[sane-devel] Backend for plustek Opticbook 3600
Hey stef, I see things seem to be progressing well with the lide 110, good times! I have recently had a problem with the account I set up the mailing list with so I have added this account instead. I doubt there will have been any progress on the plustek drivers, but as I've missed some of the emails I wanted to check in with you. Any updates? Chris
[sane-devel] Backend for plustek Opticbook 3600
Chris, please go ahead. I think you could remove some of my Chris why comments where you know. Should the scanner be added also to the genesys.conf Definitely here and libsane.rules files? I have never entered in this file, what was the reason behind adding to it? Or at least what does it do? Thanks for the support you have given here, I think we can realistically look at getting this committed within a fortnight, not sure when the next official release is, hopefully it is scheduled for entry into Fedora 15, im sure their feature freeze is a while off. Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Thanks Stef, a week is no problem for me, i'm pretty busy at work at the minute so a weeks time might work out best for all of us. Chris On 06/12/10 19:53, stef wrote: Le Monday 06 December 2010 15:54:44 Chris Berry, vous avez ?crit : Chris, please go ahead. I think you could remove some of my Chris why comments where you know. Should the scanner be added also to the genesys.conf Definitely here and libsane.rules files? I have never entered in this file, what was the reason behind adding to it? Or at least what does it do? Thanks for the support you have given here, I think we can realistically look at getting this committed within a fortnight, not sure when the next official release is, hopefully it is scheduled for entry into Fedora 15, im sure their feature freeze is a while off. Chris Hello, the mail was sitting in my inbox. I am quite busy currently finishing LiDE 110 support, and I planned to review it thoroughly in about a week. Is it OK ? Regards, Stef -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20101206/e7ce502d/attachment.htm -- next part -- An embedded and charset-unspecified text was scrubbed... Name: not available URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20101206/e7ce502d/attachment.asc
[sane-devel] Backend for plustek Opticbook 3600
Cheers Allan, Stef said hes gonna give the code a thorough look through in a week or so, should be plenty of time for the next release if it is around your proposed time. Thanks Chris On 06/12/10 15:08, m. allan noah wrote: On Mon, Dec 6, 2010 at 9:54 AM, Chris Berrys0457957 at sms.ed.ac.uk wrote: Chris, please go ahead. I think you could remove some of my Chris why comments where you know. Should the scanner be added also to the genesys.conf Definitely here and libsane.rules files? I have never entered in this file, what was the reason behind adding to it? Or at least what does it do? It controls udev permissions. These files (and their hal/policyKit/et al equivalents) are generated from the backends .desc file. Thanks for the support you have given here, I think we can realistically look at getting this committed within a fortnight, not sure when the next official release is, hopefully it is scheduled for entry into Fedora 15, im sure their feature freeze is a while off. We'll leave it to the authors of the existing backend to OK the code changes. I'm leaning toward doing the next release at the end of January, I'll post a proposed timeline shortly. allan -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20101206/ba1c3abf/attachment-0001.htm -- next part -- An embedded and charset-unspecified text was scrubbed... Name: not available URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20101206/ba1c3abf/attachment-0001.asc
[sane-devel] Backend for plustek Opticbook 3600
Hi Stef, I take it that you have now received my preliminary copy of my report. I will send you a completed copy when I am done. I have a question about the image scans at 1200+ dpi. When attempting scans at these resolutions I only get half an image. A little investigation has led me to find that the max dpi register is set to 1200 (from windows driver) BUT i still get half an image when half_ccd is set to true *and* false when i do a scan. As I am very close to handing in my report any information you can give me would be helpful if I can include it :) Thanks again for all the help you have given me with this, i definitely owe you a beer or two :) Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Next release update
I know this was discussed about 6 weeks ago but is there any update on when the next SANE release will be? Would be nice to know a rough date or at least if progress is being made Thanks Chris
[sane-devel] Backend for plustek Opticbook 3600
stef wrote: Le lundi 22 mars 2010 07:26:23 Chris Berry, vous avez ?crit : Hi guys, I have included a patch for the plustek 3600. It isnt quite there yet but at least you can apply it to the current development version and should be able to test (for those of you with the scanner) to see how it goes at this stage. I have also included Pierre and Stef on this if they would like to have a look at the revisions i have made in order to add support for this scanner. Thanks for your help with this project Chris Hello, it'll take me some days to review correctly. I'll let you know by the end of this week. Regards, Stef Cheers Stef, I tried scanning at 600dpi (which is highest available on plustek application) so I thought that was full sensor resolution and it resulted in no difference of the strange colours. I entered a custom size into the windows version of the driver (tried 1200dpi and 2400dpi) and both work but im not sure if that is hardware or software interpolated so I dunno what the max resolution is for this. I tried 1200 dpi with the sane driver and got a half image (like it needs half ccd but it is enabled). Either way I have made no head way with correcting the shading. Any ideas what im doing wrong? Thanks Chris
[sane-devel] Plustek OpticPro S28
Le jeudi 18 mars 2010 07:21:10 Yevgeny Gromov, vous avez ?crit : Hi everyone! In archive I've read that some years ago someone have developed backend for Plustek OpticPro S28 based on GL841. Is there any alpha support for this scanner or should I write modification of sane-genesys based on different GL841 models? -- sane-devel mailing list: sane-devel at lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/sane-devel Unsubscribe: Send mail with subject unsubscribe your_password to sane-devel-request at lists.alioth.debian.org Hello, I think the latter is the best option for you. All the mails about the OpticBook 3600 development will give you many interesting information on how to do it. Regards, Stef Hi Yevgeny, I have been developing support for the opticbook 3600 with the help of stef so if you would like any information I will do my best to help you, hopefully some of the difficulties I faced will be of benefit to you. If you are looking for advice on how to start the first thing you need to do is get USB snooping software and some scripts to decode it (I have a copy of some scripts from stef but if you ask him nicely he may have newer versions :) ). USB snoops will show the traffic between the scanner and the computer, after the captured snoop has been parsed by the script files a lof of vital information will be available. Good luck and if you would like more information on the 3600 for comparison the project site is here: http://sites.google.com/site/bez625/ Chris
[sane-devel] Backend for plustek Opticbook 3600
stef wrote: Hello, for the image quality, I think you could concentrate on on full sensor resolution and 'half ccd' resolution. I'm not sure what you mean but I guess full resolution is the maximum hardware resolution of the scanner? If so then what is 'half ccd' resolution? In these modes AVEENB is of no influence since no averaging or deletion occurs. Then you can adjust the offset to be used for shading data. To find the correct offset some trial and error is needed. You'll have to change it's value until bright vertical lines match the closest dark vertical line in scans (you could scan only the white cover of the scanner, lines are more visible). You may try first the offset values from other GL841 scanners. So I just do scans at full resolution of the white background and alter the shading offset until it evens out? Should be no problem Once done, you may choose to keep averaging, but in this case you'll have to tune compute_coefficients to do the same averaging than the planar method used for the CCD_CANONLIDE35. Is this method you are talking about the large function added under canonlide35 in the send_shading_coefficients? If so I will try if i get time but as my hand in date is coming up I may not get time... Regards, Stef I should have the patch sent out to you by the end of the day, i'll include pierre, you and the people who also have the scanner for a bit of testing. Also does this mean I can add my name to the developer list at the top of the genesys files :) ? Thanks Chris
[sane-devel] Backend for plustek Opticbook 3600
Hello, in order to properly review the result of your hard work, we need the changes you done as patches. Either against the version you started to modify, or as a git patchset against latest version, which is the best solution. On your site, I only found the full sources, may you have a look into it ? Regards, Stef Ok Stef I will hopefully have a patch to you by the end of the weekend. Looking through the git changes and such I really feel the best thing for me to do is get the latest version, start over adding things back in from my current version and then I can have one commit of just my changes and not any other rubbish (white space etc). As for the quality of the scans is there anything you can advise on maybe what I need to do to improve it? I have been a ill this week and had other things to work on but after tomorrow i'm back on this project full time (there is about a week left until I hand in my written report so hopefully I can get some testing done next week to include in the report). Thanks Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
On 03/09/2010 05:15 AM, stef wrote: Le lundi 8 mars 2010 17:20:23 Chris Berry, vous avez ?crit : Ok Stef I have a bit more time to concentrate on this at the minute. I have updated the white_shading.pnm and included it at http://sites.google.com/site/bez625/updates-1/update080310 there is also a copy of the latest scan there too. I only have 20 shading lines in the shading area because there is a lot of blue in that area. Should I widen it (more shading lines) or should those blue areas be included? Also how do I disable the averaging? Is it a software thing or is it on one of the registers? I'm also not too sure what you mean about the offset, is it the DAC offset? Thanks Chris Hello, the white shading picture must be all white. The picture you posted is mostly OK, but there is still some blue on the right. This is not a blocker, but it will certainly have to be fixed to reach best image quality. By averaging, I was thinking of the AVEENB bit of register 3 which tells the ASIC to delete extra data points or average them to produce data from higher dpi. Regards, Stef Hey guys, I have experimented with disabling average and altering the offset as Stef suggested. Disabling averaging seems to make things worse (dark and light lines where solid colour appears with AVEENB) and changing the offset seems to make no real difference so i'm not sure what is happening. Also I am unsure about what target code I should use for the shading compute_coefficients (assumed 0xdc00 as it seems popular). At the project site (update ) there are a few images in a bzip file to show what I mean about toggling the AVEENB. It certainly seems better to have it on but this might just be misleading images. The main issue now seems to be the different coloured vertical lines down the document. In addition I performed higer dpi scans (300, 600) and the whole image seems darker and the vertical lines are more prominent, i cant say im entirely sure whats causing them but im guessing its the shading function. I have also attached my source on the project update if this will make it easier to advise on what im doing wrong. I will get around to tidying up my code soon so i can commit to git without it looking messy. Has there been anymore information about a next release date? I have been skimming the mailing list and not seen anything mentioned. Thanks Chris -- next part -- An HTML attachment was scrubbed... URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100316/a35b0887/attachment.htm -- next part -- An embedded and charset-unspecified text was scrubbed... Name: not available URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100316/a35b0887/attachment.asc
[sane-devel] Backend for plustek Opticbook 3600
On 03/03/2010 06:03 AM, stef wrote: Le mardi 2 mars 2010 16:40:21 Chris Berry, vous avez ?crit : Hey guys, I resolved the issue I was having with the dark_shading calibration. The lamp was turning off fine but the | REG03_LAMPPWR in gl842_begin_scan was turning the light back on for the scan :/ . Having resolved this there is still a black line down the center of the page. The black_shading.pnm file is now completely black and the white_shading.pnm file is mostly white with some dark edges. Any thoughts on how I sort this out now? Is it related to the number of sensor pixels or the shading area or something? Chris Hello, the black patterns are often used to detect the exact geometry of the scanner. What is reliable is these black/white areas, while the casing may vary. For instance the HP2400/HP2300 and HP3670 have a black corner which is aligned on the exact vertical margin of the visible scan area.So the dark edges may be harmless if they are outside the effective scan area used for final scan. For the big black vertical line (which is a third of the width of the picture), it seems that shading coefficients sent aren't send in the correct format. It's hard to tell without the code, but the results looks like 'planar' format (for CIS) was used when 'chunky' format (for CCD) was needed. Surely an issue in the genesys_send_shading_coefficient() where compute_coefficient() is for chunky, compute_planar_coefficient() is for planar and the CCD_CANONLIDE35 case looks for the planar case to me. Pierre may confirm this last point. Regards, Stef Ok thanks guys, this seems to have done the trick! There is now a black line down one side of the page (not sure if that is supposed to be there or not) and the colours seem to be a bit grainy. I have attached photos to the site page (http://sites.google.com/site/bez625/updates-1/update040310), these show the shading calibration and the latest current scan. If possible could you take a look and advise on how to sort out the grainy parts of colour, im completely unsure myself :) Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Hey guys, I resolved the issue I was having with the dark_shading calibration. The lamp was turning off fine but the | REG03_LAMPPWR in gl842_begin_scan was turning the light back on for the scan :/ . Having resolved this there is still a black line down the center of the page. The black_shading.pnm file is now completely black and the white_shading.pnm file is mostly white with some dark edges. Any thoughts on how I sort this out now? Is it related to the number of sensor pixels or the shading area or something? Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Hey Stef, I think I have sorted out the white shading data but the black is still a mystery. I have uploaded some files to the project site (here http://sites.google.com/site/bez625/updates-1/update240210 ) these are at shading lines set to 400 but for the scan to work even a little bit I needed to change the lines to 20 (this removes the light blue lines at the bottom of the white_shading.pnm). As you will be able to see the black_shading400.pnm is all white, I have no idea how to alter this. You can see from the white_shading400.pnm the full calibration area so I am at a loss really. Have you encountered this problem before? Thanks Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
On 02/24/2010 10:33 AM, Pierre Willenbrock wrote: stef schrieb: Le mercredi 24 f?vrier 2010 03:00:16 Chris Berry, vous avez ?crit : Hey Stef, I think I have sorted out the white shading data but the black is still a mystery. I have uploaded some files to the project site (here http://sites.google.com/site/bez625/updates-1/update240210 ) these are at shading lines set to 400 but for the scan to work even a little bit I needed to change the lines to 20 (this removes the light blue lines at the bottom of the white_shading.pnm). As you will be able to see the black_shading400.pnm is all white, I have no idea how to alter this. You can see from the white_shading400.pnm the full calibration area so I am at a loss really. Have you encountered this problem before? Thanks Chris Hello, judging from the 2 pictures, there is no black area to scan. You should then rather use the GENESYS_FLAG_DARK_WHITE_CALIBRATION (like the lide 35) flag instead of GENESYS_FLAG_DARK_CALIBRATION. The shading data will be taken from an all white area and black data will be inferred from the small black areas still in the picture. GENESYS_FLAG_DARK_WHITE_CALIBRATION is just a way to scan the black and white strip in one go. Chris, you should try to figure out if the backend goes through genesys_dark_shading_calibration with GENESYS_FLAG_DARK_CALIBRATION, and if so, why the lamp does not switch off. Regards, Pierre Hi Pierre / Stef, I was wondering the same thing but the lamp does indeed go off. I'm not sure why the scan is so light, I had no idea which area it was scanning from I have been playing with the dark_white calibration and i seem to be getting somewhere, here ( http://sites.google.com/site/bez625/updates-1/update240210lateron ) are some scans. Should I abandon the work with DARK_WHITE and see if I can fix the issue with DARK_CALIBRATION? Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
stef wrote: Le lundi 22 f?vrier 2010 15:09:11, vous avez ?crit : On 02/22/2010 06:43 AM, stef wrote: Le lundi 22 f?vrier 2010 02:09:14 Chris Berry, vous avez ?crit : ... Hey Stef, Thanks for the awk script, helped a lot. I put the values into a spreadsheet and got gamma values all around 1.0 so I didnt really need to alter them at all, but it gives me something to talk about in my report :). I have tweaked the gain and offset (offset is largely ignored for AD type frontend it seems so I added a similar AD_OFFSET_CALIB function to gl841) and now the picture seems right with the original shading issues. So now I need to alter the shading as the shading files are still showing really odd things. How do I adjust these? I imagine I need to find the calibration area in some way but im not sure how I do this? After the shading is corrected I think I should be ready to test, is there any formal testing procedure with SANE? There are a couple of people who are awaiting this scanner backend and have agreed to help testing so I will get it to them in some way. Im also going to sort my patch out and get rid of any whitespace or other rubbish ;) Cheers Chris Hello, The shading calibration area is 'under the roof', it is located inside the scanner. You can have a picture of it by doing a scan after setting physical y offset to 0. ie you scan from parking position. Sorry I wasnt very clear with this, I know where the shading area is I just dont know how to alter where the backend reads the shading data from. At first I thought it was in the genesys_device struct in the white strip/ black mark section but that seem to make no affect. Hopefully I can fix this by tomorrow and start testing Chris
[sane-devel] Backend for plustek Opticbook 3600
Hey Stef, I think I have isolated the gamma tables from the logs but there is a significant difference; the gamma tables all seem to be sent at once to the scanner. Regs 0x5b and 0x5c are both set to 0x00 and 1536 bytes are sent to the scanner. I have attached the log file appearances of the gamma tables to this email, there are five occurrences in my log file so all sets are here. I have split them up myself into three blocks but in the original snoop they all appear together. Will I need to modify the code base to account for all the gamma tables being sent at once? Thanks Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. -- next part -- An embedded and charset-unspecified text was scrubbed... Name: gammatables.log URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100222/32d80fc2/attachment-0001.asc
[sane-devel] Backend for plustek Opticbook 3600
On 02/22/2010 06:43 AM, stef wrote:
Le lundi 22 f?vrier 2010 02:09:14 Chris Berry, vous avez ?crit :
Hey Stef,
I think I have isolated the gamma tables from the logs but there is a
significant difference; the gamma tables all seem to be sent at once to
the scanner. Regs 0x5b and 0x5c are both set to 0x00 and 1536 bytes are
sent to the scanner. I have attached the log file appearances of the
gamma tables to this email, there are five occurrences in my log file so
all sets are here. I have split them up myself into three blocks but in
the original snoop they all appear together.
Will I need to modify the code base to account for all the gamma tables
being sent at once?
Thanks
Chris
Hello,
no need to modify the backend. Both way of writing data give the same
result.
Since all 3 curves are the same, you can pick one (from the '1' write, others
are simple ramps) and turn it into a list of 256 16 bits words. You can then
import it in a spreadsheet.
You may use this kind of awk script on a 1 line file with all the hex
values:
{
for(i=1;iNF;i+=2)
{
print strtonum($i)+256*strtonum($(i+1))
}
}
Regards,
Stef
Hey Stef,
Thanks for the awk script, helped a lot. I put the values into a
spreadsheet and got gamma values all around 1.0 so I didnt really need
to alter them at all, but it gives me something to talk about in my
report :). I have tweaked the gain and offset (offset is largely ignored
for AD type frontend it seems so I added a similar AD_OFFSET_CALIB
function to gl841) and now the picture seems right with the original
shading issues.
So now I need to alter the shading as the shading files are still
showing really odd things. How do I adjust these? I imagine I need to
find the calibration area in some way but im not sure how I do this?
After the shading is corrected I think I should be ready to test, is
there any formal testing procedure with SANE? There are a couple of
people who are awaiting this scanner backend and have agreed to help
testing so I will get it to them in some way. Im also going to sort my
patch out and get rid of any whitespace or other rubbish ;)
Cheers
Chris
--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
On 02/12/2010 08:47 AM, stef wrote: Le jeudi 11 f?vrier 2010 19:50:25, vous avez ?crit : Hello, I am facing the same issues than you while adding support for the HP3670. I eventually found that one of my decoding script was buggy and hide most of AFE accesses. I have appended the fixed version. By decoding usbsnoop again, you may find more AFE writes to be coded in backend. That may fix the red tint issue you have. I wouldn't be surprised if only write to red gain was showing in decoded log, leaving other channels untouched. Ok I have decoded another log file using this revision, you are correct there are many more writes to the fe showing up but they do not really mean anything more to me :(. I have attached the fe_writes.log file, which is just a straight grep of the fe writes with the new decoded log file. If you could look at it and tell me what you think is happening, the only reasonable things I can gather is that there are: * init style writes (0x00 - 0x70, 0x01 - 0x80, everything else - 0x00) * possible offset settings (0x05-0x07 seem to be set to 0x80 and 0xff repeatedly) * offset (0x05-0x07) and gain (0x02-0x04) seem to be set arbitrarily at various times. I'm not sure if any of this is relevant but whenever I set the offset/gain values in genesys_devices they seem to make little (if any) difference to the scans. You'll also have to look at the base gamma coefficient in the Genesys_Sensor entry for your model. I've had a look at this and I take it you mean this section: float red_gamma; float green_gamma; float blue_gamma; uint16_t *red_gamma_table; uint16_t *green_gamma_table; uint16_t *blue_gamma_table; I've tried changing the floats but there doesn't seem to be much difference, maybe I should wait until I fix the shading issue then tweak this? I assume the gamma_table entries are just initialisers for the tables. When doing a calibrated scan, the REG01_DVDSET bit is set to activate shading calibration. Clearing it may side step issues with shading calibration, but defeat the purpose of calibration. I was slightly worried that's what I had done here, i've restored it to the original values :P When tuning shading calibration, you must look at the black/white_shading.pnm and black/white_average pnm files that are written when full debug is enabled. With proper offset and gain values, you must have a full black picture (with some values slightly above 0) and a full white one. If not, the computed shading calibration coefficient will be bogus and you'll get black scans. Regards, Ste Ok i've also attached the black/white_shading/average files. If I understood you correctly the black_shading.pnm should be all white and not the pretty purple colour it is now and conversely the white should not look as it does. I have tried altering the offset and gain in gen_devices but this doesnt seem to do anything :(. Any other suggestions about how I move forward with the shading? I think as soon as the shading is sorted I might be there with this project and ready to commit to the dev version. Also I currently have OFFSET_CALIBRATION and DARK_CALIBRATION flags set for the scanner, how do I know if I should be using these (as opposed to COARSE_CALIB and DARK_WHITE_CALIB)? Sorry I have given you quite an essay to read there but at the minute I have a few assignments to deal with so i'm just trying to juggle everything but again thanks for all the help and time you have given me. One other thing I wanted to ask is about a new release? I've seen a lot of chat on the mailing list about this but I just wondered if you knew any more specifics? I only ask because I am taking this project on as part of my University course and the ideal end result for me is to have the drivers present in a SANE stable release,so if I could do this by the next release that would be perfect :) Thanks Chris Hello, regarding the AFE writes, judging by the AD9826 data sheet the writes to addresses 0x02-0x04 are R, G and B gains. The interesting thing is that red gain value 0x32 is higher than the 2 others (0x1e). So a red tint seems sensible. Personally, I would try to set them all to 0x32 (preferably) or 0x1e and map the gain array to these addresses. The other writes at 0x05, 0x06 and 0x07 are for offsets. For instance the sequence of AFE writes matches it, wee see gains set to 0 and various offsets tried. Then gains are set half range (0x80) and new offsets are tried. Looks like offset calibration for me. The 2 remaining addresses are for AFE configuration. For shading calibration, GENESYS_FLAG_DARK_WHITE_CALIBRATION would allow do calibrate in one pass and generate 'black' data using dark areas. To my opinion, GL842 scanners need really good black data to do proper shading calibration, so I'd recommend
[sane-devel] Backend for plustek Opticbook 3600
Hello, regarding the gamma values you can extract the gamma tables sent to your scanner by the windows driver from USB logs. Put them in a spreadsheet, then you can insert the formula used by the genesys backend and compare the reference and generated curves. I am not sure at all how to get the gamma values from the snoop. After looking at some parsed logs there are no references to gamma values, but looking in the awk files (specifically the bulk_write_register) there should be some parsed output for them. Maybe I am using the wrong version of your awk files for this? For gain and offset, until calibration work I have nothing else to advise you than trial and error. But by choosing a black/white target, you may refine your tests. Scanned black areas should give values slightly above zero. If no raise the offset. Once tweaked, you can adjust gain until white areas give values of about 240 (0xF0). Ok ill try this once I have the correct gamma settings. For the release date I can't tell. Next release should be close. When your work should be terminated for your University course ? Regards, Stef I have about 3 weeks to finish the implementation really, after that the sooner the better for the release :). Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Hello, I am facing the same issues than you while adding support for the HP3670. I eventually found that one of my decoding script was buggy and hide most of AFE accesses. I have appended the fixed version. By decoding usbsnoop again, you may find more AFE writes to be coded in backend. That may fix the red tint issue you have. I wouldn't be surprised if only write to red gain was showing in decoded log, leaving other channels untouched. Ok I have decoded another log file using this revision, you are correct there are many more writes to the fe showing up but they do not really mean anything more to me :(. I have attached the fe_writes.log file, which is just a straight grep of the fe writes with the new decoded log file. If you could look at it and tell me what you think is happening, the only reasonable things I can gather is that there are: * init style writes (0x00 - 0x70, 0x01 - 0x80, everything else - 0x00) * possible offset settings (0x05-0x07 seem to be set to 0x80 and 0xff repeatedly) * offset (0x05-0x07) and gain (0x02-0x04) seem to be set arbitrarily at various times. I'm not sure if any of this is relevant but whenever I set the offset/gain values in genesys_devices they seem to make little (if any) difference to the scans. You'll also have to look at the base gamma coefficient in the Genesys_Sensor entry for your model. I've had a look at this and I take it you mean this section: float red_gamma; float green_gamma; float blue_gamma; uint16_t *red_gamma_table; uint16_t *green_gamma_table; uint16_t *blue_gamma_table; I've tried changing the floats but there doesn't seem to be much difference, maybe I should wait until I fix the shading issue then tweak this? I assume the gamma_table entries are just initialisers for the tables. When doing a calibrated scan, the REG01_DVDSET bit is set to activate shading calibration. Clearing it may side step issues with shading calibration, but defeat the purpose of calibration. I was slightly worried that's what I had done here, i've restored it to the original values :P When tuning shading calibration, you must look at the black/white_shading.pnm and black/white_average pnm files that are written when full debug is enabled. With proper offset and gain values, you must have a full black picture (with some values slightly above 0) and a full white one. If not, the computed shading calibration coefficient will be bogus and you'll get black scans. Regards, Ste Ok i've also attached the black/white_shading/average files. If I understood you correctly the black_shading.pnm should be all white and not the pretty purple colour it is now and conversely the white should not look as it does. I have tried altering the offset and gain in gen_devices but this doesnt seem to do anything :(. Any other suggestions about how I move forward with the shading? I think as soon as the shading is sorted I might be there with this project and ready to commit to the dev version. Also I currently have OFFSET_CALIBRATION and DARK_CALIBRATION flags set for the scanner, how do I know if I should be using these (as opposed to COARSE_CALIB and DARK_WHITE_CALIB)? Sorry I have given you quite an essay to read there but at the minute I have a few assignments to deal with so i'm just trying to juggle everything but again thanks for all the help and time you have given me. One other thing I wanted to ask is about a new release? I've seen a lot of chat on the mailing list about this but I just wondered if you knew any more specifics? I only ask because I am taking this project on as part of my University course and the ideal end result for me is to have the drivers present in a SANE stable release,so if I could do this by the next release that would be perfect :) Thanks Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Ok so ive managed to fix the problem with the FLAG being set, the REG01_DVDSET part was screwing me up, so im not sure what that was about... More details are here http://sites.google.com/site/bez625/updates-1/update090210 with a new attachment of the scanning image. The scans now have a red tint which is a bit weird. Im still not sure how i am supposed to turn the brightness up, do you think there might be something specialist because of the AD AFE and its not wolfson? Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Yea thanks guys for all the support so far, if you are interested there are scans available on the project website. As soon as the brightness/red colour is sorted i should (hopefully) be in a position to release to the development. From reading the lists I think we are probably going to be looking at a release soon so who knows, it could be in the stable release before long. Jan and Dominik, when I have a working version available would you guys mind helping me test it? Ill send you both out a copy as soon as its ready if you want. Thanks Chris On 02/09/2010 05:24 PM, Jan-Benedict Glaw wrote: On Tue, 2010-02-09 17:05:15 +, Chris Berrys0457957 at sms.ed.ac.uk wrote: Ok so ive managed to fix the problem with the FLAG being set, the REG01_DVDSET part was screwing me up, so im not sure what that was about... More details are here http://sites.google.com/site/bez625/updates-1/update090210 with a new attachment of the scanning image. Oh, great! Nice to see, that you managed to take the first and probably largest hurdle, getting an image (without looking at the quality.) The scans now have a red tint which is a bit weird. Im still not sure how i am supposed to turn the brightness up, do you think there might be something specialist because of the AD AFE and its not wolfson? I guess that'll be solved over time, at least I hope so! MfG, JBG -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Hey Stef, did you manage to have a look at the patch I sent you? I havent had much time to look at the project over the last few days so not much has happened, i'm still getting black scans from the scanner when I remove the NO_CALIBRATION flag. Have you any idea which function will be causing this? Im guessing its going to be something trivial but I havent had the time to look through the extra functions that are being called when the flag is removed :/ Anyways i'l update you when I know more and if you have any advice please let me know Thanks Chris
[sane-devel] Backend for plustek Opticbook 3600
On 02/02/2010 05:29 AM, stef wrote: Le lundi 1 f?vrier 2010 13:52:30 Chris Berry, vous avez ?crit : On 01/31/2010 09:21 PM, stef wrote: Le dimanche 31 janvier 2010 20:30:58 Chris Berry, vous avez ?crit : Ok Stef, ignore that last email, I just got a scan! its far from perfect but im on the right road at least (mostly your advice ;) ). One of my GPIO wasnt set correctly and you were right about the bit 8 being set on the control reg of the AFE. An upload of the image is at http://sites.google.com/site/bez625/updates-1 . Its under Update 31/01/10 as an attachment. Im not sure what the next step here is but ill have a play myself. Any pointers from here are always welcome of course. Since things are progressing well I would like to know what the best procedure would be for adding this to the stable version of the sane api? I know we are a little far from this but it might be good for me to have an idea as this is obviously the intended outcome. Thanks Chris Hello, good job. 'first light' is always a gratifying moment. For future inclusion of your work, you might consider using a local git tree with a dedicated branch. So that you will be able to generate patches. Regards, Stef Thanks Stef, how would I go about using the git tree? I've used SVN before but not git so it shouldn't take long for me to wrap my head around it. So far I just downloaded a copy of the development version from git and i've been working from that copy, i'm not sure how to make a new branch (or even if I have permission?). Also have you had a chance to look at the scan and suggest what i need to change? Im guessing that it might be a problem with the CCD, maybe it thinks the ccd is larger than it is? Thanks again Chris Hello, once you have cloned the master branch of SANE. You can create any number of local branches. I have a few of them, about one by backend I touch. The few commands I use are: git branch to create or list branches git pull [--rebase] to get latest version git rebase to rebase a branch on lastest version git merge to merge a branch on master version git format-patch to generate patches from my local history git am to apply patches generated by by format patch You can do all these locally without write access to SANE's git tree. At some point, you'll need it. Then you'll have to ask for an account on alioth and a project administrator will add you to the project. I have looked at the picture. What we see might relates to 'half ccd' mode. It is a mode where coordinates on the CCD are expressed in a resolution, but the timings are such that effective resolution is only the half. See flag GENESYS_FLAG_HALF_CCD_MODE. Another idea would be to compare registers 52 to 57 (AFE timings) with ones from an usb log. Regards, Stef Cheers for the advice again stef, I just added the half ccd flag and it sorted out the image as it should look. The only issue now is that the scan seems a little dark, again ill upload the latest scan to the website (http://sites.google.com/site/bez625/updates-1/update020210). For the time being ill work with git to make sure im developing in the correct way ;), if you have a suggestion on the brightness issue let me know. Also I played with the offset and gain a little and neither seem to make any difference. Not sure what that is about. Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
On 01/31/2010 09:21 PM, stef wrote: Le dimanche 31 janvier 2010 20:30:58 Chris Berry, vous avez ?crit : Ok Stef, ignore that last email, I just got a scan! its far from perfect but im on the right road at least (mostly your advice ;) ). One of my GPIO wasnt set correctly and you were right about the bit 8 being set on the control reg of the AFE. An upload of the image is at http://sites.google.com/site/bez625/updates-1 . Its under Update 31/01/10 as an attachment. Im not sure what the next step here is but ill have a play myself. Any pointers from here are always welcome of course. Since things are progressing well I would like to know what the best procedure would be for adding this to the stable version of the sane api? I know we are a little far from this but it might be good for me to have an idea as this is obviously the intended outcome. Thanks Chris Hello, good job. 'first light' is always a gratifying moment. For future inclusion of your work, you might consider using a local git tree with a dedicated branch. So that you will be able to generate patches. Regards, Stef Thanks Stef, how would I go about using the git tree? I've used SVN before but not git so it shouldn't take long for me to wrap my head around it. So far I just downloaded a copy of the development version from git and i've been working from that copy, i'm not sure how to make a new branch (or even if I have permission?). Also have you had a chance to look at the scan and suggest what i need to change? Im guessing that it might be a problem with the CCD, maybe it thinks the ccd is larger than it is? Thanks again Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
On 01/31/2010 01:11 PM, stef wrote: Le samedi 30 janvier 2010 15:29:50 Chris Berry, vous avez ?crit : Hello, the function you wrote is fine. What I wanted to share is that it is easier to do experiments from within a really simple test program than the full blown scanimage or xsane. For instance, to get data from the HP3670, I had to set up registers that aren't set up under windows. I also found that I had to use different values for offset and gain than the ones that could found in usbsnoop logs. I think you'll need much trial and errors before having it work. You must also keep in mind that if GPIO aren't correctly set up, you may not get any data whatever you do with the AFE. Regards, Stef Ah ok, is there any way I can tell if im starting to get any reasonable data from this scan? Currently I am only checking the pnm image so I have no real idea if im getting close or not. Are there any good ways to see if im getting any data at all? Thanks Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Ok Stef, ignore that last email, I just got a scan! its far from perfect but im on the right road at least (mostly your advice ;) ). One of my GPIO wasnt set correctly and you were right about the bit 8 being set on the control reg of the AFE. An upload of the image is at http://sites.google.com/site/bez625/updates-1 . Its under Update 31/01/10 as an attachment. Im not sure what the next step here is but ill have a play myself. Any pointers from here are always welcome of course. Since things are progressing well I would like to know what the best procedure would be for adding this to the stable version of the sane api? I know we are a little far from this but it might be good for me to have an idea as this is obviously the intended outcome. Thanks Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
On 01/30/2010 01:40 PM, stef wrote: Le jeudi 28 janvier 2010 11:26:56 Chris Berry, vous avez ?crit : No stef im almost certain this isnt a Wolfson, from an old project page I think it is an Analog Digital 9826 Link: http://www.analog.com/en/audiovideo-products/cameracamcorder-analog-front-e nds/ad9826/products/product.html I too am puzzled why there are so many writes to the frontend, the XP200 in the gl686 I think uses an AD frontend and even that only has a few fe writes, although the XP200 has separate functions for AD coarse grain calibration and AD offset calibration so im not sure if these extra writes are done then? Any advice? Chris Hello, I had a quick look at the data-sheet. It is coherent with the writes you see. One thing that I find strange is that bit 7 of register 0 is set to 1, while it is recommended no to do so. Maybe it is a bug, or maybe your scanner hasn't the same exact component. During calibration there maybe a lots of writes to frontends. To go further, I think it could be a good idea to have a test program to does a simple color scan with a fixed area. You could code directly in it various write to AFE until you get sensible data. I have appended the main C file and the build file as a possible example. I build like this make -DUNIT_TESTING make -f mk . The wm.c is how I experimented with HP3470's AFE. Regards, Stef Ok Stef, at the minute I have a separate function in gl841.c that is called during set_fe to skip the standard front end writes in that file and use the ones defined in my function. I have attached the function for you to look at. My question is: is there any benefit to using the method you just recommended over the method im using now? If this function is required it is likely to implemented in the way i have so will it help me much to do what you have asked? Or is what i'm doing pretty similar. Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336. -- next part -- An embedded and charset-unspecified text was scrubbed... Name: set_ad_fe URL: http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20100130/a9bf9ad8/attachment-0001.asc
[sane-devel] Backend for plustek Opticbook 3600
Hey stef, i'm really struggling to work out what all these frontend writes are for. Below is a grep of the frontends from a scan, I have done several scans and they all roughly follow this format. I'm not sure why the offset (regs 05 - 07) have values 0x80 and 0xff and the values for both offset and gain (0x02 - 0x04) have some values that change per scan? Any light you could shed on this would be great, Cheers Chris genesys_fe_write_data(0x000x00f8) genesys_fe_write_data(0x010x0080) genesys_fe_write_data(0x020x) AFEINIT genesys_fe_write_data(0x030x) genesys_fe_write_data(0x040x) genesys_fe_write_data(0x050x) genesys_fe_write_data(0x060x) genesys_fe_write_data(0x070x) genesys_fe_write_data(0x020x) genesys_fe_write_data(0x030x) genesys_fe_write_data(0x040x) AFESET? genesys_fe_write_data(0x050x0080) genesys_fe_write_data(0x060x0080) genesys_fe_write_data(0x070x0080) genesys_fe_write_data(0x050x00ff) genesys_fe_write_data(0x060x00ff) ?? genesys_fe_write_data(0x070x00ff) genesys_fe_write_data(0x050x0127) \ genesys_fe_write_data(0x060x0126) These values seem to change slightly per scan genesys_fe_write_data(0x070x011b) / genesys_fe_write_data(0x020x) genesys_fe_write_data(0x030x) genesys_fe_write_data(0x040x) genesys_fe_write_data(0x020x002d) genesys_fe_write_data(0x030x0019) genesys_fe_write_data(0x040x001e) - This value also varies slightly genesys_fe_write_data(0x050x0080) genesys_fe_write_data(0x060x0080) genesys_fe_write_data(0x070x0080) genesys_fe_write_data(0x050x00ff) genesys_fe_write_data(0x060x00ff) genesys_fe_write_data(0x070x00ff) genesys_fe_write_data(0x050x0121) \ genesys_fe_write_data(0x060x0124) These values all seem to vary slightly genesys_fe_write_data(0x070x0117) / genesys_fe_write_data(0x050x011a) \ genesys_fe_write_data(0x060x0118) These values all seem to vary slightly genesys_fe_write_data(0x070x010d) / genesys_fe_write_data(0x050x0109) \ genesys_fe_write_data(0x060x0004) These values all seem to vary slightly genesys_fe_write_data(0x070x000c) /
[sane-devel] Backend for plustek Opticbook 3600
No stef im almost certain this isnt a Wolfson, from an old project page I think it is an Analog Digital 9826 Link: http://www.analog.com/en/audiovideo-products/cameracamcorder-analog-front-ends/ad9826/products/product.html I too am puzzled why there are so many writes to the frontend, the XP200 in the gl686 I think uses an AD frontend and even that only has a few fe writes, although the XP200 has separate functions for AD coarse grain calibration and AD offset calibration so im not sure if these extra writes are done then? Any advice? Chris
[sane-devel] Backend for plustek Opticbook 3600
stef wrote: Le vendredi 22 janvier 2010 17:57:06 Chris Berry, vous avez ?crit : Hey Stef, I had to take a little time off the front end stuff to do some documentation but Im looking at it again now. Like you im fairly certain everything else is set, could you share with me a snoop of the HP3670? I'm curious to see the front end entries and whether they follow a similar pattern to the opticbook 3600. Cheers Chris Hello, I still can't get data for HP3670. After much tinkering, everything when from all black to all white... Here's all the frontend accesses (a grep result) for a 100 dpi color scan. 0x01 and 0x03 are control registers, 0x23 is offset and 0x28 is gain. I opened the HP3670 Jack donated me, but I couldn't find a discrete chip which could be the frontend. With the exact Wolfson chip, I'd be able to have a datasheet that surely explain some things. sanei_genesys_fe_write_data(0x01,0x0007) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x01,0x0007) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x03,0x0002) sanei_genesys_fe_write_data(0x23,0x) sanei_genesys_fe_write_data(0x28,0x0002) sanei_genesys_fe_write_data(0x23,0x) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x23,0x0014) sanei_genesys_fe_write_data(0x23,0x00c2) sanei_genesys_fe_write_data(0x23,0x00be) sanei_genesys_fe_write_data(0x23,0x00be) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x28,0x0002) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x28,0x0002) sanei_genesys_fe_write_data(0x03,0x0002) sanei_genesys_fe_write_data(0x23,0x0080) sanei_genesys_fe_write_data(0x28,0x0002) sanei_genesys_fe_write_data(0x23,0x) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x23,0x0014) sanei_genesys_fe_write_data(0x23,0x00c0) sanei_genesys_fe_write_data(0x23,0x00be) sanei_genesys_fe_write_data(0x23,0x00be) sanei_genesys_fe_write_data(0x28,0x0008) sanei_genesys_fe_write_data(0x28,0x0004) sanei_genesys_fe_write_data(0x28,0x0006) sanei_genesys_fe_write_data(0x28,0x0005) sanei_genesys_fe_write_data(0x28,0x0005) sanei_genesys_fe_write_data(0x23,0x) sanei_genesys_fe_write_data(0x23,0x0014) sanei_genesys_fe_write_data(0x23,0x00c0) sanei_genesys_fe_write_data(0x23,0x00bc) sanei_genesys_fe_write_data(0x23,0x00bc) sanei_genesys_fe_write_data(0x03,0x0002) sanei_genesys_fe_write_data(0x23,0x00bc) sanei_genesys_fe_write_data(0x28,0x0005) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x03,0x0002) sanei_genesys_fe_write_data(0x23,0x00be) sanei_genesys_fe_write_data(0x28,0x0002) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x03,0x0002) sanei_genesys_fe_write_data(0x23,0x0080) sanei_genesys_fe_write_data(0x28,0x0002) sanei_genesys_fe_write_data(0x23,0x) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x23,0x0014) sanei_genesys_fe_write_data(0x23,0x00be) sanei_genesys_fe_write_data(0x23,0x00be) sanei_genesys_fe_write_data(0x23,0x00be) sanei_genesys_fe_write_data(0x28,0x0008) sanei_genesys_fe_write_data(0x28,0x0004) sanei_genesys_fe_write_data(0x28,0x0006) sanei_genesys_fe_write_data(0x28,0x0007) sanei_genesys_fe_write_data(0x28,0x0007) sanei_genesys_fe_write_data(0x23,0x) sanei_genesys_fe_write_data(0x23,0x0014) sanei_genesys_fe_write_data(0x23,0x00be) sanei_genesys_fe_write_data(0x23,0x00bc) sanei_genesys_fe_write_data(0x23,0x00bc) sanei_genesys_fe_write_data(0x03,0x0002) sanei_genesys_fe_write_data(0x23,0x00bc) sanei_genesys_fe_write_data(0x28,0x0007) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x01,0x0003) sanei_genesys_fe_write_data(0x01,0x0003) Regards, Stef Yea that's a strange one Stef, are you sure this is a Wolfson AFE chip? Looking here http://www.wolfsonmicro.com/productListings/imaging you can see there are only about 6 AFE's that are three channel and reg values start at 0x01 (as opposed to one which starts at 0x00). The strange thing about this grep (or at least i find it strange) is that it only writes to PGA Red Gain and not Blue and Green (going off these 6 datasheets from wolfson as they are all roughly the same), which is what makes me ask are you sure its a Wolfson. As there are only 6 I suppose you can just work through them one at a time... I discovered where my problem was, I definitely DO NOT have a wolfson afe. I tracked the datasheet down from an old project page to be a AD9826 and it certainly marries up to the scans snoops. The only problem I have now is that I'm not really sure how to change the genesys files as my offsets are on reg 0x05, 0x06, 0x07 and my gain are on 0x02, 0x03, 0x04 (as opposed to 0x20-22 and 0x28-2a respectively). Do you have any
[sane-devel] Backend for plustek Opticbook 3600
Hi Stef, in the course of trying to get the lamp working I was playing with the settings and after I removed my changes the lamp functions as i expected too so that issue is resolved. If my understanding is correct the code should select whether to set cis or not based on the variable in genesys_driver.c. As it is currently hard coded I will try and put an if statement in to select based on the /* Is this a CIS scanner? */ sane boolean. I am still not getting any picture data, after observing some of the front end logs I see that the windows driver only writes to *FRONT END* addresses 0x00 - 0x07, but the genesys driver is writing to 0x00 - 0x09 and many others (0x20, 0x24 etc...). I am guessing this is very bad for the recovery of data from the DAC. I will attach a new genesys log and a windows sniff to the project website today and maybe some pictures. As usual any pointers are greatly appreciated Thanks Chris On 01/15/2010 01:10 PM, stef wrote: Le jeudi 14 janvier 2010 17:47:41 Chris Berry, vous avez ?crit : Ok quick update, I have managed to get the lamp on during scans! The reason it was failing is because reg[0x01] was hard coded to CIS type and this is a CCD scanner. Hopefully this means that all I have to do is polish up the remaining register values at init time. Currently the light is on when the scan head moves down but turns itself off immediately before it comes back, is this normal behaviour for the genesys driver? Chris Hello, nice job. Do you get any picture data ? I don't think that the genesys backend switch off lamp after you scan on purpose. The planned behaviour is to switch off CCD only when the defined waiting time has elapsed. Maybe what you are seeing is the power saving function kicking in and switching off light by modifying GPIOs (see gl841_save_power function in genesys_gl841.c). While this doesn't hurt for CIS scanners, CCD may have to warm up each time they light on. So this behaviour might not be desirable. Regards, Stef -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Ok I have set the values for the 0x6b and im still getting a completely black scan but the lamp is being turned off at the start of a scan. When the scanner is plugged in the light is ON and when i start a scan the light is immediately turned off. The init reg is definitely sending the correct data to the scanner so I am assuming another part of the code is turning the lamp off. I took a USB sniff of the lamp being turned on and off repeatedly which I will add to the project site (http://sites.google.com/site/bez625/home) and I am certain all that the lamp needs is to set register 0x03 to the value of 0x5f or 0xdf (difference here is lampdog on and off). I have tried various other settings including setting the lamp power to off and to no avail. Please bear in mind that I have only been changing the setting in the reg_init and no where else. I cant see how I can test anything else is working until the lamp is on and why it wont turn on baffles me. Thanks in advance Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Ok quick update, I have managed to get the lamp on during scans! The reason it was failing is because reg[0x01] was hard coded to CIS type and this is a CCD scanner. Hopefully this means that all I have to do is polish up the remaining register values at init time. Currently the light is on when the scan head moves down but turns itself off immediately before it comes back, is this normal behaviour for the genesys driver? Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Thanks for the quick reply stef, hope you had a good christmas/new year :). I have made changes to the GPIO settings as appropriate (i.e. matched the settings to what I believe is correct from register writes) but I could only find values for 0x6b, 0x6d, 0x6e and 0x6f. I couldn't find a write to 0x6c so I assumed that to be 0x00 and also I am unsure in which file the 0x6b register is set. Is 0x6b set in genesys_devices.c? If so which section? Another concern is that in the logs there are no URB's of the form: control 0x40 0x0c 0x89 0x00 len (Output enable) control 0xc0 0x0c 0x8a 0x00 len (read) control 0x40 0x0c 0x8b 0x00 len (write) I checked these after running the initial perl scripts and before the awk scripts you forwarded to me. Is it maybe possible that the control URBs for this have happened at a time that hasn't been caught by the logs or is it more likely that there is a different value being used as opposed to 0x89, 0x8a and 0x8b? Would incorrect GPO affect the light being on and off? I know for sure that the value for 0x6b from the logs is different to the one set by the genesys driver at the moment so I think maybe correcting this could be a big break through for this project :) Chris stef wrote: Le jeudi 7 janvier 2010 15:14:20 Chris Berry, vous avez ?crit : Hi Stef, I have updates on this project at http://sites.google.com/site/bez625/updates-1/update070110 with some attachments. I am still struggling to work out why the scan lamp is off during scanning. My thoughts are that maybe the genesys backend is saying the lamp sleep mode if on before the scan as opposed to off. The log files seem to support this with the register values being set but any advice you could offer on where i should go from here would be appreciated. Thanks Chris Hello, have you made changes related to GPIO ? They often change meaning from a scanner to another. Sometimes you won't get any data until they are set up correctly. Locate writes to registers 6c/6d and 6b for values and 6e/6f for mask in the USB logs you have. You'll see quite a few places where we test for the model- gpo_type and write to GPIO registers. Try to adapt these to match what you'll see in logs. Regards, Stef
[sane-devel] Backend for plustek Opticbook 3600
Hi Stef, I have updates on this project at http://sites.google.com/site/bez625/updates-1/update070110 with some attachments. I am still struggling to work out why the scan lamp is off during scanning. My thoughts are that maybe the genesys backend is saying the lamp sleep mode if on before the scan as opposed to off. The log files seem to support this with the register values being set but any advice you could offer on where i should go from here would be appreciated. Thanks Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
lol no worries Jan, hopefully I can get the scanner working soon for you :) Chris On 01/07/2010 04:25 PM, Jan-Benedict Glaw wrote: On Thu, 2010-01-07 14:14:20 +, Chris Berrys0457957 at sms.ed.ac.uk wrote: I have updates on this project at http://sites.google.com/site/bez625/updates-1/update070110 with some attachments. I am still struggling to work out why the scan lamp is off during scanning. My thoughts are that maybe the genesys backend is saying the lamp sleep mode if on before the scan as opposed to off. The log files seem to support this with the register values being set but any advice you could offer on where i should go from here would be appreciated. It's at least great to see someone working on a backend for the '3600. That will be a nice scanner once it's somewhat supported. So I'm having some hope that, at some time, I can use my scanner myself, without asking my girlfriend to scan stuff for me :) MfG, JBG -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Ok thanks stef, i've been experimenting a little with the logs, there
was a small error/issue with the parse1.awk script you sent me. I don't
know if you have encountered this but around line 113 there is an if
statement along the lines of:
if ($5 == 0x002])
this fails because of what I suspect is an end of line character in $5,
so i replaced it with:
if (index($5, 0x002]) 0) endpoint = BULK_OUT
and so on for the following values. I'm not very good with awk so I
don't know if this is the best way of fixing this but it works for now.
After some playing I think I have the CCD settings right and the motor
seems to be working so I havent really changed that much. I also think
the GPIO settings are right which leads me to believe the DAC is the
main issue at the moment.
From my understanding the front end is accessed through registers
0x51(address), 0x3a and 0x3b (values) but im not sure how this
translates to the settings in genesys_devices.c:
{DAC_PLUSTEK_3600,{0x00, 0x3d, 0x08, 0x00}
, {0x00, 0x00, 0x00}
, {0xe1, 0xe1, 0xe1}
, {0x93, 0x93, 0x93}
, {0x00, 0x19, 0x06}
}
Now I think the first 4 values correspond to addresses 0x00 - 0x03 of
the DAC, but the following arrays are a bit of a mystery to me. I have
had a look at the log from run-genesys and it seems that they are trying
to set many DAC registers but the Snoop reveals that only fe_writes to
addresses 0x00 - 0x07 are made.
As when I try a scan everything (aside from the light not being on)
seems to be reasonable but results in a garbled image I think it might
just be a case of getting the light on and setting up the DAC properly.
If you could help me understand the DAC settings I might be able to
investigate the matter further.
Thanks
Chris
On 12/29/2009 07:46 AM, stef wrote:
Le jeudi 24 d?cembre 2009 06:07:23 Chris Berry, vous avez ?crit :
Just a quick update, I had to take a little time off from the project
but I should be able to work on it more now. I have an update at
http://sites.google.com/site/bez625/updates-1/update with new log files
for smaller scan and preview snoops.
Stef is there any chance you could explain the layout of your script to
me a little, for instance
URB 7278 control 0xc0 0x0c 0x8e 0x20 len 1 read 0x01
I take it this means:
URBURB Num URB Type 4 hex values lenLength r/w data
What are the 4 hex values for? Also anything else helpful will be
appreciated. I'm not certain how I can better find the values I need for
genesys_devices.c, I look for the hex values of the registers that need
to be set and cant find them in the new reduced log files. Maybe some of
these values are only set at power up and not before a scan or preview?
Chris
Hello,
these numbers are to be converted in sanei_usb_control_msg() function
calls.
You can have a look at the defines in genesys_low.h around line 113 for the
known values. The 0x8e value isn't defined and currently used in the genesys
backend. However, after looking at the logs you posted, I think it is some
data reading. The trouble is that currently there is no such access to data
for the supported models. While I think it could be turned to something like
sanei_genesys_bulk_read(), you'll need to experiment. As far I understand,
these are vendor specific accesses. I couldn't find these values in any USB
documents.
You are right that some register values don't change much and are set up
early, and not at every scan.
Regards,
Stef
-- next part --
An HTML attachment was scrubbed...
URL:
http://lists.alioth.debian.org/pipermail/sane-devel/attachments/20091230/eb05b604/attachment-0001.htm
[sane-devel] Backend for plustek Opticbook 3600
Just a quick update, I had to take a little time off from the project but I should be able to work on it more now. I have an update at http://sites.google.com/site/bez625/updates-1/update with new log files for smaller scan and preview snoops. Stef is there any chance you could explain the layout of your script to me a little, for instance URB 7278 control 0xc0 0x0c 0x8e 0x20 len 1 read 0x01 I take it this means: URB URB Num URB Type 4 hex values len Length r/w data What are the 4 hex values for? Also anything else helpful will be appreciated. I'm not certain how I can better find the values I need for genesys_devices.c, I look for the hex values of the registers that need to be set and cant find them in the new reduced log files. Maybe some of these values are only set at power up and not before a scan or preview? Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Hey Stef, the script you forwarded to me (decode-pl.sh) seems to be reliant on some .awk files (filtre.awk, swap.awk and refait.awk). I had a quick look online but could not find them, is there any chance you could send them to me :) Chris On 12/07/2009 08:24 PM, stef wrote: Le lundi 7 d?cembre 2009 19:30:15 Chris Berry, vous avez ?crit : Ok I have managed to use SniffUSB to extract some data during the calibration and preview stages, there is too much to post here but im not sure how to interpret it. I have tried usb-robot but am struggling to get it working (it tells me there is a problem finding libusb header which is installed?). Any advice of where to go from here would be great :) Thanks Chris Hello, once you have the usb log, you can parse it with the appended scripts (from Pierre Willenbrock). You can then compare the processed output with the debug output of scanimage ran with full debug. A starting point is to add the USB ids in genesys.conf.in, then add a model description in genesys_devices.c . Cloning the Lide 35 entry seems to be a sensible start. Then you'll have to set the flags to the same values than the HP2400 entry so that the backend doesn't try to do calibration and does an uncalibrated scan. With a script like the attached 'run-genesys' (to be put in backend subdir) you can try the code you have just compiled without installing it system-wide Regards, Stef -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Ok I have managed to use SniffUSB to extract some data during the calibration and preview stages, there is too much to post here but im not sure how to interpret it. I have tried usb-robot but am struggling to get it working (it tells me there is a problem finding libusb header which is installed?). Any advice of where to go from here would be great :) Thanks Chris On 12/06/2009 10:07 PM, stef wrote: Le dimanche 6 d?cembre 2009 20:18:17 Chris Berry, vous avez ?crit : Thanks for the info Stef, I have run sane-find-scanner on the scanner and it confirms that the chipset is indeed GL482. I was under the impression the Genesys backend was for GL481, does it support GL482 as well? Thanks Chris Hello, yes, it already supports several gl842 scanners. You may start to add support for the Opticbook 3600 by adding an entry in genesys_devices.c and in genesys.conf.in . You'll may have to tweak GPIO handling. In any case, windows usbsnoop of calibration and preview will be useful. Regards, Stef -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Thanks for the info Stef, I have run sane-find-scanner on the scanner and it confirms that the chipset is indeed GL482. I was under the impression the Genesys backend was for GL481, does it support GL482 as well? Thanks Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
[sane-devel] Backend for plustek Opticbook 3600
Hello, As part of a project I am currently undertaking I am creating support for the opticbook 3600 under linux. It seems the best approach to this is to create a SANE backend for the scanner. After reading the available documentation I am of the understanding that I should post to this mailing list to prevent duplicate code. I am assuming there are no other people working on this project so far (if so please let me know!) and I was wondering if anyone could give me a few pointers as where to start with this task? Any information supplied to me would be very gratefully appreciated :-) Thanks Chris -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.
