My New SOP

[-mjk-]

Wow, what a difference, working with the DP-32 as opposed to a 24 track analog machine. After all the information I've read on this forum, I've developed a whole new way of doing things. Recording itself hasn't changed, but how I interact with the machine sure has. Basically, I keep tracks whole and delete any and all fragmented sections.

I do whatever it takes to get the track how I need it: punching, editing, bouncing, etc. Then, I export the track, do a cleanout, Delete Unused, and import back to the cleaned track. Now, every track on my machine is a single, complete track, start to finish. Anything that requires multiple takes gets exported and imported.

Since I implemented this procedure I have not experienced a single error. I'm convinced that this method will maintain the health of each song.

 

[-mjk-]

Well, so much for that.

With all the tracks being continuous single tracks either imported or recorded straight through in a single pass, I did experience a file error yesterday, when laying down six tracks of harmony guitars. I spent about 45 minutes on it, and punched in maybe a couple of dozen times. But I was only working on a small section of the song about one minute in length. After the machine locked up, I had to pull the power and, on start up it reverted to my Reset template and would not load the song I was working on. I had to copy my back up files to the SD card to go back to where I started. Doing the exact same six tracks in the exact same spot on the song, I was successful the second time around. As far as I can tell nothing was different. I did a submix of the six guitars the way I wanted it to sound, cleaned out the tracks, and reimported it back into clean tracks. I honestly have no idea why that file error occurred because, I didn't get it doing the exact same thing on the second time. That being said, I decided to back up the entire card and reformat it.

In my opinion, the single most important operation to be performed on these machines, is backing up your data every single time you finish a session. Without fail.

 

[Phil Tipping]

Interesting... just a few questions if you have the time.
1. Had the card been fully formatted since your last firmware upgrade?
2. How many files were on the card for that song? Each time you punch-in, you'll create a new zz file for each track which is set to record - doesn't matter if it's only a few seconds, a file is a file.
3. Were there any other songs on the card? If so, how many, and how many files for each one? Maybe the bottleneck is related to the total no. for the whole card.
Even if we can't pin down the reason, your advice re. regular backups is very sound.

 

[-mjk-]

hi Phil. here are the answers. they are long but i tend to try and give complete answers:

1. yes. i bought a new card after the 2G supplied card proved problematic, and i formatted it before use. i've also formatted it once or twice during my project. i had also formatted the original card after i upgraded the firmware.

2. about the song in question: my backup copy of the folder after it's subsequent corruption shows 42 files (and i think i should add checking that at each failure point to my SOP), and currently there are 23 files in that song directory, and that is after finishing the operations i was doing when i had the above-mentioned failure. all these numbers includes the 2 .sys files, and 2 mixdown files. i use mixdown/mastering pretty much every time i use the machine. i make complicated submixes in mixdown mode, then apply mastering tools to the submix before importing to a clean stereo track pair. on Nov. 10, my backup has 65 files in that song directory. since i wrote this OP on the 14th, that makes sense. On Nov. 14, my backup has 19 files - same song, but more developed in spite of the lesser number of files. Nov 15. has 23 files. so it looks like implementing my new SOP has significantly reduced my file numbers, and is keeping it down to a minimum.

when i experienced the error, yes i was making new files, but very small ones. however your contention must be correct - each file must be referenced and dealt with on the fly, and at some point the machine chokes on the sheer volume of data that must be managed. i just didn't expect that to happen from (what is considered normal on analog tape machines) short duration punch-ins on those 6 tracks. maybe i need to stop at the end of each track and fix it before moving on to the next part. but that does so stifle creativity!

3. i had a pre-production version of the same tune with several experiments. instead of cleaning that up i just created a new song and started over. i also had 2 small projects (basically just imported files to tracks), with 4, and 6 files respectively.

what i've also been doing for the sake of convenience is, keeping the last backup copy in a song folder with the suffix "BU" (for backup) on the SD card just so i don't have to stop and copy those files to the card to start over in case of a failure. perhaps i should not add to the files on the card?

yesterday, i backed up everything, reformatted the card and put back only the song i was working on, and the smallest of the other project, which is now finished. i'm going to remove that from the card and only work with the single song and see how that goes. i'm keeping my Reset template on the card though.

I want to thank you Phil, and also David Porter for sharing invaluable insight and experience with these wonderful standalone digital recorders. without that insight and experience i would pretty much be an analog man lost in a digital world by now....

 

[Phil Tipping]

Thanks for the detailed response mj. I wouldn't have thought a few dozen files would be a problem... but would the pre-production version bump the number up significantly? I also forgot to ask about the AudioDepot folder.
You can easily see the total no. of folders & files from the 'properties' of the card itself (or a copy of it if you've archived the whole card). If we're up in the hundreds then maybe there's something in this. The size of each file is not important as the FAT is just a list of file & folder names with pointers to the actual data elsewhere on the card.
My idea was that the FAT has no knowledge of the current song so the firmware would always have to scan the list looking for matching filenames each time a file is required. This list could be enormous if other songs (or even AudioDepot) contain large nos. of files, so would be an overhead even if the current song only contains a couple of files. The overhead doesn't matter the instant you press play or record as you wouldn't notice a split-second delay... but if even one of the 'play' tracks is split across 2 or more zz files, the machine will have to find them on-the-fly while the song is playing/recording. This may all be a naive & incorrect view of how the machine really works, but I'm just trying to reason it through from the various symptoms & cures that have been posted over many months.

 

[-mjk-]

Phil, I'm going to address what you just posted. shortly. But first, I must tell you about a File Error message I just got. VERY perplexing. I'm using MIDI to cut drum tracks on a totally new song, made after my Reset Template. Single pass, start to end, no stops (naturally, since it's a drum machine programmed song). Second pass, recorded the bass track from the drum machine, again, single pass, no stops.

I did a mixdown of the drums with the effects I wanted, and then went into Mastering Mode to apply my multi-band compression User Preset. Pressed record, and at exactly 50:00:00 I got the File Error message. I had to pull the plug.

When I restarted it was as if nothing ever happened. The mixdown file was intact and I turned on the compressor and successfully recorded a mastered track.

There are 3 zz files, a mixdown file, associated zz file, and the 2 .sys files.

Next, I'm going to import the mixdown file into a stereo track pair for my drum track, after I do a cleanout and Delete Unused.

I'm beginning to suspect the SD card. It's 100 MB/S micro SD with adapter.

What do you think? Thanks.

 

[Phil Tipping]

That's a shame - seems you've just confirmed what I've been testing today - large nos. of files and fragmentation does not seem to be the culprit
I made several test songs with 100, 500 and 1000 files (just copied & pasted the zz files from an existing song). The whole card had about 4000 files on it. I created a new song and tried repeated recording on several tracks to force them to be split over multiple zz files - worked perfectly. By the way, I noticed that Track Edits do not create additional zz files - they only seem to be created when the Record button is pressed, i.e. when recording or bouncing.
I then tried loading the song with 1000 files. It took about 50secs to load and played ok, but when I tried recording, it immediately said 'Writing failed'. This was not a lockup as I could just press F1 to acknowledge the message. I tried the 500 file song and this was ok with recording - repeated the same tests as the new song and all worked perfectly, so sounds like although there is a limit with the no. of files in a song, it fails gracefully so I don't think this is the same as a 'lockup glitch'. The response did seem a bit sluggish - there was a definite delay between pressing Stop after recording and the spinning wheels appearing, but that was the only difference.
Bottom line is I'm back to square one - keeping things tidy and re-formatting seems to help, but as you've just found, this isn't the whole answer. My card is a 90MB/s class 10 Sandisk, and have only had 3 non-destructive glitches in over 2 years of use, but it's difficult to compare 'usage' - I don't normally edit & mixdown on the machine so maybe I'm a light user compared to some.
The UK tech support guy recommended 16GB cards on the grounds that 32GB was 'pushing the limits' and he didn't think it was worth the risk, but several users have had no (or very little) trouble with 32GB, so it's difficult to advise.
So overall, an interesting, but ultimately disappointing, exercise. Wonder if the Model 24 has any issues - I believe it can write 24 (22?) tracks simultaneously.

 

[Mark Richards]

"...Wonder if the Model 24 has any issues - I believe it can write 24 (22?) tracks simultaneously..."

Guys, I use only 32gb sd cards, with no issues, on my DP-24 and I do some pretty sophisticated mixing and editing; and will often use multiple virtual tracks for one song. I do all my work on the DP-24 and a typical song can end up with 1.5 gbs or more before I clean it up.

I always try to make sure the card I use is on the tested/approved list. The card I'm using currently is a 32gb Sandisk Extreme 45mb/s SDHC I 1/10.
mjk, in case you havent come across it, for the 32:
https://tascam.com/downloads/tascam/775/dp-32_tested_media_list_r4_20140523.pdf
(not sure if the 24 and 32 share any of the same tested cards, but there are a lot more tested cards for the 24:https://tascam.com/downloads/tascam/723/dp-24_tested media_list_r5_20140523.pdf)


Past experience has taught me to use only tested/approved cards, and ~not~ to use ~any~ micro SD cards in an adapter.

 

[David Porter]

Man... thanks to all of you guys who are sharing detailed experience/info and helping all of us to get a better grasp on these beloved digital recorders. These archives are invaluable!

 

[-mjk-]

"...Wonder if the Model 24 has any issues - I believe it can write 24 (22?) tracks simultaneously..."

Guys, I use only 32gb sd cards, with no issues, on my DP-24 and I do some pretty sophisticated mixing and editing; and will often use multiple virtual tracks for one song. I do all my work on the DP-24 and a typical song can end up with 1.5 gbs or more before I clean it up.

I always try to make sure the card I use is on the tested/approved list. The card I'm using currently is a 32gb Sandisk Extreme 45mb/sec SDHC I 1/10.
mjk, in case you havent come across it, for the 32:
https://tascam.com/downloads/tascam/775/dp-32_tested_media_list_r4_20140523.pdf
(not sure if the 24 and 32 share any of the same tested cards, but there are a lot more tested cards for the 24:https://tascam.com/downloads/tascam/723/dp-24_tested media_list_r5_20140523.pdf)


Past experience has taught me to use only tested/approved cards, and ~not~ to use ~any~ micro SD cards in an adapter.

Good info Mark, thanks. I admit that I didn't really pay attention to the approved list, because the list has no date on it and certainly, the cards have significantly improved since release of my older DP-32. But at the top of the list are some new cards, so the list apparently was updated since release (I really hate bad documentation).

In addition to being Class 10 SDXC II (2) what the recommended Sandisk cards all have in common is:

• Read Speed: up to 300 MB/s*
• Write Speed: up to 260 MB/s*

My card is 100 MB/s. It's an Extreme model, but not the Pro. So it looks like I'll be shopping for a new SD card later today as it sure looks like the card is the issue. But, I cannot bring myself to believe that the form factor makes any difference, if the specs are the same.

Edit: What I do not understand is how a slower read/write card can out-perform a faster card. Mark's card is rated at 45 MB/s and that is considered quite slow these days. But it is a Class 10 and apparently works. I'm not sure what the actual criteria is for a good card to work with these machines.

 

[Mark Richards]

"...no date..."
Yeah, just one more example of TASCAM not thinking things through enough, but in their defense, I posted the direct link to the pdf file. On the support page that leads to that file, the date is stated (07-15-2015 for the DP-24/24SD; 07/28/2014 for the DP-32/32SD). Have they been updated since? Who knows. Both file names are 20140523.pdf

"...form factor makes any difference, if the specs are the same..."
I don't know enough about the technical side of SD cards to make any guesses about what should work, so I look at the SD cards the same way I look at legacy software/hardware: when I'm running Windows7 on a computer that's 7 years old I try to find programs designed for Windows7/2011 hardware. The problem is finding legacy stuff. Ditto for older SD cards.

"...not sure what the actual criteria is for a good card ..."
I don't think TASCAM is quite sure either!

Remember that the DP24/32 first appeared in 2012, and some of the firmware updates messed with SD cards that had been known to work, requiring those cards to be reformatted after the firmware update was installed, and then crossing fingers. I'm still using firmware 1.16 on my DP-24 (latest US version for DP-24 is 1.19 released November 1, 2018*), and I still have a few of the 45mb/s 32 gb cards, so I'll just keep on using them until they die, I guess.

*BTW, notes for US v1.19 state:
V1.19 changes
- It shows "File error" on display during recording and the recording ends in failure sometimes.
This has been fixed."
(May be, but I'm done doing QC for TASCAM.)

EDIT:
Look what I just found on the TASCAM site after choosing US English language (the USA site) on the home page. (The default is international):
Path (start with a cleared browser cache):(or you can cheat and just click on the last link )
tascam.com
click on US English
links to tascam.com/us
click on SUPPORT at the top and then in the drop down click "discontinued products"
links to tascam.com/us/support/discon
Scroll down to and Click on DP-24 (or DP-32)
Scroll down to and click on Supported Media
and you'll see a listing that includes:
"SanDisk Extreme Pro SD UHS-II card SDSDXPK-032G-JNJIP 32GB New"

 

[-mjk-]

Edit: (My apologies!) Mark, once again, thank you so much for this invaluable information. Printed, and put in my man bag (all men in Asia carry one, lol). Going shopping now!

btw, did you happen to notice a 64GB card listed?

 

[Phil Tipping]

Interesting info thanks Mark. When I started looking for cards a couple of years ago, the exact card nos. on the approved list didn't exist on my usual shopping places, so I gambled on an unlisted one with a comparable/greater spec - I'm sure many people did the same. I just noticed there's a footnote on the list which suggests that using cards with similar specs should be ok:-
Notice: The above cards were tested successfully with this model. Over time, these specific cards may become unavailable. At that time, it would be best to contact the CF/SD card manufacturer directly and have them suggest the model with the most similar specifications that is currently available which remains within the maximum capacity limit.

Would be interesting to know how Tascam test the cards which end up on the list - if they have a stress test program it would certainly help if they shared it. Given the rare & intermittent nature of these glitches though, I wonder if their tests are rigorous enough. As far as I know, no-one has managed to repeatedly re-create a situation which causes these lockups/glitches, so diagnosis is still speculation.

Good point re. legacy gear - maybe that's why the tech support guy recommended 16GB cards? btw mj - 64GB cards won't be listed as the machine spec says 32GB is the max size.

 

[Arjan P]

64GB will not be listed for another reason, as the SD-HC spec allows sizes up to 32GB and its file system is FAT32. For bigger sizes the SD-XC spec is used and the file system is different: exFAT. Here's a very informative link:
https://kb.sandisk.com/app/answers/detail/a_id/2520/~/sd/sdhc/sdxc-specifications-and-compatibility

 

[-mjk-]

Here it is gentlemen. No wonder people are confused.

 

[Phil Tipping]

Well spotted - I give up!

 

[Mark Richards]

On the DP-32/DP-32SD list but not the DP-24/DP-24SD list!
If you go for it mjk just remember:
In the Old West it was easy to spot the pioneers; they were the ones with the arrows sticking out of them.

 

[-mjk-]

Phil, the whole mess is not easy to digest. In the first place, the 2 GB card that is supplied with the unit isn't even listed on the approved media list! Clearly it doesn't match the specs, because the card crapped out on me on the first day.

Here in Taiwan, the land where all this stuff is made, I went through several SanDisk SD cards at a couple different places. In every case, I found cards that matched the verbiage of the specification, things like Class 10, USH-1, etc., but none of the part numbers matched at all. So I didn't buy any of those cards, even though they had the exact same description as the cards in the list. it's cryptic because the description doesn't tell the speed of the card, which I believe is a non-trivial part of the specification. Btw, Chinese is a fun language for talking tech.

So I've decided to drop about US$70 on the UHS-2 version of the SanDisk Extreme Pro, which is rated at 300 MP/s, U3, Class 10. The part number on that card is somewhat close to the one listed.

I believe that the reason all of this matters is because the card has to simultaneously playback as well as record. It's not like a digital camera or video camera, where all it does is record, because nobody plays back at the same time while recording video. But we might be playing 32 tracks of digital audio back while we're trying to record on one! That takes a big pipe!

 

[-mjk-]

Mark, I have a good mind to stick a 64 gig card in there and just see if the darn thing will format!

 

[-mjk-]

Interesting info thanks Mark. When I started looking for cards a couple of years ago, the exact card nos. on the approved list didn't exist on my usual shopping places, so I gambled on an unlisted one with a comparable/greater spec - I'm sure many people did the same. I just noticed there's a footnote on the list which suggests that using cards with similar specs should be ok:-
Notice: The above cards were tested successfully with this model. Over time, these specific cards may become unavailable. At that time, it would be best to contact the CF/SD card manufacturer directly and have them suggest the model with the most similar specifications that is currently available which remains within the maximum capacity limit.

Would be interesting to know how Tascam test the cards which end up on the list - if they have a stress test program it would certainly help if they shared it. Given the rare & intermittent nature of these glitches though, I wonder if their tests are rigorous enough. As far as I know, no-one has managed to repeatedly re-create a situation which causes these lockups/glitches, so diagnosis is still speculation.

Good point re. legacy gear - maybe that's why the tech support guy recommended 16GB cards? btw mj - 64GB cards won't be listed as the machine spec says 32GB is the max size.

Phil, I am totally with you on the testing. I would very much like to know the criteria. Plus, I really want a complete description of the specification, including the read/write speed, class and video grade, etc. - everything that manufacturers use to identify the capability of an SD card. The omission of the speed requirement is glaring! When Mark stated the type of card that he used, I thought that the speed was rather slow, but it is a class 10 card. Today I could've gotten a similar class 10 card for cheap money, but I'm not going to do it because I'm unsure of the singular importance of speed in the equation. The reason is that my current card has a 100 MB/S read/write speed and it's failing. The truth is, every one of the cards that I saw today (all Sandisk) that matched the description but not the part number, were 90 MB/S or under and still class 10. Is that fast enough? I don't know! There isn't a speed specification listed anywhere in the description of the cards. So I'm gonna spring for the 300.

Sorry if I Sound like I'm complaining. But if I released a document like that, I probably would've been fired.