cattivik66
New Printer Member
- Joined
- Mar 27, 2026
- Messages
- 1
- Reaction score
- 1
- Points
- 0
- Printer Model
- ET-18100
Hi everyone,
I'm a new user but was reading the forum without actively participating since some time.
I’ve been working on a personal project lately and I thought the community here might find it interesting—or at least helpful for some quick experimentation.
I’ve built a web-based utility designed to process measurement data +generate custom ICM profiles and to analyze pre-created ICM files. The goal was to create a streamlined way to go from a .ti3 file to a finished, mathematically smoothed profile, along with a detailed HTML quality report.
Why did I build this?
Adding a feature after another and now the tool tests a lot of parameters to detect issues like metamerism risks, over-inking, etc.
Evaluating profile quality and printer "stress" used to be a struggle for me, as was managing complex settings like FWA compensation. I developed this utility to visualize "ICM DNA" and get instant improvement hints, improving it over time with features like metamerism prediction and over-inking. Honestly I'm not sure how good is this tool, perhaps it is just a big loss of my time but I am open to any feedback!
The "Beta" Reality Check (Please Read!):
This tool is very much in Beta, it does not even have a name. I’m not entirely sure what the long-term plan is for it yet, but I wanted to get it into the hands of some real-world users to have some feedback. I hope this can be somehow useful, perhaps the whole project doesn't have any sense. Let's see how it will evolve.
There are a few major limitations you should be aware of:
I’d love for some of you to run some data through it and see if the results match your expectations. I’ve created a specific beta code for PrinterKnowledge users:
I’ve implemented these usage limits to prevent my modest server from being overwhelmed. Since the project is still in a very preliminary stage, these 'guardrails' help ensure the service remains stable for everyone.
You can find the tool here: https://profile.sboscolo.it
After uploading, you’ll receive a unique tracking code and password to monitor your job's progress. If a process fails, it’s typically because the input wasn’t a standard Argyll dataset and in ICC V4 standard —though, being a beta, a 'good old-fashioned bug' is always a possibility.
I’m looking forward to hearing your thoughts, feedback, or even just "it broke my file" reports.
Cheers!
PS: If you'd like to see a sample report before uploading your own data, feel free to use these demo credentials on the status page:
What's next?
At the moment, the system calculates a single global score by analyzing both the measurement data and the resulting ICM. I plan to eventually split this into two distinct metrics: Source Data Integrity (.ti3) and Profile Performance (ICM). This should allow for a better comparison between ICM made with different profiling engines and provide clearer feedback on whether a bottleneck is caused by the raw scan data or the profile's own mathematical optimization.
As for the presentation: the HTML report is currently very utilitarian (and admittedly a bit rough around the edges), but it serves its purpose for my internal testing. I’m also planning to integrate graphical visualizations soon to make the data analysis more intuitive and rapid to digest at a glance.
I'm a new user but was reading the forum without actively participating since some time.
I’ve been working on a personal project lately and I thought the community here might find it interesting—or at least helpful for some quick experimentation.
I’ve built a web-based utility designed to process measurement data +generate custom ICM profiles and to analyze pre-created ICM files. The goal was to create a streamlined way to go from a .ti3 file to a finished, mathematically smoothed profile, along with a detailed HTML quality report.
Why did I build this?
Adding a feature after another and now the tool tests a lot of parameters to detect issues like metamerism risks, over-inking, etc.
Evaluating profile quality and printer "stress" used to be a struggle for me, as was managing complex settings like FWA compensation. I developed this utility to visualize "ICM DNA" and get instant improvement hints, improving it over time with features like metamerism prediction and over-inking. Honestly I'm not sure how good is this tool, perhaps it is just a big loss of my time but I am open to any feedback!
The "Beta" Reality Check (Please Read!):
This tool is very much in Beta, it does not even have a name. I’m not entirely sure what the long-term plan is for it yet, but I wanted to get it into the hands of some real-world users to have some feedback. I hope this can be somehow useful, perhaps the whole project doesn't have any sense. Let's see how it will evolve.
There are a few major limitations you should be aware of:
- Argyll Optimized: The system is strictly tuned for Argyll datasets (.ti3) and Argyll-generated profiles (.icm/.icc). You can upload other v2.x profiles, but they will likely receive lower scores because they lack the specific metadata Argyll includes which I partially use for the calculations.
- Dye-Ink Focus: Testing has been limited to a single dye-based printer (Epson ET-18100). It has not yet been optimized for pigment inks or other printer architectures.
- Hardware Bias: Since I only have an i1Pro 2 spectrophotometer, that is the only device used for testing. Final scores may vary if you are using different measurement hardware (ex: different gamut size).
- Unoptimized: This is raw, unoptimized code. Expect bugs!
- Processing Speed: The server is modest. Processing a .ti3 can take a few minutes, and a queue will form if multiple people upload at once. Pro-tip: Uploading an Argyll .icm file is significantly faster than a .ti3 and provides the same analysis output.
- The Final Score: I’ve tried to make the scoring algorithm as impartial as possible, but it is still experimental. Ultimately, the most important metric is whether you like the final print! Think of the score as a measure of "ICM overhead"—how hard the profile has to work to manage your specific printer/paper combination.
I’d love for some of you to run some data through it and see if the results match your expectations. I’ve created a specific beta code for PrinterKnowledge users:
I’ve implemented these usage limits to prevent my modest server from being overwhelmed. Since the project is still in a very preliminary stage, these 'guardrails' help ensure the service remains stable for everyone.
You can find the tool here: https://profile.sboscolo.it
After uploading, you’ll receive a unique tracking code and password to monitor your job's progress. If a process fails, it’s typically because the input wasn’t a standard Argyll dataset and in ICC V4 standard —though, being a beta, a 'good old-fashioned bug' is always a possibility.
I’m looking forward to hearing your thoughts, feedback, or even just "it broke my file" reports.
Cheers!
PS: If you'd like to see a sample report before uploading your own data, feel free to use these demo credentials on the status page:
- Tracking: 13e799b4abe2f4bb
- Password: 2BnSJwbH
What's next?
At the moment, the system calculates a single global score by analyzing both the measurement data and the resulting ICM. I plan to eventually split this into two distinct metrics: Source Data Integrity (.ti3) and Profile Performance (ICM). This should allow for a better comparison between ICM made with different profiling engines and provide clearer feedback on whether a bottleneck is caused by the raw scan data or the profile's own mathematical optimization.
As for the presentation: the HTML report is currently very utilitarian (and admittedly a bit rough around the edges), but it serves its purpose for my internal testing. I’m also planning to integrate graphical visualizations soon to make the data analysis more intuitive and rapid to digest at a glance.
