Dark colors muddy/little detail

[IIDave]

Ok, so here's the story:
I am trying to move this shop away from Wasatch and have everything run off of ONYX. (Right now we have two RIPS... ugh...) My custom profiles through ONYX have been pretty consistently better than Wasatch and the dot patterns are smoother through ONYX.

Problem is:
We do some art repro work and my canvas profile sucks... I've tried to profile the media a couple of times now and my darks always seem to come out very muddy or they lack the detail of the Wasatch profile for the same media.

Help? Anybody? Should I try more patches on my calibration swatch step (normally I do 19 -- 5% difference).

Thanks!

[Correct Color]

Since Wasatch doesn't have a built-in profile-making engine, the first question would be what software you used to make your Wasatch profiles. Or are you comparing against a stock Wasatch profile?

Next question--if you made your Wasatch profile--would be if you're using the same software to make your profiles in Onyx.

Frankly, Wasatch isn't a bad RIP at all but their included stock profiles are pretty weak, and their standard input settings are pretty close to a disaster. If your comparison all along has been to Wasatch stock profiles, then it's not altogether a certainty the profiles you've been making are all that good; it could be that you're just not comparing them to a very high standard, and you've simply come upon a media that Wasatch had a good profile for, or that is exposing a weakness in your process/knowledge.

But here are a couple tips: The problem isn't in the number of steps in your linearization. Not only would that not cause your problem, but in Wasatch the number is fixed--off the top of my head I think it's twenty.

Second tip is this: If you really want first-rate profiles, hire someone to come and do them for you; if you really want to learn, hire someone to come teach you.

In the end, it'll be the best money you ever spent.


Mike Adams
Correct Color

[IIDave]

(02-08-2010 01:18 PM)Correct Color Wrote:  Since Wasatch doesn't have a built-in profile-making engine, the first question would be what software you used to make your Wasatch profiles. Or are you comparing against a stock Wasatch profile?

Next question--if you made your Wasatch profile--would be if you're using the same software to make your profiles in Onyx.

Frankly, Wasatch isn't a bad RIP at all but their included stock profiles are pretty weak, and their standard input settings are pretty close to a disaster. If your comparison all along has been to Wasatch stock profiles, then it's not altogether a certainty the profiles you've been making are all that good; it could be that you're just not comparing them to a very high standard, and you've simply come upon a media that Wasatch had a good profile for, or that is exposing a weakness in your process/knowledge.

But here are a couple tips: The problem isn't in the number of steps in your linearization. Not only would that not cause your problem, but in Wasatch the number is fixed--off the top of my head I think it's twenty.

Second tip is this: If you really want first-rate profiles, hire someone to come and do them for you; if you really want to learn, hire someone to come teach you.

In the end, it'll be the best money you ever spent.


Mike Adams
Correct Color

The Wasatch profiles are stock -- I haven't bothered to try to work through those to make anything custom. So yes, I am comparing my custom ONYX profiles to the stock Wasatch profiles.

I am aware that hiring someone in to profile would really be the best thing to do, but the way the economy is right now it is a hard thing to convince the powers that be to spend the money. As far as training goes I have been learning as I go and asking questions when I seem to hit a wall (my profiles have been improving so far).

Any thoughts on what may be making my profile muddy? Ink Restrictions? Ink Limits?

I have started from scratch on this media and we'll see how things turn out.

[Correct Color]

Y'know...it could be all of the above.

And I hope you'll understand and not take offense, but I've spent an entire lifetime learning what I know about color. And I live in the same economy as you. Do you give away your service for free just because times are bad?

Have the powers that be fired their lawyers or their doctors or their accountants? Have they asked you to take out their kids' tonsils? Have they quit buying media and ink?

They sell color for a living. It's all they've got to sell and yet they cut corners on the absolute basics of their business?

Makes very little sense to me.

Fact is anything could be causing your issue. The only way to know is to be there to see. And anyone chiming in on a discussion board telling you different is just sending you down the primrose path.

[Scott]

(02-08-2010 01:54 PM)Correct Color Wrote:  And I hope you'll understand and not take offense, but I've spent an entire lifetime learning what I know about color. And I live in the same economy as you. Do you give away your service for free just because times are bad?

Have the powers that be fired their lawyers or their doctors or their accountants? Have they quit buying media and ink?

They sell color for a living. It's all they've got to sell and yet they cut corners on the absolute basics of their business?

Makes very little sense to me.

Fact is anything could be causing your issue. The only way to know is to be there to see. And anyone chiming in on a discussion board telling you different is just sending you down the primrose path.

Unfortunately, not every shop owner out there values and/or cares about color consistency enough to pay for proper training and service. This is how it's always been at my shop. Colors are "good enough" and when there's a problem we just deal with it. I'm not saying it makes sense, and we've definitely wasted enough media, ink and labor to pay for a service such as yours, but that's the way it goes in my world.

I completely understand that you can't give away your service for free, and we don't expect you to by any means. That is why I started this site so that as Onyx peers we can lean on each other for help. We've been spoiled here however with some very experienced users when it comes to profiling & color management (thanks Hotspur & Marco). They're not giving away the milk for free, as they say, but they're definitely giving us a taste and for that we are grateful.

I personally appreciate your presence here Mike and welcome and any all advice you are willing to share with us poor print operators

[Correct Color]

I know. I know.

And please forgive me if I sounded brusque or harsh. It's hard to do tone online sometimes.

And...we're all dealing with an economy that would put anyone on edge.

Wasn't my intent to offend anyone, if that came across. And if I can offer a solution or suggestion on any range of issues I'm glad to do it.

But my suggestion to anyone who really, really, really wants to master color is to get professional help. It's worth it.

Anyway, hopefully I didn't step on any toes.

Wasn't my intent.



Mike

[Scott]

My reply was suffering the same limitations, so no worries. Conversing online is definitely a challenge when you can't hear inflections and/or read body language. I wish that the big dog understood how critical proper color management is. They get frustrated when there are issues, but just assume that we should be able to just figure it out. I'm a pretty smart guy, but profiling has been a PITA for me, and I for one would REALLY benefit from having someone like you sit next to me at MY shop and walk through everything. I'd be good to go from there, but doing it all on my own has been a true challenge.

[Hotspur]

So now you guys have kissed & made up can we help IIDave? ;-)

IIDave can you explain "muddy"? - lacking in gamut? too dark overall? smearing & wet?

Try and give a good description of the problem so we might stand a chance.

Also what sort of canvas is it? natural or synthetic, large weave? etc

And what of your profiling workflow? How do you do your ink restrictions? density or colourant? do you make adjustments to your curves when you calibrate? where are you calling off your ink limits? and how do you do your icc's - inside the rip? if so what options do you set? What measurement device do you use? etc etc

If you give us something to work with we may be able to steer you in the right direction.

PS Marco only posts because Dutch TV is terrible :-)

[IIDave]

(02-08-2010 03:22 PM)Hotspur Wrote:  IIDave can you explain "muddy"? - lacking in gamut? too dark overall? smearing & wet?

Dull. Not a good black, not enough contrast between dark mid-tones and shadows.

(02-08-2010 03:22 PM)Hotspur Wrote:  Also what sort of canvas is it? natural or synthetic, large weave? etc

LexJet Sunset Select Matte canvas. It is a coated tight weave canvas.

(02-08-2010 03:22 PM)Hotspur Wrote:  And what of your profiling workflow? How do you do your ink restrictions? density or colourant? do you make adjustments to your curves when you calibrate? where are you calling off your ink limits? and how do you do your icc's - inside the rip? if so what options do you set? What measurement device do you use? etc etc

Ink restrictions: I have been doing a combination of Colormetric and Densitometric. For Cyan and Magenta I use the Colormetric and pick the point where the chroma hooks to the left. When I try to limit Yellow this way the line seems to be almost horizontal to the left, so I have taken to switching to denisty for Yellow and Black. I will pick the point just before the plateau starts to drop off. (Black is sometimes a visual descision -- I will pick the point just before I start to see it gray out.)

My process for Ink Restrictions is to "Use Darkest Levels Only" and set them to 100%. After I have picked my limit (I have found that generally I end up around: 65%-75% C, 70-80% M, 75%-85% Y and 65%-75% K for this HP5500), I "Set Lighter Levels" to 80%. So far I haven't mess with the transitions and will leave them at 0.

ICCs: I am running PosterShop right now, so no internal ICC engine... I have been generating my ICC through I1 Match and importing it into media manager after I have finished my Ink Restrictions and Ink Limits. Using an I1 Pro -- they don't really give me any options... it just generates the profile.


The biggest place I get stuck right now is on the Linearization... It seems that I almost always have "Significant Spikes" when I hit "Analyze". I have tried using few patches and more patches for calibration, but it doesn't seem to affect it much. I generally don't change anything in the linearization window. There never seems to be the plateaus or hooks that the ONYX white paper on profiling describes...

Does this help?

---

Ok... Slightly different subject, but possibly related:

I was just checking through the stock profiles from ONYX for my printers and noticed a strange thing.

In the Ink Restrictions setting my solvent printer (HP DesignJet 9000) has fewer settings than my Dye and Pigment aqueous ink printers (HP Designjet 5000 and 5500). Solvent has lc and C, lm and M, Y, and K with one set of transition controls. The 5000/5500 have lc C1 and C2, lm M1 and M2, Y1 and Y2, and K1 and K2 with two sets of transition controls.

The real big flag for me that stands out as odd is that the Solvent settings seems to go in line with what I have learned about Ink Restrictions -- the light ink is approx 80% of the dark ink restriction -- but, the Aqueous restrictions are flipped around. On the Auqueous the light inks are set at a higher percentage than the dark ink.

Example settings for the aqueous:
lc 90%
C1 80%
C2 30%

Example settings for the solvent:
lc 76%
C 86%

Again, these are the stock profiles in ONYX (both for different kinds of paper), not my own.

[Hotspur]

HPs are a bit different as you have discovered (your HP9000 is really a Seiko)

I'll use Cyan here as my example but it applies to the others too.

The C1 and C2 etc you refer to are what they call "double strike" ie two ink strikes to build up the required density (HPs do not have variable dot sizes)

This double strike applies to the dark inks only in your case. Thus to get 100% of your dark Cyan head firing you must select 100% for both C1 and C2 (think of them as contributing 50% each to the maximum amount of Cyan you can possibly put down at this pass rate)

Thus set your light inks to zero and set both your C1 and C2 to 100% each, which together gives you the maximum dark Cyan output.

Run your ink restrictions as above and draw your conclusions as normal.

Now for the tricky bit. Say your results tell you that you need to set Cyan to 60%.

In theory you can either set both C1 and C2 to 60% or set C1 to 100% and C2 to 20% - both will give you an overall 60% restriction.

Onyx suggest you use the second route - use C1 first and whatever is left (if any) for C2. This is because it means you get at least one full dot spread on your page to give good coverage, with a smaller dot following to give the final density.

In our case a random selection of 20% of the dots from the first C1 strike will be struck a second time. It is often assumed wrongly that the printer will send a 20% sized dot down (in our example) following on top of the 100% dot. HPs are fixed dot printers and as such they have to use a full 100% second dot but only do so on top of 20% of the exisiting dots.

Sorry if this sounds complicated but it's why they don't recommend you do two dots at 60% each - the total coverage would suffer and it would look more grainy.

Now you can set your light inks.

In this example say our total dark Cyan restriction is set at 60% - this means our light Cyan should be 80% of 60% - ie 48%.

This assumes that 80% of your dark ink restriction is correct in the first place. It sounds quite high to me although this is the suggested start point for aqueous ink printers.

I would start at around 50% of my dark ink restrictions and see how it goes. One thing to note the auto-setting buttons don't work for double strike printers - eg set all to 100% doesn't! just set everything manually.

Solvents usually go between 30 and 50% for the lights so this is why you see the difference in the stock profiles - but don't be afraid to be down at these levels if need be.

Also be aware that Onyx's advice changes over the years as they change the software and learn more about the process. The HP5000 / 5500 are quite old now and the profile libraries were built a long time ago so you often find contradictions in older profiles Vs current advice / training - go with the advice - 90% for light ink is way too high!

Note - there is no speed penalty / advantage regarding double-strike. It is not making two passes to create the second strike.

Try a profile using this workflow and see if it makes a difference.

You don't mention how you are setting your ink limits so this could be an issue.

I have a feeling in my water you may be using a bit too much ink - stay on the conservative side and make sure your black ink compensation is set to 1 unless you have to go very low (which you shouldn't on a good canvas)

I have tried the eye1 software & didn't like the results I have to say but that's all you have so no point making any changes to the icc side of things I guess.

Regarding the linearization there is a fix for the "significant spikes" issue.

Don't hit the button

Let us know how you get on.

[Douglas]

IIDave....
I'd recommend not using the stock 5500 profiles as a baseline for you own profiles. They are so old, that....
1. Our software has changed significantly in how it handles the data provided by the user and color device.
2. Because of changes in the software many of the approaches used in the past will actually cause problems. Particularly recalibration (Ex: we used to recommend dragging target density to 100%)

- Your approach to ink restrictions is sound in my opinion.
- The spikes you mention in the linearization are common with textured materials. I'd recommend you print two separate copies of the calibration swatch, set the number of patches from 31 to 24, set the color devices to make two readings. (read each swatch individually, don't read the same swatch).
This will result in the same patches on each swatch having different placement of the pits in the texture. The readings will be averaged to get a better representation of the actual coverage through the range of the swatches.
- Regarding muddy color. Are you setting a Black Ink Compensation in the Ink Limit? If not, you may want to consider doing so. What this will do, is add color back into each combo patch, relative to the amount of K ink it would take to obtain the same level of darkness.
In other words, the ICC will use the black generation settings to otherwise provide the same makeup. (Sorry if this explanation is weird, it's one of the stranger concepts of the Ink Limit)
- You should also do two readings of the ICC swatch for textured materials, however you can likely get away with printing one copy as a short cut.
Lastly, the number of patches used and the build options of the ICC will have a direct impact on muddiness vs clarity. Particularly with regard to what I said about the ink limit.

Hope this helps and my apologies if this reply is a bit rushed.

[IIDave]

Using all of the advice here I seem to be in a much better place with my profiles. Thanks everybody!

Just one more question (for now) in regards to the BIC.

What is the "standard" BIC? It is default set to 1, when I need to increase it should I just move by tenths or whole numbers? (When one clicks on the up arrow it moves by tenths...)

Thanks!

[Douglas]

With regards to the BIC. While there is not a specific ratio for setting it in relation to the selected Ink Limits (1-7).
The following examples are typical.

If the average Ink Limits (1-7) are:

* 1.5 - 2.0 / BIC = 4-5
* 2.0 - 2.5 / BIC = 3-4
* 2.5 - 3.0 / BIC = 2-3
* 3.0 - 4.0 / BIC = 1-2

[IIDave]

(02-24-2010 01:15 PM)Douglas Wrote:  With regards to the BIC. While there is not a specific ratio for setting it in relation to the selected Ink Limits (1-7).
The following examples are typical.

If the average Ink Limits (1-7) are:

* 1.5 - 2.0 / BIC = 4-5
* 2.0 - 2.5 / BIC = 3-4
* 2.5 - 3.0 / BIC = 2-3
* 3.0 - 4.0 / BIC = 1-2

Wow... maybe that is my problem there... If I were to average my ink limits it they it would probably end up around 2.25. I have never set my BIC above 1.3. I guess I will have to give 3 a try.

The one time I cranked it up to see what it did it wasn't good... but that was before I had any clue about what I was doing.

My other big fear about turning the BIC up too high was having the printer shade with black too much... One of the issues I have been trying to solve here is getting good colors without have a strong black dot pattern.