Discussion in 'Firmware Wishlist' started by evoixmr, Mar 25, 2017.
I wholeheartadly agree!
At least if someone at SJCam made clear what mics work with the external jack, that would be a step forward. It'd be helpful if someone who has mics used with other gear tried them on the SJ7
Video recording is limited to a file size of 4 gigabytes on exfat file system.
There's no problem with 4Gb files. The camera shoots about 3.9 Gb (about 14 minutes with 1080p@60 fps) and starts a new file that picks up from the old file. Butt splice them in an editor and there's no jump in the image. There is a dropout in the audio (one or two frames at 60 fps) that makes a small click - that's been reported.
Going over roughly 15 minutes only makes editing somewhat slower. I usually edit within each 14 minute clip and then merge the clips towards the end of project.
Really? That is, regardless of the file system, the camera writes files of 4 gigabytes?
In the dark the camera can not shoot the video, does not see anything, but in the photo mode the camera sees everything and takes pictures at night. Will they fix it? Autilight is ON.
The file system doesn't change file size. Remember that the camera formats all SD's to exFAT FS. If you're going to format a chip with your PC/Mac, stick with exFAT.
I haven't worked very much with low light situations. General photo experience tells me low light video is much harder to do because the time for the exposure for a single frame limited to a very short time. Fast exposures and low light don't get along well. If the camera sensor worked well in infrared (many sensors do work into IR but have filters to block that to get a better visible light image), video might still work relatively well. AFAIK the SJ7 doesn't work into the IR, so low light isn't going to give good video. Life is like that some days. [/grin]
Can't fix physics with firmware. In very low light levels there just isn't enough light to create an image in the 1/50 or 1/60 of a second that each frame is exposed. Auto Low Light only works at the higher frame rates because in low light it actually drops the frame rates down to 25 or 30 fps rather than 50 or 60 thereby giving each frame twice as much exposure.
In photo mode the shutter is adjusted so it is open longer giving each photo much longer exposure than a video frame.
The camera does pick up IR light - just point a remote control at the camera and press the buttons. I do this to check the batteries in my remotes.
At the back of the lens there will probably be an IR-cut filter (a square of glass coated to block IR) glued on - this blocks the majority of IR. Even without an IR-cut filter you need an IR light source - that's why IR cameras usually have a ring of IR LEDs around the lens.
The fact is that many people return the camera back to the store, because in the dark the camera shoots much worse sj4000.
How did it happen that the top camera shoots worse than the first camera from the sj4000.
This is a limitation of the Ambarella chipset..... even SJ5000+ also did not support webcam...
Hello, thanks for this forum, it is really interresting to read.
I wish you could add a low cut filter (on/off) on the internal microphone recording. Like a -15 db at 100 hz to filter unwanted rumble.
A thermal protection or warning on the display when it is becoming too hot inside.
Enable FOV selection for 1080p 30fps and 60fps 16:9 video resolution with gyro on.
Firmware v1.15 had a workaround... You could select the desired FOV at a higher resolution and go back to 1080p 30 or 60fps 16:9 and everything worked fine.
Is it possible to downgrade to 1.15?
What about a "1 second" option in the Time Lapse settings?
Please add the opportunity to lock the touch screen. I use the cam often without a case an mostly one finger is on the screen an it takes any attitude.
Any other lock mechanism like to press 2 buttons for 5 seconds or so will also very nice Thank you
FEATURE REQUEST: White Balance Native/Off option
Description/Reason: There are times wen it is preferable to have the native camera white balnce without any correction performed in camera ad perform selective correction in post.
For example in videoing night time street scenes the street lighting my have a yellow tint whilst street signage may incude blues. In white balancing the over-riding yellow light the camera would add blue to the scene which over saturates the existing blue signage.
This is then compounded by the colour profile which is added to the video which adds contrast and saturation making the blues go over gamut and turning into a blue blog lacking any detail.
A Native/none option would allow selective white balance in post based on hue/saturation values.
Here's a visual example of when a flat colour would be better
In videoing night time street scenes the street lighting my have a yellow tint whilst street signage may incude blues. In white balancing the over-riding yellow light the camera would add blue to the scene which over saturates the existing blue signage.
This is then compounded by the colour profile which is added to the video which adds contrast and saturation making the blues go over gamut and turning into a blue blob lacking any detail.
A flat neutral colour profile would allow selective colour correction in post to take into account saturation and hue of the different RGB channels
Would it be possible in the next update to include Image Rotation in other video resolutions other than 4K
IMAGE ROTATION for all video resolutions
HDMI TV OUT playback...so we can watch our footage direct from camera to television
Yes. You can install an firmware rev. including reloading the rev. currently loaded. It's a good way to completely reset the camera.
I'm running the firmware 1.20, photo lapse does not have long exposure. I love to shoot night sky timelapse, I wish Long exposure in photo lapse will be available in coming firmware.
The problem with long time, low light exposures is the noise level climbs. A lot.
Wish item: For very long time lapse exposures, take 5-10-20 exposures and combine them to get one image with far less noise than a single image.
This is a common practice in astro-photography, for example. The process puts many noise images on top of each other, with good bits of data in one image covering up bad bits of data in the overall image. There is a point of diminishing returns and, more importantly, there's some possibility of the object(s) smearing as they move from frame to frame. So, a lower number of frames (10-20) will keep noise down and minimize smearing. If the noise level of the finished image is still too high, let an editor app fix it later.
Separate names with a comma.