I tested EAstudio on D1 timeframe from 2009 but it created strategies not so robust on a earlier period. I think I'll resign for now on D1.

Then I passed to H4, where .json file have around 25.000 bars from 2001-01-02 20:00, so a very far start time.

I kept 2009 as start time, and after few optimizations your tool gave me around +7.000$ net balance with 100% from 2009 febrary to Dicember 2016.

I went on Mt4 and everytick backtest gave around 4.309$ net profit. So a difference of almost -3.000$.

I'm sorry I cannot  upload images becasue the button postimage doesn't seem to accept images created in .png format, and the button "upload image" doesn't appear.

Should Mt4 give a similar result to the worst Montecarlo test net profit, right?

Later you can test how the strategy behaves on different periods by using the Data Horizon tool. You can easily set the data series starting date there:

After 2009.02.01 there are 2 days missing every 5.

Please see image, there's still something I can do or is it a bug of Mt4 Period Converter not working well on D1 timeframe?

Your script now works well after changing the line code as you shown.

The script exports D1 bars since 2007.01.01 because most older data are broken (missing High and Low ).

You can change that in the script. Please find the following code and modify the year. Set 2000 or 1900 for example.

    if (period == PERIOD_D1)
    {
       datetime from = D'2007.01.01 00:00';
       datetime to   = TimeCurrent();
       bars = CopyRates(symbol, period, from, to, rates);
    }

Before to the correction, scroll your D1 chart to the beginning and check the data quality.

As you can see, on EurUsd and timeframe D1 there are 4576 records (collected after using Period Converter script with 1440), but your script creates only a file with 2707 bars. All other timeframes data are exported correctly.

Please what can I do? Thank you,
Angelo