Ad
Ad
Ad
Pages: « 1 [2] 3 4 »   Bottom of Page
Print
Author Topic: Canon x300 Media Configuration Tool  (Read 13665 times)
shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #20 on: August 02, 2010, 07:56:10 PM »
ReplyReply

Alrighty then... DLing the ColorBurst demo now.

To Peninsula... yeah, I'm a little suspect of Hahnemuhle's recommended media settings. I was hoping that perhaps they had made their own media types, like Crane Museo has done but that appears to not be the case. I just did a comparison of the recommended Special 9 setting for HFA Photo Rag Ultra Smooth vs. the media type I created myself (which is either using low or medium low in). My own little test target has a strip with a black background that has 9 patches on it in increments of 2 going from 0,0,0 to 16,16,16 and then something similar with white patches going back from 255, 255, 255. I also have a b+w gradient and a granger chart. Special 9 doesn't resolve the shadows as well as my media type with the low ink setting. It also exhibits a severe hotspot of ink in the... I think purple, magenta (don't want to turn the lights on right now to verify... the Spectroscan is huffing along in the corner).

Cheers, Joe
Logged
shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #21 on: August 02, 2010, 08:28:54 PM »
ReplyReply

Alright... Spectroscan is still huffing and puffing but when it's done... basically I want to create a few different media types. I figure maybe divide an conquer-ish approach. I'll do a low, medium and high media type, and print the 20 patch linearization chart with each (Hmmm... looks like ColorBurst is using the Monaco Profiler engine?Huh) .From there I fire up SpectralVision, scan in the targets and then compare the graphs??? (this is from all of 1 minute and 30 seconds of skimming through the manual... go easy on me if I'm way off base).

Cheers, Joe
Logged
shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #22 on: August 02, 2010, 08:55:21 PM »
ReplyReply

Quote from: Onsight
Yes, it's all there. What I'm not completely certain about is if the demo version let's you do it all without a dongle. I think so...

Still waiting for the Spectroscan to finish up. Anyhow... I think this verifies what you're saying:

From the SpetralVision manual:
Quote
A ColorBurst dongle is required to build an ICC Profile. SpectralVision Pro will create a lin- earization file without a dongle, but the Profile Build window is disabled when a ColorBurst dongle is not plugged in to a USB port. This means that SpectralVision Pro can not be used with the ColorBurst demo to create an ICC profile—a full license of ColorBurst with the dongle is required.
Logged
shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #23 on: August 02, 2010, 11:19:16 PM »
ReplyReply

Oy... So SpectralVision has CMYK targets. In order to test this stuff with I bought a pack of 8.5x11 sheets of matte paper so I didn't have to waste a ton of empty space on my rolls. Well, the Photoshop pluggin which does print to 8.5x11 cut sheets just fine, will only accept RGB input. The 8 bit driver via Photoshop won't let me select "cut sheet" as the source, complaining that I need to feed the printer a valid paper size in order to use a cut sheet. Sigh. This ain't gonna happen tonight I guess.

ETA - Never mind... in the 8 bit driver (which is kind of painful to use after using the PS plugin) you can't just select "8.5x11" for page size... there's a tear-off menu that gives you sub item selections and one of them is "8.5x11 cut sheet". So why is it not enough to select 8.5x11 and specify cut sheet in the check box (which you have to do anyway)Huh Sigh. Okay... let's see what happens here.

Cheers, Joe
« Last Edit: August 02, 2010, 11:24:10 PM by shewhorn » Logged
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1274


WWW
« Reply #24 on: August 03, 2010, 04:25:02 PM »
ReplyReply

Quote
A ColorBurst dongle is required to build an ICC Profile
Right, but we're not interested in CB making a profile, just helping us determine which media setting will give optimal results. Here's how it works: Print the CB linearization image with the MCT Tool, read the Linearization target in SpecralVision, load it in CB and have it graph the results with the Chroma graphing option. You'll want to look for a chroma (saturation) plateau. As density increases so does chroma - up to a certain point. After hitting a sweet spot, saturation actually decreases with increased density. Our eyes can't see that - we perceive increased density as increased saturation and we shouldn't trust them in this way! Looking for and hitting those chroma sweet spots is the key to maximizing DMax and color gamut! This is what advanced RIP installation experts have been doing for years. Perhaps someone will write a more elegant tool for scientifically determining optimal media settings with regular printer drivers....
Logged

jgbowerman
Full Member
***
Offline Offline

Posts: 168


Where it all started


WWW
« Reply #25 on: August 03, 2010, 07:15:22 PM »
ReplyReply

Quote from: shewhorn
Alrighty then... DLing the ColorBurst demo now.

To Peninsula... yeah, I'm a little suspect of Hahnemuhle's recommended media settings. I was hoping that perhaps they had made their own media types, like Crane Museo has done but that appears to not be the case. I just did a comparison of the recommended Special 9 setting for HFA Photo Rag Ultra Smooth vs. the media type I created myself (which is either using low or medium low in). My own little test target has a strip with a black background that has 9 patches on it in increments of 2 going from 0,0,0 to 16,16,16 and then something similar with white patches going back from 255, 255, 255. I also have a b+w gradient and a granger chart. Special 9 doesn't resolve the shadows as well as my media type with the low ink setting. It also exhibits a severe hotspot of ink in the... I think purple, magenta (don't want to turn the lights on right now to verify... the Spectroscan is huffing along in the corner).

Cheers, Joe

Thanks for the update. I plan to profile the same Ultra Sm next week. It sounds like you are onto a good ink setting. I started a post on the subject when I could not, for some odd reason, relocate this post... go figure.

Is there any way I can get a tiff of your test target? It sounds WAY better than the one that comes with the MCT.

Regards, Greg
Logged

jgbowerman
Full Member
***
Offline Offline

Posts: 168


Where it all started


WWW
« Reply #26 on: August 03, 2010, 07:20:31 PM »
ReplyReply

Quote from: Onsight
Right, but we're not interested in CB making a profile, just helping us determine which media setting will give optimal results. Here's how it works: Print the CB linearization image with the MCT Tool, read the Linearization target in SpecralVision, load it in CB and have it graph the results with the Chroma graphing option. You'll want to look for a chroma (saturation) plateau. As density increases so does chroma - up to a certain point. After hitting a sweet spot, saturation actually decreases with increased density. Our eyes can't see that - we perceive increased density as increased saturation and we shouldn't trust them in this way! Looking for and hitting those chroma sweet spots is the key to maximizing DMax and color gamut! This is what advanced RIP installation experts have been doing for years. Perhaps someone will write a more elegant tool for scientifically determining optimal media settings with regular printer drivers....

Thanks for those details, I'm going to copy and paste as a convenient reference in the case I go there. Sounds pretty simple, too.
Logged

jgbowerman
Full Member
***
Offline Offline

Posts: 168


Where it all started


WWW
« Reply #27 on: August 03, 2010, 07:32:07 PM »
ReplyReply

My take on H. Photo Rag Baryta 315:

Hahnemuhle recommends Special 5 for all of their glossy/semi-glossy papers. I figured we can do better, and I believe I have. I'll share my method (crude as it may be):

After evaluating ink usage choices with the under-adequate test-print target that comes with the MCT, I perceived a drop in saturation at the High setting relative to the Medium-high setting, and better saturation for Medium-high relative to Standard (which should be called Medium IMO). This was done by looking at the color patch aspect of the test print with an 8x magnifying loop and under a D50 lamp. It was subtle, these differences, but I could reproduce my findings without any question. Otherwise, I find the MCT Test print worthless when attempting to perceive a difference among the three highest settings. I then configured a profile, entering "Photo/Art/Proofing Paper (Glossy), clicked the "Assist" button for choosing the best Media Type name, enter the paper weight at 315, and went with the recommended HW Gloss Photo for the Standard Paper choice. I left both head height and Vacuum Strength at auto, and Max Ink Usage at Medium-high.

Next, I created paper profiles using both the Special 5 configuration and my custom configuration by printing a two-page 1860 patch profile target. I took these 1860 test prints and further evaluated them under the D50 lamp, and I could see darker color, perhaps with a tad more pop, viewing the Special 5 target, it was subtle. However, the Custom target showed, only in a few hues, clearly better tonal separation between patches. Using ProfileMaker, I made the profiles and I took both into ColorThink and compared them. I'd view one in the Wireframe mode and the other Flat, and switch the modes back and forth. It was subtle, but the custom configuration had a slight edge with a larger gamut.

I know this is not scientifically precise as I understand one can accomplish using other methods like ColorBurst software, but I have no question in my eye the Custom configuration using Medium-high for maximum ink volume is an improvement over the canned Special 5 setting when used on this paper. I'm going to build a profile using H. Ultra-Smooth 305 next.
« Last Edit: August 05, 2010, 08:36:23 AM by peninsula » Logged

shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #28 on: August 03, 2010, 08:33:43 PM »
ReplyReply

Quote from: Onsight
Here's how it works: Print the CB linearization image with the MCT Tool, read the Linearization target in SpecralVision, load it in CB and have it graph the results with the Chroma graphing option

Thanks Scott,

I know we don't want to build a profile BUT... are these features part of the module to build a profile? I read in the target and then saved it via SV but I can't find any way to load it into CB. When the dongle is in place, SV will automatically transfer the data to CB if I understood the manual correctly.The other method is to open it from within CB. In CB itself if you click on ink and color, Ink Curves, Linearization, and Ink Limit are greyed out. If I'm not mistaken, in order to open that Linearization file you need to click on the Linearization tab in the Ink and Colors module and then click the open button.

There's a high probability that I'm doing something wrong but in the mean time I created my own test file. They had some color blocks (0,0,255; 0,255,0; 0,255,255; 255,0,0; 255,0,255; 255,255,0; 255,255,255; 128,128,128). Then I had two strips, one with a black background and one with a white. On each there were 9 blocks inside each strip. In the black strip the first block would be 0,0,0, the second 2,2,2, the third, 4,4,4, etc. I did the same thing with the white only in reverse. Finally I also had a Granger Chart on there.

I'd created 3 media types to start out with using Hahnemuhle Photo Rag Satin, low, medium low, and medium. Looking at the prints with a 4700ºk Solux bulb, the low and medium low ink limits I can see the blocks down to 2,2,2. Overall the distinction on the medium low is better than on medium. When you get to medium you can no longer easily see the 2,2,2 block. If you tilt the print there's some gloss differential there and you can see the blocks but that doesn't really help for normal viewing of the print.

Unfortunately Monaco Profiler doesn't have anything I can use to measure dMax but I do have an Eye One Extreme as well so I can at least measure dE between patches using Eye One Share. Measuring the distance between the low black patch and the medium patch I get a difference of 1.5 - 1.8 dE. Between the medium low and medium I get 0.1 dE which is within the noise floor of the i1 Pro. Another thing I noted is that if I measured multiple spots on the low setting the dE would vary by up to 0.5 - 0.6 dE. On the medium patch it was about  0.3 dE. the Medium Low was about 0.1 dE between multple spots, again within the noise floor of the i1 Pro.

Don't know that this proves anything or not but visually the medium low did appear to be the best.

Another observation, on the granger chart with the medium print I noticed what appeared to be heavy pooling in areas of purple and blue.

I haven't measured the other patches yet so I'm not sure how they're responding.

Cheers, Joe
Logged
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1274


WWW
« Reply #29 on: August 04, 2010, 10:34:57 AM »
ReplyReply

Quote from: shewhorn
I know we don't want to build a profile BUT... are these features part of the module to build a profile?
You'll use SV to measure and save the Lin file. Then load the Lin file in CB (Ink&Color>Linearization>LinearizationFile>Open), select the Chroma linearization method and then press the "Advanced Lin Control" to see the results graphed out.

Quote from: shewhorn
In CB itself if you click on ink and color, Ink Curves, Linearization, and Ink Limit are greyed out. If I'm not mistaken, in order to open that Linearization file you need to click on the Linearization tab in the Ink and Colors module and then click the open button.
Hmmmm, yes, that could be because 1) you've got an Epson x900 printer selected or 2) a limitation of the latest demo version. Can you add a 9880 printer? Use a fake IP address just to get it to load.

Quote from: shewhorn
There's a high probability that I'm doing something wrong but in the mean time I created my own test file. They had some color blocks (0,0,255; 0,255,0; 0,255,255; 255,0,0; 255,0,255; 255,255,0; 255,255,255; 128,128,128). Then I had two strips, one with a black background and one with a white. On each there were 9 blocks inside each strip. In the black strip the first block would be 0,0,0, the second 2,2,2, the third, 4,4,4, etc. I did the same thing with the white only in reverse. Finally I also had a Granger Chart on there.
Sounds like a good visual analysis file. If you haven't already, see my "Media Selection Image" at http://www.on-sight.com/downloads/

Quote from: shewhorn
Another observation, on the granger chart with the medium print I noticed what appeared to be heavy pooling in areas of purple and blue.
Good eye - I've noticed that too on the x300 printers. This kind of symptom is where, were we printing through a RIP, we would want to set the total ink limit and linearization for each individual ink - particularly watching the blue, cyan and magentas in this case. Seems like Canon's linearization (that's baked into the driver) is a bit heavy on this part of the gamut.

In this situation we've got too choices. 1) Lower the overall total ink limit until we prevent pooling in these darkest purples or 2) Ignore this area, let it pool a little and set the total ink limit for the rest of the colors.  The granger in your test file represents the absolute wide rage of colors the printer is capable of, but not necessarily what you'll ever use. Canon's x300 inks "dive" deeply into the purples - so deep that it's kind of ludicrous to think we'd ever use that part of the gamut when printing photography. Although not ideal, I vote for the second option since, after all, it's extremely rare that we would actually have a photograph that had such deep purples to begin with. One could deal with the problem later should the need present itself. I've been watching those deep purples myself and have been looking for images that can exploit that area. I've found some deep purple flower images that do quiet well but don't print with the pooling I've seen prior to profiling.

Good thoughts Joe. It's nice to see comments on this level.
Logged

shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #30 on: August 04, 2010, 11:02:15 AM »
ReplyReply

Quote from: Onsight
You'll use SV to measure and save the Lin file. Then load the Lin file in CB (Ink&Color>Linearization>LinearizationFile>Open), select the Chroma linearization method and then press the "Advanced Lin Control" to see the results graphed out.


Hmmmm, yes, that could be because 1) you've got an Epson x900 printer selected or

Bingo!!! Okay, added the 9880 and was able to get to the linearization tab, opened the target that I'd previously scanned in using SV, hit "reset baseline", was then able to select chroma for lin method and now I can see the graphs. AWESOME!!! Okay... back to the salt mines. First I have to disassembled my Spectroscan table and locate whatever component got toasted (crossing fingers and hoping for a fan... it still works just fine so... errr... it can't be something critical but anyone who has worked in an R&D department KNOWS that smell by heart... and also knows that it's never a good thing LOL).

Be back sometime soon with some quantifiable data!

Cheers, Joe

Logged
shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #31 on: August 04, 2010, 05:40:37 PM »
ReplyReply

Okay, here we go.

HFA William Turner. Base paper suggested by the MCT was HW Textured. I created 5 media types, Low, Medium Low, Medium, Medium High, and High (referring to the amount of ink). I'd like to say I know what I'm looking at but errr... ummm... Well.. okay I know what I'm looking at but I'm not sure what it means. I can definitely say that something "BIG" happens between medium low and medium as the yellow and magenta channels look a bit more linear.

The targets were scanned in with a Spectroscan/Spectrolino table, no UV filter, in a completely dark room (aside from the glow of the 8300's LCD).

Here's some screen shots:

Logged
shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #32 on: August 04, 2010, 09:03:35 PM »
ReplyReply

More data... although this set is probably flawed as I forgot to put the UV filter on (Hahnemuhle indicates that Photo Rag Ultra Smooth has "moderate" OBAs).

Logged
shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #33 on: August 04, 2010, 10:00:55 PM »
ReplyReply

Alright... Hahnemuhle Photo Rag Ultra Smooth with the UV filter:

Logged
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1274


WWW
« Reply #34 on: August 04, 2010, 10:44:34 PM »
ReplyReply

Which Linearization method is this? Also have you played with any non-matte papers yet? For the process of getting to know this procedure, non-matte papers would be helpful.
Logged

shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #35 on: August 04, 2010, 11:41:17 PM »
ReplyReply

Quote from: Onsight
Which Linearization method is this? Also have you played with any non-matte papers yet? For the process of getting to know this procedure, non-matte papers would be helpful.

Hey Scott,

The linearization method was set to chroma. I just printed out a set of targets on Crane Museo Silver Rag but I didn't have time to scan the targets in before I left tonight. I'll do that tomorrow.

Cheers, Joe
Logged
shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #36 on: August 05, 2010, 12:26:45 PM »
ReplyReply

Here's Crane Museo Silver Rag. This set also includes the Factory provided custom media type downloaded from Crane Museo's website. Something kinda interesting going on with the Cyan channel there.

Logged
shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #37 on: August 05, 2010, 06:14:30 PM »
ReplyReply

One more to add... Proofline Metallic. Scanned the target without the UV filter. Kinda hard to say what the OBA content is. Blacklight says some. 'Tis a little brighter than Crane Museo Silver Rag (which claims to have no OBAs but then again, HFA German Etching claims to have no OBAs either and I do not thing that's true at all) but not as bad as Epson Exhibition Fiber or HFA Photo Rag Ultra Smooth.

Logged
shewhorn
Sr. Member
****
Offline Offline

Posts: 533


WWW
« Reply #38 on: August 05, 2010, 06:42:22 PM »
ReplyReply

FWIW... I posed a question to someone at Canon around the time when I started this thread and they just got some new information. Canon is indeed working on improving the MCT. There weren't many details but the implication was that the next version would have a more methodical and scientific method of determining the optimal ink parameters. He had no information on the timeline. Either way though, this is good news as it looks like Canon is really taking things a step further in terms of making it easier to use 3rd party papers.

Cheers, Joe
Logged
Scott Martin
Sr. Member
****
Offline Offline

Posts: 1274


WWW
« Reply #39 on: August 06, 2010, 09:45:40 AM »
ReplyReply

These graphs don't look right, and that's probably my fault by not suggesting you zero out the settings. Under the "Ink Curves" tab, set all the values to 100%. Under the Linearization tab, hit the "Reset Baseline" button, reload your Lin file and then re graph by going to "Advanced Lin Control". Then you'll see the desired plateaus.
Logged

Pages: « 1 [2] 3 4 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad