on each of the large files - towards the end (54 minutes on one, 57 minutes on the other), where I get exactly the same problem reported by the OP.

The card is a Class 4, and reputable brand.

The card was formatted (on windows) the afternoon before the concert.

At the start of the second file the glitch occurs for 10 secs before clearing and the rest of the recording, well up to the 57 minute, is spot-on perfect. This seems to be some form of hangover from the first glitch.

Strange.

Did anyone do any more testing and find an answer to this?

It does look to me like some problem in writing to the SD as each file becomes larger.

Fatah Ruark (aka MIKE B) · « **Reply #28 on:** May 28, 2013, 11:49:16 AM »

I think people have found that if you format the card on Windows (ie not in the device) that you'll be more likely to run into trouble. Not sure if that is the source of your trouble though.

yates7592 · « **Reply #29 on:** May 28, 2013, 12:29:00 PM »

I've used Class 4 16GB Kingston SD cards many times at 24/96 without any issues at all (going into R-09HR not M10). I always format the card every time in the recorder.

MrMuse · « **Reply #30 on:** May 28, 2013, 01:11:53 PM »

Thanks for the replies.

It seems that a format on the device is a logical precaution - wish I'd have thought of this beforehand...

I'm tempted to see if I can recreate this problem so at least I know what is causing it. I still have a hunch about it being related to the length of the recording with it appearing in similar places on both files. I guess if I can recreate it and then eradicate it through on-device formatting then we will all be better off.

Author Topic: PCM-M10 Glitch--and of course it was during Radiohead.... (Read 9893 times)

adrianf74

adrianf74