BMPCC 6K Pro stops recording on Samsung T7 SSD

Hi everyone!

I'm really stuck so I decided to post my question here. Today I was shooting a video for a client in Marrakech on my BMPCC 6K Pro, Constant Bitrate, 8:1, 60 FPS onto my Samsung SSD T7.

I've never had any issues with this setup, but all of a sudden today it kept stopping with recording after 8 seconds, 10 seconds, sometimes 2 seconds. And if I would quickly press record again, I could film for up to like 40 seconds if I was lucky. But then the same trend would happen.

I had another Samsung SSD T7 with me, switched to that one and tried recording. Tried switching the (default) cables, but still the same issue.

I was thinking it was maybe due to the weather, it was about 30 degrees Celcius. But when I shot in Jordan a couple months back, minutes on end there was no issue. And I decided to take a break and live it in a room with AC but that didn't help. And when I switched to 6K full frame and tried recording 24 FPS seconds there was no issue.

I read in the forums that might card might be too slow, but I double checked the specs and it can write up to 500 mbit/s - 1000 mbit/s so this shouldn't be the issue right?

Also, I have never had this problem before. I also formatted both cards before the shoot so there wasn't anything on it.

I saw some people recommend the Angel CFast cards, but they are out of stock and also crazy expensive.

I don't know what I am doing wrong, or if you have any suggestions. But all help is greatly appreciated!

How long and how intense has the SSD been used?

Frequent rewrites using NAND (or other similar technologies) come at a performance cost.

You should consider using a different SSD, e.g. Samsung T7 Shield which is there on the list of officially supported media.

How long and how intense has the SSD been used?

Frequent rewrites using NAND (or other similar technologies) come at a performance cost.

I think I have used it for about a year, the other one maybe 6 months. And I don't have shoots very often, but if I do have one (maybe 2-3 per month) I think I shoot on in for about 3-6 hours but it is mostly 24 fps, and then 2-3 minutes 60 fps.

But indeed, in the beginning not an issue (with the older and newer one that I bought). But even with the new one, I had the same issue yesterday.

You should consider using a different SSD, e.g. Samsung T7 Shield which is there on the list of officially supported media.

Hmmm, I thought all Samsung SSD were the same so I will have a look into that thank you!

Hi everyone,

Thank you so much for your help and feedback, greatly appreciated

After the comments and doing more research, I figured out that even though my Samsung SSD T7 could work doesn't means it is meant for my specific workflow/workload.

So I ordered the Angelbird AV PRO SE (CFexpress type B, 512 GB), I saw in comments that people trusted and used these the most. The website states it can record up to 8K RAW and I shoot in 6K. Plus, I also bought the native card reader just to speed up the process Angelbird CFexpress Type B MK2 (USB-C).

I'm confident that now I shouldn't run into this problem hopefully, hopefully, it will be worth the investment

Thank you so much to everyone, your comments really helped!

Hey.
The same thing happened to me today. I was recording 5.7k 60p to a T7 Shield and it stopped recording after like an hour of shooting. Before it started doing this, I'd recorded 15 minutes of continuous footage. Then it stopped and when I pressed the record button again it stopped after a second. And again, and again. I switched to 24p as I really had to capture the moment there, and it didn't stop recording. But then it stopped again at 24p a few minutes later. Currently looking to find the issue, I'm guessing it's the SSD but I don't know why. It's stated on the site that it's fine to use this SSD for external recording.

Same issue on two shoots. Ive done testing on cables for the second shoot and it was fine, but on the event, its stopped recording. Switched from 6k to 4k and it was ok.

After editing, ill try to format t7 on exFat, but its already formatted correctly id presume. I dont know what the issue is right now.

Hi everyone,

Thank you so much for your help and feedback, greatly appreciated

After the comments and doing more research, I figured out that even though my Samsung SSD T7 could work doesn't means it is meant for my specific workflow/workload.

So I ordered the Angelbird AV PRO SE (CFexpress type B, 512 GB), I saw in comments that people trusted and used these the most. The website states it can record up to 8K RAW and I shoot in 6K. Plus, I also bought the native card reader just to speed up the process Angelbird CFexpress Type B MK2 (USB-C).

I'm confident that now I shouldn't run into this problem hopefully, hopefully, it will be worth the investment

Thank you so much to everyone, your comments really helped!

Im personally not investing into 512GB CF card after buying 3 T7 Shields at 4TB, T7 issues were fixed since T5 days (long long time ago), and Ive been using them on 10 shoots since i have BMCC 6K so theres some new issue now. Ill buy some cables I guess, maybe T7 ones are not good enough

I also remember on 1st shoot, switching to 4k or even 1080 didnt work, it just kept stopping. So its weird.

Hey.
The same thing happened to me today. I was recording 5.7k 60p to a T7 Shield and it stopped recording after like an hour of shooting. Before it started doing this, I'd recorded 15 minutes of continuous footage. Then it stopped and when I pressed the record button again it stopped after a second. And again, and again. I switched to 24p as I really had to capture the moment there, and it didn't stop recording. But then it stopped again at 24p a few minutes later. Currently looking to find the issue, I'm guessing it's the SSD but I don't know why. It's stated on the site that it's fine to use this SSD for external recording.

Try testing your SSD with the “Samsung Magician” found here:

https://semiconductor.samsung.com/consu ... /magician/ (updated link)

The software will also check, if you need to upgrade the SSD firmware.

Regards

Disk sustained write test example

I decided to create a simple “disk performance torture test”, that can be easily reproduced in nature by anyone interested. It is meant as a supplement to manufacturer test software, like Samsungs Magician (available here:https://semiconductor.samsung.com/consumer-storage/magician/), that has a plethora of tests, options and configuration methods coming handy, when problems arise. Even alerting you to and performing critical firmware updates if required.

Crucial and other SSD manufacturers have similar, specialized tools available for free. Use them.

There are good and thorough tests available online - the best and most thorough from Toms Hardware and Anandtech. The only specification, that really matters (in addition to power and temperature in use) is “sustained write speed”, and especially if the speed is constant for large (to full disk) writes. Some “lookers” have phenomenal speeds, while writing the first 30, 50 or 100 Gigabyte, but then - suddenly - speeds drop to even below “rotating rust drives” at around d 100-125 megabyte/second (instead of initially 900 to 1000 megabyte/second), when using 10 Gigabit/sec USB-C connections.

Luckily, sustained write is very easy to test in real life.

The test

The test involves copy from a known fast source. In my case a Samsung 990 Pro 4TB mounted in a Thunderbolt 4 case reading at around or above 3 Gigabyte/second. Fast enough to not have any significant influence on test results.

My source data set contains 338 Gigabyte of mixed info, text, Lightroom, FCPX projects, Resolve Studio projects and RAW images plus ProRES and h265 videos. Read from one folder and copied to another folder in full on the target drive to be measured. You will get higher data rates, if you’re using one, single 338 video file Gigabyte in size. Alas… I had this folder handy, not the 1TB video file.

The data set is exactly 338.053.067.389 bytes in size in the following. You choose, what you have at hand.

The data is placed in the source folder:

Code: Select all

/Volumes/980Work2TB/Temp

The data is copied into the target drive and folder:

Code: Select all

/Volumes/Lexar/Temp1

Here a Lexar SSD. If I want to test sustained writes of 1TB (just above) I repeat the copy 3 times - into the target drive folders temp1, temp2 and temp3. Writing into the same folder will not do.

I’m lazy, so keeping track of time is not my thing. Computer does that very well, if told to.

So… I ask the computer to note time and date of pertinent states in the following script (here the Lexar SSD) run in Terminal on my MacBook M1 Pro 14 Pro:

Code: Select all

Echo “Start” > /Volumes/990Pro4TB/Testresults/date.txt
date >> /Volumes/990Pro4TB/Testresults/date.txt
CP -R /Volumes/980Work2TB/Temp /Volumes/Lexar/Temp1
date >> /Volumes/990Pro4TB/Testresults/date.txt
CP -R /Volumes/980Work2TB/Temp /Volumes/Lexar/Temp2
date >> /Volumes/990Pro4TB/Testresults/date.txt
CP -R /Volumes/980Work2TB/Temp /Volumes/Lexar/Temp3
date >> /Volumes/990Pro4TB/Testresults/date.txt
Echo “End” >> /Volumes/990Pro4TB/Testresults/date.txt


This produces the following “date.txt” data in a folder (on a third disk): /Volumes/990Pro4TB/Testresults/date.txt

Code: Select all

“Start”
Tir 22 Okt 2024 14:27:49 CEST
Tir 22 Okt 2024 14:45:50 CEST
Tir 22 Okt 2024 15:06:13 CEST
Tir 22 Okt 2024 15:26:35 CEST
“End”


These data are entered into an excel spreadsheet (you can download the above zip’d spreadsheet (LexarTestresults.xlsx.zip) as a template from here: https://drive.google.com/file/d/18aJKio ... sp=sharing)

The tiny Lexar Go 2TB SSD delivered these results:

Sustained write stress test results (including temperatures)

Lexar 2TB results.png (53.62 KiB) Viewed 1709 times

The results are more than enough for iPhone use, but the drive will do fine for most use cases up to around 200 Megabyte/second or thereabouts. Just stick into your camera USB-C socket on the side (have tried some), and it will work.

I decided to test temperatures too (measuring points shown):

Lexar temperature measuring points

Lexar Temperature Measuring Spots.jpg (519.41 KiB) Viewed 1709 times

These are not really alarming for such a tiny critter. The interesting thing is, that temperatures stay constant as well as speed (after initially showing very high write to higher speeds the first 30-150 GByte or so).

The more “normal” sized Samsung T7 Shield delivered these values:

Samsung T7 Shield 4TB sustained write results

Samsung T7 Shield Results.png (42.12 KiB) Viewed 1709 times

Only tested 1TB sustained write here (I KNOW, it will work for all the disk, but I got lazy).

So... this is a poor mans easy SSD sustained write stress test, that delivers plausible results. If many files are involved, as here, the results, you get, tend to be a bit lower, than actual performance obtained by writing one, single 1TB video file to the target drive. You can do that, if you like.

This will ALSO allow you to detect drives with impressive specifications, that do not hold up for sustained writes. That's where Toms Hardware or Anandtech or other websites can help you decide, whether your drive is a marketing wonder ("dud" in anyone elses book) or may have acquired defects. In that case contact the manufacturer support department.

Regards

Size comparison between Lexar 2TB SSD, Samsung T7 Shield 4TB SSD and iPhone 15 Pro:

Size comparison

Size comparison.jpg (422.21 KiB) Viewed 1707 times

Regards