Tuesday, July 7, 2009

The 4GB file limit test - retest and updates

This is in responce to one reader's query on the gap between the 4GB files if a different class of SDHC card is used. A class 6 card has higher read/write properties than a class 4 card so in theory, the time taken to start and store a file should be faster on a class 6 card compared to a class 4 card; the only question is whether it is noticeable in real life. (Note: A class 2 card should not be used with the HD2000 since it cannot handle the videos at the higher resolutions.)

Since the last time I used a Kingston class 4 card for my test, I did the same test again using a Transcend class 6 card. The result I got proved the above theory since the gap was only 5 seconds compared to 10 seconds the last time. Out of curiosity, I repeated the test again using the class 4 card and the results surprised me.

No, it is not because the gap was shorter than using the class 6 card but because it was shorter than what I measured the last time. This time it was only 7 seconds compared to 10 seconds previously! Not sure if this was the SDHC card itself that affected the results, I repeated the test again with my 2 other Kingston class 4 cards. The results were very similar, one also gave 7 seconds and the other, 8 seconds. For some unknown reasons, the gap is somehow shorter than when I tested it the last occasion. I can only suspect that the cards are now more seasoned after being in used many times and the read/write is a bit faster. Who knows?

To confirm the results, I tested using the Transcend class 6 card again. The result was, thank goodness, consistent at a 5 seconds gap.

For those of you wondering, this is how I did the test; nothing elaborate but something very simple and which I think is fairly accurate. I left the camcorder to run on recording and when the end of file countdown come on, I video the Windows clock. Thus it will show the time the recording stop and also the time when the next video starts again. The difference in the two times captured on video will be the gap. There is a possible error of +- one second but good enough for this purpose.

BTW, there are also slight differences for the length of the video of the 4GB video among the cards used and it ranges from 23:39 to 23:48 using just the length of the video shown on the video playback listing screen.

To sum it up, these are my findings. You'll be the judge to say whether it is conclusive or not.

1. There is a gap between the end of the first 4GB file and the start of the next 4GB file.

2. Using the higher class 6 SDHC card will reduce this gap as compared to a class 4 card which is about 5 seconds as compared to about 7 seconds from the test carried out

3. The length of the video at 4GB also varies slightly depending on the card in use and ranges from about 23 minutes 39 seconds to about 23 minutes 48 seconds.

4. Use a class 6 SDHC card as far as possible.

If you are recording a long video that may exceed the 4GB limit and knowing the above figures, it is best to stop the recording manually at a suitable time before it stops automatically in order not to miss out any critical moments that you wish to capture, should they occur during the gap.
I believe this gap will also occur in all camcorders that uses the FAT32 file system but it may vary for the different brands.

Happy shooting.

Ronald Kwok


  1. This comment has been removed by the author.

  2. Thank you so much. I really enjoy your accurate reports/tests. They help us to understand better this new sanyo camcorder.
    I am wondering what would happen if we use this card :


    it says that records at 30 mb/s.

    Maybe the 7 to 10 seconds of delay can go even more down if we use this new class 10 card. If we think that the camcorder records videos at 24 mb/s....maybe 30 mb/s of the card would be perfect.
    Who knows if in the future someone that own that card can experiment and let us know.
    Thank you

  3. Hi macro,
    thanks for the kind words. If anybody is generous enough to send me one of that class 10 SDHC card (non-returnable), I would gladly run the test again with it! I guess the gap will probably be reduced but I suppose there will be a physical limit and we have to work around it until they change the file system. The 5 seconds gap with a class 6 card is OK for me but it may be too long for some users. Cheers.

  4. Are u sure that the hd2000 can not read NTFS formatted cards ?

    Can u pls try to format your SD card in NTSF and see what happen ?

    Thank you so much.
    I really appreciate your time and help.
    The pnly thingthat is holding mne back in buying the camcorder is the 23 minutes limit splitting file.

  5. Hi macro,
    I tried to format my SDHC card in NTFS via Windows but it would not do it, don't know why. Even if it cane, I believe it would not work since all camcorders (as far as I know) are using FAT32 system to record the videos. So they will also have this 4GB file limit. Sorry.

  6. The reason the maximum recording time differs (23:39 and 23:48 in your case) has to do with the bitrate being dynamic.

    One night, I set up my tripod and let the camera record until the battery was depleted. This was on the highest quality setting so I was expecting a recording time of appx. 23 minutes/file. The first video clocked in at 24 minutes, the second video was an incredible 36 minutes long. The file size was identical. The bitrate in the second video must have been 67 per cent lower than in the first video. This makes sense because as the night was closing in, the picture turned almost entirely black (manual exposure and ISO settings used). The black portions of the picture probably got compressed so effectively that a lower bitrate was sufficient, hence the recording time was longer.

    Nice blog btw. I'm currently thinking about getting a new lens for the Sanyo VPC-HD2000 that allows me to adjust focus and exposure while the camera is recording. It should be possible if I set the HD2000 to manual focus (infinite) and open the aperture all the way. Need to get a step up ring first.

    -simpucker from the AVS forums

  7. Hi simpucker,
    thanks for the feedback and the interesting results. You are probably right about the bitrate since that sounds logical. Anyway do share your experience in using an attached lens with the Sanyo HD2000 either here or in the AVS forum for the benefit of other users. Cheers.