The Red obviously is a nice portable size so it’s what I like to tote around to shoot with. I have the older LumePad1 and I can take move the images over to it just fine and they retain their 3D aspects.
So far when I have tried that with my LumePad2 it treats them like 2D JPG files. If I load the same images on the LumePad1 first then send it as a LIF file to the LumePad2 it works like a champ, BUT that is a lot of steps and devices just to get that done.
What SHOULD I be doing? I want to use the newer version to show off photos taken on my Red Hydrogen.
Hmm. Well it’s good to hear that is supposed to work without extra steps. I was just using Googles photo cloud service to upload and download the files. It actually worked fine from Red to Google to Lume1, but didn’t appear to work from Red to Google to Lume2. Could be user error. I can try OneDrive as well.
I made progress by completely bypassing Google Images and just using OneDrive to move the files. That let me bypass the middle-man thing.
But one more related issue.
I also want to export/send these out as SBS Images from my LumePad2 for display on My 3D TV and for editing in Stereo Photo Maker. When I do this the resolution is reduced from 4056x3040. At least when sending to OneDrive I get no option for the output size.
to 1280x720 per image. I can handle a little image reduction, but that is a lot!
Yes, the resolution should be MUCH higher. Should easily output dual 5K (nearly 10K horizontal resolution) when using SBS in LeiaPlayer.
Are you on Lume Pad 2 and is your LeiaPlayer app fully updated? Have you tried other sources besides OneDrive? Because it’s possible that’s the maximum the app accepts or it scales the resolution down upon upload.
Its fully updated. I’m assuming the way i’m doing this, OneDrive is just treating it as a file and doesn’t care what resolution it is. When I export out the LIF to OneDrive it works fine. Here is my little test share of the same image. The first is the raw file from the Red phone, then the LIF export which works fine (file size increases), then the SBS.
One additional datapoint from testing today.
If I take a photo on the LumePad2 itself, the SBS export is fine. It appears to be only happening with images that originated on the Red Hydrogen.
I heard somewhere there was multiple Left and Right images embedded in that original format. Could it be pulling the wrong ones? I tried a standalone script someone wrote and got interesting results as well. The Left image was full resolution but the Right image s hight was 720pix. I’m not sure if any of this sheds more light on the issue, but I thought I would add it.
So I confirmed with my team. The reason this is happening is because we used to encode the stereo 3D images down to the maximum resolution visible on the RED Hydrogen One display, which was 720p per eye. The 2D view remains at high resolution, but that only works in 2D. We did this because we wanted to save storage space for users.
As we started making products with higher resolution screens, we started to encode all images at maximum resolution, knowing that even if it’s not useful for the screens of today, it will be useful for the screens of tomorrow.
If you want to circumvent the issue on RED Hydrogen One, take the photos using “Pro 4K SBS” mode.
For these images, however, they will always be 720p when exported as SBS because that is the 3D image resolution of the file internally.
Thanks so much for the follow up. I tried to find that setting “Pro 4K SBS” but had no luck. Where is it?
If i’m understanding how all this works (very possible I don’t) there is still a full resolution image along with a depth map in the file and that is why the LumePad2 displays it so nicely correct?
Assuming this is correct, do any of your tools allow “conversion” using the existing depth map? Your online converter tool is fantastic, but it would try to create a depth map from scratch instead of using the more accurate one already in the image.
Stereo Photo Maker doesn’t do anything with these formats right now.
It’s in the Camera app on RED Hydrogen One under the Gear icon!
No, there are a bunch of different kind of LIF files. The kind from Hydrogen have one high-res 2D image and then have two lower-res views for a 3D image, and optionally have a depth map for each of the 3D images (but not the 2D image).
Yes, if the image is a LIF with single image and a depth map, or simply a standalone 2D image and depth map. However, that’s not what the images from Hydrogen are.
No, it will use the the depth map if it’s the right kind of image. It won’t in cases of legacy files like those from RED Hydrogen One. IF the Hydrogen image has a depth map for each image, you’ll want to pull the left depth map and match it with the high-res 2D image. You can upload the 2D image to LeiaPix Converter then manually upload the depth map using the depth map upload feature. But that’s only if it already has that. If not, you may need to use DMAG with SPM or in a different way to generate a depth map for you and map that to the left image, but it may be off a bit as it’ll be based on a central view, not the left view as needed.
Masuji Suto has a Lume Pad 2 now, so maybe this will change in the future.
Yup, that worked. I’m not sure how I missed that those updates existed.
Is there any downside other than file size to using that setting? Does it still capture the depth maps, or does it just become two traditional stereo pairs?
Thanks again for all your help. This has really helped me get things going. Right now I have been using GigaPixel (Topaz labs) to blow up the 720 pairs I already took and with some tweaking its working quite well.
It just captures them as a standard SBS .jpg, so there’s nowhere to put the depth map.
Still, there’s no worries, we can compute a superior depth map after the fact. Just hit the “Edit” button on the image in LeiaPlayer on Lume Pad 2, make your adjustments, and save it: it’ll turn the SBS into a LIF with a modern, high-quality depth map. The stuff on Hydrogen is now half a decade old disparity estimation tech, it’s obsolete!
That is interesting. I had some shots from my W3 (talk about old tech) that I made adjustments to and wasn’t sure what was actually happening behind the scenes. It does appear to do some magic and make it display much better.
One image recently it pukes on and creates some scrambled result. It was a shot I edited in SPM and it displays OK with the _2X1 added, but when I edit it, something goes wrong.
I haven’t found the DMAG thing to be very good. Very blocky levels and weird mistakes.