Hi team,
We noticed the issue when EMF image is created and PDF is published from Excel 3-D graph
with gradient fill in data series.
The gradient fill colors do not have same look and fill as in the input Excel graph.
Attached is zip archive illustrating the issue, it contains:
test3D.docx - input excel template with 3D graph named range “Chart1”;
test3D.xlsx-v24.10.0.emf - output EMF image;
test3DChart1-v24.10.0.pdf - output after save as PDF;
Gradient_glow_in_PDF_vs_Excel.png - comparison between input Excel and output PDF, illustrating the issue;
TestEMFImageAndPdfFromExcelRange.java - sample test program to reproduce the issue
This issue occurs under both Linux 7 and Windows 11 OS.
Are there any options that can be applied so the output gradient fill in EMF image and PDF would match input Excel graph?
@oraspose
By testing with sample files and code on the latest version v24.10, we can reproduce the issue. Found that gradient fill in 3D graph data series does not show properly when converting file to PDF or EMF image.
We have opened the following new ticket(s) in our internal issue tracking system and will deliver their fixes according to the terms mentioned in Free Support Policies.
Issue ID(s): CELLSJAVA-46157
You can obtain Paid Support Services if you need support on a priority basis, along with the direct access to our Paid Support management team.
Hi @oraspose
We have made some optimizations and adjustments to the images. Please confirm if they meet your requirements. Thank you. 241101.zip (17.5 KB)
Sorry, but attached CELLSNET57054_new_241101.png is almost exactly the same as CELLSNET57054_old_241101.png.
You can see the difference if you copy input excel graph and simply paste it into new Word document.
Thank you.
We apologize, but it is difficult to identify the difference. Could you please provide additional screenshots with detailed information to highlight the specific problematic areas profoundly by comparing the CELLSNET57054_new_241101.png file with the Microsoft Excel view of the chart? This will help us understand your needs precisely and address the issue accordingly and promptly.
Attached is the description of the problem.
Please let us know if something is not clear and needs clarification.
Thank you. CELLSJAVA-46157_issueDescription.docx (151.0 KB)
Thanks for providing the document containing screenshots to describe and highlight the issues. We have logged it against your existing ticket “CELLSJAVA-46157” into our database. This will help us understand your issue and then to evaluate your problems. We will try to figure them out soon.
Thanks for letting us know. However, your issue is “In Progress” and it could be fixed in the next few days. Once we have an update on it, we will let you know here.
Hi @oraspose
Regarding the issue of gradient colors, we have conducted research and our current findings are as follows, Please download the zip package first: 241108.zip (36.3 KB)
CELLSNET57054_old.png: The drawing effect of the current Aspose.cells version.
CELLSNET57054_Excel.png: The drawing effect in Excel.
CELLSNET57054_241101.png: The result after we making some adjustments.
CELLSNET57054_241108.png: The result after further adjustments.
Please compare these pictures carefully, you can find the changes after adjustment.
Unfortunately, Both Java and C# graphics libraries will get results of “CELLSNET57054_old.png” when attempting to draw the gradient color parameters set in Excel.
Excel might have made additional adjustments, resulting in more yellow color and less black color.
Please note that “CELLSNET57054_241101.png” and “CELLSNET57054_241108.png” are just the results from our tests and do not fully follow the original parameters in excel (we also add additional adjustments), but still differ from the results of excel.
Since the adjustment method of Excel is not public, this issue still needs more research. If you find “CELLSNET57054_241108.png” performance better than “CELLSNET57054_old.png”, please let us know, it will be helpful to our research and adjustment, thank you.
In order to make the contrast more obvious, I wrote a simple project: GradientTest.zip (649.3 KB)
The project creates an output file “gradient_square.png”.
It can be seen that the graphics library’s handling of gradient colors differs from that of the Excel interface. Compare241108.png (28.7 KB)
Additionally, we noticed that if the black color on the left side of the gradient in the Excel interface settings is adjusted from 0% to 1%, there is a sudden change in the effect. The display effect of OpenOffice and LibreOffice is also closer to the Windows Graphics Library. Therefore, Excel must be doing some unique processing.
The above are our current findings, and this issue may require more time for further research.
Thanks for taking time researching this issue. Yes, CELLSNET57054_241108.png looks better than the old graph (CELLSNET57054_old.png) and not darker than the input Excel graph. This result is much closer to the expected output.
Thank you.