Using Python: Encoding MKV to MP4 with 2 pass LOUDNORM

[RogerTango]

Folks, Ive worked hard to build this tested script, and Ill help the best I can if you are trying to adapt it... but no promises implied or expressed.

Short story: DOS Batch is all I ever used, and when it came to something more advanced than what (limited) DOS could provide, I had to bang out an executable in QB or VB.NET to do a special task, and even then it as a mix match of code and script... I got tired of all this and discovered Python can... do everything I want to do and no longer, EVER, use DOS again.

Below is the current script, lets call it "Version 1.0" which I use for personal application and is tailored to me. Feel free to play with it as you wish, but I cant baby step you if you know nothing about FFMPEG, command line arguments *or* Python.

I do hope this helps, if anyone else wants to chime in and help anyone that has questions, feel free to..

This project will *not* be going to github or other sites, I do not have time for all that.

Use of my script is at your own risk. Developed and running on Windows 10 and Python 3.11 (no other modules are add-ins used)

Enjoy-
Andrew

"""
Encode BD Archive to Media Player MP4 with Loudnorm
Version 1.0
"""
# import required module
from pathlib import Path
import os
import subprocess
from os import system
import json


#setup some colors
CEND      = '\33[0m'
CRED    = '\33[31m'
CYELLOW = '\33[33m'
CBOLD     = '\33[1m'
CITALIC   = '\33[3m'

#setup some variables
crlf = "\r\n"
infile = ""
outfile = ""


os.system('cls')

print("Encode BD Archive to Media Player MP4 with Loudnorm")
print(crlf)

#initial setup
pathhere = str((os.path.dirname(os.path.realpath(__file__))))
sourcedirectory = "H:\BD-HEVC_ARCHIVE"
destdirectory = "W:\\NEW_VIDEOS\\BD-HEVC-HB"
pattern = "*.mkv"

#cleanup
if os.path.exists(pathhere + "\encodethis.mkv"):
    os.remove(pathhere + "\encodethis.mkv")

if os.path.exists(pathhere + "\loudnorm.txt"):
    os.remove(pathhere + "\loudnorm.txt")


# Python does not like trailing \ in string, so make sure the
# directories are defined without one, but is needed so it is added
sourcedirectory = sourcedirectory + chr(92)
destdirectory = destdirectory + chr(92)
    

print("The source directory is: ", sourcedirectory)
print("The destination directory is: ", destdirectory)
print("The file pattern is:", pattern)
print(crlf)

    
print(CITALIC + CBOLD + "The following source files were discovered:" + CEND)

files = Path(sourcedirectory).glob(pattern)
for file in files:
    print(file)

print(crlf)


files = Path(sourcedirectory).glob(pattern)
for file in files:
#    print(file)
#    print(file.name)
#    print(file.suffix)
#    justfile = file.name.replace(file.suffix, "")
#    print(justfile)
    infile = (str(file))
    # outfile = destdirectory + str(file.name.replace(file.suffix, "")) + "-BD.MKV"
    outfile = pathhere + "\encodethis.mkv"
    runstring = '!!PATHHERE!!\handbrakecli --input "!!INFILE!!" --output "!!OUTFILE!!" --encoder nvenc_h265 --quality 31 --crop 0:0:0:0 --non-anamorphic --aencoder eac3 --ab 448k --mixdown 5point1 '

    # build out run string
    runstring = runstring.replace("!!INFILE!!", str(infile))
    runstring = runstring.replace("!!OUTFILE!!", str(outfile))
    runstring = runstring.replace("!!PATHHERE!!", pathhere)
    #print(runstring)
    #print(crlf)
    
    print(crlf)
    print(CRED + "Step 1/3) Media Player Encoding " + CYELLOW + str(infile) + " to " + str(outfile) + CEND)
    system("title " + "Step 1/3) Media Player Encoding   [" + str(infile) + "]   to   [" + str(outfile) + "]")
    print(CRED + "Using: " + CEND + CITALIC + runstring + CEND)
    print(crlf)
    
    if os.path.exists(pathhere + "\encodethis.mkv"):
        os.remove(pathhere + "\encodethis.mkv")
    subprocess.call(runstring,shell=True)

    #The base file is encoded with the desired video, lets run FFMPEG on it and determine the LOUDNORM settings
    if os.path.exists(pathhere + "\loudnorm.txt"):
        os.remove(pathhere + "\loudnorm.txt")
    
              

    runstring = '!!PATHHERE!!\\ffmpeg -hide_banner -i "!!PATHHERE!!\encodethis.mkv" -nostats -af "loudnorm=print_format=json" -f null nul 2> "!!PATHHERE!!\loudnorm.txt"'
    runstring = runstring.replace("!!PATHHERE!!", pathhere)
    
    print(crlf)
    print(CRED + "Step 2/3) Reading LOUDNORM... " + CYELLOW + str(infile) + " to " + str(outfile) + CEND)
    system("title " + "Step 2/3) Reading LOUDNORM   [" + str(infile) + "]   to   [" + str(outfile) + "]")
    print(CRED + "Using: " + CEND + CITALIC + runstring + CEND)
    print(crlf)
    subprocess.call(runstring,shell=True)


    #read JSON loudnorm.txt
    #open text file in read mode
    text_file = open(pathhere + "\\loudnorm.txt", "r")

    #read whole file to a string
    data = text_file.read()

    #close file
    text_file.close()

    #remove all but JSON text
    before, sep, after = data.partition('{')
    data = "{" + after
    

    #parse
    
    print(crlf)
    print("Parsing Data: " + data)
    print(crlf)
    

    y = json.loads(data)

    input_i = y["input_i"]
    input_tp = y["input_tp"]
    input_lra = y["input_lra"]
    input_thresh = y["input_thresh"]

    #print(input_i)
    #print(input_tp)
    #print(input_lra)
    #print(input_thresh)

    outfile = destdirectory + str(file.name.replace(file.suffix, "")) + "-HEVC.MP4"

    runstring = '!!PATHHERE!!\\ffmpeg -hide_banner -fflags +genpts -i "!!PATHHERE!!\\encodethis.mkv" -filter:a "loudnorm=linear=true:i=-23.0:lra=7.0:tp=-2.0:offset=0.39:measured_I=!!input_i!!:measured_tp=!!input_tp!!:measured_LRA=!!input_lra!!:measured_thresh=!!input_thresh!!"  -c:a aac -ar 48000 -b:a 192k -ac 2 -c:v copy "!!OUTFILE!!"'

    runstring = runstring.replace("!!input_i!!", input_i)
    runstring = runstring.replace("!!input_tp!!", input_tp)
    runstring = runstring.replace("!!input_lra!!", input_lra)
    runstring = runstring.replace("!!input_thresh!!", input_thresh)
    runstring = runstring.replace("!!OUTFILE!!", str(outfile))
    runstring = runstring.replace("!!PATHHERE!!", pathhere)


    print(crlf)
    print(CRED + "Step 3/3) Applying LOUDNORM... " + CYELLOW + str(infile) + " to " + str(outfile) + CEND)
    system("title " + "Step 3/3) Applying LOUDNORM   [" + str(infile) + "]   to   [" + str(outfile) + "]")
    print(CRED + "Using: " + CEND + CITALIC + runstring + CEND)
    print(crlf)
 
    subprocess.call(runstring,shell=True)

#cleanup
if os.path.exists(pathhere + "\encodethis.mkv"):
    os.remove(pathhere + "\encodethis.mkv")

if os.path.exists(pathhere + "\loudnorm.txt"):
    os.remove(pathhere + "\loudnorm.txt")
        

#pause
input("Press Enter to continue...")

[RogerTango]

Version 1.1

Added:
filename added to title metadata, some programs & DLNA software use it
date and time added to comment metadata, to track when a file was encoded

"""
Encode BD Archive to Media Player MP4 with Loudnorm
Version 1.1
"""
# import required module
from pathlib import Path
import os
import subprocess
from os import system
import json
from datetime import datetime


#setup some colors
CEND      = '\33[0m'
CRED    = '\33[31m'
CYELLOW = '\33[33m'
CBOLD     = '\33[1m'
CITALIC   = '\33[3m'

#setup some variables
crlf = "\r\n"
infile = ""
outfile = ""


os.system('cls')

print("Encode BD Archive to Media Player MP4 with Loudnorm")
print(crlf)

#initial setup
pathhere = str((os.path.dirname(os.path.realpath(__file__))))
#sourcedirectory = "Y:\\Test-BD_MKV_TRIMMED"
sourcedirectory = "H:\\BD-HEVC_ARCHIVE"
destdirectory = "W:\\NEW_VIDEOS\\BD-HEVC"
pattern = "*.mkv"

#cleanup
if os.path.exists(pathhere + "\encodethis.mkv"):
    os.remove(pathhere + "\encodethis.mkv")

if os.path.exists(pathhere + "\loudnorm.txt"):
    os.remove(pathhere + "\loudnorm.txt")


# Python does not like trailing \ in string, so make sure the
# directories are defined without one, but is needed so it is added
sourcedirectory = sourcedirectory + chr(92) + chr(92)
destdirectory = destdirectory + chr(92) + chr(92)
    

print("The source directory is: ", sourcedirectory)
print("The destination directory is: ", destdirectory)
print("The file pattern is:", pattern)
print(crlf)

    
print(CITALIC + CBOLD + "The following source files were discovered:" + CEND)

files = Path(sourcedirectory).glob(pattern)
for file in files:
    print(file)

print(crlf)


files = Path(sourcedirectory).glob(pattern)
for file in files:
#    print(file)
#    print(file.name)
#    print(file.suffix)
    justfile = file.name.replace(file.suffix, "")
#    print(justfile)
    infile = (str(file))
    # outfile = destdirectory + str(file.name.replace(file.suffix, "")) + "-BD.MKV"
    outfile = pathhere + "\encodethis.mkv"
    runstring = '!!PATHHERE!!\handbrakecli --input "!!INFILE!!" --output "!!OUTFILE!!" --encoder nvenc_h265 --quality 30 --crop 0:0:0:0 --non-anamorphic --aencoder eac3 --ab 448k --mixdown 5point1 '

    # build out run string
    runstring = runstring.replace("!!INFILE!!", str(infile))
    runstring = runstring.replace("!!OUTFILE!!", str(outfile))
    runstring = runstring.replace("!!PATHHERE!!", pathhere)
    #print(runstring)
    #print(crlf)
    
    print(crlf)
    print(CRED + "Step 1/3) Media Player Encoding " + CYELLOW + str(infile) + " to " + str(outfile) + CEND)
    system("title " + "Step 1/3) Media Player Encoding   [" + str(infile) + "]   to   [" + str(outfile) + "]")
    print(CRED + "Using: " + CEND + CITALIC + runstring + CEND)
    print(crlf)
    
    if os.path.exists(pathhere + "\encodethis.mkv"):
        os.remove(pathhere + "\encodethis.mkv")
    subprocess.call(runstring,shell=True)

    #The base file is encoded with the desired video, lets run FFMPEG on it and determine the LOUDNORM settings
    if os.path.exists(pathhere + "\loudnorm.txt"):
        os.remove(pathhere + "\loudnorm.txt")
    
              

    runstring = '!!PATHHERE!!\\ffmpeg -hide_banner -i "!!PATHHERE!!\encodethis.mkv" -nostats -af "loudnorm=print_format=json" -f null nul 2> "!!PATHHERE!!\loudnorm.txt"'
    runstring = runstring.replace("!!PATHHERE!!", pathhere)
    
    print(crlf)
    print(CRED + "Step 2/3) Reading LOUDNORM... " + CYELLOW + str(infile) + " to " + str(outfile) + CEND)
    system("title " + "Step 2/3) Reading LOUDNORM   [" + str(infile) + "]   to   [" + str(outfile) + "]")
    print(CRED + "Using: " + CEND + CITALIC + runstring + CEND)
    print(crlf)
    subprocess.call(runstring,shell=True)


    #read JSON loudnorm.txt
    #open text file in read mode
    text_file = open(pathhere + "\\loudnorm.txt", "r")

    #read whole file to a string
    data = text_file.read()

    #close file
    text_file.close()

    #remove all but JSON text
    before, sep, after = data.partition('{')
    data = "{" + after
    

    #parse
    
    print(crlf)
    print("Parsing Data: " + data)
    print(crlf)
    

    y = json.loads(data)

    input_i = y["input_i"]
    input_tp = y["input_tp"]
    input_lra = y["input_lra"]
    input_thresh = y["input_thresh"]

    #print(input_i)
    #print(input_tp)
    #print(input_lra)
    #print(input_thresh)
    
    #Get current date and time
    time_now = datetime.now()
    current_time = time_now.strftime("%H:%M:%S")
    #print("The current time is", current_time)
    current_date = time_now.strftime("%m-%d-%Y")
    #print("The current date is", current_date)
    current_date_time = current_date + " " + current_time
    

    outfile = destdirectory + str(file.name.replace(file.suffix, "")) + "-HEVC.MP4"

    runstring = '!!PATHHERE!!\\ffmpeg -hide_banner -fflags +genpts -i "!!PATHHERE!!\\encodethis.mkv" -metadata title="!!JUSTFILE!!-HEVC" -metadata comment="!!CURRENT_DATE_TIME!!" -filter:a "loudnorm=linear=true:i=-23.0:lra=7.0:tp=-2.0:offset=0.39:measured_I=!!input_i!!:measured_tp=!!input_tp!!:measured_LRA=!!input_lra!!:measured_thresh=!!input_thresh!!"  -c:a aac -ar 48000 -b:a 192k -ac 2 -c:v copy "!!OUTFILE!!"'

    runstring = runstring.replace("!!input_i!!", input_i)
    runstring = runstring.replace("!!input_tp!!", input_tp)
    runstring = runstring.replace("!!input_lra!!", input_lra)
    runstring = runstring.replace("!!input_thresh!!", input_thresh)
    runstring = runstring.replace("!!OUTFILE!!", str(outfile))
    runstring = runstring.replace("!!PATHHERE!!", pathhere)
    runstring = runstring.replace("!!JUSTFILE!!", justfile)
    runstring = runstring.replace("!!CURRENT_DATE_TIME!!", current_date_time)

    print(crlf)
    print(CRED + "Step 3/3) Applying LOUDNORM... " + CYELLOW + str(infile) + " to " + str(outfile) + CEND)
    system("title " + "Step 3/3) Applying LOUDNORM   [" + str(infile) + "]   to   [" + str(outfile) + "]")
    print(CRED + "Using: " + CEND + CITALIC + runstring + CEND)
    print(crlf)
 
    subprocess.call(runstring,shell=True)

#cleanup
if os.path.exists(pathhere + "\encodethis.mkv"):
    os.remove(pathhere + "\encodethis.mkv")

if os.path.exists(pathhere + "\loudnorm.txt"):
    os.remove(pathhere + "\loudnorm.txt")
        

#pause
input("Press Enter to continue...")

[_Al_]

I agree, definitely automatizing works better,

I simplified and cleared your version for better reading:
-paths are manipulated as Path objects
-ffmpeg does not need to create "loudnorm.txt", text is just directly passed in Python from ffmpeg
-last cmd was hanging for me, but f used subprocess.Popen instead subprocess.call, it was working, but I left it there as you wrote it using subprocess.call

"""
Encode BD Archive to Media Player MP4 with Loudnorm
Version 1.1
"""
# import required module
from pathlib import Path
import os
import subprocess
import json
from datetime import datetime

print("Encode BD Archive to Media Player MP4 with Loudnorm")

#initial setup
pathhere          = Path(__file__).parent
sourcedirectory   = Path("H:\\BD-HEVC_ARCHIVE")
destdirectory     = Path("W:\\NEW_VIDEOS\\BD-HEVC")
pattern           = "*.mkv"
outfile_handbrake = pathhere / "encodethis.mkv"
outfile_final     = lambda file: destdirectory / f"{file.stem}-HEVC.MP4"
handbrake         = pathhere / "handbrakecli"
ffmpeg            = pathhere / "ffmpeg"



files = list(sourcedirectory.glob(pattern))
for file in files:
    print(file)

for file in files:
    outfile_handbrake.unlink(missing_ok=True)   #delete file if exists

##HANDBRAKE CMD
    runstring = f'"{handbrake}" --input "{file}" --output "{outfile_handbrake}"'\
                f' --encoder nvenc_h265 --quality 30 --crop 0:0:0:0 --non-anamorphic --aencoder eac3 --ab 448k --mixdown 5point1 '    
    subprocess.call(runstring,shell=True)


##FFMPEG CMD
    runstring = f'"{ffmpeg}" -hide_banner -i "{outfile_handbrake}"'\
                f' -nostats -af "loudnorm=print_format=json" -f null null'
    p = subprocess.Popen(runstring, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    stdout, stderr = p.communicate() #ffmpeg writes into stderr
    #print(stderr.decode())
    #remove all but JSON text
    before, sep, after = stderr.decode().partition('{')
    data = "{" + after
    y = json.loads(data)

    time_now = datetime.now()
    current_time = time_now.strftime("%H:%M:%S")
    current_date = time_now.strftime("%m-%d-%Y")
    current_date_time = current_date + " " + current_time

##FFMPEG CMD
    runstring = f'"{ffmpeg}"  -hide_banner -fflags +genpts -i "{outfile_handbrake}"'\
                f' -metadata title="{file.stem}-HEVC" -metadata comment="{current_date_time}"'\
                f' -filter:a "loudnorm=linear=true:i=-23.0:lra=7.0:tp=-2.0:offset=0.39:'\
                f'measured_I={y["input_i"]}:measured_tp={y["input_tp"]}:measured_LRA={y["input_lra"]}:measured_thresh={y["input_thresh"]}"'\
                f' -c:a aac -ar 48000 -b:a 192k -ac 2 -c:v copy "{outfile_final(file)}"'
    subprocess.call(runstring,shell=True)

#cleanup
outfile_handbrake.unlink(missing_ok=True)

The whole script with your printing lines:

"""
Encode BD Archive to Media Player MP4 with Loudnorm
Version 1.1
"""
# import required module
from pathlib import Path
import os
import subprocess
from os import system
import json
from datetime import datetime


#setup some colors
CEND      = '\33[0m'
CRED    = '\33[31m'
CYELLOW = '\33[33m'
CBOLD     = '\33[1m'
CITALIC   = '\33[3m'

#setup some variables
crlf = "\r\n"

os.system('cls')

print("Encode BD Archive to Media Player MP4 with Loudnorm")
print(crlf)

#initial setup
pathhere          = Path(__file__).parent
sourcedirectory   = Path("H:\\BD-HEVC_ARCHIVE")
destdirectory     = Path("W:\\NEW_VIDEOS\\BD-HEVC")
pattern           = "*.mkv"
outfile_handbrake = pathhere / "encodethis.mkv"
outfile_final     = lambda file: destdirectory / f"{file.stem}-HEVC.MP4"
handbrake         = pathhere / "handbrakecli"
ffmpeg            = pathhere / "ffmpeg"



 
print("The source directory is: ", sourcedirectory)
print("The destination directory is: ", destdirectory)
print("The file pattern is:", pattern)
print(crlf)
print(CITALIC + CBOLD + "The following source files were discovered:" + CEND)

files = list(sourcedirectory.glob(pattern))
for file in files:
    print(file)

print(crlf)

for file in files:
    outfile_handbrake.unlink(missing_ok=True)
    runstring = f'"{handbrake}" --input "{file}" --output "{outfile_handbrake}"'\
                f' --encoder nvenc_h265 --quality 30 --crop 0:0:0:0 --non-anamorphic --aencoder eac3 --ab 448k --mixdown 5point1 '
    
    print(crlf)
    print(CRED + "Step 1/3) Media Player Encoding " + CYELLOW + str(file) + " to " + str(outfile) + CEND)
    system("title " + "Step 1/3) Media Player Encoding   [" + str(file) + "]   to   [" + str(outfile) + "]")
    print(CRED + "Using: " + CEND + CITALIC + runstring + CEND)
    print(crlf)
    
    subprocess.call(runstring,shell=True)

    #The base file is encoded with the desired video, lets run FFMPEG on it and determine the LOUDNORM settings
    runstring = f'"{ffmpeg}" -hide_banner -i "{outfile_handbrake}"'\
                f' -nostats -af "loudnorm=print_format=json" -f null null'
    
    print(crlf)
    print(CRED + "Step 2/3) Reading LOUDNORM... " + CYELLOW + str(file) + " to " + str(outfile) + CEND)
    system("title " + "Step 2/3) Reading LOUDNORM   [" + str(file) + "]   to   [" + str(outfile) + "]")
    print(CRED + "Using: " + CEND + CITALIC + runstring + CEND)
    print(crlf)
    p = subprocess.Popen(runstring, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
    stdout, stderr = p.communicate() #ffmpeg writes into stderr
    #print(stderr.decode())
 
    #remove all but JSON text
    before, sep, after = stderr.decode().partition('{')
    data = "{" + after

    #parse
    print(crlf)
    print("Parsing Data: " + data)
    print(crlf)

    y = json.loads(data)

    input_i = y["input_i"]
    input_tp = y["input_tp"]
    input_lra = y["input_lra"]
    input_thresh = y["input_thresh"]

    #print(input_i)
    #print(input_tp)
    #print(input_lra)
    #print(input_thresh)
    
    #Get current date and time
    time_now = datetime.now()
    current_time = time_now.strftime("%H:%M:%S")
    #print("The current time is", current_time)
    current_date = time_now.strftime("%m-%d-%Y")
    #print("The current date is", current_date)
    current_date_time = current_date + " " + current_time

    runstring = f'"{ffmpeg}"  -hide_banner -fflags +genpts -i "{outfile_handbrake}"'\
                f' -metadata title="{file.stem}-HEVC" -metadata comment="{current_date_time}"'\
                f' -filter:a "loudnorm=linear=true:i=-23.0:lra=7.0:tp=-2.0:offset=0.39:'\
                f'measured_I={input_i}:measured_tp={input_tp}:measured_LRA={input_lra}:measured_thresh={input_thresh}"'\
                f' -c:a aac -ar 48000 -b:a 192k -ac 2 -c:v copy "{outfile_final(file)}"'

    print(crlf)
    print(CRED + "Step 3/3) Applying LOUDNORM... " + CYELLOW + str(file) + " to " + str(outfile) + CEND)
    system("title " + "Step 3/3) Applying LOUDNORM   [" + str(file) + "]   to   [" + str(outfile) + "]")
    print(CRED + "Using: " + CEND + CITALIC + runstring + CEND)
    print(crlf)
 
    subprocess.call(runstring,shell=True)

#cleanup
outfile_handbrake.unlink(missing_ok=True)
        

#pause
input("Press Enter to continue...")

[RogerTango]

Wow, Al! I am not the only Phython person here!

Nice work, I didn't know how to capture stdout and parse it, you taught me a few things! I may have some questions for you later..

Thanks sir!
Andrew

[RogerTango]

Al,
Ive been reading your code and there is a lot I still have to learn, one thing is Python is way more powerful than I first thought!!

I have a question, I did read up on lamda but I cannot grasp the concept. Could you put in layman's terms what you did here?

outfile_final     = lambda file: destdirectory / f"{file.stem}-HEVC.MP4"

Thanks,
Andrew

[_Al_]

lambda is anonymous function, it is a function basically

outfile_final = lambda file: destdirectory /  f"{file.stem}-HEVC.MP4"
path = outfile_final(file)

is the same as:

def outfile_final(file):
    return  destdirectory  / f"{file.stem}-HEVC.MP4"
path = outfile_final(file)

So the latter example, there is a function named: outfile_final, but lambda function does not have a name

[_Al_]

Also that f" ....." is called f-string, introduced in some python3 version, very powerful also, if you google for it.

If using it , like f"....{variable} ....", it just executes whatever is inside of those complex brackets and looks what is a __str__ representation for that object class, that is what I think, it substitutes that object to a string and pastes that string in that whole string.
It is much faster than: something + " " + other_something
Better: f"{something} {other_something}"

So if you have a pathlib Path object included in those f-strings, it gets turned into a string, no problem

sourcedirectory   = Path("H:\\BD-HEVC_ARCHIVE")
string  = f'"{sourcedirectory}"'

Some apps might not accept Path as a path variable (they are expecting a string only), but that is not our case in your code. Then you'd just pass str(sourcedirectory) to be sure.
Using pathlib clears the code immensely.

[RogerTango]

Is file.stem the file name without the extension?

[_Al_]

yes
https://docs.python.org/3/library/pathlib.html#pathlib.PurePath.stem
https://docs.python.org/3/library/pathlib.html#correspondence-to-tools-in-the-os-module

[RogerTango]

yes
https://docs.python.org/3/library/pathlib.html#pathlib.PurePath.stem
https://docs.python.org/3/library/pathlib.html#correspondence-to-tools-in-the-os-module

Very good! I have added those two links to my bookmarks, many thanks!

Andrew