Hi all,
please have a look onto the given problem. I tried to generate a profile for the Casio Exilim P700 digicam. By this, I found out the ISO level is not shown by NeatImage. This leads to confusion during batch profile generation for this digicam. If you want, I can provide you with the images. BTW the Casio tool for downloading the images to the PC shows the ISO info correctly
Well, this is another example of a camera manufacturer producing a product that does not follow the recommendations of the EXIF specifications. Instead of saving the ISO value in the standard tag in the EXIF data fields, this camera encodes it inside the manufacturer's proprietory makernotes field. The immediate result is that most third-party software tools (e.g., Photoshop) will not see the ISO value.
So this is a bug, but a bug in the camera firmware. I recommend to ask the camera manufacturer to update the firmware to make it write the ISO value to the standard tag suggested by the EXIF specifications.
Before the bugfix is implemented, you can use Auto Profile feature in Neat Image. It will take a bit longer to process each image but will help to get round that firmware bug.
Hi Vlad,
can you provide some details (bits, bytes, URLs) ? Casio UK didn't respond on my yesterday email yet. I think they will do nothing for firmware update. I'm a little bit experienced in C/C++ and perl programming. So I want to make a litte tool for myself, but I didn't find the Casio specific exif information (i.e. directories, offsets - you know, the ISO is the most important part for me).
Thank you in advance,
Karl
NITeam wrote:...Instead of saving the ISO value in the standard tag in the EXIF data fields, this camera encodes it inside the manufacturer's proprietory makernotes field. ...
To see that the ISO value is indeed saved in a non-standard location, try this free tool.
To work with the Casio makernotes, start with this classic page. There is not much but it is good start in my opinion. Unfortunately I don't have more specific details on the level of bytes.
There must be more resources regarding Casio camera metadata in the Internet but I don't have direct links at the moment. If you find anything please let me know.
Hi Vlad,
I'm back again with a bug fixing tool for casio p700 post-processing. My tool allowes to repair the images (jpeg as well as tiff) after transfer to the PC.
The key idea is nothing new as retrieving the CCD ISO Sensitivity Information from the casio makernote section and put it into the common Exif section. For keeping this tool as easy as possible I avoided the full insertion of the ISO equivalent tag. Doing this would cause me te re-calculate a lot of value offset inside the jpegs/tiffs Exif section. So I looked for an un-importand Exif tag with matching type format for overwriting purposes- and I found the Exposure program tag. So this information (from my point of view redundant anyway) gets lost during image processing.
The tool itself is C in full. I found a very nice reference under this URL (and I re-used it for my approach - last but not least thanks a lot to the originator): http://www.sentex.net/~mwandel/jhead/
I can provide my tool on request from any interest side (Vlad ?).
Regards,
Karl
NITeam wrote:...
There must be more resources regarding Casio camera metadata in the Internet but I don't have direct links at the moment. If you find anything please let me know.
Thank you for the update and links to makernote documentation. I will check the documentation first to see if we can directly use it to read the ISO from the makernotes using tools currently available in NI. I guess it shouldn't be difficult. I will contact you if we have any trouble with this.
when I tried to re-create the profiles with the new NeatImage Version, the effect (I know it from older versions too, but I didn't had the focus) of 100% CPU raised again. This is very incomfortable, since the PC is completely blocked for app. 10 minutes (right, the procedure of processing 60 images of 4MByte image size takes a long time). Do you know this effect or is there anything wrong with my installation ?
You may try to decrease the priority of the Neat Image process using Windows Task Manager during this long profiling process. This should make Windows more responsive.
However, it is pretty strange anyway. - When I converted a large batch of profiles recently, my computer worked just fine with other applications.
Which Windows type do you run? Which priority is given to Neat Image in its options (Tools | Options... | Filtration | Filter process priority)?
follow the facts:
o ASUS A7N8X Mainb. plus 3000XP plus 2GB RAM plus SATA HDs
o Windows XP prof.
o NeatImage Filtration Prio is idle (I started with 'below', went to 'idle' now)
o NI Windows task prio. changed to lowest possible
o almost no further application running (only WEB browsing is possible - seames to be consuming less resources only; no chance for office or other image processing with adobe tools)
I restarted NI after each filtration options change. I remember this effect from further installation (as well the demo version).
I did the job with the large image sized Casio files as well as for the smaller Nikon files - with the same result. No change of observation - the NI claimes the CPU with 100% (memory is below 500MBytes - no problem). In the Casio case also the cross bar update stopped - only the explorer did show the progess by providing further dnp files. In all cases the procedure comes to end - but the PC is under water completely.
Any idea ? However, it's really not a totaly blocking issue, since profiling works in general (isn't an every day task as well) ...
The configuration looks very good. There shouldn't be any problem for Windows XP to do nice multi-tasking when Neat Image is set on normal, below normal, or low priority in Task Manager.
Neat Image is doing batch profiling now as I am writing. It runs on normal priority, all defaults. It does take ~95-99% of CPU time as reported by Task Manager (which is normal because profiling is CPU intensive) but only if nothing else is using CPU time, in that case NI CPU utilization goes down. When other programs need time, they get it.
So, the effect you describe is not what I see on my computer. Probably there is something else that makes your computer behave like that. Do you have any other PC to test NI profiling? - I will try to run Batch Profiler on all our PCs here to make sure it works correctly as well.
I did some testing this morning with another PC (ECS K7S5a, 1800XP, RAM 256MB, IDE HD system). The behaviour was almost the same, when I started profile creation on large camera image data: 100% CPU, no NI GUI update (NI windows task responds "no reaction of this window ..."). NI filtering options are on no effect. Unfortunately there is no tooling for task state analysis available for windows. I'm very familier with this stuff from solaris/linux side. You know truss, strace, top, vmstat, gcore and all these thinks from Unix ... Actually it's difficult for me to get more knowledge on what happens here.
A significant impact comes from networking as far as I saw: Whenever I tried to look for other PCs of my home network (tcp based windows network) during profiling, the GUIs (NI as well as win explorer) did hang completely with white screens over minutes. Windows tasking went on normaly after finishing NI profiling job. But this is another story.
Actually I can live with this phenomenon. Next days I will refresh installation of my "work station" PC. May be it will clean up some think. I let you know.
This is more or less normal. NI does need a lot of CPU to do profiling (and it does not pay too much attention to redrawing the Batch Profiler window during profiling) but my point is that Windows is normally able to very nicely multi-task active NI and other programs. Yes, NI takes CPU time but it is running on normal (not higher than any other application) priority so Windows should be able to nicely share CPU resources with other programs. It does this well on our computers here. No locking of the whole system, no mouse freezing. It all works very nicely. "the PC is completely blocked" - this is something I have never seen with NI on our computers. Must be something with your computers or Windows setup. Please let me know when you reinstall it.
I am very interested in getting a copy of the utility you have written. I am sure it would also be quite popular with the folks in the Casio forum at DPReview.com. I am happy to host the file for you if you would like to share it with the online community, or otherwise you could setup your own free account at a site like topcities.com or fortunecity.com etc.
My email is sotterraneo@mail.com and I would appreciate very much if you could email me your utility for fixing the ISO info in the EXIF of JPEGs created by Casio cameras (I have an Exilim Z750)
Thanks
Elia
Karl wrote:Hi Vlad,
I'm back again with a bug fixing tool for casio p700 post-processing. My tool allowes to repair the images (jpeg as well as tiff) after transfer to the PC.
The key idea is nothing new as retrieving the CCD ISO Sensitivity Information from the casio makernote section and put it into the common Exif section. For keeping this tool as easy as possible I avoided the full insertion of the ISO equivalent tag. Doing this would cause me te re-calculate a lot of value offset inside the jpegs/tiffs Exif section. So I looked for an un-importand Exif tag with matching type format for overwriting purposes- and I found the Exposure program tag. So this information (from my point of view redundant anyway) gets lost during image processing.
The tool itself is C in full. I found a very nice reference under this URL (and I re-used it for my approach - last but not least thanks a lot to the originator): http://www.sentex.net/~mwandel/jhead/
I can provide my tool on request from any interest side (Vlad ?).
Regards,
Karl
NITeam wrote:...
There must be more resources regarding Casio camera metadata in the Internet but I don't have direct links at the moment. If you find anything please let me know.
