Active Topics
-
The N900 is now 15 years old! (9)
to Nokia N900 by Arno_11 - 12 hrs, 3 mins ago -
Sailfish OS for the Motorola Moto G7 Power (XT1955-5) - (ocean) (4)
to SailfishOS by Maemish - 6 days ago - more...
 |
|
2010-07-29
, 15:32
|
|
Posts: 601 |
Thanked: 549 times |
Joined on Mar 2010
@ Redditch, UK
|
#102
|
Originally Posted by Bingley Joe
Nope. every time I try to update the fcam drivers it simply tells me that the operation to be performed is the removal of power-kernel and also power-kernel-settings... Guess not!
By the way.. are the new FCam drivers 1.0.5-2 compatible with Titan's Kernel?
| The Following User Says Thank You to Switch_ For This Useful Post: | ||
 |
|
2010-07-29
, 15:44
|
|
Administrator |
Posts: 1,036 |
Thanked: 2,019 times |
Joined on Sep 2009
@ Germany
|
#103
|
Originally Posted by Flandry
Gimp 2.6.8-3 did open the dng files I shot right after first release of fcam and fcamera... wrong size and wrong colours due to cyan bug and stuff but it did open the files!
Does anyone else think it's kinda silly that this amazing open source, collaborative industry/academic, revolutionary camera project that was first released on a linux phone only supports a raw format that proprietary ($@%#) Adobe windows software can handle?
For the complaints about the quality... please stop comparing with dslr or stuff it is still a handheld and not a professional camera.
Comparing fcam-api using fcamera or hdr or lowlight with native cam app complaints I had to test it myself, super lowlight conditions (only my monitor was on) I took several pictures of my right control key and best to worst was native cam with flash, lowlight, fcamera, HDRcapture and last native cam without flash. To give a clear picture I should have kept the shots (maybe I can do some again tonight) but just to tell, the flashed shots are well flashed but the lowlight-app and fcamera-app shots were just the scene a bit grainy but amazing shots for that light condition, HDRcapture made a pattern out of 3 where you were able to see the patterns edges repeating and last the native without flash... black picture where you could guess what is shown but nothing more.
My handycam got a realy nice feature which is just another kind of processing or kind of taking a picture. Some like lowlight but with flash, it flashes takes a picture with short shutter time and then it takes another with longer exposure and puts them together. Notice the opaque hand as it was moved for the second picture.
http://dostortugas.org/gallery/galle....090517_13.JPG
__________________
useful links for newcomers: New members say hello, New users start here, Community subforum, Beginners' wiki page, Maemo5 101, FAQ
useful links for newcomers: New members say hello, New users start here, Community subforum, Beginners' wiki page, Maemo5 101, FAQ
|
|
2010-07-29
, 16:52
|
|
Posts: 356 |
Thanked: 172 times |
Joined on Jan 2010
@ Canada
|
#104
|
Originally Posted by romanianusa
That's because Photoshop's "Merge to HDR" feature is far from automatic. It is however very powerful if you know a bit about how to use it..
NO need to buy CS5..i got it FREE..hehe. Anyway, when i used the merge to HDR in CS5 with the raw images or the jpg from N900 gallery..it doesn't look nice like the one captured by the standalone HDR Capture. I try playing with different tuning and settings...it's just never the same and in situation, it's worst. I got 3 pics that i combined....and it didn't look good to me.
Here are a few basic tips:
- When you first load a set of images into Merge to HDR, you are presented with a dialogue showing your source images down the left, a large preview window in the centre, and a histogram over on the right. The only thing to really fiddle with here is turning on/off the various source images to see their effect on the HDR image. Otherwise, just click OK in this dialogue -- adjusting other settings here isn't really very productive.
- After a moment, you'll be presented with your HDR image. You'll notice it basically looks like crap. This is because it's a 32-bit-per-channel image, and your monitor almost certainly cannot show you all the image data it contains accurately. You need to down-sample it to 16-bits-per-channel.
- Go to Image --> Mode --> 16 Bits/Channel... you'll have a new dialogue. The pick-list at the top allows you to specify a variety of methods for Photoshop to use in the conversion. If you select "Local Adaptation", you'll be able to adjust the curves directly from this dialogue (you may need to expand the dialogue to display the tone/histogram controls). Pick whichever looks best -- I generally find either Highlight Compression or Local Adaptation to give me the initial results I'm looking for, but YMMV. Adjust the curves if applicable (you can always do this later), and click OK.
- Now you will have a 16-bit-per-channel image that is much MUCH easier to work on. It still contains an enormous amount of image data, so filesize will be quite large, and adding layers can really add up quickly, so keep that in mind if your machine is short on RAM. You can, however, do some very dramatic edits to the image without breaking up the histogram, which is the beauty of working in high-bit.
- Tone the image any way you see fit until you've got something you're happy with. Don't forget that you'll need to down-sample it again to 8-bits-per-channel if you want to save it as a JPEG
There's obviously MUCH more to know about all of this, but I hope this helps with getting an HDR image that looks the way you imagined it should in the first place
|
|
2010-07-29
, 18:21
|
|
Posts: 739 |
Thanked: 114 times |
Joined on Sep 2009
|
#105
|
Originally Posted by Bingley Joe
I don't know what version of Photoshop you're using, I didn't have to go through the step of down-sampling to 16bit from 32bit. It already present me with 16-bit per channel HDR. And all I have to do then is fiddle with the tools such as ghosting, details...Local adaptation...ect.. But no matter what i do....i am not getting the same result that i am getting from HDR Capture app itself. In fact, i even took out the 3 same input images from the phone and use photoshop to see if it give the same result that the phone give me....NOPE. Not the same fine print like the phone give me.
That's because Photoshop's "Merge to HDR" feature is far from automatic. It is however very powerful if you know a bit about how to use it..
Here are a few basic tips:
- When you first load a set of images into Merge to HDR, you are presented with a dialogue showing your source images down the left, a large preview window in the centre, and a histogram over on the right. The only thing to really fiddle with here is turning on/off the various source images to see their effect on the HDR image. Otherwise, just click OK in this dialogue -- adjusting other settings here isn't really very productive.
- After a moment, you'll be presented with your HDR image. You'll notice it basically looks like crap. This is because it's a 32-bit-per-channel image, and your monitor almost certainly cannot show you all the image data it contains accurately. You need to down-sample it to 16-bits-per-channel.
- Go to Image --> Mode --> 16 Bits/Channel... you'll have a new dialogue. The pick-list at the top allows you to specify a variety of methods for Photoshop to use in the conversion. If you select "Local Adaptation", you'll be able to adjust the curves directly from this dialogue (you may need to expand the dialogue to display the tone/histogram controls). Pick whichever looks best -- I generally find either Highlight Compression or Local Adaptation to give me the initial results I'm looking for, but YMMV. Adjust the curves if applicable (you can always do this later), and click OK.
- Now you will have a 16-bit-per-channel image that is much MUCH easier to work on. It still contains an enormous amount of image data, so filesize will be quite large, and adding layers can really add up quickly, so keep that in mind if your machine is short on RAM. You can, however, do some very dramatic edits to the image without breaking up the histogram, which is the beauty of working in high-bit.
- Tone the image any way you see fit until you've got something you're happy with. Don't forget that you'll need to down-sample it again to 8-bits-per-channel if you want to save it as a JPEG
There's obviously MUCH more to know about all of this, but I hope this helps with getting an HDR image that looks the way you imagined it should in the first place
You mention down-sampling to 8bit before saving to jpeg, it didn't matter i save it under 16-bit and it looks no different to me. Size is about the same as the phone would give me 1.2mb.
You say, "when you first load a set of images into Merge to HDR, you are presented with a dialogue showing your source images down the left, a large preview window in the centre, and a histogram over on the right." Well, when i first load a set of images Merge to HDR, i am presented with the EV manual dialogue settings. This is the part i don't understand because i expect the program to know all the values and calculate by itself. I expect it just load these 3 images and merge them into HDR.
Don't get me wrong...the HDR that photoshop produce is okay but not the one i expected. I can not get the fine smoothness of the one you get from HDR Capture on N900. The one produce in Photoshop, i feel like i can just take one picture (rather than 3) and fine-tune it and still come out the same as with 3 picture merged. That's why i am not impressed with the HDR in Photoshop.
Last edited by romanianusa; 2010-07-29 at 18:30.
|
|
2010-07-29
, 20:48
|
|
Posts: 356 |
Thanked: 172 times |
Joined on Jan 2010
@ Canada
|
#106
|
Erm.. ok, probably going to get well into tl;dr territory on this one, but here goes..
Doing that conversion later affords you the extra control over the method Photoshop will use to down-sample the image, which is why I recommend simply clicking OK.
To the best of my knowledge, this is the way it works for Photoshop CS3 through CS5, although I'm at work right now where I only have CS3 to confirm this with.. so maybe I'm forgetting how CS5 handles it.
The same theory should still apply even if the controls have moved a bit though.
You wouldn't.. that was kind of the point I was trying to make about using Photoshop to make HDRs -- you can get amazing results from it, but you will need to work on the image quite a bit to get it where you want it to be. You almost always need to work the curves for a while to get the tones to look right, for example.
The HDR capture app does a very good job of blending the source images into something that looks natural. If you like the results, why fiddle with Photoshop at all?
This is where I start to get confused because I was under the assumption that you were already trying to use Photoshop to merge the 3 input images from the camera..
Further confusion here. You can not save a JPEG in 16bit/channel. That mode is not supported by the JPEG file format, so I'm not sure what's happened to your file along the way, but you must have converted it to 8bit/channel at some point.
Regardless, there really shouldn't be any visible difference between the 16 and 8 bit/channel images -- in the simplest terms, the difference is more of a mathematical one in that the 16bit/channel version allows for a much higher degree of precision in the calculations involved in image adjustment, and therefore a much smaller degree of data-loss due to rounding. Any data-loss that does occur is well within the 8-bit space your monitor uses, and so you'll never see it.
BUT -- that greater flexibility is why I didn't recommend down-sampling to 8bit/channel until the very end of your editing process when all you want to do is pump out a JPEG to put on Flickr or something.
This almost sounds to me like it's loading the JPEGs into Camera RAW first, but like I say -- I haven't got CS5 handy right now to verify what's going on.. sorry. I'll give it a look when I get home and see if I can offer some insight here.
Yeah the HDR app on the N900 does a really good job all on its own, and the results look very natural, which is something I find lacking in a lot of HDR images I see.
I happen to find that Photoshop can be used to produce some truly spectacular results from images merged to HDR, but that doesn't mean you have to
HDR Capture only gives you JPEGs to work with anyway, so if you like the results you get with the merged image straight out of the camera, I don't see any reason to muck around in Photoshop at all..
Originally Posted by romanianusa
You can indeed choose 16-bit-per-channel in the initial Merge To HDR dialogue you're presented with (the pick-list is right under the histogram. You could even go straight to 8-bit if you want to, but it's not really a good idea).
I don't know what version of Photoshop you're using, I didn't have to go through the step of down-sampling to 16bit from 32bit. It already present me with 16-bit per channel HDR.
Doing that conversion later affords you the extra control over the method Photoshop will use to down-sample the image, which is why I recommend simply clicking OK.
To the best of my knowledge, this is the way it works for Photoshop CS3 through CS5, although I'm at work right now where I only have CS3 to confirm this with.. so maybe I'm forgetting how CS5 handles it.
The same theory should still apply even if the controls have moved a bit though.
And all I have to do then is fiddle with the tools such as ghosting, details...Local adaptation...ect.. But no matter what i do....i am not getting the same result that i am getting from HDR Capture app itself.
The HDR capture app does a very good job of blending the source images into something that looks natural. If you like the results, why fiddle with Photoshop at all?
In fact, i even took out the 3 same input images from the phone and use photoshop to see if it give the same result that the phone give me....NOPE. Not the same fine print like the phone give me.
You mention down-sampling to 8bit before saving to jpeg, it didn't matter i save it under 16-bit and it looks no different to me. Size is about the same as the phone would give me 1.2mb.
Regardless, there really shouldn't be any visible difference between the 16 and 8 bit/channel images -- in the simplest terms, the difference is more of a mathematical one in that the 16bit/channel version allows for a much higher degree of precision in the calculations involved in image adjustment, and therefore a much smaller degree of data-loss due to rounding. Any data-loss that does occur is well within the 8-bit space your monitor uses, and so you'll never see it.
BUT -- that greater flexibility is why I didn't recommend down-sampling to 8bit/channel until the very end of your editing process when all you want to do is pump out a JPEG to put on Flickr or something.
You say, "when you first load a set of images into Merge to HDR, you are presented with a dialogue showing your source images down the left, a large preview window in the centre, and a histogram over on the right." Well, when i first load a set of images Merge to HDR, i am presented with the EV manual dialogue settings. This is the part i don't understand because i expect the program to know all the values and calculate by itself. I expect it just load these 3 images and merge them into HDR.
Don't get me wrong...the HDR that photoshop produce is okay but not the one i expected. I can not get the fine smoothness of the one you get from HDR Capture on N900. The one produce in Photoshop, i feel like i can just take one picture (rather than 3) and fine-tune it and still come out the same as with 3 picture merged. That's why i am not impressed with the HDR in Photoshop.
I happen to find that Photoshop can be used to produce some truly spectacular results from images merged to HDR, but that doesn't mean you have to
HDR Capture only gives you JPEGs to work with anyway, so if you like the results you get with the merged image straight out of the camera, I don't see any reason to muck around in Photoshop at all..
| The Following User Says Thank You to Bingley Joe For This Useful Post: | ||
|
|
2010-07-30
, 01:00
|
|
Posts: 739 |
Thanked: 114 times |
Joined on Sep 2009
|
#107
|
Originally Posted by Bingley Joe
It takes awhile to process under N900...that's why i try to find an alternative.
HDR Capture only gives you JPEGs to work with anyway, so if you like the results you get with the merged image straight out of the camera, I don't see any reason to muck around in Photoshop at all..
| The Following User Says Thank You to romanianusa For This Useful Post: | ||
|
|
2010-07-30
, 02:32
|
|
Posts: 515 |
Thanked: 259 times |
Joined on Jan 2010
|
#108
|
Does any know how to make fCam the default camera for the N900? I'd like to have all hardware actions / events trigger fCam instead of the built-in camera app. Thanks.
|
|
2010-07-30
, 03:38
|
|
Posts: 31 |
Thanked: 35 times |
Joined on Jun 2010
|
#109
|
I have played with fcam a bit and think this application has a lot of potential, but I have a few comments
Delay Timer. Badly needed.
The FCam app needs a delay timer badly, especially with all the newly unlocked abilities to do long shutter speeds and multiple exposures that need to be exactly aligned. Pushing the shutter button moves the camera ever so slightly so it is difficult to avoid camera shake.
I would like to have the standard 2s/10s delay shutter, so I can focus and trigger the shutter, then take my hands off the shutter and keep it stable so it can take pics undisturbed.
As a more advanced option I would propose a accelerometer-triggered shutter, I press the shutter but the camera waits for the shake to stop first.
Histogram
There is really no reason to have the histogram stretch across the entire image. A small icon-sized histogram would be sufficient (like in normal cameras).
Smart Focus
Accelerometer-aided focus: if the camera is pointed in the same direction and has focused once then is moved forward or backwards relative to the sensor plane, it should add the movement to the focus distance before focusing again.
Last edited by wotevah; 2010-07-30 at 03:40.
Delay Timer. Badly needed.
The FCam app needs a delay timer badly, especially with all the newly unlocked abilities to do long shutter speeds and multiple exposures that need to be exactly aligned. Pushing the shutter button moves the camera ever so slightly so it is difficult to avoid camera shake.
I would like to have the standard 2s/10s delay shutter, so I can focus and trigger the shutter, then take my hands off the shutter and keep it stable so it can take pics undisturbed.
As a more advanced option I would propose a accelerometer-triggered shutter, I press the shutter but the camera waits for the shake to stop first.
Histogram
There is really no reason to have the histogram stretch across the entire image. A small icon-sized histogram would be sufficient (like in normal cameras).
Smart Focus
Accelerometer-aided focus: if the camera is pointed in the same direction and has focused once then is moved forward or backwards relative to the sensor plane, it should add the movement to the focus distance before focusing again.
Last edited by wotevah; 2010-07-30 at 03:40.
| The Following User Says Thank You to wotevah For This Useful Post: | ||
 |
|
2010-07-30
, 03:46
|
|
Posts: 3,159 |
Thanked: 2,023 times |
Joined on Feb 2008
@ Finland
|
#110
|
Originally Posted by wotevah
what are you talking about? fcam is the API!
I have played with fcam a bit and think this application has a lot of potential, but I have a few comments
Delay Timer. Badly needed.
The FCam app needs a delay timer badly, especially with all the newly unlocked abilities to do long shutter speeds and multiple exposures that need to be exactly aligned. Pushing the shutter button moves the camera ever so slightly so it is difficult to avoid camera shake.
I would like to have the standard 2s/10s delay shutter, so I can focus and trigger the shutter, then take my hands off the shutter and keep it stable so it can take pics undisturbed.
As a more advanced option I would propose a accelerometer-triggered shutter, I press the shutter but the camera waits for the shake to stop first.
Histogram
There is really no reason to have the histogram stretch across the entire image. A small icon-sized histogram would be sufficient (like in normal cameras).
Smart Focus
Accelerometer-aided focus: if the camera is pointed in the same direction and has focused once then is moved forward or backwards relative to the sensor plane, it should add the movement to the focus distance before focusing again.
you probably mean fcamera?
camera shake - don't shoot in so murky places. or choose best of 8 ....
histogram- the bigger the better.
smart focus- and how in hell did you think that you can measure distance with accelometer anyway?
__________________
Want to know something?
K.I.S.S. approach:
wiki category:beginners. Browse it through and you'll be much wiser!
If the link doesn't help, just use
Google Custom Search
Want to know something?
K.I.S.S. approach:
wiki category:beginners. Browse it through and you'll be much wiser!
If the link doesn't help, just use
Google Custom Search
Don't get me wrong, I'm not a fanboy of Adobe, but I certainly do enjoy working within their application suites, albeit in a Micro$haft Windoze environment, and when you get to learn the applications inside-out you realise the incredible effects you can get out of them.
Most people on deviantART work within these applications and put out incredible works, which just goes to show what the benchmark actually is.....